version 8.2.0
@@ -0,0 +1,1394 @@ | ||
1 | +<?xml version="1.0" encoding="UTF-8" ?> | |
2 | +<!-- | |
3 | + Licensed to the Apache Software Foundation (ASF) under one or more | |
4 | + contributor license agreements. See the NOTICE file distributed with | |
5 | + this work for additional information regarding copyright ownership. | |
6 | + The ASF licenses this file to You under the Apache License, Version 2.0 | |
7 | + (the "License"); you may not use this file except in compliance with | |
8 | + the License. You may obtain a copy of the License at | |
9 | + | |
10 | + http://www.apache.org/licenses/LICENSE-2.0 | |
11 | + | |
12 | + Unless required by applicable law or agreed to in writing, software | |
13 | + distributed under the License is distributed on an "AS IS" BASIS, | |
14 | + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
15 | + See the License for the specific language governing permissions and | |
16 | + limitations under the License. | |
17 | +--> | |
18 | + | |
19 | +<!-- | |
20 | + This is the Solr schema file. This file should be named "schema.xml" and | |
21 | + should be in the conf directory under the solr home | |
22 | + (i.e. ./solr/conf/schema.xml by default) | |
23 | + or located where the classloader for the Solr webapp can find it. | |
24 | + | |
25 | + This example schema is the recommended starting point for users. | |
26 | + It should be kept correct and concise, usable out-of-the-box. | |
27 | + | |
28 | + For more information, on how to customize this file, please see | |
29 | + http://wiki.apache.org/solr/SchemaXml | |
30 | + | |
31 | + PERFORMANCE NOTE: this schema includes many optional features and should not | |
32 | + be used for benchmarking. To improve performance one could | |
33 | + - set stored="false" for all fields possible (esp large fields) when you | |
34 | + only need to search on the field but don't need to return the original | |
35 | + value. | |
36 | + - set indexed="false" if you don't need to search on the field, but only | |
37 | + return the field as a result of searching on other indexed fields. | |
38 | + - remove all unneeded copyField statements | |
39 | + - for best index size and searching performance, set "index" to false | |
40 | + for all general text fields, use copyField to copy them to the | |
41 | + catchall "text" field, and use that for searching. | |
42 | + - For maximum indexing performance, use the StreamingUpdateSolrServer | |
43 | + java client. | |
44 | + - Remember to run the JVM in server mode, and use a higher logging level | |
45 | + that avoids logging every request | |
46 | +--> | |
47 | + | |
48 | +<schema name="fess" version="1.5"> | |
49 | + <!-- attribute "name" is the name of this schema and is only used for display purposes. | |
50 | + version="x.y" is Solr's version number for the schema syntax and | |
51 | + semantics. It should not normally be changed by applications. | |
52 | + | |
53 | + 1.0: multiValued attribute did not exist, all fields are multiValued | |
54 | + by nature | |
55 | + 1.1: multiValued attribute introduced, false by default | |
56 | + 1.2: omitTermFreqAndPositions attribute introduced, true by default | |
57 | + except for text fields. | |
58 | + 1.3: removed optional field compress feature | |
59 | + 1.4: autoGeneratePhraseQueries attribute introduced to drive QueryParser | |
60 | + behavior when a single string produces multiple tokens. Defaults | |
61 | + to off for version >= 1.4 | |
62 | + 1.5: omitNorms defaults to true for primitive field types | |
63 | + (int, float, boolean, string...) | |
64 | + --> | |
65 | + | |
66 | + <fields> | |
67 | + <!-- Valid attributes for fields: | |
68 | + name: mandatory - the name for the field | |
69 | + type: mandatory - the name of a field type from the | |
70 | + <types> fieldType section | |
71 | + indexed: true if this field should be indexed (searchable or sortable) | |
72 | + stored: true if this field should be retrievable | |
73 | + docValues: true if this field should have doc values. Doc values are | |
74 | + useful for faceting, grouping, sorting and function queries. Although not | |
75 | + required, doc values will make the index faster to load, more | |
76 | + NRT-friendly and more memory-efficient. They however come with some | |
77 | + limitations: they are currently only supported by StrField, UUIDField | |
78 | + and all Trie*Fields, and depending on the field type, they might | |
79 | + require the field to be single-valued, be required or have a default | |
80 | + value (check the documentation of the field type you're interested in | |
81 | + for more information) | |
82 | + multiValued: true if this field may contain multiple values per document | |
83 | + omitNorms: (expert) set to true to omit the norms associated with | |
84 | + this field (this disables length normalization and index-time | |
85 | + boosting for the field, and saves some memory). Only full-text | |
86 | + fields or fields that need an index-time boost need norms. | |
87 | + Norms are omitted for primitive (non-analyzed) types by default. | |
88 | + termVectors: [false] set to true to store the term vector for a | |
89 | + given field. | |
90 | + When using MoreLikeThis, fields used for similarity should be | |
91 | + stored for best performance. | |
92 | + termPositions: Store position information with the term vector. | |
93 | + This will increase storage costs. | |
94 | + termOffsets: Store offset information with the term vector. This | |
95 | + will increase storage costs. | |
96 | + required: The field is required. It will throw an error if the | |
97 | + value does not exist | |
98 | + default: a value that should be used if no value is specified | |
99 | + when adding a document. | |
100 | + --> | |
101 | + | |
102 | + <field name="id" type="string" stored="true" indexed="true" required="true" multiValued="false" docValues="true" /> | |
103 | + <!-- core fields --> | |
104 | + <field name="parentId" type="string" stored="true" indexed="true"/> | |
105 | + <field name="segment" type="string" stored="true" indexed="true"/> | |
106 | + <field name="digest" type="string" stored="true" indexed="false"/> | |
107 | + <field name="boost" type="float" stored="true" indexed="false"/> | |
108 | + <field name="host" type="domain_name" stored="true" indexed="true"/> | |
109 | + <field name="site" type="string" stored="true" indexed="false"/> | |
110 | + <field name="url" type="string" stored="true" indexed="true" required="true"/> | |
111 | + <field name="content" type="text" stored="true" indexed="true" termVectors="true"/> | |
112 | + <field name="title" type="text" stored="true" indexed="true" termVectors="true"/> | |
113 | + <field name="cache" type="string" stored="true" indexed="false"/> | |
114 | + <field name="tstamp" type="tdate" stored="true" indexed="true"/> | |
115 | + <field name="anchor" type="string" stored="true" indexed="true" multiValued="true"/> | |
116 | + <field name="contentLength" type="tlong" stored="true" indexed="true"/> | |
117 | + <field name="lastModified" type="tdate" stored="true" indexed="true"/> | |
118 | + <field name="lang" type="string" stored="true" indexed="true"/> | |
119 | + <field name="mimetype" type="string" stored="true" indexed="true"/> | |
120 | + <field name="_version_" type="long" indexed="true" stored="true"/> | |
121 | + <!-- multi values --> | |
122 | + <field name="type" type="string" stored="true" indexed="true" multiValued="true"/> | |
123 | + <field name="label" type="string" stored="true" indexed="true" multiValued="true"/> | |
124 | + <field name="role" type="string" stored="true" indexed="true" multiValued="true"/> | |
125 | + <!-- geo search --> | |
126 | + <field name="location" type="location" indexed="true" stored="true"/> | |
127 | + <!-- Language Dynamic field definitions --> | |
128 | + <dynamicField name="*_ar" type="text_ar" indexed="true" stored="true" termVectors="true"/> | |
129 | + <dynamicField name="*_bg" type="text_bg" indexed="true" stored="true" termVectors="true"/> | |
130 | + <dynamicField name="*_ca" type="text_ca" indexed="true" stored="true" termVectors="true"/> | |
131 | + <dynamicField name="*_cjk" type="text_cjk" indexed="true" stored="true" termVectors="true"/> | |
132 | + <dynamicField name="*_cz" type="text_cz" indexed="true" stored="true" termVectors="true"/> | |
133 | + <dynamicField name="*_da" type="text_da" indexed="true" stored="true" termVectors="true"/> | |
134 | + <dynamicField name="*_de" type="text_de" indexed="true" stored="true" termVectors="true"/> | |
135 | + <dynamicField name="*_el" type="text_el" indexed="true" stored="true" termVectors="true"/> | |
136 | + <dynamicField name="*_en" type="text_en" indexed="true" stored="true" termVectors="true"/> | |
137 | + <dynamicField name="*_es" type="text_es" indexed="true" stored="true" termVectors="true"/> | |
138 | + <dynamicField name="*_eu" type="text_eu" indexed="true" stored="true" termVectors="true"/> | |
139 | + <dynamicField name="*_fa" type="text_fa" indexed="true" stored="true" termVectors="true"/> | |
140 | + <dynamicField name="*_fi" type="text_fi" indexed="true" stored="true" termVectors="true"/> | |
141 | + <dynamicField name="*_fr" type="text_fr" indexed="true" stored="true" termVectors="true"/> | |
142 | + <dynamicField name="*_ga" type="text_ga" indexed="true" stored="true" termVectors="true"/> | |
143 | + <dynamicField name="*_gl" type="text_gl" indexed="true" stored="true" termVectors="true"/> | |
144 | + <dynamicField name="*_hi" type="text_hi" indexed="true" stored="true" termVectors="true"/> | |
145 | + <dynamicField name="*_hu" type="text_hu" indexed="true" stored="true" termVectors="true"/> | |
146 | + <dynamicField name="*_hy" type="text_hy" indexed="true" stored="true" termVectors="true"/> | |
147 | + <dynamicField name="*_id" type="text_id" indexed="true" stored="true" termVectors="true"/> | |
148 | + <dynamicField name="*_it" type="text_it" indexed="true" stored="true" termVectors="true"/> | |
149 | + <dynamicField name="*_ja" type="text_ja" indexed="true" stored="true" termVectors="true"/> | |
150 | + <dynamicField name="*_lv" type="text_lv" indexed="true" stored="true" termVectors="true"/> | |
151 | + <dynamicField name="*_nl" type="text_nl" indexed="true" stored="true" termVectors="true"/> | |
152 | + <dynamicField name="*_no" type="text_no" indexed="true" stored="true" termVectors="true"/> | |
153 | + <dynamicField name="*_pt" type="text_pt" indexed="true" stored="true" termVectors="true"/> | |
154 | + <dynamicField name="*_ro" type="text_ro" indexed="true" stored="true" termVectors="true"/> | |
155 | + <dynamicField name="*_ru" type="text_ru" indexed="true" stored="true" termVectors="true"/> | |
156 | + <dynamicField name="*_sv" type="text_sv" indexed="true" stored="true" termVectors="true"/> | |
157 | + <dynamicField name="*_th" type="text_th" indexed="true" stored="true" termVectors="true"/> | |
158 | + <dynamicField name="*_tr" type="text_tr" indexed="true" stored="true" termVectors="true"/> | |
159 | + <dynamicField name="*_ws" type="text_ws" indexed="true" stored="true" termVectors="true"/> | |
160 | + <dynamicField name="*_gosen" type="text_gosen" indexed="true" stored="true" termVectors="true"/> | |
161 | + <!-- Dynamic field definitions --> | |
162 | + <dynamicField name="*_s" type="string" indexed="true" stored="true"/> | |
163 | + <dynamicField name="*_t" type="text" indexed="true" stored="true"/> | |
164 | + <dynamicField name="*_b" type="boolean" indexed="true" stored="true"/> | |
165 | + <dynamicField name="*_i" type="int" indexed="true" stored="true"/> | |
166 | + <dynamicField name="*_l" type="long" indexed="true" stored="true"/> | |
167 | + <dynamicField name="*_f" type="float" indexed="true" stored="true"/> | |
168 | + <dynamicField name="*_d" type="double" indexed="true" stored="true"/> | |
169 | + <dynamicField name="*_dt" type="date" indexed="true" stored="true"/> | |
170 | + <dynamicField name="*_ti" type="tint" indexed="true" stored="true"/> | |
171 | + <dynamicField name="*_tl" type="tlong" indexed="true" stored="true"/> | |
172 | + <dynamicField name="*_tf" type="tfloat" indexed="true" stored="true"/> | |
173 | + <dynamicField name="*_td" type="tdouble" indexed="true" stored="true"/> | |
174 | + <dynamicField name="*_tdt" type="tdate" indexed="true" stored="true"/> | |
175 | + <dynamicField name="*_coordinate" type="tdouble" indexed="true" stored="false"/> | |
176 | + <dynamicField name="ignored_*" type="ignored" multiValued="true"/> | |
177 | + <dynamicField name="random_*" type="random"/> | |
178 | + <dynamicField name="*_t_ar" type="text_ar" indexed="true" stored="true"/> | |
179 | + <dynamicField name="*_t_bg" type="text_bg" indexed="true" stored="true"/> | |
180 | + <dynamicField name="*_t_ca" type="text_ca" indexed="true" stored="true"/> | |
181 | + <dynamicField name="*_t_char_norm" type="text_char_norm" indexed="true" stored="true"/> | |
182 | + <dynamicField name="*_t_cjk" type="text_cjk" indexed="true" stored="true"/> | |
183 | + <dynamicField name="*_t_cz" type="text_cz" indexed="true" stored="true"/> | |
184 | + <dynamicField name="*_t_da" type="text_da" indexed="true" stored="true"/> | |
185 | + <dynamicField name="*_t_de" type="text_de" indexed="true" stored="true"/> | |
186 | + <dynamicField name="*_t_el" type="text_el" indexed="true" stored="true"/> | |
187 | + <dynamicField name="*_t_en" type="text_en" indexed="true" stored="true"/> | |
188 | + <dynamicField name="*_t_en_splitting" type="text_en_splitting" indexed="true" stored="true"/> | |
189 | + <dynamicField name="*_t_en_splitting_tight" type="text_en_splitting_tight" indexed="true" stored="true"/> | |
190 | + <dynamicField name="*_t_es" type="text_es" indexed="true" stored="true"/> | |
191 | + <dynamicField name="*_t_eu" type="text_eu" indexed="true" stored="true"/> | |
192 | + <dynamicField name="*_t_fa" type="text_fa" indexed="true" stored="true"/> | |
193 | + <dynamicField name="*_t_fi" type="text_fi" indexed="true" stored="true"/> | |
194 | + <dynamicField name="*_t_fr" type="text_fr" indexed="true" stored="true"/> | |
195 | + <dynamicField name="*_t_ga" type="text_ga" indexed="true" stored="true"/> | |
196 | + <dynamicField name="*_t_general" type="text_general" indexed="true" stored="true"/> | |
197 | + <dynamicField name="*_t_general_rev" type="text_general_rev" indexed="true" stored="true"/> | |
198 | + <dynamicField name="*_t_gl" type="text_gl" indexed="true" stored="true"/> | |
199 | + <dynamicField name="*_t_gosen" type="text_gosen" indexed="true" stored="true"/> | |
200 | + <dynamicField name="*_t_greek" type="text_greek" indexed="true" stored="true"/> | |
201 | + <dynamicField name="*_t_hi" type="text_hi" indexed="true" stored="true"/> | |
202 | + <dynamicField name="*_t_hu" type="text_hu" indexed="true" stored="true"/> | |
203 | + <dynamicField name="*_t_hy" type="text_hy" indexed="true" stored="true"/> | |
204 | + <dynamicField name="*_t_id" type="text_id" indexed="true" stored="true"/> | |
205 | + <dynamicField name="*_t_it" type="text_it" indexed="true" stored="true"/> | |
206 | + <dynamicField name="*_t_ja" type="text_ja" indexed="true" stored="true"/> | |
207 | + <dynamicField name="*_t_lv" type="text_lv" indexed="true" stored="true"/> | |
208 | + <dynamicField name="*_t_nl" type="text_nl" indexed="true" stored="true"/> | |
209 | + <dynamicField name="*_t_no" type="text_no" indexed="true" stored="true"/> | |
210 | + <dynamicField name="*_t_pt" type="text_pt" indexed="true" stored="true"/> | |
211 | + <dynamicField name="*_t_ro" type="text_ro" indexed="true" stored="true"/> | |
212 | + <dynamicField name="*_t_ru" type="text_ru" indexed="true" stored="true"/> | |
213 | + <dynamicField name="*_t_sv" type="text_sv" indexed="true" stored="true"/> | |
214 | + <dynamicField name="*_t_th" type="text_th" indexed="true" stored="true"/> | |
215 | + <dynamicField name="*_t_tr" type="text_tr" indexed="true" stored="true"/> | |
216 | + <dynamicField name="*_t_ws" type="text_ws" indexed="true" stored="true"/> | |
217 | + <dynamicField name="*_a_path" type="ancestor_path" indexed="true" stored="true"/> | |
218 | + <dynamicField name="*_d_path" type="descendent_path" indexed="true" stored="true"/> | |
219 | + <!-- Dynamic field definitions (Index only) --> | |
220 | + <dynamicField name="*_s_i" type="string" indexed="true" stored="false"/> | |
221 | + <dynamicField name="*_t_i" type="text" indexed="true" stored="false"/> | |
222 | + <dynamicField name="*_b_i" type="boolean" indexed="true" stored="false"/> | |
223 | + <dynamicField name="*_i_i" type="int" indexed="true" stored="false"/> | |
224 | + <dynamicField name="*_l_i" type="long" indexed="true" stored="false"/> | |
225 | + <dynamicField name="*_f_i" type="float" indexed="true" stored="false"/> | |
226 | + <dynamicField name="*_d_i" type="double" indexed="true" stored="false"/> | |
227 | + <dynamicField name="*_ti_i" type="tint" indexed="true" stored="false"/> | |
228 | + <dynamicField name="*_tl_i" type="tlong" indexed="true" stored="false"/> | |
229 | + <dynamicField name="*_tf_i" type="tfloat" indexed="true" stored="false"/> | |
230 | + <dynamicField name="*_td_i" type="tdouble" indexed="true" stored="false"/> | |
231 | + <dynamicField name="*_tdt_i" type="tdate" indexed="true" stored="false"/> | |
232 | + <dynamicField name="*_t_ar_i" type="text_ar" indexed="true" stored="false"/> | |
233 | + <dynamicField name="*_t_bg_i" type="text_bg" indexed="true" stored="false"/> | |
234 | + <dynamicField name="*_t_ca_i" type="text_ca" indexed="true" stored="false"/> | |
235 | + <dynamicField name="*_t_char_norm_i" type="text_char_norm" indexed="true" stored="false"/> | |
236 | + <dynamicField name="*_t_cjk_i" type="text_cjk" indexed="true" stored="false"/> | |
237 | + <dynamicField name="*_t_cz_i" type="text_cz" indexed="true" stored="false"/> | |
238 | + <dynamicField name="*_t_da_i" type="text_da" indexed="true" stored="false"/> | |
239 | + <dynamicField name="*_t_de_i" type="text_de" indexed="true" stored="false"/> | |
240 | + <dynamicField name="*_t_el_i" type="text_el" indexed="true" stored="false"/> | |
241 | + <dynamicField name="*_t_en_i" type="text_en" indexed="true" stored="false"/> | |
242 | + <dynamicField name="*_t_en_splitting_i" type="text_en_splitting" indexed="true" stored="false"/> | |
243 | + <dynamicField name="*_t_en_splitting_tight_i" type="text_en_splitting_tight" indexed="true" stored="false"/> | |
244 | + <dynamicField name="*_t_es_i" type="text_es" indexed="true" stored="false"/> | |
245 | + <dynamicField name="*_t_eu_i" type="text_eu" indexed="true" stored="false"/> | |
246 | + <dynamicField name="*_t_fa_i" type="text_fa" indexed="true" stored="false"/> | |
247 | + <dynamicField name="*_t_fi_i" type="text_fi" indexed="true" stored="false"/> | |
248 | + <dynamicField name="*_t_fr_i" type="text_fr" indexed="true" stored="false"/> | |
249 | + <dynamicField name="*_t_ga_i" type="text_ga" indexed="true" stored="false"/> | |
250 | + <dynamicField name="*_t_general_i" type="text_general" indexed="true" stored="false"/> | |
251 | + <dynamicField name="*_t_general_rev_i" type="text_general_rev" indexed="true" stored="false"/> | |
252 | + <dynamicField name="*_t_gl_i" type="text_gl" indexed="true" stored="false"/> | |
253 | + <dynamicField name="*_t_gosen_i" type="text_gosen" indexed="true" stored="false"/> | |
254 | + <dynamicField name="*_t_greek_i" type="text_greek" indexed="true" stored="false"/> | |
255 | + <dynamicField name="*_t_hi_i" type="text_hi" indexed="true" stored="false"/> | |
256 | + <dynamicField name="*_t_hu_i" type="text_hu" indexed="true" stored="false"/> | |
257 | + <dynamicField name="*_t_hy_i" type="text_hy" indexed="true" stored="false"/> | |
258 | + <dynamicField name="*_t_id_i" type="text_id" indexed="true" stored="false"/> | |
259 | + <dynamicField name="*_t_it_i" type="text_it" indexed="true" stored="false"/> | |
260 | + <dynamicField name="*_t_ja_i" type="text_ja" indexed="true" stored="false"/> | |
261 | + <dynamicField name="*_t_lv_i" type="text_lv" indexed="true" stored="false"/> | |
262 | + <dynamicField name="*_t_nl_i" type="text_nl" indexed="true" stored="false"/> | |
263 | + <dynamicField name="*_t_no_i" type="text_no" indexed="true" stored="false"/> | |
264 | + <dynamicField name="*_t_pt_i" type="text_pt" indexed="true" stored="false"/> | |
265 | + <dynamicField name="*_t_ro_i" type="text_ro" indexed="true" stored="false"/> | |
266 | + <dynamicField name="*_t_ru_i" type="text_ru" indexed="true" stored="false"/> | |
267 | + <dynamicField name="*_t_sv_i" type="text_sv" indexed="true" stored="false"/> | |
268 | + <dynamicField name="*_t_th_i" type="text_th" indexed="true" stored="false"/> | |
269 | + <dynamicField name="*_t_tr_i" type="text_tr" indexed="true" stored="false"/> | |
270 | + <dynamicField name="*_t_ws_i" type="text_ws" indexed="true" stored="false"/> | |
271 | + <dynamicField name="*_i_a_path" type="ancestor_path" indexed="true" stored="false"/> | |
272 | + <dynamicField name="*_i_d_path" type="descendent_path" indexed="true" stored="false"/> | |
273 | + <!-- Dynamic field definitions (Store only) --> | |
274 | + <dynamicField name="*_s_s" type="string" indexed="false" stored="true"/> | |
275 | + <dynamicField name="*_t_s" type="text" indexed="false" stored="true"/> | |
276 | + <dynamicField name="*_b_s" type="boolean" indexed="false" stored="true"/> | |
277 | + <dynamicField name="*_i_s" type="int" indexed="false" stored="true"/> | |
278 | + <dynamicField name="*_l_s" type="long" indexed="false" stored="true"/> | |
279 | + <dynamicField name="*_f_s" type="float" indexed="false" stored="true"/> | |
280 | + <dynamicField name="*_d_s" type="double" indexed="false" stored="true"/> | |
281 | + <dynamicField name="*_ti_s" type="tint" indexed="false" stored="true"/> | |
282 | + <dynamicField name="*_tl_s" type="tlong" indexed="false" stored="true"/> | |
283 | + <dynamicField name="*_tf_s" type="tfloat" indexed="false" stored="true"/> | |
284 | + <dynamicField name="*_td_s" type="tdouble" indexed="false" stored="true"/> | |
285 | + <dynamicField name="*_tdt_s" type="tdate" indexed="false" stored="true"/> | |
286 | + <!-- Dynamic multi-valued field definitions --> | |
287 | + <dynamicField name="*_s_m" type="string" indexed="true" stored="true" multiValued="true"/> | |
288 | + <dynamicField name="*_t_m" type="text" indexed="true" stored="true" multiValued="true"/> | |
289 | + <dynamicField name="*_b_m" type="boolean" indexed="true" stored="true" multiValued="true"/> | |
290 | + <dynamicField name="*_i_m" type="int" indexed="true" stored="true" multiValued="true"/> | |
291 | + <dynamicField name="*_l_m" type="long" indexed="true" stored="true" multiValued="true"/> | |
292 | + <dynamicField name="*_f_m" type="float" indexed="true" stored="true" multiValued="true"/> | |
293 | + <dynamicField name="*_d_m" type="double" indexed="true" stored="true" multiValued="true"/> | |
294 | + <dynamicField name="*_dt_m" type="date" indexed="true" stored="true" multiValued="true"/> | |
295 | + <dynamicField name="*_ti_m" type="tint" indexed="true" stored="true" multiValued="true"/> | |
296 | + <dynamicField name="*_tl_m" type="tlong" indexed="true" stored="true" multiValued="true"/> | |
297 | + <dynamicField name="*_tf_m" type="tfloat" indexed="true" stored="true" multiValued="true"/> | |
298 | + <dynamicField name="*_td_m" type="tdouble" indexed="true" stored="true" multiValued="true"/> | |
299 | + <dynamicField name="*_tdt_m" type="tdate" indexed="true" stored="true" multiValued="true"/> | |
300 | + <dynamicField name="*_t_ar_m" type="text_ar" indexed="true" stored="true" multiValued="true"/> | |
301 | + <dynamicField name="*_t_bg_m" type="text_bg" indexed="true" stored="true" multiValued="true"/> | |
302 | + <dynamicField name="*_t_ca_m" type="text_ca" indexed="true" stored="true" multiValued="true"/> | |
303 | + <dynamicField name="*_t_char_norm_m" type="text_char_norm" indexed="true" stored="true" multiValued="true"/> | |
304 | + <dynamicField name="*_t_cjk_m" type="text_cjk" indexed="true" stored="true" multiValued="true"/> | |
305 | + <dynamicField name="*_t_cz_m" type="text_cz" indexed="true" stored="true" multiValued="true"/> | |
306 | + <dynamicField name="*_t_da_m" type="text_da" indexed="true" stored="true" multiValued="true"/> | |
307 | + <dynamicField name="*_t_de_m" type="text_de" indexed="true" stored="true" multiValued="true"/> | |
308 | + <dynamicField name="*_t_el_m" type="text_el" indexed="true" stored="true" multiValued="true"/> | |
309 | + <dynamicField name="*_t_en_m" type="text_en" indexed="true" stored="true" multiValued="true"/> | |
310 | + <dynamicField name="*_t_en_splitting_m" type="text_en_splitting" indexed="true" stored="true" multiValued="true"/> | |
311 | + <dynamicField name="*_t_en_splitting_tight_m" type="text_en_splitting_tight" indexed="true" stored="true" multiValued="true"/> | |
312 | + <dynamicField name="*_t_es_m" type="text_es" indexed="true" stored="true" multiValued="true"/> | |
313 | + <dynamicField name="*_t_eu_m" type="text_eu" indexed="true" stored="true" multiValued="true"/> | |
314 | + <dynamicField name="*_t_fa_m" type="text_fa" indexed="true" stored="true" multiValued="true"/> | |
315 | + <dynamicField name="*_t_fi_m" type="text_fi" indexed="true" stored="true" multiValued="true"/> | |
316 | + <dynamicField name="*_t_fr_m" type="text_fr" indexed="true" stored="true" multiValued="true"/> | |
317 | + <dynamicField name="*_t_ga_m" type="text_ga" indexed="true" stored="true" multiValued="true"/> | |
318 | + <dynamicField name="*_t_general_m" type="text_general" indexed="true" stored="true" multiValued="true"/> | |
319 | + <dynamicField name="*_t_general_rev_m" type="text_general_rev" indexed="true" stored="true" multiValued="true"/> | |
320 | + <dynamicField name="*_t_gl_m" type="text_gl" indexed="true" stored="true" multiValued="true"/> | |
321 | + <dynamicField name="*_t_gosen_m" type="text_gosen" indexed="true" stored="true" multiValued="true"/> | |
322 | + <dynamicField name="*_t_greek_m" type="text_greek" indexed="true" stored="true" multiValued="true"/> | |
323 | + <dynamicField name="*_t_hi_m" type="text_hi" indexed="true" stored="true" multiValued="true"/> | |
324 | + <dynamicField name="*_t_hu_m" type="text_hu" indexed="true" stored="true" multiValued="true"/> | |
325 | + <dynamicField name="*_t_hy_m" type="text_hy" indexed="true" stored="true" multiValued="true"/> | |
326 | + <dynamicField name="*_t_id_m" type="text_id" indexed="true" stored="true" multiValued="true"/> | |
327 | + <dynamicField name="*_t_it_m" type="text_it" indexed="true" stored="true" multiValued="true"/> | |
328 | + <dynamicField name="*_t_ja_m" type="text_ja" indexed="true" stored="true" multiValued="true"/> | |
329 | + <dynamicField name="*_t_lv_m" type="text_lv" indexed="true" stored="true" multiValued="true"/> | |
330 | + <dynamicField name="*_t_nl_m" type="text_nl" indexed="true" stored="true" multiValued="true"/> | |
331 | + <dynamicField name="*_t_no_m" type="text_no" indexed="true" stored="true" multiValued="true"/> | |
332 | + <dynamicField name="*_t_pt_m" type="text_pt" indexed="true" stored="true" multiValued="true"/> | |
333 | + <dynamicField name="*_t_ro_m" type="text_ro" indexed="true" stored="true" multiValued="true"/> | |
334 | + <dynamicField name="*_t_ru_m" type="text_ru" indexed="true" stored="true" multiValued="true"/> | |
335 | + <dynamicField name="*_t_sv_m" type="text_sv" indexed="true" stored="true" multiValued="true"/> | |
336 | + <dynamicField name="*_t_th_m" type="text_th" indexed="true" stored="true" multiValued="true"/> | |
337 | + <dynamicField name="*_t_tr_m" type="text_tr" indexed="true" stored="true" multiValued="true"/> | |
338 | + <dynamicField name="*_t_ws_m" type="text_ws" indexed="true" stored="true" multiValued="true"/> | |
339 | + <dynamicField name="*_a_path_m" type="ancestor_path" indexed="true" stored="true" multiValued="true"/> | |
340 | + <dynamicField name="*_d_path_m" type="descendent_path" indexed="true" stored="true" multiValued="true"/> | |
341 | + <!-- Dynamic multi-valued field definitions (Index only) --> | |
342 | + <dynamicField name="*_s_i_m" type="string" indexed="true" stored="false" multiValued="true"/> | |
343 | + <dynamicField name="*_t_i_m" type="text" indexed="true" stored="false" multiValued="true"/> | |
344 | + <dynamicField name="*_b_i_m" type="boolean" indexed="true" stored="false" multiValued="true"/> | |
345 | + <dynamicField name="*_i_i_m" type="int" indexed="true" stored="false" multiValued="true"/> | |
346 | + <dynamicField name="*_l_i_m" type="long" indexed="true" stored="false" multiValued="true"/> | |
347 | + <dynamicField name="*_f_i_m" type="float" indexed="true" stored="false" multiValued="true"/> | |
348 | + <dynamicField name="*_d_i_m" type="double" indexed="true" stored="false" multiValued="true"/> | |
349 | + <dynamicField name="*_ti_i_m" type="tint" indexed="true" stored="false" multiValued="true"/> | |
350 | + <dynamicField name="*_tl_i_m" type="tlong" indexed="true" stored="false" multiValued="true"/> | |
351 | + <dynamicField name="*_tf_i_m" type="tfloat" indexed="true" stored="false" multiValued="true"/> | |
352 | + <dynamicField name="*_td_i_m" type="tdouble" indexed="true" stored="false" multiValued="true"/> | |
353 | + <dynamicField name="*_tdt_i_m" type="tdate" indexed="true" stored="false" multiValued="true"/> | |
354 | + <dynamicField name="*_t_ar_i_m" type="text_ar" indexed="true" stored="false" multiValued="true"/> | |
355 | + <dynamicField name="*_t_bg_i_m" type="text_bg" indexed="true" stored="false" multiValued="true"/> | |
356 | + <dynamicField name="*_t_ca_i_m" type="text_ca" indexed="true" stored="false" multiValued="true"/> | |
357 | + <dynamicField name="*_t_char_norm_i_m" type="text_char_norm" indexed="true" stored="false" multiValued="true"/> | |
358 | + <dynamicField name="*_t_cjk_i_m" type="text_cjk" indexed="true" stored="false" multiValued="true"/> | |
359 | + <dynamicField name="*_t_cz_i_m" type="text_cz" indexed="true" stored="false" multiValued="true"/> | |
360 | + <dynamicField name="*_t_da_i_m" type="text_da" indexed="true" stored="false" multiValued="true"/> | |
361 | + <dynamicField name="*_t_de_i_m" type="text_de" indexed="true" stored="false" multiValued="true"/> | |
362 | + <dynamicField name="*_t_el_i_m" type="text_el" indexed="true" stored="false" multiValued="true"/> | |
363 | + <dynamicField name="*_t_en_i_m" type="text_en" indexed="true" stored="false" multiValued="true"/> | |
364 | + <dynamicField name="*_t_en_splitting_i_m" type="text_en_splitting" indexed="true" stored="false" multiValued="true"/> | |
365 | + <dynamicField name="*_t_en_splitting_tight_i_m" type="text_en_splitting_tight" indexed="true" stored="false" multiValued="true"/> | |
366 | + <dynamicField name="*_t_es_i_m" type="text_es" indexed="true" stored="false" multiValued="true"/> | |
367 | + <dynamicField name="*_t_eu_i_m" type="text_eu" indexed="true" stored="false" multiValued="true"/> | |
368 | + <dynamicField name="*_t_fa_i_m" type="text_fa" indexed="true" stored="false" multiValued="true"/> | |
369 | + <dynamicField name="*_t_fi_i_m" type="text_fi" indexed="true" stored="false" multiValued="true"/> | |
370 | + <dynamicField name="*_t_fr_i_m" type="text_fr" indexed="true" stored="false" multiValued="true"/> | |
371 | + <dynamicField name="*_t_ga_i_m" type="text_ga" indexed="true" stored="false" multiValued="true"/> | |
372 | + <dynamicField name="*_t_general_i_m" type="text_general" indexed="true" stored="false" multiValued="true"/> | |
373 | + <dynamicField name="*_t_general_rev_i_m" type="text_general_rev" indexed="true" stored="false" multiValued="true"/> | |
374 | + <dynamicField name="*_t_gl_i_m" type="text_gl" indexed="true" stored="false" multiValued="true"/> | |
375 | + <dynamicField name="*_t_gosen_i_m" type="text_gosen" indexed="true" stored="false" multiValued="true"/> | |
376 | + <dynamicField name="*_t_greek_i_m" type="text_greek" indexed="true" stored="false" multiValued="true"/> | |
377 | + <dynamicField name="*_t_hi_i_m" type="text_hi" indexed="true" stored="false" multiValued="true"/> | |
378 | + <dynamicField name="*_t_hu_i_m" type="text_hu" indexed="true" stored="false" multiValued="true"/> | |
379 | + <dynamicField name="*_t_hy_i_m" type="text_hy" indexed="true" stored="false" multiValued="true"/> | |
380 | + <dynamicField name="*_t_id_i_m" type="text_id" indexed="true" stored="false" multiValued="true"/> | |
381 | + <dynamicField name="*_t_it_i_m" type="text_it" indexed="true" stored="false" multiValued="true"/> | |
382 | + <dynamicField name="*_t_ja_i_m" type="text_ja" indexed="true" stored="false" multiValued="true"/> | |
383 | + <dynamicField name="*_t_lv_i_m" type="text_lv" indexed="true" stored="false" multiValued="true"/> | |
384 | + <dynamicField name="*_t_nl_i_m" type="text_nl" indexed="true" stored="false" multiValued="true"/> | |
385 | + <dynamicField name="*_t_no_i_m" type="text_no" indexed="true" stored="false" multiValued="true"/> | |
386 | + <dynamicField name="*_t_pt_i_m" type="text_pt" indexed="true" stored="false" multiValued="true"/> | |
387 | + <dynamicField name="*_t_ro_i_m" type="text_ro" indexed="true" stored="false" multiValued="true"/> | |
388 | + <dynamicField name="*_t_ru_i_m" type="text_ru" indexed="true" stored="false" multiValued="true"/> | |
389 | + <dynamicField name="*_t_sv_i_m" type="text_sv" indexed="true" stored="false" multiValued="true"/> | |
390 | + <dynamicField name="*_t_th_i_m" type="text_th" indexed="true" stored="false" multiValued="true"/> | |
391 | + <dynamicField name="*_t_tr_i_m" type="text_tr" indexed="true" stored="false" multiValued="true"/> | |
392 | + <dynamicField name="*_t_ws_i_m" type="text_ws" indexed="true" stored="false" multiValued="true"/> | |
393 | + <dynamicField name="*_a_path_i_m" type="ancestor_path" indexed="true" stored="false" multiValued="true"/> | |
394 | + <dynamicField name="*_d_path_i_m" type="descendent_path" indexed="true" stored="false" multiValued="true"/> | |
395 | + <!-- Dynamic multi-valued field definitions (Store only) --> | |
396 | + <dynamicField name="*_s_s_m" type="string" indexed="false" stored="true" multiValued="true"/> | |
397 | + <dynamicField name="*_t_s_m" type="text" indexed="false" stored="true" multiValued="true"/> | |
398 | + <dynamicField name="*_b_s_m" type="boolean" indexed="false" stored="true" multiValued="true"/> | |
399 | + <dynamicField name="*_i_s_m" type="int" indexed="false" stored="true" multiValued="true"/> | |
400 | + <dynamicField name="*_l_s_m" type="long" indexed="false" stored="true" multiValued="true"/> | |
401 | + <dynamicField name="*_f_s_m" type="float" indexed="false" stored="true" multiValued="true"/> | |
402 | + <dynamicField name="*_d_s_m" type="double" indexed="false" stored="true" multiValued="true"/> | |
403 | + <dynamicField name="*_ti_s_m" type="tint" indexed="false" stored="true" multiValued="true"/> | |
404 | + <dynamicField name="*_tl_s_m" type="tlong" indexed="false" stored="true" multiValued="true"/> | |
405 | + <dynamicField name="*_tf_s_m" type="tfloat" indexed="false" stored="true" multiValued="true"/> | |
406 | + <dynamicField name="*_td_s_m" type="tdouble" indexed="false" stored="true" multiValued="true"/> | |
407 | + <dynamicField name="*_tdt_s_m" type="tdate" indexed="false" stored="true" multiValued="true"/> | |
408 | + <!-- Dynamic field definitions (compatibility) --> | |
409 | + <dynamicField name="*_pi" type="pint" indexed="true" stored="true"/> | |
410 | + <dynamicField name="*_pl" type="plong" indexed="true" stored="true"/> | |
411 | + <dynamicField name="*_pf" type="pfloat" indexed="true" stored="true"/> | |
412 | + <dynamicField name="*_pd" type="pdouble" indexed="true" stored="true"/> | |
413 | + <dynamicField name="*_pdt" type="pdate" indexed="true" stored="true"/> | |
414 | + <dynamicField name="*_tg" type="text_gosen" indexed="true" stored="true"/> | |
415 | + <dynamicField name="*_tg_i" type="text_gosen" indexed="true" stored="false"/> | |
416 | + <dynamicField name="*_tg_s" type="text_gosen" indexed="false" stored="true"/> | |
417 | + <dynamicField name="*_tg_m" type="text_gosen" indexed="true" stored="true" multiValued="true"/> | |
418 | + <dynamicField name="*_tg_i_m" type="text_gosen" indexed="true" stored="false" multiValued="true"/> | |
419 | + <dynamicField name="*_tg_s_m" type="text_gosen" indexed="false" stored="true" multiValued="true"/> | |
420 | + | |
421 | + <!-- Suggest --> | |
422 | + <dynamicField name="*_ts" type="text_suggest" indexed="true" stored="true"/> | |
423 | + <dynamicField name="*_ss" type="string_suggest" indexed="true" stored="true"/> | |
424 | +<!-- | |
425 | + <copyField source="content" dest="content_ts"/> | |
426 | + <copyField source="title" dest="title_ss"/> | |
427 | +--> | |
428 | + | |
429 | + </fields> | |
430 | + | |
431 | + <!-- Field to use to determine and enforce document uniqueness. | |
432 | + Unless this field is marked with required="false", it will be a required field | |
433 | + --> | |
434 | + <uniqueKey>id</uniqueKey> | |
435 | + | |
436 | + <types> | |
437 | + <!-- field type definitions. The "name" attribute is | |
438 | + just a label to be used by field definitions. The "class" | |
439 | + attribute and any other attributes determine the real | |
440 | + behavior of the fieldType. | |
441 | + Class names starting with "solr" refer to java classes in a | |
442 | + standard package such as org.apache.solr.analysis | |
443 | + --> | |
444 | + | |
445 | + <!-- The StrField type is not analyzed, but indexed/stored verbatim. | |
446 | + It supports doc values but in that case the field needs to be | |
447 | + single-valued and either required or have a default value. | |
448 | + --> | |
449 | + <fieldType name="string" class="solr.StrField" sortMissingLast="true" /> | |
450 | + | |
451 | + <!-- boolean type: "true" or "false" --> | |
452 | + <fieldType name="boolean" class="solr.BoolField" sortMissingLast="true"/> | |
453 | + | |
454 | + <!-- sortMissingLast and sortMissingFirst attributes are optional attributes are | |
455 | + currently supported on types that are sorted internally as strings | |
456 | + and on numeric types. | |
457 | + This includes "string","boolean", and, as of 3.5 (and 4.x), | |
458 | + int, float, long, date, double, including the "Trie" variants. | |
459 | + - If sortMissingLast="true", then a sort on this field will cause documents | |
460 | + without the field to come after documents with the field, | |
461 | + regardless of the requested sort order (asc or desc). | |
462 | + - If sortMissingFirst="true", then a sort on this field will cause documents | |
463 | + without the field to come before documents with the field, | |
464 | + regardless of the requested sort order. | |
465 | + - If sortMissingLast="false" and sortMissingFirst="false" (the default), | |
466 | + then default lucene sorting will be used which places docs without the | |
467 | + field first in an ascending sort and last in a descending sort. | |
468 | + --> | |
469 | + | |
470 | + <!-- | |
471 | + Default numeric field types. For faster range queries, consider the tint/tfloat/tlong/tdouble types. | |
472 | + | |
473 | + These fields support doc values, but they require the field to be | |
474 | + single-valued and either be required or have a default value. | |
475 | + --> | |
476 | + <fieldType name="int" class="solr.TrieIntField" precisionStep="0" positionIncrementGap="0"/> | |
477 | + <fieldType name="float" class="solr.TrieFloatField" precisionStep="0" positionIncrementGap="0"/> | |
478 | + <fieldType name="long" class="solr.TrieLongField" precisionStep="0" positionIncrementGap="0"/> | |
479 | + <fieldType name="double" class="solr.TrieDoubleField" precisionStep="0" positionIncrementGap="0"/> | |
480 | + | |
481 | + <!-- | |
482 | + Numeric field types that index each value at various levels of precision | |
483 | + to accelerate range queries when the number of values between the range | |
484 | + endpoints is large. See the javadoc for NumericRangeQuery for internal | |
485 | + implementation details. | |
486 | + | |
487 | + Smaller precisionStep values (specified in bits) will lead to more tokens | |
488 | + indexed per value, slightly larger index size, and faster range queries. | |
489 | + A precisionStep of 0 disables indexing at different precision levels. | |
490 | + --> | |
491 | + <fieldType name="tint" class="solr.TrieIntField" precisionStep="8" positionIncrementGap="0"/> | |
492 | + <fieldType name="tfloat" class="solr.TrieFloatField" precisionStep="8" positionIncrementGap="0"/> | |
493 | + <fieldType name="tlong" class="solr.TrieLongField" precisionStep="8" positionIncrementGap="0"/> | |
494 | + <fieldType name="tdouble" class="solr.TrieDoubleField" precisionStep="8" positionIncrementGap="0"/> | |
495 | + | |
496 | + <!-- The format for this date field is of the form 1995-12-31T23:59:59Z, and | |
497 | + is a more restricted form of the canonical representation of dateTime | |
498 | + http://www.w3.org/TR/xmlschema-2/#dateTime | |
499 | + The trailing "Z" designates UTC time and is mandatory. | |
500 | + Optional fractional seconds are allowed: 1995-12-31T23:59:59.999Z | |
501 | + All other components are mandatory. | |
502 | + | |
503 | + Expressions can also be used to denote calculations that should be | |
504 | + performed relative to "NOW" to determine the value, ie... | |
505 | + | |
506 | + NOW/HOUR | |
507 | + ... Round to the start of the current hour | |
508 | + NOW-1DAY | |
509 | + ... Exactly 1 day prior to now | |
510 | + NOW/DAY+6MONTHS+3DAYS | |
511 | + ... 6 months and 3 days in the future from the start of | |
512 | + the current day | |
513 | + | |
514 | + Consult the DateField javadocs for more information. | |
515 | + | |
516 | + Note: For faster range queries, consider the tdate type | |
517 | + --> | |
518 | + <fieldType name="date" class="solr.TrieDateField" precisionStep="0" positionIncrementGap="0"/> | |
519 | + | |
520 | + <!-- A Trie based date field for faster date range queries and date faceting. --> | |
521 | + <fieldType name="tdate" class="solr.TrieDateField" precisionStep="6" positionIncrementGap="0"/> | |
522 | + | |
523 | + | |
524 | + <!--Binary data type. The data should be sent/retrieved in as Base64 encoded Strings --> | |
525 | + <fieldtype name="binary" class="solr.BinaryField"/> | |
526 | + | |
527 | + <!-- | |
528 | + Note: | |
529 | + These should only be used for compatibility with existing indexes (created with lucene or older Solr versions). | |
530 | + Use Trie based fields instead. As of Solr 3.5 and 4.x, Trie based fields support sortMissingFirst/Last | |
531 | + | |
532 | + Plain numeric field types that store and index the text | |
533 | + value verbatim (and hence don't correctly support range queries, since the | |
534 | + lexicographic ordering isn't equal to the numeric ordering) | |
535 | + --> | |
536 | + <fieldType name="pint" class="solr.IntField"/> | |
537 | + <fieldType name="plong" class="solr.LongField"/> | |
538 | + <fieldType name="pfloat" class="solr.FloatField"/> | |
539 | + <fieldType name="pdouble" class="solr.DoubleField"/> | |
540 | + <fieldType name="pdate" class="solr.DateField" sortMissingLast="true"/> | |
541 | + | |
542 | + <!-- The "RandomSortField" is not used to store or search any | |
543 | + data. You can declare fields of this type it in your schema | |
544 | + to generate pseudo-random orderings of your docs for sorting | |
545 | + or function purposes. The ordering is generated based on the field | |
546 | + name and the version of the index. As long as the index version | |
547 | + remains unchanged, and the same field name is reused, | |
548 | + the ordering of the docs will be consistent. | |
549 | + If you want different psuedo-random orderings of documents, | |
550 | + for the same version of the index, use a dynamicField and | |
551 | + change the field name in the request. | |
552 | + --> | |
553 | + <fieldType name="random" class="solr.RandomSortField" indexed="true" /> | |
554 | + | |
555 | + <!-- solr.TextField allows the specification of custom text analyzers | |
556 | + specified as a tokenizer and a list of token filters. Different | |
557 | + analyzers may be specified for indexing and querying. | |
558 | + | |
559 | + The optional positionIncrementGap puts space between multiple fields of | |
560 | + this type on the same document, with the purpose of preventing false phrase | |
561 | + matching across fields. | |
562 | + | |
563 | + For more info on customizing your analyzer chain, please see | |
564 | + http://wiki.apache.org/solr/AnalyzersTokenizersTokenFilters | |
565 | + --> | |
566 | + | |
567 | + <!-- One can also specify an existing Analyzer class that has a | |
568 | + default constructor via the class attribute on the analyzer element. | |
569 | + Example: | |
570 | + --> | |
571 | + <fieldType name="text_greek" class="solr.TextField"> | |
572 | + <analyzer class="org.apache.lucene.analysis.el.GreekAnalyzer"/> | |
573 | + </fieldType> | |
574 | + | |
575 | + <!-- A text field that only splits on whitespace for exact matching of words --> | |
576 | + <fieldType name="text_ws" class="solr.TextField" positionIncrementGap="100"> | |
577 | + <analyzer> | |
578 | + <tokenizer class="solr.WhitespaceTokenizerFactory"/> | |
579 | + </analyzer> | |
580 | + </fieldType> | |
581 | + | |
582 | + <!-- A general text field that has reasonable, generic | |
583 | + cross-language defaults: it tokenizes with StandardTokenizer, | |
584 | + removes stop words from case-insensitive "stopwords.txt" | |
585 | + (empty by default), and down cases. At query time only, it | |
586 | + also applies synonyms. --> | |
587 | + <fieldType name="text_general" class="solr.TextField" positionIncrementGap="100"> | |
588 | + <analyzer type="index"> | |
589 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
590 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="stopwords.txt" /> | |
591 | + <!-- in this example, we will only use synonyms at query time | |
592 | + <filter class="solr.SynonymFilterFactory" synonyms="index_synonyms.txt" ignoreCase="true" expand="false"/> | |
593 | + --> | |
594 | + <filter class="solr.LowerCaseFilterFactory"/> | |
595 | + </analyzer> | |
596 | + <analyzer type="query"> | |
597 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
598 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="stopwords.txt" /> | |
599 | + <filter class="solr.SynonymFilterFactory" synonyms="synonyms.txt" ignoreCase="true" expand="true"/> | |
600 | + <filter class="solr.LowerCaseFilterFactory"/> | |
601 | + </analyzer> | |
602 | + </fieldType> | |
603 | + | |
604 | + <!-- A text field with defaults appropriate for English: it | |
605 | + tokenizes with StandardTokenizer, removes English stop words | |
606 | + (lang/stopwords_en.txt), down cases, protects words from protwords.txt, and | |
607 | + finally applies Porter's stemming. The query time analyzer | |
608 | + also applies synonyms from synonyms.txt. --> | |
609 | + <fieldType name="text_en" class="solr.TextField" positionIncrementGap="100"> | |
610 | + <analyzer type="index"> | |
611 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
612 | + <!-- in this example, we will only use synonyms at query time | |
613 | + <filter class="solr.SynonymFilterFactory" synonyms="index_synonyms.txt" ignoreCase="true" expand="false"/> | |
614 | + --> | |
615 | + <!-- Case insensitive stop word removal. | |
616 | + --> | |
617 | + <filter class="solr.StopFilterFactory" | |
618 | + ignoreCase="true" | |
619 | + words="lang/stopwords_en.txt" | |
620 | + /> | |
621 | + <filter class="solr.LowerCaseFilterFactory"/> | |
622 | + <filter class="solr.EnglishPossessiveFilterFactory"/> | |
623 | + <filter class="solr.KeywordMarkerFilterFactory" protected="protwords.txt"/> | |
624 | + <!-- Optionally you may want to use this less aggressive stemmer instead of PorterStemFilterFactory: | |
625 | + <filter class="solr.EnglishMinimalStemFilterFactory"/> | |
626 | + --> | |
627 | + <filter class="solr.PorterStemFilterFactory"/> | |
628 | + </analyzer> | |
629 | + <analyzer type="query"> | |
630 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
631 | + <filter class="solr.SynonymFilterFactory" synonyms="synonyms.txt" ignoreCase="true" expand="true"/> | |
632 | + <filter class="solr.StopFilterFactory" | |
633 | + ignoreCase="true" | |
634 | + words="lang/stopwords_en.txt" | |
635 | + /> | |
636 | + <filter class="solr.LowerCaseFilterFactory"/> | |
637 | + <filter class="solr.EnglishPossessiveFilterFactory"/> | |
638 | + <filter class="solr.KeywordMarkerFilterFactory" protected="protwords.txt"/> | |
639 | + <!-- Optionally you may want to use this less aggressive stemmer instead of PorterStemFilterFactory: | |
640 | + <filter class="solr.EnglishMinimalStemFilterFactory"/> | |
641 | + --> | |
642 | + <filter class="solr.PorterStemFilterFactory"/> | |
643 | + </analyzer> | |
644 | + </fieldType> | |
645 | + | |
646 | + <!-- A text field with defaults appropriate for English, plus | |
647 | + aggressive word-splitting and autophrase features enabled. | |
648 | + This field is just like text_en, except it adds | |
649 | + WordDelimiterFilter to enable splitting and matching of | |
650 | + words on case-change, alpha numeric boundaries, and | |
651 | + non-alphanumeric chars. This means certain compound word | |
652 | + cases will work, for example query "wi fi" will match | |
653 | + document "WiFi" or "wi-fi". | |
654 | + --> | |
655 | + <fieldType name="text_en_splitting" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true"> | |
656 | + <analyzer type="index"> | |
657 | + <tokenizer class="solr.WhitespaceTokenizerFactory"/> | |
658 | + <!-- in this example, we will only use synonyms at query time | |
659 | + <filter class="solr.SynonymFilterFactory" synonyms="index_synonyms.txt" ignoreCase="true" expand="false"/> | |
660 | + --> | |
661 | + <!-- Case insensitive stop word removal. | |
662 | + --> | |
663 | + <filter class="solr.StopFilterFactory" | |
664 | + ignoreCase="true" | |
665 | + words="lang/stopwords_en.txt" | |
666 | + /> | |
667 | + <filter class="solr.WordDelimiterFilterFactory" generateWordParts="1" generateNumberParts="1" catenateWords="1" catenateNumbers="1" catenateAll="0" splitOnCaseChange="1"/> | |
668 | + <filter class="solr.LowerCaseFilterFactory"/> | |
669 | + <filter class="solr.KeywordMarkerFilterFactory" protected="protwords.txt"/> | |
670 | + <filter class="solr.PorterStemFilterFactory"/> | |
671 | + </analyzer> | |
672 | + <analyzer type="query"> | |
673 | + <tokenizer class="solr.WhitespaceTokenizerFactory"/> | |
674 | + <filter class="solr.SynonymFilterFactory" synonyms="synonyms.txt" ignoreCase="true" expand="true"/> | |
675 | + <filter class="solr.StopFilterFactory" | |
676 | + ignoreCase="true" | |
677 | + words="lang/stopwords_en.txt" | |
678 | + /> | |
679 | + <filter class="solr.WordDelimiterFilterFactory" generateWordParts="1" generateNumberParts="1" catenateWords="0" catenateNumbers="0" catenateAll="0" splitOnCaseChange="1"/> | |
680 | + <filter class="solr.LowerCaseFilterFactory"/> | |
681 | + <filter class="solr.KeywordMarkerFilterFactory" protected="protwords.txt"/> | |
682 | + <filter class="solr.PorterStemFilterFactory"/> | |
683 | + </analyzer> | |
684 | + </fieldType> | |
685 | + | |
686 | + <!-- Less flexible matching, but less false matches. Probably not ideal for product names, | |
687 | + but may be good for SKUs. Can insert dashes in the wrong place and still match. --> | |
688 | + <fieldType name="text_en_splitting_tight" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true"> | |
689 | + <analyzer> | |
690 | + <tokenizer class="solr.WhitespaceTokenizerFactory"/> | |
691 | + <filter class="solr.SynonymFilterFactory" synonyms="synonyms.txt" ignoreCase="true" expand="false"/> | |
692 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_en.txt"/> | |
693 | + <filter class="solr.WordDelimiterFilterFactory" generateWordParts="0" generateNumberParts="0" catenateWords="1" catenateNumbers="1" catenateAll="0"/> | |
694 | + <filter class="solr.LowerCaseFilterFactory"/> | |
695 | + <filter class="solr.KeywordMarkerFilterFactory" protected="protwords.txt"/> | |
696 | + <filter class="solr.EnglishMinimalStemFilterFactory"/> | |
697 | + <!-- this filter can remove any duplicate tokens that appear at the same position - sometimes | |
698 | + possible with WordDelimiterFilter in conjuncton with stemming. --> | |
699 | + <filter class="solr.RemoveDuplicatesTokenFilterFactory"/> | |
700 | + </analyzer> | |
701 | + </fieldType> | |
702 | + | |
703 | + <!-- Just like text_general except it reverses the characters of | |
704 | + each token, to enable more efficient leading wildcard queries. --> | |
705 | + <fieldType name="text_general_rev" class="solr.TextField" positionIncrementGap="100"> | |
706 | + <analyzer type="index"> | |
707 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
708 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="stopwords.txt" /> | |
709 | + <filter class="solr.LowerCaseFilterFactory"/> | |
710 | + <filter class="solr.ReversedWildcardFilterFactory" withOriginal="true" | |
711 | + maxPosAsterisk="3" maxPosQuestion="2" maxFractionAsterisk="0.33"/> | |
712 | + </analyzer> | |
713 | + <analyzer type="query"> | |
714 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
715 | + <filter class="solr.SynonymFilterFactory" synonyms="synonyms.txt" ignoreCase="true" expand="true"/> | |
716 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="stopwords.txt" /> | |
717 | + <filter class="solr.LowerCaseFilterFactory"/> | |
718 | + </analyzer> | |
719 | + </fieldType> | |
720 | + | |
721 | + <!-- charFilter + WhitespaceTokenizer --> | |
722 | + <fieldType name="text_char_norm" class="solr.TextField" positionIncrementGap="100" > | |
723 | + <analyzer> | |
724 | + <charFilter class="solr.MappingCharFilterFactory" mapping="mapping-ISOLatin1Accent.txt"/> | |
725 | + <tokenizer class="solr.WhitespaceTokenizerFactory"/> | |
726 | + </analyzer> | |
727 | + </fieldType> | |
728 | + | |
729 | + <!-- This is an example of using the KeywordTokenizer along | |
730 | + With various TokenFilterFactories to produce a sortable field | |
731 | + that does not include some properties of the source text | |
732 | + --> | |
733 | + <fieldType name="alphaOnlySort" class="solr.TextField" sortMissingLast="true" omitNorms="true"> | |
734 | + <analyzer> | |
735 | + <!-- KeywordTokenizer does no actual tokenizing, so the entire | |
736 | + input string is preserved as a single token | |
737 | + --> | |
738 | + <tokenizer class="solr.KeywordTokenizerFactory"/> | |
739 | + <!-- The LowerCase TokenFilter does what you expect, which can be | |
740 | + when you want your sorting to be case insensitive | |
741 | + --> | |
742 | + <filter class="solr.LowerCaseFilterFactory" /> | |
743 | + <!-- The TrimFilter removes any leading or trailing whitespace --> | |
744 | + <filter class="solr.TrimFilterFactory" /> | |
745 | + <!-- The PatternReplaceFilter gives you the flexibility to use | |
746 | + Java Regular expression to replace any sequence of characters | |
747 | + matching a pattern with an arbitrary replacement string, | |
748 | + which may include back references to portions of the original | |
749 | + string matched by the pattern. | |
750 | + | |
751 | + See the Java Regular Expression documentation for more | |
752 | + information on pattern and replacement string syntax. | |
753 | + | |
754 | + http://java.sun.com/j2se/1.6.0/docs/api/java/util/regex/package-summary.html | |
755 | + --> | |
756 | + <filter class="solr.PatternReplaceFilterFactory" | |
757 | + pattern="([^a-z])" replacement="" replace="all" | |
758 | + /> | |
759 | + </analyzer> | |
760 | + </fieldType> | |
761 | + | |
762 | + <fieldtype name="phonetic" stored="false" indexed="true" class="solr.TextField" > | |
763 | + <analyzer> | |
764 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
765 | + <filter class="solr.DoubleMetaphoneFilterFactory" inject="false"/> | |
766 | + </analyzer> | |
767 | + </fieldtype> | |
768 | + | |
769 | + <fieldtype name="payloads" stored="false" indexed="true" class="solr.TextField" > | |
770 | + <analyzer> | |
771 | + <tokenizer class="solr.WhitespaceTokenizerFactory"/> | |
772 | + <!-- | |
773 | + The DelimitedPayloadTokenFilter can put payloads on tokens... for example, | |
774 | + a token of "foo|1.4" would be indexed as "foo" with a payload of 1.4f | |
775 | + Attributes of the DelimitedPayloadTokenFilterFactory : | |
776 | + "delimiter" - a one character delimiter. Default is | (pipe) | |
777 | + "encoder" - how to encode the following value into a playload | |
778 | + float -> org.apache.lucene.analysis.payloads.FloatEncoder, | |
779 | + integer -> o.a.l.a.p.IntegerEncoder | |
780 | + identity -> o.a.l.a.p.IdentityEncoder | |
781 | + Fully Qualified class name implementing PayloadEncoder, Encoder must have a no arg constructor. | |
782 | + --> | |
783 | + <filter class="solr.DelimitedPayloadTokenFilterFactory" encoder="float"/> | |
784 | + </analyzer> | |
785 | + </fieldtype> | |
786 | + | |
787 | + <!-- lowercases the entire field value, keeping it as a single token. --> | |
788 | + <fieldType name="lowercase" class="solr.TextField" positionIncrementGap="100"> | |
789 | + <analyzer> | |
790 | + <tokenizer class="solr.KeywordTokenizerFactory"/> | |
791 | + <filter class="solr.LowerCaseFilterFactory" /> | |
792 | + </analyzer> | |
793 | + </fieldType> | |
794 | + | |
795 | + <!-- | |
796 | + Example of using PathHierarchyTokenizerFactory at index time, so | |
797 | + queries for paths match documents at that path, or in descendent paths | |
798 | + --> | |
799 | + <fieldType name="descendent_path" class="solr.TextField"> | |
800 | + <analyzer type="index"> | |
801 | + <tokenizer class="solr.PathHierarchyTokenizerFactory" delimiter="/" /> | |
802 | + </analyzer> | |
803 | + <analyzer type="query"> | |
804 | + <tokenizer class="solr.KeywordTokenizerFactory" /> | |
805 | + </analyzer> | |
806 | + </fieldType> | |
807 | + <!-- | |
808 | + Example of using PathHierarchyTokenizerFactory at query time, so | |
809 | + queries for paths match documents at that path, or in ancestor paths | |
810 | + --> | |
811 | + <fieldType name="ancestor_path" class="solr.TextField"> | |
812 | + <analyzer type="index"> | |
813 | + <tokenizer class="solr.KeywordTokenizerFactory" /> | |
814 | + </analyzer> | |
815 | + <analyzer type="query"> | |
816 | + <tokenizer class="solr.PathHierarchyTokenizerFactory" delimiter="/" /> | |
817 | + </analyzer> | |
818 | + </fieldType> | |
819 | + | |
820 | + <!-- since fields of this type are by default not stored or indexed, | |
821 | + any data added to them will be ignored outright. --> | |
822 | + <fieldtype name="ignored" stored="false" indexed="false" multiValued="true" class="solr.StrField" /> | |
823 | + | |
824 | + <!-- This point type indexes the coordinates as separate fields (subFields) | |
825 | + If subFieldType is defined, it references a type, and a dynamic field | |
826 | + definition is created matching *___<typename>. Alternately, if | |
827 | + subFieldSuffix is defined, that is used to create the subFields. | |
828 | + Example: if subFieldType="double", then the coordinates would be | |
829 | + indexed in fields myloc_0___double,myloc_1___double. | |
830 | + Example: if subFieldSuffix="_d" then the coordinates would be indexed | |
831 | + in fields myloc_0_d,myloc_1_d | |
832 | + The subFields are an implementation detail of the fieldType, and end | |
833 | + users normally should not need to know about them. | |
834 | + --> | |
835 | + <fieldType name="point" class="solr.PointType" dimension="2" subFieldSuffix="_d"/> | |
836 | + | |
837 | + <!-- A specialized field for geospatial search. If indexed, this fieldType must not be multivalued. --> | |
838 | + <fieldType name="location" class="solr.LatLonType" subFieldSuffix="_coordinate"/> | |
839 | + | |
840 | + <!-- An alternative geospatial field type new to Solr 4. It supports multiValued and polygon shapes. | |
841 | + For more information about this and other Spatial fields new to Solr 4, see: | |
842 | + http://wiki.apache.org/solr/SolrAdaptersForLuceneSpatial4 | |
843 | + --> | |
844 | + <fieldType name="location_rpt" class="solr.SpatialRecursivePrefixTreeFieldType" | |
845 | + geo="true" distErrPct="0.025" maxDistErr="0.000009" units="degrees" /> | |
846 | + | |
847 | + <!-- Money/currency field type. See http://wiki.apache.org/solr/MoneyFieldType | |
848 | + Parameters: | |
849 | + defaultCurrency: Specifies the default currency if none specified. Defaults to "USD" | |
850 | + precisionStep: Specifies the precisionStep for the TrieLong field used for the amount | |
851 | + providerClass: Lets you plug in other exchange provider backend: | |
852 | + solr.FileExchangeRateProvider is the default and takes one parameter: | |
853 | + currencyConfig: name of an xml file holding exchange rates | |
854 | + solr.OpenExchangeRatesOrgProvider uses rates from openexchangerates.org: | |
855 | + ratesFileLocation: URL or path to rates JSON file (default latest.json on the web) | |
856 | + refreshInterval: Number of minutes between each rates fetch (default: 1440, min: 60) | |
857 | + --> | |
858 | + <fieldType name="currency" class="solr.CurrencyField" precisionStep="8" defaultCurrency="USD" currencyConfig="currency.xml" /> | |
859 | + | |
860 | + | |
861 | + | |
862 | + <!-- some examples for different languages (generally ordered by ISO code) --> | |
863 | + | |
864 | + <!-- Arabic --> | |
865 | + <fieldType name="text_ar" class="solr.TextField" positionIncrementGap="100"> | |
866 | + <analyzer> | |
867 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
868 | + <!-- for any non-arabic --> | |
869 | + <filter class="solr.LowerCaseFilterFactory"/> | |
870 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_ar.txt" /> | |
871 | + <!-- normalizes ﻯ to ﻱ, etc --> | |
872 | + <filter class="solr.ArabicNormalizationFilterFactory"/> | |
873 | + <filter class="solr.ArabicStemFilterFactory"/> | |
874 | + </analyzer> | |
875 | + </fieldType> | |
876 | + | |
877 | + <!-- Bulgarian --> | |
878 | + <fieldType name="text_bg" class="solr.TextField" positionIncrementGap="100"> | |
879 | + <analyzer> | |
880 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
881 | + <filter class="solr.LowerCaseFilterFactory"/> | |
882 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_bg.txt" /> | |
883 | + <filter class="solr.BulgarianStemFilterFactory"/> | |
884 | + </analyzer> | |
885 | + </fieldType> | |
886 | + | |
887 | + <!-- Catalan --> | |
888 | + <fieldType name="text_ca" class="solr.TextField" positionIncrementGap="100"> | |
889 | + <analyzer> | |
890 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
891 | + <!-- removes l', etc --> | |
892 | + <filter class="solr.ElisionFilterFactory" ignoreCase="true" articles="lang/contractions_ca.txt"/> | |
893 | + <filter class="solr.LowerCaseFilterFactory"/> | |
894 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_ca.txt" /> | |
895 | + <filter class="solr.SnowballPorterFilterFactory" language="Catalan"/> | |
896 | + </analyzer> | |
897 | + </fieldType> | |
898 | + | |
899 | + <!-- CJK bigram (see text_ja for a Japanese configuration using morphological analysis) --> | |
900 | + <fieldType name="text_cjk" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true"> | |
901 | + <analyzer> | |
902 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
903 | + <!-- normalize width before bigram, as e.g. half-width dakuten combine --> | |
904 | + <filter class="solr.CJKWidthFilterFactory"/> | |
905 | + <!-- for any non-CJK --> | |
906 | + <filter class="solr.LowerCaseFilterFactory"/> | |
907 | + <filter class="solr.CJKBigramFilterFactory"/> | |
908 | + </analyzer> | |
909 | + </fieldType> | |
910 | + | |
911 | + <!-- Czech --> | |
912 | + <fieldType name="text_cz" class="solr.TextField" positionIncrementGap="100"> | |
913 | + <analyzer> | |
914 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
915 | + <filter class="solr.LowerCaseFilterFactory"/> | |
916 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_cz.txt" /> | |
917 | + <filter class="solr.CzechStemFilterFactory"/> | |
918 | + </analyzer> | |
919 | + </fieldType> | |
920 | + | |
921 | + <!-- Danish --> | |
922 | + <fieldType name="text_da" class="solr.TextField" positionIncrementGap="100"> | |
923 | + <analyzer> | |
924 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
925 | + <filter class="solr.LowerCaseFilterFactory"/> | |
926 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_da.txt" format="snowball" /> | |
927 | + <filter class="solr.SnowballPorterFilterFactory" language="Danish"/> | |
928 | + </analyzer> | |
929 | + </fieldType> | |
930 | + | |
931 | + <!-- German --> | |
932 | + <fieldType name="text_de" class="solr.TextField" positionIncrementGap="100"> | |
933 | + <analyzer> | |
934 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
935 | + <filter class="solr.LowerCaseFilterFactory"/> | |
936 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_de.txt" format="snowball" /> | |
937 | + <filter class="solr.GermanNormalizationFilterFactory"/> | |
938 | + <filter class="solr.GermanLightStemFilterFactory"/> | |
939 | + <!-- less aggressive: <filter class="solr.GermanMinimalStemFilterFactory"/> --> | |
940 | + <!-- more aggressive: <filter class="solr.SnowballPorterFilterFactory" language="German2"/> --> | |
941 | + </analyzer> | |
942 | + </fieldType> | |
943 | + | |
944 | + <!-- Greek --> | |
945 | + <fieldType name="text_el" class="solr.TextField" positionIncrementGap="100"> | |
946 | + <analyzer> | |
947 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
948 | + <!-- greek specific lowercase for sigma --> | |
949 | + <filter class="solr.GreekLowerCaseFilterFactory"/> | |
950 | + <filter class="solr.StopFilterFactory" ignoreCase="false" words="lang/stopwords_el.txt" /> | |
951 | + <filter class="solr.GreekStemFilterFactory"/> | |
952 | + </analyzer> | |
953 | + </fieldType> | |
954 | + | |
955 | + <!-- Spanish --> | |
956 | + <fieldType name="text_es" class="solr.TextField" positionIncrementGap="100"> | |
957 | + <analyzer> | |
958 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
959 | + <filter class="solr.LowerCaseFilterFactory"/> | |
960 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_es.txt" format="snowball" /> | |
961 | + <filter class="solr.SpanishLightStemFilterFactory"/> | |
962 | + <!-- more aggressive: <filter class="solr.SnowballPorterFilterFactory" language="Spanish"/> --> | |
963 | + </analyzer> | |
964 | + </fieldType> | |
965 | + | |
966 | + <!-- Basque --> | |
967 | + <fieldType name="text_eu" class="solr.TextField" positionIncrementGap="100"> | |
968 | + <analyzer> | |
969 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
970 | + <filter class="solr.LowerCaseFilterFactory"/> | |
971 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_eu.txt" /> | |
972 | + <filter class="solr.SnowballPorterFilterFactory" language="Basque"/> | |
973 | + </analyzer> | |
974 | + </fieldType> | |
975 | + | |
976 | + <!-- Persian --> | |
977 | + <fieldType name="text_fa" class="solr.TextField" positionIncrementGap="100"> | |
978 | + <analyzer> | |
979 | + <!-- for ZWNJ --> | |
980 | + <charFilter class="solr.PersianCharFilterFactory"/> | |
981 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
982 | + <filter class="solr.LowerCaseFilterFactory"/> | |
983 | + <filter class="solr.ArabicNormalizationFilterFactory"/> | |
984 | + <filter class="solr.PersianNormalizationFilterFactory"/> | |
985 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_fa.txt" /> | |
986 | + </analyzer> | |
987 | + </fieldType> | |
988 | + | |
989 | + <!-- Finnish --> | |
990 | + <fieldType name="text_fi" class="solr.TextField" positionIncrementGap="100"> | |
991 | + <analyzer> | |
992 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
993 | + <filter class="solr.LowerCaseFilterFactory"/> | |
994 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_fi.txt" format="snowball" /> | |
995 | + <filter class="solr.SnowballPorterFilterFactory" language="Finnish"/> | |
996 | + <!-- less aggressive: <filter class="solr.FinnishLightStemFilterFactory"/> --> | |
997 | + </analyzer> | |
998 | + </fieldType> | |
999 | + | |
1000 | + <!-- French --> | |
1001 | + <fieldType name="text_fr" class="solr.TextField" positionIncrementGap="100"> | |
1002 | + <analyzer> | |
1003 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1004 | + <!-- removes l', etc --> | |
1005 | + <filter class="solr.ElisionFilterFactory" ignoreCase="true" articles="lang/contractions_fr.txt"/> | |
1006 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1007 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_fr.txt" format="snowball" /> | |
1008 | + <filter class="solr.FrenchLightStemFilterFactory"/> | |
1009 | + <!-- less aggressive: <filter class="solr.FrenchMinimalStemFilterFactory"/> --> | |
1010 | + <!-- more aggressive: <filter class="solr.SnowballPorterFilterFactory" language="French"/> --> | |
1011 | + </analyzer> | |
1012 | + </fieldType> | |
1013 | + | |
1014 | + <!-- Irish --> | |
1015 | + <fieldType name="text_ga" class="solr.TextField" positionIncrementGap="100"> | |
1016 | + <analyzer> | |
1017 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1018 | + <!-- removes d', etc --> | |
1019 | + <filter class="solr.ElisionFilterFactory" ignoreCase="true" articles="lang/contractions_ga.txt"/> | |
1020 | + <!-- removes n-, etc. position increments is intentionally false! --> | |
1021 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/hyphenations_ga.txt"/> | |
1022 | + <filter class="solr.IrishLowerCaseFilterFactory"/> | |
1023 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_ga.txt"/> | |
1024 | + <filter class="solr.SnowballPorterFilterFactory" language="Irish"/> | |
1025 | + </analyzer> | |
1026 | + </fieldType> | |
1027 | + | |
1028 | + <!-- Galician --> | |
1029 | + <fieldType name="text_gl" class="solr.TextField" positionIncrementGap="100"> | |
1030 | + <analyzer> | |
1031 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1032 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1033 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_gl.txt" /> | |
1034 | + <filter class="solr.GalicianStemFilterFactory"/> | |
1035 | + <!-- less aggressive: <filter class="solr.GalicianMinimalStemFilterFactory"/> --> | |
1036 | + </analyzer> | |
1037 | + </fieldType> | |
1038 | + | |
1039 | + <!-- Hindi --> | |
1040 | + <fieldType name="text_hi" class="solr.TextField" positionIncrementGap="100"> | |
1041 | + <analyzer> | |
1042 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1043 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1044 | + <!-- normalizes unicode representation --> | |
1045 | + <filter class="solr.IndicNormalizationFilterFactory"/> | |
1046 | + <!-- normalizes variation in spelling --> | |
1047 | + <filter class="solr.HindiNormalizationFilterFactory"/> | |
1048 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_hi.txt" /> | |
1049 | + <filter class="solr.HindiStemFilterFactory"/> | |
1050 | + </analyzer> | |
1051 | + </fieldType> | |
1052 | + | |
1053 | + <!-- Hungarian --> | |
1054 | + <fieldType name="text_hu" class="solr.TextField" positionIncrementGap="100"> | |
1055 | + <analyzer> | |
1056 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1057 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1058 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_hu.txt" format="snowball" /> | |
1059 | + <filter class="solr.SnowballPorterFilterFactory" language="Hungarian"/> | |
1060 | + <!-- less aggressive: <filter class="solr.HungarianLightStemFilterFactory"/> --> | |
1061 | + </analyzer> | |
1062 | + </fieldType> | |
1063 | + | |
1064 | + <!-- Armenian --> | |
1065 | + <fieldType name="text_hy" class="solr.TextField" positionIncrementGap="100"> | |
1066 | + <analyzer> | |
1067 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1068 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1069 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_hy.txt" /> | |
1070 | + <filter class="solr.SnowballPorterFilterFactory" language="Armenian"/> | |
1071 | + </analyzer> | |
1072 | + </fieldType> | |
1073 | + | |
1074 | + <!-- Indonesian --> | |
1075 | + <fieldType name="text_id" class="solr.TextField" positionIncrementGap="100"> | |
1076 | + <analyzer> | |
1077 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1078 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1079 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_id.txt" /> | |
1080 | + <!-- for a less aggressive approach (only inflectional suffixes), set stemDerivational to false --> | |
1081 | + <filter class="solr.IndonesianStemFilterFactory" stemDerivational="true"/> | |
1082 | + </analyzer> | |
1083 | + </fieldType> | |
1084 | + | |
1085 | + <!-- Italian --> | |
1086 | + <fieldType name="text_it" class="solr.TextField" positionIncrementGap="100"> | |
1087 | + <analyzer> | |
1088 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1089 | + <!-- removes l', etc --> | |
1090 | + <filter class="solr.ElisionFilterFactory" ignoreCase="true" articles="lang/contractions_it.txt"/> | |
1091 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1092 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_it.txt" format="snowball" /> | |
1093 | + <filter class="solr.ItalianLightStemFilterFactory"/> | |
1094 | + <!-- more aggressive: <filter class="solr.SnowballPorterFilterFactory" language="Italian"/> --> | |
1095 | + </analyzer> | |
1096 | + </fieldType> | |
1097 | + | |
1098 | + <!-- Japanese using morphological analysis (see text_cjk for a configuration using bigramming) | |
1099 | + | |
1100 | + NOTE: If you want to optimize search for precision, use default operator AND in your query | |
1101 | + parser config with <solrQueryParser defaultOperator="AND"/> further down in this file. Use | |
1102 | + OR if you would like to optimize for recall (default). | |
1103 | + --> | |
1104 | + <fieldType name="text_ja" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true"> | |
1105 | + <analyzer> | |
1106 | + <!-- Kuromoji Japanese morphological analyzer/tokenizer (JapaneseTokenizer) | |
1107 | + | |
1108 | + Kuromoji has a search mode (default) that does segmentation useful for search. A heuristic | |
1109 | + is used to segment compounds into its parts and the compound itself is kept as synonym. | |
1110 | + | |
1111 | + Valid values for attribute mode are: | |
1112 | + normal: regular segmentation | |
1113 | + search: segmentation useful for search with synonyms compounds (default) | |
1114 | + extended: same as search mode, but unigrams unknown words (experimental) | |
1115 | + | |
1116 | + For some applications it might be good to use search mode for indexing and normal mode for | |
1117 | + queries to reduce recall and prevent parts of compounds from being matched and highlighted. | |
1118 | + Use <analyzer type="index"> and <analyzer type="query"> for this and mode normal in query. | |
1119 | + | |
1120 | + Kuromoji also has a convenient user dictionary feature that allows overriding the statistical | |
1121 | + model with your own entries for segmentation, part-of-speech tags and readings without a need | |
1122 | + to specify weights. Notice that user dictionaries have not been subject to extensive testing. | |
1123 | + | |
1124 | + User dictionary attributes are: | |
1125 | + userDictionary: user dictionary filename | |
1126 | + userDictionaryEncoding: user dictionary encoding (default is UTF-8) | |
1127 | + | |
1128 | + See lang/userdict_ja.txt for a sample user dictionary file. | |
1129 | + | |
1130 | + Punctuation characters are discarded by default. Use discardPunctuation="false" to keep them. | |
1131 | + | |
1132 | + See http://wiki.apache.org/solr/JapaneseLanguageSupport for more on Japanese language support. | |
1133 | + --> | |
1134 | + <tokenizer class="solr.JapaneseTokenizerFactory" mode="search"/> | |
1135 | + <!--<tokenizer class="solr.JapaneseTokenizerFactory" mode="search" userDictionary="lang/userdict_ja.txt"/>--> | |
1136 | + <!-- Reduces inflected verbs and adjectives to their base/dictionary forms (辞書形) --> | |
1137 | + <filter class="solr.JapaneseBaseFormFilterFactory"/> | |
1138 | + <!-- Removes tokens with certain part-of-speech tags --> | |
1139 | + <filter class="solr.JapanesePartOfSpeechStopFilterFactory" tags="lang/stoptags_ja.txt" /> | |
1140 | + <!-- Normalizes full-width romaji to half-width and half-width kana to full-width (Unicode NFKC subset) --> | |
1141 | + <filter class="solr.CJKWidthFilterFactory"/> | |
1142 | + <!-- Removes common tokens typically not useful for search, but have a negative effect on ranking --> | |
1143 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_ja.txt" /> | |
1144 | + <!-- Normalizes common katakana spelling variations by removing any last long sound character (U+30FC) --> | |
1145 | + <filter class="solr.JapaneseKatakanaStemFilterFactory" minimumLength="4"/> | |
1146 | + <!-- Lower-cases romaji characters --> | |
1147 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1148 | + </analyzer> | |
1149 | + </fieldType> | |
1150 | + | |
1151 | + <!-- Latvian --> | |
1152 | + <fieldType name="text_lv" class="solr.TextField" positionIncrementGap="100"> | |
1153 | + <analyzer> | |
1154 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1155 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1156 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_lv.txt" /> | |
1157 | + <filter class="solr.LatvianStemFilterFactory"/> | |
1158 | + </analyzer> | |
1159 | + </fieldType> | |
1160 | + | |
1161 | + <!-- Dutch --> | |
1162 | + <fieldType name="text_nl" class="solr.TextField" positionIncrementGap="100"> | |
1163 | + <analyzer> | |
1164 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1165 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1166 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_nl.txt" format="snowball" /> | |
1167 | + <filter class="solr.StemmerOverrideFilterFactory" dictionary="lang/stemdict_nl.txt" ignoreCase="false"/> | |
1168 | + <filter class="solr.SnowballPorterFilterFactory" language="Dutch"/> | |
1169 | + </analyzer> | |
1170 | + </fieldType> | |
1171 | + | |
1172 | + <!-- Norwegian --> | |
1173 | + <fieldType name="text_no" class="solr.TextField" positionIncrementGap="100"> | |
1174 | + <analyzer> | |
1175 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1176 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1177 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_no.txt" format="snowball" /> | |
1178 | + <filter class="solr.SnowballPorterFilterFactory" language="Norwegian"/> | |
1179 | + <!-- less aggressive: <filter class="solr.NorwegianLightStemFilterFactory" variant="nb"/> --> | |
1180 | + <!-- singular/plural: <filter class="solr.NorwegianMinimalStemFilterFactory" variant="nb"/> --> | |
1181 | + <!-- The "light" and "minimal" stemmers support variants: nb=Bokmål, nn=Nynorsk, no=Both --> | |
1182 | + </analyzer> | |
1183 | + </fieldType> | |
1184 | + | |
1185 | + <!-- Portuguese --> | |
1186 | + <fieldType name="text_pt" class="solr.TextField" positionIncrementGap="100"> | |
1187 | + <analyzer> | |
1188 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1189 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1190 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_pt.txt" format="snowball" /> | |
1191 | + <filter class="solr.PortugueseLightStemFilterFactory"/> | |
1192 | + <!-- less aggressive: <filter class="solr.PortugueseMinimalStemFilterFactory"/> --> | |
1193 | + <!-- more aggressive: <filter class="solr.SnowballPorterFilterFactory" language="Portuguese"/> --> | |
1194 | + <!-- most aggressive: <filter class="solr.PortugueseStemFilterFactory"/> --> | |
1195 | + </analyzer> | |
1196 | + </fieldType> | |
1197 | + | |
1198 | + <!-- Romanian --> | |
1199 | + <fieldType name="text_ro" class="solr.TextField" positionIncrementGap="100"> | |
1200 | + <analyzer> | |
1201 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1202 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1203 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_ro.txt" /> | |
1204 | + <filter class="solr.SnowballPorterFilterFactory" language="Romanian"/> | |
1205 | + </analyzer> | |
1206 | + </fieldType> | |
1207 | + | |
1208 | + <!-- Russian --> | |
1209 | + <fieldType name="text_ru" class="solr.TextField" positionIncrementGap="100"> | |
1210 | + <analyzer> | |
1211 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1212 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1213 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_ru.txt" format="snowball" /> | |
1214 | + <filter class="solr.SnowballPorterFilterFactory" language="Russian"/> | |
1215 | + <!-- less aggressive: <filter class="solr.RussianLightStemFilterFactory"/> --> | |
1216 | + </analyzer> | |
1217 | + </fieldType> | |
1218 | + | |
1219 | + <!-- Swedish --> | |
1220 | + <fieldType name="text_sv" class="solr.TextField" positionIncrementGap="100"> | |
1221 | + <analyzer> | |
1222 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1223 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1224 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_sv.txt" format="snowball" /> | |
1225 | + <filter class="solr.SnowballPorterFilterFactory" language="Swedish"/> | |
1226 | + <!-- less aggressive: <filter class="solr.SwedishLightStemFilterFactory"/> --> | |
1227 | + </analyzer> | |
1228 | + </fieldType> | |
1229 | + | |
1230 | + <!-- Thai --> | |
1231 | + <fieldType name="text_th" class="solr.TextField" positionIncrementGap="100"> | |
1232 | + <analyzer> | |
1233 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1234 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1235 | + <filter class="solr.ThaiWordFilterFactory"/> | |
1236 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="lang/stopwords_th.txt" /> | |
1237 | + </analyzer> | |
1238 | + </fieldType> | |
1239 | + | |
1240 | + <!-- Turkish --> | |
1241 | + <fieldType name="text_tr" class="solr.TextField" positionIncrementGap="100"> | |
1242 | + <analyzer> | |
1243 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1244 | + <filter class="solr.TurkishLowerCaseFilterFactory"/> | |
1245 | + <filter class="solr.StopFilterFactory" ignoreCase="false" words="lang/stopwords_tr.txt" /> | |
1246 | + <filter class="solr.SnowballPorterFilterFactory" language="Turkish"/> | |
1247 | + </analyzer> | |
1248 | + </fieldType> | |
1249 | + | |
1250 | + <!-- gosen --> | |
1251 | + <!-- configuration for japanese text, using a morphological analyzer | |
1252 | + Most possibilities for customization are specified here in the schema. | |
1253 | + | |
1254 | + Note: you can set the default query operator to be OR, AND, or PHRASE: | |
1255 | + OR: Use these defaults (autoGeneratePhraseQueries="false", <solrQueryParser defaultOperator="OR"/> | |
1256 | + In this case Solr works like it does with the English language. The default query is OR, | |
1257 | + but documents that contain more of the query terms get a special boost. You can probably | |
1258 | + use a less aggressive stopwords/stoptags in this case, and its probably a good idea to use | |
1259 | + enablePositionIncrements=true, so that if a user puts a query in quotes, they get a much more | |
1260 | + exact phrase query. | |
1261 | + AND: Set autoGeneratePhraseQueries=false, but set <solrQueryParser defaultOperator="AND"/> in | |
1262 | + your schema.xml. Note if you do this, you should use a more aggressive stopwords/stoptags | |
1263 | + list (at least at query-time), otherwise a document might not match simply because it does | |
1264 | + not contain a prefix or particle. As in the above case, its probably a good idea to use | |
1265 | + enablePositionIncrements=true for explicit phrase queries from the user. | |
1266 | + PHRASE: Set autoGeneratePhraseQueries=true. If you do this, you should probably use both a very | |
1267 | + aggressive stopwords list, and you should probably also set enablePositionIncrements=false | |
1268 | + everywhere. Otherwise, even documents that contain the query's phrase in exact order will | |
1269 | + not match because of slightly different grammatical structure. | |
1270 | + --> | |
1271 | + <fieldType name="text_gosen" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true"> | |
1272 | + <analyzer> | |
1273 | + <!-- map characters before the tokenizer: | |
1274 | + Optionally, instead of the GosenWidthFactory, you can choose to do the width | |
1275 | + mappings before the text is sent to the tokenizer. | |
1276 | + <charFilter class="solr.MappingCharFilterFactory" mapping="mapping-japanese.txt"/> | |
1277 | + --> | |
1278 | + | |
1279 | + <!-- morphological tokenizer: sets the SURFACE form as the token, but also sets these attributes: | |
1280 | + BasicFormAttribute, ConjugationAttribute, PartOfSpeechAttribute, PronunciationsAttribute, | |
1281 | + ReadingsAttribute, and SentenceStartAttribute. | |
1282 | + Option setting | |
1283 | + compositePOS : merge multiple similar tokens by a part-of-speech based configrations | |
1284 | + dictionaryDir : specified a dictionary directory. you can use multiple dictionary for each tokenizer. | |
1285 | + --> | |
1286 | + <tokenizer class="solr.GosenTokenizerFactory" compositePOS="compositePOS.txt" dictionaryDir="dictionary/naist-chasen"/> | |
1287 | + | |
1288 | + <!-- normalizes CJK width differences: | |
1289 | + 1. Folds fullwidth ASCII variants into the equivalent basic latin | |
1290 | + 2. Folds halfwidth Katakana variants into the equivalent kana | |
1291 | + | |
1292 | + Note: alternatively you can use a MappingCharFilter before the tokenizer for this, but please note | |
1293 | + that mapping characters can change how Sen tokenizes text. | |
1294 | + --> | |
1295 | + <filter class="solr.GosenWidthFilterFactory"/> | |
1296 | + | |
1297 | + <!-- the punctuation filter removes all-punctuation tokens base on Unicode properties. | |
1298 | + punctuation tokens are tagged as "unknown", and its better to do this than to remove | |
1299 | + tokens with an unknown pos (as they might be valuable!). Because this punctuation | |
1300 | + usually signifies a phrase or sentence boundary, enablePositionIncrements can be | |
1301 | + used to prevent phrase queries from matching across natural phrase/sentence boundaries --> | |
1302 | + <filter class="solr.GosenPunctuationFilterFactory" enablePositionIncrements="false"/> | |
1303 | + | |
1304 | + <!-- this is a part-of-speech based stopfilter, it removes any tokens that have a certain | |
1305 | + of speech. you can set enablePositionIncrements for tighter phrase queries --> | |
1306 | + <filter class="solr.GosenPartOfSpeechStopFilterFactory" tags="stoptags_ja.txt" enablePositionIncrements="false"/> | |
1307 | + | |
1308 | + <!-- a standard stopfilter, to specify specific stopwords. --> | |
1309 | + <filter class="solr.StopFilterFactory" ignoreCase="true" words="stopwords_ja.txt" enablePositionIncrements="false"/> | |
1310 | + | |
1311 | + <!-- alternatively, instead of using a part-of-speech based stopfilter, you can use a | |
1312 | + part-of-speech based keepfilter: specifying only the parts of speech you wish to index. | |
1313 | + anything else will be removed. HOWEVER: this could be a little dangerous, because if | |
1314 | + we upgrade ipadic they might add some new tags (the tags are fairly specific), and suddenly | |
1315 | + things that you were indexing before are no longer being indexed. Its recommended to | |
1316 | + use the part-of-speech based stopfilter above if at all possible, for safety. | |
1317 | + <filter class="solr.GosenPartOfSpeechKeepFilterFactory" tags="keeptags_ja.txt" enablePositionIncrements="true"/> | |
1318 | + --> | |
1319 | + | |
1320 | + <!-- before any stemming/lemmatization, you can protect words from being modified by specifying | |
1321 | + a protwords.txt. | |
1322 | + <filter class="solr.KeywordMarkerFilterFactory" protected="protwords_ja.txt" ignoreCase="false"/> | |
1323 | + | |
1324 | + or you can also supply a custom stem dictionary for inflected forms (tab separated). No | |
1325 | + further stemming/lemmatization will modify this. | |
1326 | + <filter class="solr.StemmerOverrideFilterFactory" dictionary="dictionary.txt" ignoreCase="false"/> | |
1327 | + --> | |
1328 | + | |
1329 | + <!-- the basic form filter converts inflected verbs and adjectives to their dictionary citation form. --> | |
1330 | + <filter class="solr.GosenBasicFormFilterFactory"/> | |
1331 | + | |
1332 | + <!-- this filter heuristically normalizes katakana forms with a final prolonged sound mark --> | |
1333 | + <filter class="solr.GosenKatakanaStemFilterFactory"/> | |
1334 | + | |
1335 | + <!-- you might want to lowercase for any english text content you have --> | |
1336 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1337 | + </analyzer> | |
1338 | + </fieldType> | |
1339 | + | |
1340 | + <fieldType name="text" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true"> | |
1341 | + <analyzer> | |
1342 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1343 | + <!-- normalize width before bigram, as e.g. half-width dakuten combine --> | |
1344 | + <filter class="solr.CJKWidthFilterFactory"/> | |
1345 | + <!-- for any non-CJK --> | |
1346 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1347 | + <filter class="solr.CJKBigramFilterFactory"/> | |
1348 | + </analyzer> | |
1349 | + </fieldType> | |
1350 | + | |
1351 | + <fieldType name="url" class="solr.TextField" positionIncrementGap="100"> | |
1352 | + <analyzer> | |
1353 | + <tokenizer class="solr.StandardTokenizerFactory"/> | |
1354 | + <filter class="solr.LowerCaseFilterFactory"/> | |
1355 | + <filter class="solr.WordDelimiterFilterFactory" generateWordParts="1" generateNumberParts="1"/> | |
1356 | + <filter class="solr.RemoveDuplicatesTokenFilterFactory"/> | |
1357 | + </analyzer> | |
1358 | + </fieldType> | |
1359 | + | |
1360 | + <fieldType name="text_suggest" class="solr.TextField" positionIncrementGap="100"> | |
1361 | + <analyzer> | |
1362 | + <tokenizer class="jp.sf.fess.solr.plugin.analysis.SuggestTextTokenizerFactory" maxLength="10000" userDictionary="lang/userdict_suggest_ja.txt" userDictionaryEncoding="UTF-8" includePartOfSpeech="start:名詞,start:接頭詞,start:形容詞,middle:名詞,middle:接頭詞,middle:形容詞" excludePartOfSpeech="start:副詞可能" includeCharTerm="middle:な" preConverters="[{"class":"jp.sf.fess.suggest.converter.ICUConverter", "args":["Fullwidth-Halfwidth"]}, {"class":"jp.sf.fess.suggest.converter.ICUConverter", "args":["Any-Lower"]}, {"class":"jp.sf.fess.suggest.converter.NormalizeConverter"}]" converters="[{"class":"jp.sf.fess.suggest.converter.SymbolConverter", "method":[{"name":"addSymbol", "args":[["0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", " ", " ", "【", "】", "・", "(", ")", "(", ")", "%", "%", "~", "+", "#", "#", "\"", "”", "'", "\\", "¥", "[", "]", "「", "」", "『", "』", "<", ">", "<", ">", "/", "/", "{", "}", "&", "&", "Ⅰ", "Ⅱ", "Ⅲ", "Ⅳ", "Ⅴ", "Ⅵ", "Ⅶ", "Ⅷ", "Ⅸ", "Ⅹ", "ⅰ", "ⅱ", "ⅲ", "ⅳ", "ⅴ", "ⅵ", "ⅶ", "ⅷ", "ⅸ", "ⅹ", "-", "-", "."]]}]}]"/> | |
1363 | + </analyzer> | |
1364 | + </fieldType> | |
1365 | + | |
1366 | + <fieldType name="string_suggest" class="solr.TextField" positionIncrementGap="100"> | |
1367 | + <analyzer> | |
1368 | + <tokenizer class="jp.sf.fess.solr.plugin.analysis.SuggestStringTokenizerFactory" userDictionary="lang/userdict_suggest_ja.txt" userDictionaryEncoding="UTF-8" preConverters="[{"class":"jp.sf.fess.suggest.converter.ICUConverter", "args":["Fullwidth-Halfwidth"]}, {"class":"jp.sf.fess.suggest.converter.ICUConverter", "args":["Any-Lower"]}, {"class":"jp.sf.fess.suggest.converter.NormalizeConverter"}]" converters="[{"class":"jp.sf.fess.suggest.converter.SymbolConverter", "method":[{"name":"addSymbol", "args":[["0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", " ", " ", "【", "】", "・", "(", ")", "(", ")", "%", "%", "~", "+", "#", "#", "\"", "”", "'", "\\", "¥", "[", "]", "「", "」", "『", "』", "<", ">", "<", ">", "/", "/", "{", "}", "&", "&", "Ⅰ", "Ⅱ", "Ⅲ", "Ⅳ", "Ⅴ", "Ⅵ", "Ⅶ", "Ⅷ", "Ⅸ", "Ⅹ", "ⅰ", "ⅱ", "ⅲ", "ⅳ", "ⅴ", "ⅵ", "ⅶ", "ⅷ", "ⅸ", "ⅹ", "-", "-", "." ]]}]}]"/> | |
1369 | + </analyzer> | |
1370 | + </fieldType> | |
1371 | + | |
1372 | + <fieldType name="domain_name" class="solr.TextField"> | |
1373 | + <analyzer type="index"> | |
1374 | + <tokenizer class="solr.PathHierarchyTokenizerFactory" delimiter="." reverse="true"/> | |
1375 | + </analyzer> | |
1376 | + <analyzer type="query"> | |
1377 | + <tokenizer class="solr.KeywordTokenizerFactory" /> | |
1378 | + </analyzer> | |
1379 | + </fieldType> | |
1380 | + | |
1381 | + </types> | |
1382 | + | |
1383 | + <!-- Similarity is the scoring routine for each document vs. a query. | |
1384 | + A custom Similarity or SimilarityFactory may be specified here, but | |
1385 | + the default is fine for most applications. | |
1386 | + For more info: http://wiki.apache.org/solr/SchemaXml#Similarity | |
1387 | + --> | |
1388 | + <!-- | |
1389 | + <similarity class="com.example.solr.CustomSimilarityFactory"> | |
1390 | + <str name="paramkey">param value</str> | |
1391 | + </similarity> | |
1392 | + --> | |
1393 | + | |
1394 | +</schema> |
@@ -0,0 +1,2 @@ | ||
1 | +名詞-数 | |
2 | +未知語 記号-アルファベット |
@@ -0,0 +1,179 @@ | ||
1 | +# Half-width Katakana => Full-width Katakana | |
2 | +"ア" => "ア" | |
3 | +"イ" => "イ" | |
4 | +"ウ" => "ウ" | |
5 | +"エ" => "エ" | |
6 | +"オ" => "オ" | |
7 | +"カ" => "カ" | |
8 | +"キ" => "キ" | |
9 | +"ク" => "ク" | |
10 | +"ケ" => "ケ" | |
11 | +"コ" => "コ" | |
12 | +"サ" => "サ" | |
13 | +"シ" => "シ" | |
14 | +"ス" => "ス" | |
15 | +"セ" => "セ" | |
16 | +"ソ" => "ソ" | |
17 | +"タ" => "タ" | |
18 | +"チ" => "チ" | |
19 | +"ツ" => "ツ" | |
20 | +"テ" => "テ" | |
21 | +"ト" => "ト" | |
22 | +"ナ" => "ナ" | |
23 | +"ニ" => "ニ" | |
24 | +"ヌ" => "ヌ" | |
25 | +"ネ" => "ネ" | |
26 | +"ノ" => "ノ" | |
27 | +"ハ" => "ハ" | |
28 | +"ヒ" => "ヒ" | |
29 | +"フ" => "フ" | |
30 | +"ヘ" => "ヘ" | |
31 | +"ホ" => "ホ" | |
32 | +"マ" => "マ" | |
33 | +"ミ" => "ミ" | |
34 | +"ム" => "ム" | |
35 | +"メ" => "メ" | |
36 | +"モ" => "モ" | |
37 | +"ヤ" => "ヤ" | |
38 | +"ユ" => "ユ" | |
39 | +"ヨ" => "ヨ" | |
40 | +"ラ" => "ラ" | |
41 | +"リ" => "リ" | |
42 | +"ル" => "ル" | |
43 | +"レ" => "レ" | |
44 | +"ロ" => "ロ" | |
45 | +"ワ" => "ワ" | |
46 | +"ヲ" => "ヲ" | |
47 | +"ン" => "ン" | |
48 | +"ッ" => "ッ" | |
49 | +"ァ" => "ァ" | |
50 | +"ィ" => "ィ" | |
51 | +"ゥ" => "ゥ" | |
52 | +"ェ" => "ェ" | |
53 | +"ォ" => "ォ" | |
54 | +"ャ" => "ャ" | |
55 | +"ュ" => "ュ" | |
56 | +"ョ" => "ョ" | |
57 | +"ガ" => "ガ" | |
58 | +"ギ" => "ギ" | |
59 | +"グ" => "グ" | |
60 | +"ゲ" => "ゲ" | |
61 | +"ゴ" => "ゴ" | |
62 | +"ザ" => "ザ" | |
63 | +"ジ" => "ジ" | |
64 | +"ズ" => "ズ" | |
65 | +"ゼ" => "ゼ" | |
66 | +"ゾ" => "ゾ" | |
67 | +"ダ" => "ダ" | |
68 | +"ヂ" => "ヂ" | |
69 | +"ヅ" => "ヅ" | |
70 | +"デ" => "デ" | |
71 | +"ド" => "ド" | |
72 | +"バ" => "バ" | |
73 | +"ビ" => "ビ" | |
74 | +"ブ" => "ブ" | |
75 | +"ベ" => "ベ" | |
76 | +"ボ" => "ボ" | |
77 | +"パ" => "パ" | |
78 | +"ピ" => "ピ" | |
79 | +"プ" => "プ" | |
80 | +"ペ" => "ペ" | |
81 | +"ポ" => "ポ" | |
82 | + | |
83 | +# Full-width alpha-numeric => Half-width alpha-numeric | |
84 | +"0" => "0" | |
85 | +"1" => "1" | |
86 | +"2" => "2" | |
87 | +"3" => "3" | |
88 | +"4" => "4" | |
89 | +"5" => "5" | |
90 | +"6" => "6" | |
91 | +"7" => "7" | |
92 | +"8" => "8" | |
93 | +"9" => "9" | |
94 | + | |
95 | +"A" => "A" | |
96 | +"B" => "B" | |
97 | +"C" => "C" | |
98 | +"D" => "D" | |
99 | +"E" => "E" | |
100 | +"F" => "F" | |
101 | +"G" => "G" | |
102 | +"H" => "H" | |
103 | +"I" => "I" | |
104 | +"J" => "J" | |
105 | +"K" => "K" | |
106 | +"L" => "L" | |
107 | +"M" => "M" | |
108 | +"N" => "N" | |
109 | +"O" => "O" | |
110 | +"P" => "P" | |
111 | +"Q" => "Q" | |
112 | +"R" => "R" | |
113 | +"S" => "S" | |
114 | +"T" => "T" | |
115 | +"U" => "U" | |
116 | +"V" => "V" | |
117 | +"W" => "W" | |
118 | +"X" => "X" | |
119 | +"Y" => "Y" | |
120 | +"Z" => "Z" | |
121 | + | |
122 | +"a" => "a" | |
123 | +"b" => "b" | |
124 | +"c" => "c" | |
125 | +"d" => "d" | |
126 | +"e" => "e" | |
127 | +"f" => "f" | |
128 | +"g" => "g" | |
129 | +"h" => "h" | |
130 | +"i" => "i" | |
131 | +"j" => "j" | |
132 | +"k" => "k" | |
133 | +"l" => "l" | |
134 | +"m" => "m" | |
135 | +"n" => "n" | |
136 | +"o" => "o" | |
137 | +"p" => "p" | |
138 | +"q" => "q" | |
139 | +"r" => "r" | |
140 | +"s" => "s" | |
141 | +"t" => "t" | |
142 | +"u" => "u" | |
143 | +"v" => "v" | |
144 | +"w" => "w" | |
145 | +"x" => "x" | |
146 | +"y" => "y" | |
147 | +"z" => "z" | |
148 | + | |
149 | +# !"#$%&'()=|-^\@[`{;:]+*}/_<>? | |
150 | +"!" => "!" | |
151 | +"”" => "\"" | |
152 | +"#" => "#" | |
153 | +"$" => "$" | |
154 | +"%" => "%" | |
155 | +"&" => "&" | |
156 | +"’" => "'" | |
157 | +"(" => "(" | |
158 | +")" => ")" | |
159 | +"=" => "=" | |
160 | +"|" => "|" | |
161 | +"-" => "-" | |
162 | +"^" => "^" | |
163 | +"¥" => "\\" | |
164 | +"@" => "@" | |
165 | +"[" => "[" | |
166 | +"`" => "`" | |
167 | +"{" => "{" | |
168 | +";" => ";" | |
169 | +":" => ":" | |
170 | +"]" => "]" | |
171 | +"+" => "+" | |
172 | +"*" => "*" | |
173 | +"}" => "}" | |
174 | +"/" => "/" | |
175 | +"_" => "_" | |
176 | +"<" => "<" | |
177 | +">" => ">" | |
178 | +"?" => "?" | |
179 | +"‘" => "`" |
@@ -0,0 +1,1936 @@ | ||
1 | +<?xml version="1.0" encoding="UTF-8" ?> | |
2 | +<!-- | |
3 | + Licensed to the Apache Software Foundation (ASF) under one or more | |
4 | + contributor license agreements. See the NOTICE file distributed with | |
5 | + this work for additional information regarding copyright ownership. | |
6 | + The ASF licenses this file to You under the Apache License, Version 2.0 | |
7 | + (the "License"); you may not use this file except in compliance with | |
8 | + the License. You may obtain a copy of the License at | |
9 | + | |
10 | + http://www.apache.org/licenses/LICENSE-2.0 | |
11 | + | |
12 | + Unless required by applicable law or agreed to in writing, software | |
13 | + distributed under the License is distributed on an "AS IS" BASIS, | |
14 | + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
15 | + See the License for the specific language governing permissions and | |
16 | + limitations under the License. | |
17 | +--> | |
18 | + | |
19 | +<!-- | |
20 | + For more details about configurations options that may appear in | |
21 | + this file, see http://wiki.apache.org/solr/SolrConfigXml. | |
22 | +--> | |
23 | +<config> | |
24 | + <!-- In all configuration below, a prefix of "solr." for class names | |
25 | + is an alias that causes solr to search appropriate packages, | |
26 | + including org.apache.solr.(search|update|request|core|analysis) | |
27 | + | |
28 | + You may also specify a fully qualified Java classname if you | |
29 | + have your own custom plugins. | |
30 | + --> | |
31 | + | |
32 | + <!-- Controls what version of Lucene various components of Solr | |
33 | + adhere to. Generally, you want to use the latest version to | |
34 | + get all bug fixes and improvements. It is highly recommended | |
35 | + that you fully re-index after changing this setting as it can | |
36 | + affect both how text is indexed and queried. | |
37 | + --> | |
38 | + <luceneMatchVersion>4.4</luceneMatchVersion> | |
39 | + | |
40 | + <!-- <lib/> directives can be used to instruct Solr to load an Jars | |
41 | + identified and use them to resolve any "plugins" specified in | |
42 | + your solrconfig.xml or schema.xml (ie: Analyzers, Request | |
43 | + Handlers, etc...). | |
44 | + | |
45 | + All directories and paths are resolved relative to the | |
46 | + instanceDir. | |
47 | + | |
48 | + Please note that <lib/> directives are processed in the order | |
49 | + that they appear in your solrconfig.xml file, and are "stacked" | |
50 | + on top of each other when building a ClassLoader - so if you have | |
51 | + plugin jars with dependencies on other jars, the "lower level" | |
52 | + dependency jars should be loaded first. | |
53 | + | |
54 | + If a "./lib" directory exists in your instanceDir, all files | |
55 | + found in it are included as if you had used the following | |
56 | + syntax... | |
57 | + | |
58 | + <lib dir="./lib" /> | |
59 | + --> | |
60 | + | |
61 | + <!-- A 'dir' option by itself adds any files found in the directory | |
62 | + to the classpath, this is useful for including all jars in a | |
63 | + directory. | |
64 | + | |
65 | + When a 'regex' is specified in addition to a 'dir', only the | |
66 | + files in that directory which completely match the regex | |
67 | + (anchored on both ends) will be included. | |
68 | + | |
69 | + If a 'dir' option (with or without a regex) is used and nothing | |
70 | + is found that matches, a warning will be logged. | |
71 | + | |
72 | + The examples below can be used to load some solr-contribs along | |
73 | + with their external dependencies. | |
74 | + --> | |
75 | + <lib dir="../contrib/extraction/lib" regex=".*\.jar" /> | |
76 | + <lib dir="../dist/" regex="solr-cell-\d.*\.jar" /> | |
77 | + | |
78 | + <lib dir="../contrib/clustering/lib/" regex=".*\.jar" /> | |
79 | + <lib dir="../dist/" regex="solr-clustering-\d.*\.jar" /> | |
80 | + | |
81 | + <lib dir="../contrib/langid/lib/" regex=".*\.jar" /> | |
82 | + <lib dir="../dist/" regex="solr-langid-\d.*\.jar" /> | |
83 | + | |
84 | + <lib dir="../contrib/velocity/lib" regex=".*\.jar" /> | |
85 | + <lib dir="../dist/" regex="solr-velocity-\d.*\.jar" /> | |
86 | + | |
87 | + <!-- an exact 'path' can be used instead of a 'dir' to specify a | |
88 | + specific jar file. This will cause a serious error to be logged | |
89 | + if it can't be loaded. | |
90 | + --> | |
91 | + <!-- | |
92 | + <lib path="../a-jar-that-does-not-exist.jar" /> | |
93 | + --> | |
94 | + | |
95 | + <!-- Data Directory | |
96 | + | |
97 | + Used to specify an alternate directory to hold all index data | |
98 | + other than the default ./data under the Solr home. If | |
99 | + replication is in use, this should match the replication | |
100 | + configuration. | |
101 | + --> | |
102 | + <dataDir>${solr.solr.home}/${solr.core.name}/data</dataDir> | |
103 | + | |
104 | + | |
105 | + <!-- The DirectoryFactory to use for indexes. | |
106 | + | |
107 | + solr.StandardDirectoryFactory is filesystem | |
108 | + based and tries to pick the best implementation for the current | |
109 | + JVM and platform. solr.NRTCachingDirectoryFactory, the default, | |
110 | + wraps solr.StandardDirectoryFactory and caches small files in memory | |
111 | + for better NRT performance. | |
112 | + | |
113 | + One can force a particular implementation via solr.MMapDirectoryFactory, | |
114 | + solr.NIOFSDirectoryFactory, or solr.SimpleFSDirectoryFactory. | |
115 | + | |
116 | + solr.RAMDirectoryFactory is memory based, not | |
117 | + persistent, and doesn't work with replication. | |
118 | + --> | |
119 | + <directoryFactory name="DirectoryFactory" | |
120 | + class="${solr.directoryFactory:solr.NRTCachingDirectoryFactory}"/> | |
121 | + | |
122 | + <!-- The CodecFactory for defining the format of the inverted index. | |
123 | + The default implementation is SchemaCodecFactory, which is the official Lucene | |
124 | + index format, but hooks into the schema to provide per-field customization of | |
125 | + the postings lists and per-document values in the fieldType element | |
126 | + (postingsFormat/docValuesFormat). Note that most of the alternative implementations | |
127 | + are experimental, so if you choose to customize the index format, its a good | |
128 | + idea to convert back to the official format e.g. via IndexWriter.addIndexes(IndexReader) | |
129 | + before upgrading to a newer version to avoid unnecessary reindexing. | |
130 | + --> | |
131 | + <codecFactory class="solr.SchemaCodecFactory"/> | |
132 | + | |
133 | + <!-- To enable dynamic schema REST APIs, use the following for <schemaFactory>: | |
134 | + | |
135 | + <schemaFactory class="ManagedIndexSchemaFactory"> | |
136 | + <bool name="mutable">true</bool> | |
137 | + <str name="managedSchemaResourceName">managed-schema</str> | |
138 | + </schemaFactory> | |
139 | + | |
140 | + When ManagedIndexSchemaFactory is specified, Solr will load the schema from | |
141 | + he resource named in 'managedSchemaResourceName', rather than from schema.xml. | |
142 | + Note that the managed schema resource CANNOT be named schema.xml. If the managed | |
143 | + schema does not exist, Solr will create it after reading schema.xml, then rename | |
144 | + 'schema.xml' to 'schema.xml.bak'. | |
145 | + | |
146 | + Do NOT hand edit the managed schema - external modifications will be ignored and | |
147 | + overwritten as a result of schema modification REST API calls. | |
148 | + | |
149 | + When ManagedIndexSchemaFactory is specified with mutable = true, schema | |
150 | + modification REST API calls will be allowed; otherwise, error responses will be | |
151 | + sent back for these requests. | |
152 | + --> | |
153 | + <schemaFactory class="ClassicIndexSchemaFactory"/> | |
154 | + | |
155 | + <!-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
156 | + Index Config - These settings control low-level behavior of indexing | |
157 | + Most example settings here show the default value, but are commented | |
158 | + out, to more easily see where customizations have been made. | |
159 | + | |
160 | + Note: This replaces <indexDefaults> and <mainIndex> from older versions | |
161 | + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --> | |
162 | + <indexConfig> | |
163 | + <!-- maxFieldLength was removed in 4.0. To get similar behavior, include a | |
164 | + LimitTokenCountFilterFactory in your fieldType definition. --> | |
165 | + <filter class="solr.LimitTokenCountFilterFactory" maxTokenCount="50000"/> | |
166 | + <!-- Maximum time to wait for a write lock (ms) for an IndexWriter. Default: 1000 --> | |
167 | + <!-- <writeLockTimeout>1000</writeLockTimeout> --> | |
168 | + | |
169 | + <!-- The maximum number of simultaneous threads that may be | |
170 | + indexing documents at once in IndexWriter; if more than this | |
171 | + many threads arrive they will wait for others to finish. | |
172 | + Default in Solr/Lucene is 8. --> | |
173 | + <!-- <maxIndexingThreads>8</maxIndexingThreads> --> | |
174 | + | |
175 | + <!-- Expert: Enabling compound file will use less files for the index, | |
176 | + using fewer file descriptors on the expense of performance decrease. | |
177 | + Default in Lucene is "true". Default in Solr is "false" (since 3.6) --> | |
178 | + <!-- <useCompoundFile>false</useCompoundFile> --> | |
179 | + | |
180 | + <!-- ramBufferSizeMB sets the amount of RAM that may be used by Lucene | |
181 | + indexing for buffering added documents and deletions before they are | |
182 | + flushed to the Directory. | |
183 | + maxBufferedDocs sets a limit on the number of documents buffered | |
184 | + before flushing. | |
185 | + If both ramBufferSizeMB and maxBufferedDocs is set, then | |
186 | + Lucene will flush based on whichever limit is hit first. | |
187 | + The default is 100 MB. --> | |
188 | + <!-- <ramBufferSizeMB>100</ramBufferSizeMB> --> | |
189 | + <!-- <maxBufferedDocs>1000</maxBufferedDocs> --> | |
190 | + | |
191 | + <!-- Expert: Merge Policy | |
192 | + The Merge Policy in Lucene controls how merging of segments is done. | |
193 | + The default since Solr/Lucene 3.3 is TieredMergePolicy. | |
194 | + The default since Lucene 2.3 was the LogByteSizeMergePolicy, | |
195 | + Even older versions of Lucene used LogDocMergePolicy. | |
196 | + --> | |
197 | + <!-- | |
198 | + <mergePolicy class="org.apache.lucene.index.TieredMergePolicy"> | |
199 | + <int name="maxMergeAtOnce">10</int> | |
200 | + <int name="segmentsPerTier">10</int> | |
201 | + </mergePolicy> | |
202 | + --> | |
203 | + | |
204 | + <!-- Merge Factor | |
205 | + The merge factor controls how many segments will get merged at a time. | |
206 | + For TieredMergePolicy, mergeFactor is a convenience parameter which | |
207 | + will set both MaxMergeAtOnce and SegmentsPerTier at once. | |
208 | + For LogByteSizeMergePolicy, mergeFactor decides how many new segments | |
209 | + will be allowed before they are merged into one. | |
210 | + Default is 10 for both merge policies. | |
211 | + --> | |
212 | + <!-- | |
213 | + <mergeFactor>10</mergeFactor> | |
214 | + --> | |
215 | + | |
216 | + <!-- Expert: Merge Scheduler | |
217 | + The Merge Scheduler in Lucene controls how merges are | |
218 | + performed. The ConcurrentMergeScheduler (Lucene 2.3 default) | |
219 | + can perform merges in the background using separate threads. | |
220 | + The SerialMergeScheduler (Lucene 2.2 default) does not. | |
221 | + --> | |
222 | + <!-- | |
223 | + <mergeScheduler class="org.apache.lucene.index.ConcurrentMergeScheduler"/> | |
224 | + --> | |
225 | + | |
226 | + <!-- LockFactory | |
227 | + | |
228 | + This option specifies which Lucene LockFactory implementation | |
229 | + to use. | |
230 | + | |
231 | + single = SingleInstanceLockFactory - suggested for a | |
232 | + read-only index or when there is no possibility of | |
233 | + another process trying to modify the index. | |
234 | + native = NativeFSLockFactory - uses OS native file locking. | |
235 | + Do not use when multiple solr webapps in the same | |
236 | + JVM are attempting to share a single index. | |
237 | + simple = SimpleFSLockFactory - uses a plain file for locking | |
238 | + | |
239 | + Defaults: 'native' is default for Solr3.6 and later, otherwise | |
240 | + 'simple' is the default | |
241 | + | |
242 | + More details on the nuances of each LockFactory... | |
243 | + http://wiki.apache.org/lucene-java/AvailableLockFactories | |
244 | + --> | |
245 | + <lockType>${solr.lock.type:native}</lockType> | |
246 | + | |
247 | + <!-- Unlock On Startup | |
248 | + | |
249 | + If true, unlock any held write or commit locks on startup. | |
250 | + This defeats the locking mechanism that allows multiple | |
251 | + processes to safely access a lucene index, and should be used | |
252 | + with care. Default is "false". | |
253 | + | |
254 | + This is not needed if lock type is 'single' | |
255 | + --> | |
256 | + <!-- | |
257 | + <unlockOnStartup>false</unlockOnStartup> | |
258 | + --> | |
259 | + | |
260 | + <!-- Expert: Controls how often Lucene loads terms into memory | |
261 | + Default is 128 and is likely good for most everyone. | |
262 | + --> | |
263 | + <!-- <termIndexInterval>128</termIndexInterval> --> | |
264 | + | |
265 | + <!-- If true, IndexReaders will be reopened (often more efficient) | |
266 | + instead of closed and then opened. Default: true | |
267 | + --> | |
268 | + <!-- | |
269 | + <reopenReaders>true</reopenReaders> | |
270 | + --> | |
271 | + | |
272 | + <!-- Commit Deletion Policy | |
273 | + Custom deletion policies can be specified here. The class must | |
274 | + implement org.apache.lucene.index.IndexDeletionPolicy. | |
275 | + | |
276 | + The default Solr IndexDeletionPolicy implementation supports | |
277 | + deleting index commit points on number of commits, age of | |
278 | + commit point and optimized status. | |
279 | + | |
280 | + The latest commit point should always be preserved regardless | |
281 | + of the criteria. | |
282 | + --> | |
283 | + <!-- | |
284 | + <deletionPolicy class="solr.SolrDeletionPolicy"> | |
285 | + --> | |
286 | + <!-- The number of commit points to be kept --> | |
287 | + <!-- <str name="maxCommitsToKeep">1</str> --> | |
288 | + <!-- The number of optimized commit points to be kept --> | |
289 | + <!-- <str name="maxOptimizedCommitsToKeep">0</str> --> | |
290 | + <!-- | |
291 | + Delete all commit points once they have reached the given age. | |
292 | + Supports DateMathParser syntax e.g. | |
293 | + --> | |
294 | + <!-- | |
295 | + <str name="maxCommitAge">30MINUTES</str> | |
296 | + <str name="maxCommitAge">1DAY</str> | |
297 | + --> | |
298 | + <!-- | |
299 | + </deletionPolicy> | |
300 | + --> | |
301 | + | |
302 | + <!-- Lucene Infostream | |
303 | + | |
304 | + To aid in advanced debugging, Lucene provides an "InfoStream" | |
305 | + of detailed information when indexing. | |
306 | + | |
307 | + Setting the value to true will instruct the underlying Lucene | |
308 | + IndexWriter to write its info stream to solr's log. By default, | |
309 | + this is enabled here, and controlled through log4j.properties. | |
310 | + --> | |
311 | + <infoStream>true</infoStream> | |
312 | + </indexConfig> | |
313 | + | |
314 | + | |
315 | + <!-- JMX | |
316 | + | |
317 | + This example enables JMX if and only if an existing MBeanServer | |
318 | + is found, use this if you want to configure JMX through JVM | |
319 | + parameters. Remove this to disable exposing Solr configuration | |
320 | + and statistics to JMX. | |
321 | + | |
322 | + For more details see http://wiki.apache.org/solr/SolrJmx | |
323 | + --> | |
324 | + <jmx /> | |
325 | + <!-- If you want to connect to a particular server, specify the | |
326 | + agentId | |
327 | + --> | |
328 | + <!-- <jmx agentId="myAgent" /> --> | |
329 | + <!-- If you want to start a new MBeanServer, specify the serviceUrl --> | |
330 | + <!-- <jmx serviceUrl="service:jmx:rmi:///jndi/rmi://localhost:9999/solr"/> | |
331 | + --> | |
332 | + | |
333 | + <!-- The default high-performance update handler --> | |
334 | + <updateHandler class="solr.DirectUpdateHandler2"> | |
335 | + | |
336 | + <!-- Enables a transaction log, used for real-time get, durability, and | |
337 | + and solr cloud replica recovery. The log can grow as big as | |
338 | + uncommitted changes to the index, so use of a hard autoCommit | |
339 | + is recommended (see below). | |
340 | + "dir" - the target directory for transaction logs, defaults to the | |
341 | + solr data directory. --> | |
342 | + <updateLog> | |
343 | + <str name="dir">${solr.solr.home}/${solr.core.name}/txlog</str> | |
344 | + </updateLog> | |
345 | + | |
346 | + <!-- AutoCommit | |
347 | + | |
348 | + Perform a hard commit automatically under certain conditions. | |
349 | + Instead of enabling autoCommit, consider using "commitWithin" | |
350 | + when adding documents. | |
351 | + | |
352 | + http://wiki.apache.org/solr/UpdateXmlMessages | |
353 | + | |
354 | + maxDocs - Maximum number of documents to add since the last | |
355 | + commit before automatically triggering a new commit. | |
356 | + | |
357 | + maxTime - Maximum amount of time in ms that is allowed to pass | |
358 | + since a document was added before automatically | |
359 | + triggering a new commit. | |
360 | + openSearcher - if false, the commit causes recent index changes | |
361 | + to be flushed to stable storage, but does not cause a new | |
362 | + searcher to be opened to make those changes visible. | |
363 | + | |
364 | + If the updateLog is enabled, then it's highly recommended to | |
365 | + have some sort of hard autoCommit to limit the log size. | |
366 | + --> | |
367 | +<!-- | |
368 | + <autoCommit> | |
369 | + <maxTime>${solr.autoCommit.maxTime:15000}</maxTime> | |
370 | + <openSearcher>false</openSearcher> | |
371 | + </autoCommit> | |
372 | +--> | |
373 | + | |
374 | + <!-- softAutoCommit is like autoCommit except it causes a | |
375 | + 'soft' commit which only ensures that changes are visible | |
376 | + but does not ensure that data is synced to disk. This is | |
377 | + faster and more near-realtime friendly than a hard commit. | |
378 | + --> | |
379 | + <!-- | |
380 | + <autoSoftCommit> | |
381 | + <maxTime>${solr.autoSoftCommit.maxTime:-1}</maxTime> | |
382 | + </autoSoftCommit> | |
383 | + --> | |
384 | + | |
385 | + <!-- Update Related Event Listeners | |
386 | + | |
387 | + Various IndexWriter related events can trigger Listeners to | |
388 | + take actions. | |
389 | + | |
390 | + postCommit - fired after every commit or optimize command | |
391 | + postOptimize - fired after every optimize command | |
392 | + --> | |
393 | + <!-- The RunExecutableListener executes an external command from a | |
394 | + hook such as postCommit or postOptimize. | |
395 | + | |
396 | + exe - the name of the executable to run | |
397 | + dir - dir to use as the current working directory. (default=".") | |
398 | + wait - the calling thread waits until the executable returns. | |
399 | + (default="true") | |
400 | + args - the arguments to pass to the program. (default is none) | |
401 | + env - environment variables to set. (default is none) | |
402 | + --> | |
403 | + <!-- This example shows how RunExecutableListener could be used | |
404 | + with the script based replication... | |
405 | + http://wiki.apache.org/solr/CollectionDistribution | |
406 | + --> | |
407 | + <!-- | |
408 | + <listener event="postCommit" class="solr.RunExecutableListener"> | |
409 | + <str name="exe">solr/bin/snapshooter</str> | |
410 | + <str name="dir">.</str> | |
411 | + <bool name="wait">true</bool> | |
412 | + <arr name="args"> <str>arg1</str> <str>arg2</str> </arr> | |
413 | + <arr name="env"> <str>MYVAR=val1</str> </arr> | |
414 | + </listener> | |
415 | + --> | |
416 | + | |
417 | + </updateHandler> | |
418 | + | |
419 | + <!-- IndexReaderFactory | |
420 | + | |
421 | + Use the following format to specify a custom IndexReaderFactory, | |
422 | + which allows for alternate IndexReader implementations. | |
423 | + | |
424 | + ** Experimental Feature ** | |
425 | + | |
426 | + Please note - Using a custom IndexReaderFactory may prevent | |
427 | + certain other features from working. The API to | |
428 | + IndexReaderFactory may change without warning or may even be | |
429 | + removed from future releases if the problems cannot be | |
430 | + resolved. | |
431 | + | |
432 | + | |
433 | + ** Features that may not work with custom IndexReaderFactory ** | |
434 | + | |
435 | + The ReplicationHandler assumes a disk-resident index. Using a | |
436 | + custom IndexReader implementation may cause incompatibility | |
437 | + with ReplicationHandler and may cause replication to not work | |
438 | + correctly. See SOLR-1366 for details. | |
439 | + | |
440 | + --> | |
441 | + <!-- | |
442 | + <indexReaderFactory name="IndexReaderFactory" class="package.class"> | |
443 | + <str name="someArg">Some Value</str> | |
444 | + </indexReaderFactory > | |
445 | + --> | |
446 | + <!-- By explicitly declaring the Factory, the termIndexDivisor can | |
447 | + be specified. | |
448 | + --> | |
449 | + <!-- | |
450 | + <indexReaderFactory name="IndexReaderFactory" | |
451 | + class="solr.StandardIndexReaderFactory"> | |
452 | + <int name="setTermIndexDivisor">12</int> | |
453 | + </indexReaderFactory > | |
454 | + --> | |
455 | + | |
456 | + <!-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
457 | + Query section - these settings control query time things like caches | |
458 | + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --> | |
459 | + <query> | |
460 | + <!-- Max Boolean Clauses | |
461 | + | |
462 | + Maximum number of clauses in each BooleanQuery, an exception | |
463 | + is thrown if exceeded. | |
464 | + | |
465 | + ** WARNING ** | |
466 | + | |
467 | + This option actually modifies a global Lucene property that | |
468 | + will affect all SolrCores. If multiple solrconfig.xml files | |
469 | + disagree on this property, the value at any given moment will | |
470 | + be based on the last SolrCore to be initialized. | |
471 | + | |
472 | + --> | |
473 | + <maxBooleanClauses>1024</maxBooleanClauses> | |
474 | + | |
475 | + | |
476 | + <!-- Solr Internal Query Caches | |
477 | + | |
478 | + There are two implementations of cache available for Solr, | |
479 | + LRUCache, based on a synchronized LinkedHashMap, and | |
480 | + FastLRUCache, based on a ConcurrentHashMap. | |
481 | + | |
482 | + FastLRUCache has faster gets and slower puts in single | |
483 | + threaded operation and thus is generally faster than LRUCache | |
484 | + when the hit ratio of the cache is high (> 75%), and may be | |
485 | + faster under other scenarios on multi-cpu systems. | |
486 | + --> | |
487 | + | |
488 | + <!-- Filter Cache | |
489 | + | |
490 | + Cache used by SolrIndexSearcher for filters (DocSets), | |
491 | + unordered sets of *all* documents that match a query. When a | |
492 | + new searcher is opened, its caches may be prepopulated or | |
493 | + "autowarmed" using data from caches in the old searcher. | |
494 | + autowarmCount is the number of items to prepopulate. For | |
495 | + LRUCache, the autowarmed items will be the most recently | |
496 | + accessed items. | |
497 | + | |
498 | + Parameters: | |
499 | + class - the SolrCache implementation LRUCache or | |
500 | + (LRUCache or FastLRUCache) | |
501 | + size - the maximum number of entries in the cache | |
502 | + initialSize - the initial capacity (number of entries) of | |
503 | + the cache. (see java.util.HashMap) | |
504 | + autowarmCount - the number of entries to prepopulate from | |
505 | + and old cache. | |
506 | + --> | |
507 | + <filterCache class="solr.FastLRUCache" | |
508 | + size="512" | |
509 | + initialSize="512" | |
510 | + autowarmCount="0"/> | |
511 | + | |
512 | + <!-- Query Result Cache | |
513 | + | |
514 | + Caches results of searches - ordered lists of document ids | |
515 | + (DocList) based on a query, a sort, and the range of documents requested. | |
516 | + --> | |
517 | + <queryResultCache class="solr.LRUCache" | |
518 | + size="512" | |
519 | + initialSize="512" | |
520 | + autowarmCount="0"/> | |
521 | + | |
522 | + <!-- Document Cache | |
523 | + | |
524 | + Caches Lucene Document objects (the stored fields for each | |
525 | + document). Since Lucene internal document ids are transient, | |
526 | + this cache will not be autowarmed. | |
527 | + --> | |
528 | + <documentCache class="solr.LRUCache" | |
529 | + size="512" | |
530 | + initialSize="512" | |
531 | + autowarmCount="0"/> | |
532 | + | |
533 | + <!-- Field Value Cache | |
534 | + | |
535 | + Cache used to hold field values that are quickly accessible | |
536 | + by document id. The fieldValueCache is created by default | |
537 | + even if not configured here. | |
538 | + --> | |
539 | + <!-- | |
540 | + <fieldValueCache class="solr.FastLRUCache" | |
541 | + size="512" | |
542 | + autowarmCount="128" | |
543 | + showItems="32" /> | |
544 | + --> | |
545 | + | |
546 | + <!-- Custom Cache | |
547 | + | |
548 | + Example of a generic cache. These caches may be accessed by | |
549 | + name through SolrIndexSearcher.getCache(),cacheLookup(), and | |
550 | + cacheInsert(). The purpose is to enable easy caching of | |
551 | + user/application level data. The regenerator argument should | |
552 | + be specified as an implementation of solr.CacheRegenerator | |
553 | + if autowarming is desired. | |
554 | + --> | |
555 | + <!-- | |
556 | + <cache name="myUserCache" | |
557 | + class="solr.LRUCache" | |
558 | + size="4096" | |
559 | + initialSize="1024" | |
560 | + autowarmCount="1024" | |
561 | + regenerator="com.mycompany.MyRegenerator" | |
562 | + /> | |
563 | + --> | |
564 | + | |
565 | + | |
566 | + <!-- Lazy Field Loading | |
567 | + | |
568 | + If true, stored fields that are not requested will be loaded | |
569 | + lazily. This can result in a significant speed improvement | |
570 | + if the usual case is to not load all stored fields, | |
571 | + especially if the skipped fields are large compressed text | |
572 | + fields. | |
573 | + --> | |
574 | + <enableLazyFieldLoading>true</enableLazyFieldLoading> | |
575 | + | |
576 | + <!-- Use Filter For Sorted Query | |
577 | + | |
578 | + A possible optimization that attempts to use a filter to | |
579 | + satisfy a search. If the requested sort does not include | |
580 | + score, then the filterCache will be checked for a filter | |
581 | + matching the query. If found, the filter will be used as the | |
582 | + source of document ids, and then the sort will be applied to | |
583 | + that. | |
584 | + | |
585 | + For most situations, this will not be useful unless you | |
586 | + frequently get the same search repeatedly with different sort | |
587 | + options, and none of them ever use "score" | |
588 | + --> | |
589 | + <!-- | |
590 | + <useFilterForSortedQuery>true</useFilterForSortedQuery> | |
591 | + --> | |
592 | + | |
593 | + <!-- Result Window Size | |
594 | + | |
595 | + An optimization for use with the queryResultCache. When a search | |
596 | + is requested, a superset of the requested number of document ids | |
597 | + are collected. For example, if a search for a particular query | |
598 | + requests matching documents 10 through 19, and queryWindowSize is 50, | |
599 | + then documents 0 through 49 will be collected and cached. Any further | |
600 | + requests in that range can be satisfied via the cache. | |
601 | + --> | |
602 | + <queryResultWindowSize>20</queryResultWindowSize> | |
603 | + | |
604 | + <!-- Maximum number of documents to cache for any entry in the | |
605 | + queryResultCache. | |
606 | + --> | |
607 | + <queryResultMaxDocsCached>200</queryResultMaxDocsCached> | |
608 | + | |
609 | + <!-- Query Related Event Listeners | |
610 | + | |
611 | + Various IndexSearcher related events can trigger Listeners to | |
612 | + take actions. | |
613 | + | |
614 | + newSearcher - fired whenever a new searcher is being prepared | |
615 | + and there is a current searcher handling requests (aka | |
616 | + registered). It can be used to prime certain caches to | |
617 | + prevent long request times for certain requests. | |
618 | + | |
619 | + firstSearcher - fired whenever a new searcher is being | |
620 | + prepared but there is no current registered searcher to handle | |
621 | + requests or to gain autowarming data from. | |
622 | + | |
623 | + | |
624 | + --> | |
625 | + <!-- QuerySenderListener takes an array of NamedList and executes a | |
626 | + local query request for each NamedList in sequence. | |
627 | + --> | |
628 | + <listener event="newSearcher" class="solr.QuerySenderListener"> | |
629 | + <arr name="queries"> | |
630 | + <!-- | |
631 | + <lst><str name="q">solr</str><str name="sort">price asc</str></lst> | |
632 | + <lst><str name="q">rocks</str><str name="sort">weight asc</str></lst> | |
633 | + --> | |
634 | + </arr> | |
635 | + </listener> | |
636 | + <listener event="firstSearcher" class="solr.QuerySenderListener"> | |
637 | + <arr name="queries"> | |
638 | + <lst> | |
639 | + <str name="q">static firstSearcher warming in solrconfig.xml</str> | |
640 | + </lst> | |
641 | + </arr> | |
642 | + </listener> | |
643 | + | |
644 | + <!-- Use Cold Searcher | |
645 | + | |
646 | + If a search request comes in and there is no current | |
647 | + registered searcher, then immediately register the still | |
648 | + warming searcher and use it. If "false" then all requests | |
649 | + will block until the first searcher is done warming. | |
650 | + --> | |
651 | + <useColdSearcher>false</useColdSearcher> | |
652 | + | |
653 | + <!-- Max Warming Searchers | |
654 | + | |
655 | + Maximum number of searchers that may be warming in the | |
656 | + background concurrently. An error is returned if this limit | |
657 | + is exceeded. | |
658 | + | |
659 | + Recommend values of 1-2 for read-only slaves, higher for | |
660 | + masters w/o cache warming. | |
661 | + --> | |
662 | + <maxWarmingSearchers>2</maxWarmingSearchers> | |
663 | + | |
664 | + </query> | |
665 | + | |
666 | + | |
667 | + <!-- Request Dispatcher | |
668 | + | |
669 | + This section contains instructions for how the SolrDispatchFilter | |
670 | + should behave when processing requests for this SolrCore. | |
671 | + | |
672 | + handleSelect is a legacy option that affects the behavior of requests | |
673 | + such as /select?qt=XXX | |
674 | + | |
675 | + handleSelect="true" will cause the SolrDispatchFilter to process | |
676 | + the request and dispatch the query to a handler specified by the | |
677 | + "qt" param, assuming "/select" isn't already registered. | |
678 | + | |
679 | + handleSelect="false" will cause the SolrDispatchFilter to | |
680 | + ignore "/select" requests, resulting in a 404 unless a handler | |
681 | + is explicitly registered with the name "/select" | |
682 | + | |
683 | + handleSelect="true" is not recommended for new users, but is the default | |
684 | + for backwards compatibility | |
685 | + --> | |
686 | + <requestDispatcher handleSelect="false" > | |
687 | + <!-- Request Parsing | |
688 | + | |
689 | + These settings indicate how Solr Requests may be parsed, and | |
690 | + what restrictions may be placed on the ContentStreams from | |
691 | + those requests | |
692 | + | |
693 | + enableRemoteStreaming - enables use of the stream.file | |
694 | + and stream.url parameters for specifying remote streams. | |
695 | + | |
696 | + multipartUploadLimitInKB - specifies the max size (in KiB) of | |
697 | + Multipart File Uploads that Solr will allow in a Request. | |
698 | + | |
699 | + formdataUploadLimitInKB - specifies the max size (in KiB) of | |
700 | + form data (application/x-www-form-urlencoded) sent via | |
701 | + POST. You can use POST to pass request parameters not | |
702 | + fitting into the URL. | |
703 | + | |
704 | + addHttpRequestToContext - if set to true, it will instruct | |
705 | + the requestParsers to include the original HttpServletRequest | |
706 | + object in the context map of the SolrQueryRequest under the | |
707 | + key "httpRequest". It will not be used by any of the existing | |
708 | + Solr components, but may be useful when developing custom | |
709 | + plugins. | |
710 | + | |
711 | + *** WARNING *** | |
712 | + The settings below authorize Solr to fetch remote files, You | |
713 | + should make sure your system has some authentication before | |
714 | + using enableRemoteStreaming="true" | |
715 | + | |
716 | + --> | |
717 | + <requestParsers enableRemoteStreaming="true" | |
718 | + multipartUploadLimitInKB="2048000" | |
719 | + formdataUploadLimitInKB="2048" | |
720 | + addHttpRequestToContext="false"/> | |
721 | + | |
722 | + <!-- HTTP Caching | |
723 | + | |
724 | + Set HTTP caching related parameters (for proxy caches and clients). | |
725 | + | |
726 | + The options below instruct Solr not to output any HTTP Caching | |
727 | + related headers | |
728 | + --> | |
729 | + <httpCaching never304="true" /> | |
730 | + <!-- If you include a <cacheControl> directive, it will be used to | |
731 | + generate a Cache-Control header (as well as an Expires header | |
732 | + if the value contains "max-age=") | |
733 | + | |
734 | + By default, no Cache-Control header is generated. | |
735 | + | |
736 | + You can use the <cacheControl> option even if you have set | |
737 | + never304="true" | |
738 | + --> | |
739 | + <!-- | |
740 | + <httpCaching never304="true" > | |
741 | + <cacheControl>max-age=30, public</cacheControl> | |
742 | + </httpCaching> | |
743 | + --> | |
744 | + <!-- To enable Solr to respond with automatically generated HTTP | |
745 | + Caching headers, and to response to Cache Validation requests | |
746 | + correctly, set the value of never304="false" | |
747 | + | |
748 | + This will cause Solr to generate Last-Modified and ETag | |
749 | + headers based on the properties of the Index. | |
750 | + | |
751 | + The following options can also be specified to affect the | |
752 | + values of these headers... | |
753 | + | |
754 | + lastModFrom - the default value is "openTime" which means the | |
755 | + Last-Modified value (and validation against If-Modified-Since | |
756 | + requests) will all be relative to when the current Searcher | |
757 | + was opened. You can change it to lastModFrom="dirLastMod" if | |
758 | + you want the value to exactly correspond to when the physical | |
759 | + index was last modified. | |
760 | + | |
761 | + etagSeed="..." is an option you can change to force the ETag | |
762 | + header (and validation against If-None-Match requests) to be | |
763 | + different even if the index has not changed (ie: when making | |
764 | + significant changes to your config file) | |
765 | + | |
766 | + (lastModifiedFrom and etagSeed are both ignored if you use | |
767 | + the never304="true" option) | |
768 | + --> | |
769 | + <!-- | |
770 | + <httpCaching lastModifiedFrom="openTime" | |
771 | + etagSeed="Solr"> | |
772 | + <cacheControl>max-age=30, public</cacheControl> | |
773 | + </httpCaching> | |
774 | + --> | |
775 | + </requestDispatcher> | |
776 | + | |
777 | + <!-- Request Handlers | |
778 | + | |
779 | + http://wiki.apache.org/solr/SolrRequestHandler | |
780 | + | |
781 | + Incoming queries will be dispatched to a specific handler by name | |
782 | + based on the path specified in the request. | |
783 | + | |
784 | + Legacy behavior: If the request path uses "/select" but no Request | |
785 | + Handler has that name, and if handleSelect="true" has been specified in | |
786 | + the requestDispatcher, then the Request Handler is dispatched based on | |
787 | + the qt parameter. Handlers without a leading '/' are accessed this way | |
788 | + like so: http://host/app/[core/]select?qt=name If no qt is | |
789 | + given, then the requestHandler that declares default="true" will be | |
790 | + used or the one named "standard". | |
791 | + | |
792 | + If a Request Handler is declared with startup="lazy", then it will | |
793 | + not be initialized until the first request that uses it. | |
794 | + | |
795 | + --> | |
796 | + <!-- SearchHandler | |
797 | + | |
798 | + http://wiki.apache.org/solr/SearchHandler | |
799 | + | |
800 | + For processing Search Queries, the primary Request Handler | |
801 | + provided with Solr is "SearchHandler" It delegates to a sequent | |
802 | + of SearchComponents (see below) and supports distributed | |
803 | + queries across multiple shards | |
804 | + --> | |
805 | + <requestHandler name="/select" class="solr.SearchHandler"> | |
806 | + <!-- default values for query parameters can be specified, these | |
807 | + will be overridden by parameters in the request | |
808 | + --> | |
809 | + <lst name="defaults"> | |
810 | + <str name="echoParams">explicit</str> | |
811 | + <int name="rows">10</int> | |
812 | + <str name="df">content</str> | |
813 | + </lst> | |
814 | + <!-- In addition to defaults, "appends" params can be specified | |
815 | + to identify values which should be appended to the list of | |
816 | + multi-val params from the query (or the existing "defaults"). | |
817 | + --> | |
818 | + <!-- In this example, the param "fq=instock:true" would be appended to | |
819 | + any query time fq params the user may specify, as a mechanism for | |
820 | + partitioning the index, independent of any user selected filtering | |
821 | + that may also be desired (perhaps as a result of faceted searching). | |
822 | + | |
823 | + NOTE: there is *absolutely* nothing a client can do to prevent these | |
824 | + "appends" values from being used, so don't use this mechanism | |
825 | + unless you are sure you always want it. | |
826 | + --> | |
827 | + <!-- | |
828 | + <lst name="appends"> | |
829 | + <str name="fq">inStock:true</str> | |
830 | + </lst> | |
831 | + --> | |
832 | + <!-- "invariants" are a way of letting the Solr maintainer lock down | |
833 | + the options available to Solr clients. Any params values | |
834 | + specified here are used regardless of what values may be specified | |
835 | + in either the query, the "defaults", or the "appends" params. | |
836 | + | |
837 | + In this example, the facet.field and facet.query params would | |
838 | + be fixed, limiting the facets clients can use. Faceting is | |
839 | + not turned on by default - but if the client does specify | |
840 | + facet=true in the request, these are the only facets they | |
841 | + will be able to see counts for; regardless of what other | |
842 | + facet.field or facet.query params they may specify. | |
843 | + | |
844 | + NOTE: there is *absolutely* nothing a client can do to prevent these | |
845 | + "invariants" values from being used, so don't use this mechanism | |
846 | + unless you are sure you always want it. | |
847 | + --> | |
848 | + <!-- | |
849 | + <lst name="invariants"> | |
850 | + <str name="facet.field">cat</str> | |
851 | + <str name="facet.field">manu_exact</str> | |
852 | + <str name="facet.query">price:[* TO 500]</str> | |
853 | + <str name="facet.query">price:[500 TO *]</str> | |
854 | + </lst> | |
855 | + --> | |
856 | + <!-- If the default list of SearchComponents is not desired, that | |
857 | + list can either be overridden completely, or components can be | |
858 | + prepended or appended to the default list. (see below) | |
859 | + --> | |
860 | + <!-- | |
861 | + <arr name="components"> | |
862 | + <str>nameOfCustomComponent1</str> | |
863 | + <str>nameOfCustomComponent2</str> | |
864 | + </arr> | |
865 | + --> | |
866 | + </requestHandler> | |
867 | + | |
868 | + <!-- A request handler that returns indented JSON by default --> | |
869 | + <requestHandler name="/query" class="solr.SearchHandler"> | |
870 | + <lst name="defaults"> | |
871 | + <str name="echoParams">explicit</str> | |
872 | + <str name="wt">json</str> | |
873 | + <str name="indent">true</str> | |
874 | + <str name="df">content</str> | |
875 | + </lst> | |
876 | + </requestHandler> | |
877 | + | |
878 | + | |
879 | + <!-- realtime get handler, guaranteed to return the latest stored fields of | |
880 | + any document, without the need to commit or open a new searcher. The | |
881 | + current implementation relies on the updateLog feature being enabled. --> | |
882 | + <requestHandler name="/get" class="solr.RealTimeGetHandler"> | |
883 | + <lst name="defaults"> | |
884 | + <str name="omitHeader">true</str> | |
885 | + <str name="wt">json</str> | |
886 | + <str name="indent">true</str> | |
887 | + </lst> | |
888 | + </requestHandler> | |
889 | + | |
890 | + | |
891 | + <!-- A Robust Example | |
892 | + | |
893 | + This example SearchHandler declaration shows off usage of the | |
894 | + SearchHandler with many defaults declared | |
895 | + | |
896 | + Note that multiple instances of the same Request Handler | |
897 | + (SearchHandler) can be registered multiple times with different | |
898 | + names (and different init parameters) | |
899 | + --> | |
900 | + <requestHandler name="/browse" class="solr.SearchHandler"> | |
901 | + <lst name="defaults"> | |
902 | + <str name="echoParams">explicit</str> | |
903 | + | |
904 | + <!-- VelocityResponseWriter settings --> | |
905 | + <str name="wt">velocity</str> | |
906 | + <str name="v.template">browse</str> | |
907 | + <str name="v.layout">layout</str> | |
908 | + <str name="title">Solritas</str> | |
909 | + | |
910 | + <!-- Query settings --> | |
911 | + <str name="defType">edismax</str> | |
912 | + <str name="qf"> | |
913 | + content^0.5 title^10.0 | |
914 | + </str> | |
915 | + <str name="df">content</str> | |
916 | + <str name="mm">100%</str> | |
917 | + <str name="q.alt">*:*</str> | |
918 | + <str name="rows">10</str> | |
919 | + <str name="fl">*,score</str> | |
920 | + | |
921 | + <str name="mlt.qf"> | |
922 | + content^0.5 title^10.0 | |
923 | + </str> | |
924 | + <str name="mlt.fl">content,title</str> | |
925 | + <int name="mlt.count">3</int> | |
926 | + | |
927 | + <!-- Faceting defaults --> | |
928 | + <str name="facet">on</str> | |
929 | + <str name="facet.field">cat</str> | |
930 | + <str name="facet.field">manu_exact</str> | |
931 | + <str name="facet.field">content_type</str> | |
932 | + <str name="facet.field">author_s</str> | |
933 | + <str name="facet.query">ipod</str> | |
934 | + <str name="facet.query">GB</str> | |
935 | + <str name="facet.mincount">1</str> | |
936 | + <str name="facet.pivot">cat,inStock</str> | |
937 | + <str name="facet.range.other">after</str> | |
938 | + <str name="facet.range">price</str> | |
939 | + <int name="f.price.facet.range.start">0</int> | |
940 | + <int name="f.price.facet.range.end">600</int> | |
941 | + <int name="f.price.facet.range.gap">50</int> | |
942 | + <str name="facet.range">popularity</str> | |
943 | + <int name="f.popularity.facet.range.start">0</int> | |
944 | + <int name="f.popularity.facet.range.end">10</int> | |
945 | + <int name="f.popularity.facet.range.gap">3</int> | |
946 | + <str name="facet.range">manufacturedate_dt</str> | |
947 | + <str name="f.manufacturedate_dt.facet.range.start">NOW/YEAR-10YEARS</str> | |
948 | + <str name="f.manufacturedate_dt.facet.range.end">NOW</str> | |
949 | + <str name="f.manufacturedate_dt.facet.range.gap">+1YEAR</str> | |
950 | + <str name="f.manufacturedate_dt.facet.range.other">before</str> | |
951 | + <str name="f.manufacturedate_dt.facet.range.other">after</str> | |
952 | + | |
953 | + <!-- Highlighting defaults --> | |
954 | + <str name="hl">on</str> | |
955 | + <str name="hl.fl">content features title name</str> | |
956 | + <str name="hl.encoder">html</str> | |
957 | + <str name="hl.simple.pre"><b></str> | |
958 | + <str name="hl.simple.post"></b></str> | |
959 | + <str name="f.title.hl.fragsize">0</str> | |
960 | + <str name="f.title.hl.alternateField">title</str> | |
961 | + <str name="f.name.hl.fragsize">0</str> | |
962 | + <str name="f.name.hl.alternateField">name</str> | |
963 | + <str name="f.content.hl.snippets">3</str> | |
964 | + <str name="f.content.hl.fragsize">200</str> | |
965 | + <str name="f.content.hl.alternateField">content</str> | |
966 | + <str name="f.content.hl.maxAlternateFieldLength">750</str> | |
967 | + | |
968 | + <!-- Spell checking defaults --> | |
969 | + <str name="spellcheck">on</str> | |
970 | + <str name="spellcheck.extendedResults">false</str> | |
971 | + <str name="spellcheck.count">5</str> | |
972 | + <str name="spellcheck.alternativeTermCount">2</str> | |
973 | + <str name="spellcheck.maxResultsForSuggest">5</str> | |
974 | + <str name="spellcheck.collate">true</str> | |
975 | + <str name="spellcheck.collateExtendedResults">true</str> | |
976 | + <str name="spellcheck.maxCollationTries">5</str> | |
977 | + <str name="spellcheck.maxCollations">3</str> | |
978 | + </lst> | |
979 | + | |
980 | + <!-- append spellchecking to our list of components --> | |
981 | + <arr name="last-components"> | |
982 | + <str>spellcheck</str> | |
983 | + </arr> | |
984 | + </requestHandler> | |
985 | + | |
986 | + | |
987 | + <!-- Update Request Handler. | |
988 | + | |
989 | + http://wiki.apache.org/solr/UpdateXmlMessages | |
990 | + | |
991 | + The canonical Request Handler for Modifying the Index through | |
992 | + commands specified using XML, JSON, CSV, or JAVABIN | |
993 | + | |
994 | + Note: Since solr1.1 requestHandlers requires a valid content | |
995 | + type header if posted in the body. For example, curl now | |
996 | + requires: -H 'Content-type:text/xml; charset=utf-8' | |
997 | + | |
998 | + To override the request content type and force a specific | |
999 | + Content-type, use the request parameter: | |
1000 | + ?update.contentType=text/csv | |
1001 | + | |
1002 | + This handler will pick a response format to match the input | |
1003 | + if the 'wt' parameter is not explicit | |
1004 | + --> | |
1005 | + <requestHandler name="/update" class="solr.UpdateRequestHandler"> | |
1006 | + <!-- See below for information on defining | |
1007 | + updateRequestProcessorChains that can be used by name | |
1008 | + on each Update Request | |
1009 | + --> | |
1010 | + <!-- | |
1011 | + <lst name="defaults"> | |
1012 | + <str name="update.chain">dedupe</str> | |
1013 | + </lst> | |
1014 | + --> | |
1015 | + <lst name="defaults"> | |
1016 | + <str name="update.chain">langid</str> | |
1017 | + </lst> | |
1018 | + </requestHandler> | |
1019 | + | |
1020 | + <!-- for back compat with clients using /update/json and /update/csv --> | |
1021 | + <requestHandler name="/update/json" class="solr.JsonUpdateRequestHandler"> | |
1022 | + <lst name="defaults"> | |
1023 | + <str name="stream.contentType">application/json</str> | |
1024 | + </lst> | |
1025 | + </requestHandler> | |
1026 | + <requestHandler name="/update/csv" class="solr.CSVRequestHandler"> | |
1027 | + <lst name="defaults"> | |
1028 | + <str name="stream.contentType">application/csv</str> | |
1029 | + </lst> | |
1030 | + </requestHandler> | |
1031 | + | |
1032 | + <!-- Solr Cell Update Request Handler | |
1033 | + | |
1034 | + http://wiki.apache.org/solr/ExtractingRequestHandler | |
1035 | + | |
1036 | + --> | |
1037 | + <requestHandler name="/update/extract" | |
1038 | + startup="lazy" | |
1039 | + class="solr.extraction.ExtractingRequestHandler" > | |
1040 | + <lst name="defaults"> | |
1041 | + <str name="lowernames">true</str> | |
1042 | + <str name="uprefix">ignored_</str> | |
1043 | + | |
1044 | + <!-- capture link hrefs but ignore div attributes --> | |
1045 | + <str name="captureAttr">true</str> | |
1046 | + <str name="fmap.a">links</str> | |
1047 | + <str name="fmap.div">ignored_</str> | |
1048 | + </lst> | |
1049 | + </requestHandler> | |
1050 | + | |
1051 | + | |
1052 | + <!-- Field Analysis Request Handler | |
1053 | + | |
1054 | + RequestHandler that provides much the same functionality as | |
1055 | + analysis.jsp. Provides the ability to specify multiple field | |
1056 | + types and field names in the same request and outputs | |
1057 | + index-time and query-time analysis for each of them. | |
1058 | + | |
1059 | + Request parameters are: | |
1060 | + analysis.fieldname - field name whose analyzers are to be used | |
1061 | + | |
1062 | + analysis.fieldtype - field type whose analyzers are to be used | |
1063 | + analysis.fieldvalue - text for index-time analysis | |
1064 | + q (or analysis.q) - text for query time analysis | |
1065 | + analysis.showmatch (true|false) - When set to true and when | |
1066 | + query analysis is performed, the produced tokens of the | |
1067 | + field value analysis will be marked as "matched" for every | |
1068 | + token that is produces by the query analysis | |
1069 | + --> | |
1070 | + <requestHandler name="/analysis/field" | |
1071 | + startup="lazy" | |
1072 | + class="solr.FieldAnalysisRequestHandler" /> | |
1073 | + | |
1074 | + | |
1075 | + <!-- Document Analysis Handler | |
1076 | + | |
1077 | + http://wiki.apache.org/solr/AnalysisRequestHandler | |
1078 | + | |
1079 | + An analysis handler that provides a breakdown of the analysis | |
1080 | + process of provided documents. This handler expects a (single) | |
1081 | + content stream with the following format: | |
1082 | + | |
1083 | + <docs> | |
1084 | + <doc> | |
1085 | + <field name="id">1</field> | |
1086 | + <field name="name">The Name</field> | |
1087 | + <field name="text">The Text Value</field> | |
1088 | + </doc> | |
1089 | + <doc>...</doc> | |
1090 | + <doc>...</doc> | |
1091 | + ... | |
1092 | + </docs> | |
1093 | + | |
1094 | + Note: Each document must contain a field which serves as the | |
1095 | + unique key. This key is used in the returned response to associate | |
1096 | + an analysis breakdown to the analyzed document. | |
1097 | + | |
1098 | + Like the FieldAnalysisRequestHandler, this handler also supports | |
1099 | + query analysis by sending either an "analysis.query" or "q" | |
1100 | + request parameter that holds the query text to be analyzed. It | |
1101 | + also supports the "analysis.showmatch" parameter which when set to | |
1102 | + true, all field tokens that match the query tokens will be marked | |
1103 | + as a "match". | |
1104 | + --> | |
1105 | + <requestHandler name="/analysis/document" | |
1106 | + class="solr.DocumentAnalysisRequestHandler" | |
1107 | + startup="lazy" /> | |
1108 | + | |
1109 | + <!-- Admin Handlers | |
1110 | + | |
1111 | + Admin Handlers - This will register all the standard admin | |
1112 | + RequestHandlers. | |
1113 | + --> | |
1114 | + <requestHandler name="/admin/" | |
1115 | + class="solr.admin.AdminHandlers" /> | |
1116 | + <!-- This single handler is equivalent to the following... --> | |
1117 | + <!-- | |
1118 | + <requestHandler name="/admin/luke" class="solr.admin.LukeRequestHandler" /> | |
1119 | + <requestHandler name="/admin/system" class="solr.admin.SystemInfoHandler" /> | |
1120 | + <requestHandler name="/admin/plugins" class="solr.admin.PluginInfoHandler" /> | |
1121 | + <requestHandler name="/admin/threads" class="solr.admin.ThreadDumpHandler" /> | |
1122 | + <requestHandler name="/admin/properties" class="solr.admin.PropertiesRequestHandler" /> | |
1123 | + <requestHandler name="/admin/file" class="solr.admin.ShowFileRequestHandler" > | |
1124 | + --> | |
1125 | + <!-- If you wish to hide files under ${solr.home}/conf, explicitly | |
1126 | + register the ShowFileRequestHandler using: | |
1127 | + --> | |
1128 | + <!-- | |
1129 | + <requestHandler name="/admin/file" | |
1130 | + class="solr.admin.ShowFileRequestHandler" > | |
1131 | + <lst name="invariants"> | |
1132 | + <str name="hidden">synonyms.txt</str> | |
1133 | + <str name="hidden">anotherfile.txt</str> | |
1134 | + </lst> | |
1135 | + </requestHandler> | |
1136 | + --> | |
1137 | + | |
1138 | + <!-- ping/healthcheck --> | |
1139 | + <requestHandler name="/admin/ping" class="solr.PingRequestHandler"> | |
1140 | + <lst name="invariants"> | |
1141 | + <str name="q">solrpingquery</str> | |
1142 | + </lst> | |
1143 | + <lst name="defaults"> | |
1144 | + <str name="echoParams">all</str> | |
1145 | + </lst> | |
1146 | + <!-- An optional feature of the PingRequestHandler is to configure the | |
1147 | + handler with a "healthcheckFile" which can be used to enable/disable | |
1148 | + the PingRequestHandler. | |
1149 | + relative paths are resolved against the data dir | |
1150 | + --> | |
1151 | + <!-- <str name="healthcheckFile">server-enabled.txt</str> --> | |
1152 | + </requestHandler> | |
1153 | + | |
1154 | + <!-- Echo the request contents back to the client --> | |
1155 | + <requestHandler name="/debug/dump" class="solr.DumpRequestHandler" > | |
1156 | + <lst name="defaults"> | |
1157 | + <str name="echoParams">explicit</str> | |
1158 | + <str name="echoHandler">true</str> | |
1159 | + </lst> | |
1160 | + </requestHandler> | |
1161 | + | |
1162 | + <!-- Solr Replication | |
1163 | + | |
1164 | + The SolrReplicationHandler supports replicating indexes from a | |
1165 | + "master" used for indexing and "slaves" used for queries. | |
1166 | + | |
1167 | + http://wiki.apache.org/solr/SolrReplication | |
1168 | + | |
1169 | + It is also necessary for SolrCloud to function (in Cloud mode, the | |
1170 | + replication handler is used to bulk transfer segments when nodes | |
1171 | + are added or need to recover). | |
1172 | + | |
1173 | + https://wiki.apache.org/solr/SolrCloud/ | |
1174 | + --> | |
1175 | + <requestHandler name="/replication" class="solr.ReplicationHandler" > | |
1176 | + <!-- | |
1177 | + To enable simple master/slave replication, uncomment one of the | |
1178 | + sections below, depending on whether this solr instance should be | |
1179 | + the "master" or a "slave". If this instance is a "slave" you will | |
1180 | + also need to fill in the masterUrl to point to a real machine. | |
1181 | + --> | |
1182 | + <!-- | |
1183 | + <lst name="master"> | |
1184 | + <str name="replicateAfter">commit</str> | |
1185 | + <str name="replicateAfter">startup</str> | |
1186 | + <str name="confFiles">schema.xml,stopwords.txt</str> | |
1187 | + </lst> | |
1188 | + --> | |
1189 | + <!-- | |
1190 | + <lst name="slave"> | |
1191 | + <str name="masterUrl">http://your-master-hostname:8983/solr</str> | |
1192 | + <str name="pollInterval">00:00:60</str> | |
1193 | + </lst> | |
1194 | + --> | |
1195 | + </requestHandler> | |
1196 | + | |
1197 | + <!-- Search Components | |
1198 | + | |
1199 | + Search components are registered to SolrCore and used by | |
1200 | + instances of SearchHandler (which can access them by name) | |
1201 | + | |
1202 | + By default, the following components are available: | |
1203 | + | |
1204 | + <searchComponent name="query" class="solr.QueryComponent" /> | |
1205 | + <searchComponent name="facet" class="solr.FacetComponent" /> | |
1206 | + <searchComponent name="mlt" class="solr.MoreLikeThisComponent" /> | |
1207 | + <searchComponent name="highlight" class="solr.HighlightComponent" /> | |
1208 | + <searchComponent name="stats" class="solr.StatsComponent" /> | |
1209 | + <searchComponent name="debug" class="solr.DebugComponent" /> | |
1210 | + | |
1211 | + Default configuration in a requestHandler would look like: | |
1212 | + | |
1213 | + <arr name="components"> | |
1214 | + <str>query</str> | |
1215 | + <str>facet</str> | |
1216 | + <str>mlt</str> | |
1217 | + <str>highlight</str> | |
1218 | + <str>stats</str> | |
1219 | + <str>debug</str> | |
1220 | + </arr> | |
1221 | + | |
1222 | + If you register a searchComponent to one of the standard names, | |
1223 | + that will be used instead of the default. | |
1224 | + | |
1225 | + To insert components before or after the 'standard' components, use: | |
1226 | + | |
1227 | + <arr name="first-components"> | |
1228 | + <str>myFirstComponentName</str> | |
1229 | + </arr> | |
1230 | + | |
1231 | + <arr name="last-components"> | |
1232 | + <str>myLastComponentName</str> | |
1233 | + </arr> | |
1234 | + | |
1235 | + NOTE: The component registered with the name "debug" will | |
1236 | + always be executed after the "last-components" | |
1237 | + | |
1238 | + --> | |
1239 | + | |
1240 | + <!-- Spell Check | |
1241 | + | |
1242 | + The spell check component can return a list of alternative spelling | |
1243 | + suggestions. | |
1244 | + | |
1245 | + http://wiki.apache.org/solr/SpellCheckComponent | |
1246 | + --> | |
1247 | + <searchComponent name="spellcheck" class="solr.SpellCheckComponent"> | |
1248 | + | |
1249 | + <str name="queryAnalyzerFieldType">content</str> | |
1250 | + | |
1251 | + <!-- Multiple "Spell Checkers" can be declared and used by this | |
1252 | + component | |
1253 | + --> | |
1254 | + | |
1255 | + <!-- a spellchecker built from a field of the main index --> | |
1256 | + <lst name="spellchecker"> | |
1257 | + <str name="name">default</str> | |
1258 | + <str name="field">content</str> | |
1259 | + <str name="classname">solr.DirectSolrSpellChecker</str> | |
1260 | + <!-- the spellcheck distance measure used, the default is the internal levenshtein --> | |
1261 | + <str name="distanceMeasure">internal</str> | |
1262 | + <!-- minimum accuracy needed to be considered a valid spellcheck suggestion --> | |
1263 | + <float name="accuracy">0.5</float> | |
1264 | + <!-- the maximum #edits we consider when enumerating terms: can be 1 or 2 --> | |
1265 | + <int name="maxEdits">2</int> | |
1266 | + <!-- the minimum shared prefix when enumerating terms --> | |
1267 | + <int name="minPrefix">1</int> | |
1268 | + <!-- maximum number of inspections per result. --> | |
1269 | + <int name="maxInspections">5</int> | |
1270 | + <!-- minimum length of a query term to be considered for correction --> | |
1271 | + <int name="minQueryLength">4</int> | |
1272 | + <!-- maximum threshold of documents a query term can appear to be considered for correction --> | |
1273 | + <float name="maxQueryFrequency">0.01</float> | |
1274 | + <!-- uncomment this to require suggestions to occur in 1% of the documents | |
1275 | + <float name="thresholdTokenFrequency">.01</float> | |
1276 | + --> | |
1277 | + </lst> | |
1278 | + | |
1279 | + <!-- a spellchecker that can break or combine words. See "/spell" handler below for usage --> | |
1280 | + <lst name="spellchecker"> | |
1281 | + <str name="name">wordbreak</str> | |
1282 | + <str name="classname">solr.WordBreakSolrSpellChecker</str> | |
1283 | + <str name="field">name</str> | |
1284 | + <str name="combineWords">true</str> | |
1285 | + <str name="breakWords">true</str> | |
1286 | + <int name="maxChanges">10</int> | |
1287 | + </lst> | |
1288 | + | |
1289 | + <!-- a spellchecker that uses a different distance measure --> | |
1290 | + <!-- | |
1291 | + <lst name="spellchecker"> | |
1292 | + <str name="name">jarowinkler</str> | |
1293 | + <str name="field">spell</str> | |
1294 | + <str name="classname">solr.DirectSolrSpellChecker</str> | |
1295 | + <str name="distanceMeasure"> | |
1296 | + org.apache.lucene.search.spell.JaroWinklerDistance | |
1297 | + </str> | |
1298 | + </lst> | |
1299 | + --> | |
1300 | + | |
1301 | + <!-- a spellchecker that use an alternate comparator | |
1302 | + | |
1303 | + comparatorClass be one of: | |
1304 | + 1. score (default) | |
1305 | + 2. freq (Frequency first, then score) | |
1306 | + 3. A fully qualified class name | |
1307 | + --> | |
1308 | + <!-- | |
1309 | + <lst name="spellchecker"> | |
1310 | + <str name="name">freq</str> | |
1311 | + <str name="field">lowerfilt</str> | |
1312 | + <str name="classname">solr.DirectSolrSpellChecker</str> | |
1313 | + <str name="comparatorClass">freq</str> | |
1314 | + --> | |
1315 | + | |
1316 | + <!-- A spellchecker that reads the list of words from a file --> | |
1317 | + <!-- | |
1318 | + <lst name="spellchecker"> | |
1319 | + <str name="classname">solr.FileBasedSpellChecker</str> | |
1320 | + <str name="name">file</str> | |
1321 | + <str name="sourceLocation">spellings.txt</str> | |
1322 | + <str name="characterEncoding">UTF-8</str> | |
1323 | + <str name="spellcheckIndexDir">spellcheckerFile</str> | |
1324 | + </lst> | |
1325 | + --> | |
1326 | + </searchComponent> | |
1327 | + | |
1328 | + <!-- A request handler for demonstrating the spellcheck component. | |
1329 | + | |
1330 | + NOTE: This is purely as an example. The whole purpose of the | |
1331 | + SpellCheckComponent is to hook it into the request handler that | |
1332 | + handles your normal user queries so that a separate request is | |
1333 | + not needed to get suggestions. | |
1334 | + | |
1335 | + IN OTHER WORDS, THERE IS REALLY GOOD CHANCE THE SETUP BELOW IS | |
1336 | + NOT WHAT YOU WANT FOR YOUR PRODUCTION SYSTEM! | |
1337 | + | |
1338 | + See http://wiki.apache.org/solr/SpellCheckComponent for details | |
1339 | + on the request parameters. | |
1340 | + --> | |
1341 | + <requestHandler name="/spell" class="solr.SearchHandler" startup="lazy"> | |
1342 | + <lst name="defaults"> | |
1343 | + <str name="df">content</str> | |
1344 | + <!-- Solr will use suggestions from both the 'default' spellchecker | |
1345 | + and from the 'wordbreak' spellchecker and combine them. | |
1346 | + collations (re-written queries) can include a combination of | |
1347 | + corrections from both spellcheckers --> | |
1348 | + <str name="spellcheck.dictionary">default</str> | |
1349 | + <str name="spellcheck.dictionary">wordbreak</str> | |
1350 | + <str name="spellcheck">on</str> | |
1351 | + <str name="spellcheck.extendedResults">true</str> | |
1352 | + <str name="spellcheck.count">10</str> | |
1353 | + <str name="spellcheck.alternativeTermCount">5</str> | |
1354 | + <str name="spellcheck.maxResultsForSuggest">5</str> | |
1355 | + <str name="spellcheck.collate">true</str> | |
1356 | + <str name="spellcheck.collateExtendedResults">true</str> | |
1357 | + <str name="spellcheck.maxCollationTries">10</str> | |
1358 | + <str name="spellcheck.maxCollations">5</str> | |
1359 | + </lst> | |
1360 | + <arr name="last-components"> | |
1361 | + <str>spellcheck</str> | |
1362 | + </arr> | |
1363 | + </requestHandler> | |
1364 | + | |
1365 | + <!-- a suggest component --> | |
1366 | + <!-- content field --> | |
1367 | + <searchComponent class="solr.SpellCheckComponent" name="suggestContent"> | |
1368 | + <lst name="spellchecker"> | |
1369 | + <str name="name">suggestContent</str> | |
1370 | + <str name="classname">org.apache.solr.spelling.suggest.Suggester</str> | |
1371 | + <str name="lookupImpl">org.apache.solr.spelling.suggest.fst.WFSTLookupFactory</str> | |
1372 | + <str name="field">content</str> | |
1373 | + <float name="threshold">0.005</float> | |
1374 | + <str name="buildOnCommit">true</str> | |
1375 | + </lst> | |
1376 | + </searchComponent> | |
1377 | + <requestHandler class="org.apache.solr.handler.component.SearchHandler" name="/suggest/content"> | |
1378 | + <lst name="defaults"> | |
1379 | + <str name="spellcheck">true</str> | |
1380 | + <str name="spellcheck.dictionary">suggestContent</str> | |
1381 | + <str name="spellcheck.onlyMorePopular">true</str> | |
1382 | + <str name="spellcheck.count">5</str> | |
1383 | + <str name="spellcheck.collate">true</str> | |
1384 | + </lst> | |
1385 | + <arr name="components"> | |
1386 | + <str>suggestContent</str> | |
1387 | + </arr> | |
1388 | + </requestHandler> | |
1389 | + <!-- content_ja field --> | |
1390 | + <searchComponent class="solr.SpellCheckComponent" name="suggestContentJa"> | |
1391 | + <lst name="spellchecker"> | |
1392 | + <str name="name">suggestContentJa</str> | |
1393 | + <str name="classname">org.apache.solr.spelling.suggest.Suggester</str> | |
1394 | + <str name="lookupImpl">org.apache.solr.spelling.suggest.fst.WFSTLookupFactory</str> | |
1395 | + <str name="field">content_ja</str> | |
1396 | + <float name="threshold">0.005</float> | |
1397 | + <str name="buildOnCommit">true</str> | |
1398 | + </lst> | |
1399 | + </searchComponent> | |
1400 | + <requestHandler class="org.apache.solr.handler.component.SearchHandler" name="/suggest/content_ja"> | |
1401 | + <lst name="defaults"> | |
1402 | + <str name="spellcheck">true</str> | |
1403 | + <str name="spellcheck.dictionary">suggestContentJa</str> | |
1404 | + <str name="spellcheck.onlyMorePopular">true</str> | |
1405 | + <str name="spellcheck.count">5</str> | |
1406 | + <str name="spellcheck.collate">true</str> | |
1407 | + </lst> | |
1408 | + <arr name="components"> | |
1409 | + <str>suggestContentJa</str> | |
1410 | + </arr> | |
1411 | + </requestHandler> | |
1412 | + <!-- content_ts field --> | |
1413 | + <searchComponent class="solr.SpellCheckComponent" name="suggestContentTs"> | |
1414 | + <lst name="spellchecker"> | |
1415 | + <str name="name">suggestContentTs</str> | |
1416 | + <str name="classname">org.apache.solr.spelling.suggest.Suggester</str> | |
1417 | + <str name="lookupImpl">org.apache.solr.spelling.suggest.fst.WFSTLookupFactory</str> | |
1418 | + <str name="field">content_ts</str> | |
1419 | + <float name="threshold">0.005</float> | |
1420 | + <str name="buildOnCommit">true</str> | |
1421 | + </lst> | |
1422 | + </searchComponent> | |
1423 | + <requestHandler class="org.apache.solr.handler.component.SearchHandler" name="/suggest/content_ts"> | |
1424 | + <lst name="defaults"> | |
1425 | + <str name="spellcheck">true</str> | |
1426 | + <str name="spellcheck.dictionary">suggestContentTs</str> | |
1427 | + <str name="spellcheck.onlyMorePopular">true</str> | |
1428 | + <str name="spellcheck.count">5</str> | |
1429 | + <str name="spellcheck.collate">true</str> | |
1430 | + </lst> | |
1431 | + <arr name="components"> | |
1432 | + <str>suggestContentTs</str> | |
1433 | + </arr> | |
1434 | + </requestHandler> | |
1435 | + <!-- title_ss field --> | |
1436 | + <searchComponent class="solr.SpellCheckComponent" name="suggestTitleSs"> | |
1437 | + <lst name="spellchecker"> | |
1438 | + <str name="name">suggestTitleSs</str> | |
1439 | + <str name="classname">org.apache.solr.spelling.suggest.Suggester</str> | |
1440 | + <str name="lookupImpl">org.apache.solr.spelling.suggest.fst.WFSTLookupFactory</str> | |
1441 | + <str name="field">content_ja</str> | |
1442 | + <float name="threshold">0.005</float> | |
1443 | + <str name="buildOnCommit">true</str> | |
1444 | + </lst> | |
1445 | + </searchComponent> | |
1446 | + <requestHandler class="org.apache.solr.handler.component.SearchHandler" name="/suggest/title_ss"> | |
1447 | + <lst name="defaults"> | |
1448 | + <str name="spellcheck">true</str> | |
1449 | + <str name="spellcheck.dictionary">suggestTitleSs</str> | |
1450 | + <str name="spellcheck.onlyMorePopular">true</str> | |
1451 | + <str name="spellcheck.count">5</str> | |
1452 | + <str name="spellcheck.collate">true</str> | |
1453 | + </lst> | |
1454 | + <arr name="components"> | |
1455 | + <str>suggestTitleSs</str> | |
1456 | + </arr> | |
1457 | + </requestHandler> | |
1458 | + | |
1459 | + <!-- Term Vector Component | |
1460 | + | |
1461 | + http://wiki.apache.org/solr/TermVectorComponent | |
1462 | + --> | |
1463 | + <searchComponent name="tvComponent" class="solr.TermVectorComponent"/> | |
1464 | + | |
1465 | + <!-- A request handler for demonstrating the term vector component | |
1466 | + | |
1467 | + This is purely as an example. | |
1468 | + | |
1469 | + In reality you will likely want to add the component to your | |
1470 | + already specified request handlers. | |
1471 | + --> | |
1472 | + <requestHandler name="/tvrh" class="solr.SearchHandler" startup="lazy"> | |
1473 | + <lst name="defaults"> | |
1474 | + <str name="df">content</str> | |
1475 | + <bool name="tv">true</bool> | |
1476 | + </lst> | |
1477 | + <arr name="last-components"> | |
1478 | + <str>tvComponent</str> | |
1479 | + </arr> | |
1480 | + </requestHandler> | |
1481 | + | |
1482 | + <!-- Clustering Component | |
1483 | + | |
1484 | + http://wiki.apache.org/solr/ClusteringComponent | |
1485 | + | |
1486 | + You'll need to set the solr.clustering.enabled system property | |
1487 | + when running solr to run with clustering enabled: | |
1488 | + | |
1489 | + java -Dsolr.clustering.enabled=true -jar start.jar | |
1490 | + | |
1491 | + --> | |
1492 | + <searchComponent name="clustering" | |
1493 | + enable="${solr.clustering.enabled:false}" | |
1494 | + class="solr.clustering.ClusteringComponent" > | |
1495 | + <!-- Declare an engine --> | |
1496 | + <lst name="engine"> | |
1497 | + <!-- The name, only one can be named "default" --> | |
1498 | + <str name="name">default</str> | |
1499 | + | |
1500 | + <!-- Class name of Carrot2 clustering algorithm. | |
1501 | + | |
1502 | + Currently available algorithms are: | |
1503 | + | |
1504 | + * org.carrot2.clustering.lingo.LingoClusteringAlgorithm | |
1505 | + * org.carrot2.clustering.stc.STCClusteringAlgorithm | |
1506 | + * org.carrot2.clustering.kmeans.BisectingKMeansClusteringAlgorithm | |
1507 | + | |
1508 | + See http://project.carrot2.org/algorithms.html for the | |
1509 | + algorithm's characteristics. | |
1510 | + --> | |
1511 | + <str name="carrot.algorithm">org.carrot2.clustering.lingo.LingoClusteringAlgorithm</str> | |
1512 | + | |
1513 | + <!-- Overriding values for Carrot2 default algorithm attributes. | |
1514 | + | |
1515 | + For a description of all available attributes, see: | |
1516 | + http://download.carrot2.org/stable/manual/#chapter.components. | |
1517 | + Use attribute key as name attribute of str elements | |
1518 | + below. These can be further overridden for individual | |
1519 | + requests by specifying attribute key as request parameter | |
1520 | + name and attribute value as parameter value. | |
1521 | + --> | |
1522 | + <str name="LingoClusteringAlgorithm.desiredClusterCountBase">20</str> | |
1523 | + | |
1524 | + <!-- Location of Carrot2 lexical resources. | |
1525 | + | |
1526 | + A directory from which to load Carrot2-specific stop words | |
1527 | + and stop labels. Absolute or relative to Solr config directory. | |
1528 | + If a specific resource (e.g. stopwords.en) is present in the | |
1529 | + specified dir, it will completely override the corresponding | |
1530 | + default one that ships with Carrot2. | |
1531 | + | |
1532 | + For an overview of Carrot2 lexical resources, see: | |
1533 | + http://download.carrot2.org/head/manual/#chapter.lexical-resources | |
1534 | + --> | |
1535 | + <str name="carrot.lexicalResourcesDir">clustering/carrot2</str> | |
1536 | + | |
1537 | + <!-- The language to assume for the documents. | |
1538 | + | |
1539 | + For a list of allowed values, see: | |
1540 | + http://download.carrot2.org/stable/manual/#section.attribute.lingo.MultilingualClustering.defaultLanguage | |
1541 | + --> | |
1542 | + <str name="MultilingualClustering.defaultLanguage">ENGLISH</str> | |
1543 | + </lst> | |
1544 | + <lst name="engine"> | |
1545 | + <str name="name">stc</str> | |
1546 | + <str name="carrot.algorithm">org.carrot2.clustering.stc.STCClusteringAlgorithm</str> | |
1547 | + </lst> | |
1548 | + </searchComponent> | |
1549 | + | |
1550 | + <!-- A request handler for demonstrating the clustering component | |
1551 | + | |
1552 | + This is purely as an example. | |
1553 | + | |
1554 | + In reality you will likely want to add the component to your | |
1555 | + already specified request handlers. | |
1556 | + --> | |
1557 | + <requestHandler name="/clustering" | |
1558 | + startup="lazy" | |
1559 | + enable="${solr.clustering.enabled:false}" | |
1560 | + class="solr.SearchHandler"> | |
1561 | + <lst name="defaults"> | |
1562 | + <bool name="clustering">true</bool> | |
1563 | + <str name="clustering.engine">default</str> | |
1564 | + <bool name="clustering.results">true</bool> | |
1565 | + <!-- The title field --> | |
1566 | + <str name="carrot.title">name</str> | |
1567 | + <str name="carrot.url">id</str> | |
1568 | + <!-- The field to cluster on --> | |
1569 | + <str name="carrot.snippet">features</str> | |
1570 | + <!-- produce summaries --> | |
1571 | + <bool name="carrot.produceSummary">true</bool> | |
1572 | + <!-- the maximum number of labels per cluster --> | |
1573 | + <!--<int name="carrot.numDescriptions">5</int>--> | |
1574 | + <!-- produce sub clusters --> | |
1575 | + <bool name="carrot.outputSubClusters">false</bool> | |
1576 | + | |
1577 | + <str name="defType">edismax</str> | |
1578 | + <str name="qf"> | |
1579 | + content^0.5 | |
1580 | + </str> | |
1581 | + <str name="q.alt">*:*</str> | |
1582 | + <str name="rows">10</str> | |
1583 | + <str name="fl">*,score</str> | |
1584 | + </lst> | |
1585 | + <arr name="last-components"> | |
1586 | + <str>clustering</str> | |
1587 | + </arr> | |
1588 | + </requestHandler> | |
1589 | + | |
1590 | + <!-- Terms Component | |
1591 | + | |
1592 | + http://wiki.apache.org/solr/TermsComponent | |
1593 | + | |
1594 | + A component to return terms and document frequency of those | |
1595 | + terms | |
1596 | + --> | |
1597 | + <searchComponent name="terms" class="solr.TermsComponent"/> | |
1598 | + | |
1599 | + <!-- A request handler for demonstrating the terms component --> | |
1600 | + <requestHandler name="/terms" class="solr.SearchHandler" startup="lazy"> | |
1601 | + <lst name="defaults"> | |
1602 | + <bool name="terms">true</bool> | |
1603 | + <bool name="distrib">false</bool> | |
1604 | + </lst> | |
1605 | + <arr name="components"> | |
1606 | + <str>terms</str> | |
1607 | + </arr> | |
1608 | + </requestHandler> | |
1609 | + | |
1610 | + | |
1611 | + <!-- Query Elevation Component | |
1612 | + | |
1613 | + http://wiki.apache.org/solr/QueryElevationComponent | |
1614 | + | |
1615 | + a search component that enables you to configure the top | |
1616 | + results for a given query regardless of the normal lucene | |
1617 | + scoring. | |
1618 | + --> | |
1619 | + <searchComponent name="elevator" class="solr.QueryElevationComponent" > | |
1620 | + <!-- pick a fieldType to analyze queries --> | |
1621 | + <str name="queryFieldType">string</str> | |
1622 | + <str name="config-file">elevate.xml</str> | |
1623 | + </searchComponent> | |
1624 | + | |
1625 | + <!-- A request handler for demonstrating the elevator component --> | |
1626 | + <requestHandler name="/elevate" class="solr.SearchHandler" startup="lazy"> | |
1627 | + <lst name="defaults"> | |
1628 | + <str name="echoParams">explicit</str> | |
1629 | + <str name="df">content</str> | |
1630 | + </lst> | |
1631 | + <arr name="last-components"> | |
1632 | + <str>elevator</str> | |
1633 | + </arr> | |
1634 | + </requestHandler> | |
1635 | + | |
1636 | + <!-- Highlighting Component | |
1637 | + | |
1638 | + http://wiki.apache.org/solr/HighlightingParameters | |
1639 | + --> | |
1640 | + <searchComponent class="solr.HighlightComponent" name="highlight"> | |
1641 | + <highlighting> | |
1642 | + <!-- Configure the standard fragmenter --> | |
1643 | + <!-- This could most likely be commented out in the "default" case --> | |
1644 | + <fragmenter name="gap" | |
1645 | + default="true" | |
1646 | + class="solr.highlight.GapFragmenter"> | |
1647 | + <lst name="defaults"> | |
1648 | + <int name="hl.fragsize">100</int> | |
1649 | + </lst> | |
1650 | + </fragmenter> | |
1651 | + | |
1652 | + <!-- A regular-expression-based fragmenter | |
1653 | + (for sentence extraction) | |
1654 | + --> | |
1655 | + <fragmenter name="regex" | |
1656 | + class="solr.highlight.RegexFragmenter"> | |
1657 | + <lst name="defaults"> | |
1658 | + <!-- slightly smaller fragsizes work better because of slop --> | |
1659 | + <int name="hl.fragsize">70</int> | |
1660 | + <!-- allow 50% slop on fragment sizes --> | |
1661 | + <float name="hl.regex.slop">0.5</float> | |
1662 | + <!-- a basic sentence pattern --> | |
1663 | + <str name="hl.regex.pattern">[-\w ,/\n\"']{20,200}</str> | |
1664 | + </lst> | |
1665 | + </fragmenter> | |
1666 | + | |
1667 | + <!-- Configure the standard formatter --> | |
1668 | + <formatter name="html" | |
1669 | + default="true" | |
1670 | + class="solr.highlight.HtmlFormatter"> | |
1671 | + <lst name="defaults"> | |
1672 | + <str name="hl.simple.pre"><![CDATA[<em>]]></str> | |
1673 | + <str name="hl.simple.post"><![CDATA[</em>]]></str> | |
1674 | + </lst> | |
1675 | + </formatter> | |
1676 | + | |
1677 | + <!-- Configure the standard encoder --> | |
1678 | + <encoder name="html" | |
1679 | + class="solr.highlight.HtmlEncoder" /> | |
1680 | + | |
1681 | + <!-- Configure the standard fragListBuilder --> | |
1682 | + <fragListBuilder name="simple" | |
1683 | + class="solr.highlight.SimpleFragListBuilder"/> | |
1684 | + | |
1685 | + <!-- Configure the single fragListBuilder --> | |
1686 | + <fragListBuilder name="single" | |
1687 | + class="solr.highlight.SingleFragListBuilder"/> | |
1688 | + | |
1689 | + <!-- Configure the weighted fragListBuilder --> | |
1690 | + <fragListBuilder name="weighted" | |
1691 | + default="true" | |
1692 | + class="solr.highlight.WeightedFragListBuilder"/> | |
1693 | + | |
1694 | + <!-- default tag FragmentsBuilder --> | |
1695 | + <fragmentsBuilder name="default" | |
1696 | + default="true" | |
1697 | + class="solr.highlight.ScoreOrderFragmentsBuilder"> | |
1698 | + <!-- | |
1699 | + <lst name="defaults"> | |
1700 | + <str name="hl.multiValuedSeparatorChar">/</str> | |
1701 | + </lst> | |
1702 | + --> | |
1703 | + </fragmentsBuilder> | |
1704 | + | |
1705 | + <!-- multi-colored tag FragmentsBuilder --> | |
1706 | + <fragmentsBuilder name="colored" | |
1707 | + class="solr.highlight.ScoreOrderFragmentsBuilder"> | |
1708 | + <lst name="defaults"> | |
1709 | + <str name="hl.tag.pre"><![CDATA[ | |
1710 | + <b style="background:yellow">,<b style="background:lawgreen">, | |
1711 | + <b style="background:aquamarine">,<b style="background:magenta">, | |
1712 | + <b style="background:palegreen">,<b style="background:coral">, | |
1713 | + <b style="background:wheat">,<b style="background:khaki">, | |
1714 | + <b style="background:lime">,<b style="background:deepskyblue">]]></str> | |
1715 | + <str name="hl.tag.post"><![CDATA[</b>]]></str> | |
1716 | + </lst> | |
1717 | + </fragmentsBuilder> | |
1718 | + | |
1719 | + <boundaryScanner name="default" | |
1720 | + default="true" | |
1721 | + class="solr.highlight.SimpleBoundaryScanner"> | |
1722 | + <lst name="defaults"> | |
1723 | + <str name="hl.bs.maxScan">10</str> | |
1724 | + <str name="hl.bs.chars">.,!? 	 </str> | |
1725 | + </lst> | |
1726 | + </boundaryScanner> | |
1727 | + | |
1728 | + <boundaryScanner name="breakIterator" | |
1729 | + class="solr.highlight.BreakIteratorBoundaryScanner"> | |
1730 | + <lst name="defaults"> | |
1731 | + <!-- type should be one of CHARACTER, WORD(default), LINE and SENTENCE --> | |
1732 | + <str name="hl.bs.type">WORD</str> | |
1733 | + <!-- language and country are used when constructing Locale object. --> | |
1734 | + <!-- And the Locale object will be used when getting instance of BreakIterator --> | |
1735 | + <str name="hl.bs.language">en</str> | |
1736 | + <str name="hl.bs.country">US</str> | |
1737 | + </lst> | |
1738 | + </boundaryScanner> | |
1739 | + </highlighting> | |
1740 | + </searchComponent> | |
1741 | + | |
1742 | + <!-- Update Processors | |
1743 | + | |
1744 | + Chains of Update Processor Factories for dealing with Update | |
1745 | + Requests can be declared, and then used by name in Update | |
1746 | + Request Processors | |
1747 | + | |
1748 | + http://wiki.apache.org/solr/UpdateRequestProcessor | |
1749 | + | |
1750 | + --> | |
1751 | + <!-- Deduplication | |
1752 | + | |
1753 | + An example dedup update processor that creates the "id" field | |
1754 | + on the fly based on the hash code of some other fields. This | |
1755 | + example has overwriteDupes set to false since we are using the | |
1756 | + id field as the signatureField and Solr will maintain | |
1757 | + uniqueness based on that anyway. | |
1758 | + | |
1759 | + --> | |
1760 | + <!-- | |
1761 | + <updateRequestProcessorChain name="dedupe"> | |
1762 | + <processor class="solr.processor.SignatureUpdateProcessorFactory"> | |
1763 | + <bool name="enabled">true</bool> | |
1764 | + <str name="signatureField">id</str> | |
1765 | + <bool name="overwriteDupes">false</bool> | |
1766 | + <str name="fields">name,features,cat</str> | |
1767 | + <str name="signatureClass">solr.processor.Lookup3Signature</str> | |
1768 | + </processor> | |
1769 | + <processor class="solr.LogUpdateProcessorFactory" /> | |
1770 | + <processor class="solr.RunUpdateProcessorFactory" /> | |
1771 | + </updateRequestProcessorChain> | |
1772 | + --> | |
1773 | + | |
1774 | + <!-- Language identification | |
1775 | + | |
1776 | + This example update chain identifies the language of the incoming | |
1777 | + documents using the langid contrib. The detected language is | |
1778 | + written to field language_s. No field name mapping is done. | |
1779 | + The fields used for detection are text, title, subject and description, | |
1780 | + making this example suitable for detecting languages form full-text | |
1781 | + rich documents injected via ExtractingRequestHandler. | |
1782 | + See more about langId at http://wiki.apache.org/solr/LanguageDetection | |
1783 | + --> | |
1784 | + <!-- | |
1785 | + <updateRequestProcessorChain name="langid"> | |
1786 | + <processor class="org.apache.solr.update.processor.TikaLanguageIdentifierUpdateProcessorFactory"> | |
1787 | + <str name="langid.fl">text,title,subject,description</str> | |
1788 | + <str name="langid.langField">language_s</str> | |
1789 | + <str name="langid.fallback">en</str> | |
1790 | + </processor> | |
1791 | + <processor class="solr.LogUpdateProcessorFactory" /> | |
1792 | + <processor class="solr.RunUpdateProcessorFactory" /> | |
1793 | + </updateRequestProcessorChain> | |
1794 | + --> | |
1795 | + <updateRequestProcessorChain name="langid"> | |
1796 | + <processor class="org.apache.solr.update.processor.LangDetectLanguageIdentifierUpdateProcessorFactory"> | |
1797 | + <str name="langid.fl">content</str> | |
1798 | + <str name="langid.langField">solrlang_s</str> | |
1799 | + <str name="langid.whitelist">ar,bg,ca,cz,da,de,el,es,eu,fa,fi,fr,ga,gl,hi,hu,hy,id,it,ja,lv,ko,nl,no,pt,ro,ru,sv,th,tr,zh,zh_CN,zh_TW</str> | |
1800 | + <bool name="langid.map">true</bool> | |
1801 | + <bool name="langid.map.keepOrig">true</bool> | |
1802 | + <str name="langid.map.lcmap">zh:cjk zh_CN:cjk zh_TW:cjk ko:cjk</str> | |
1803 | + <!-- If you want to run on English environment, replace with en --> | |
1804 | + <str name="langid.fallback">ja</str> | |
1805 | + </processor> | |
1806 | + <processor class="solr.LogUpdateProcessorFactory" /> | |
1807 | + <processor class="solr.RunUpdateProcessorFactory" /> | |
1808 | + </updateRequestProcessorChain> | |
1809 | + | |
1810 | + <!-- Script update processor | |
1811 | + | |
1812 | + This example hooks in an update processor implemented using JavaScript. | |
1813 | + | |
1814 | + See more about the script update processor at http://wiki.apache.org/solr/ScriptUpdateProcessor | |
1815 | + --> | |
1816 | + <!-- | |
1817 | + <updateRequestProcessorChain name="script"> | |
1818 | + <processor class="solr.StatelessScriptUpdateProcessorFactory"> | |
1819 | + <str name="script">update-script.js</str> | |
1820 | + <lst name="params"> | |
1821 | + <str name="config_param">example config parameter</str> | |
1822 | + </lst> | |
1823 | + </processor> | |
1824 | + <processor class="solr.RunUpdateProcessorFactory" /> | |
1825 | + </updateRequestProcessorChain> | |
1826 | + --> | |
1827 | + | |
1828 | + <!-- Response Writers | |
1829 | + | |
1830 | + http://wiki.apache.org/solr/QueryResponseWriter | |
1831 | + | |
1832 | + Request responses will be written using the writer specified by | |
1833 | + the 'wt' request parameter matching the name of a registered | |
1834 | + writer. | |
1835 | + | |
1836 | + The "default" writer is the default and will be used if 'wt' is | |
1837 | + not specified in the request. | |
1838 | + --> | |
1839 | + <!-- The following response writers are implicitly configured unless | |
1840 | + overridden... | |
1841 | + --> | |
1842 | + <!-- | |
1843 | + <queryResponseWriter name="xml" | |
1844 | + default="true" | |
1845 | + class="solr.XMLResponseWriter" /> | |
1846 | + <queryResponseWriter name="json" class="solr.JSONResponseWriter"/> | |
1847 | + <queryResponseWriter name="python" class="solr.PythonResponseWriter"/> | |
1848 | + <queryResponseWriter name="ruby" class="solr.RubyResponseWriter"/> | |
1849 | + <queryResponseWriter name="php" class="solr.PHPResponseWriter"/> | |
1850 | + <queryResponseWriter name="phps" class="solr.PHPSerializedResponseWriter"/> | |
1851 | + <queryResponseWriter name="csv" class="solr.CSVResponseWriter"/> | |
1852 | + <queryResponseWriter name="schema.xml" class="solr.SchemaXmlResponseWriter"/> | |
1853 | + --> | |
1854 | + | |
1855 | + <queryResponseWriter name="json" class="solr.JSONResponseWriter"> | |
1856 | + <!-- For the purposes of the tutorial, JSON responses are written as | |
1857 | + plain text so that they are easy to read in *any* browser. | |
1858 | + If you expect a MIME type of "application/json" just remove this override. | |
1859 | + --> | |
1860 | + <str name="content-type">text/plain; charset=UTF-8</str> | |
1861 | + </queryResponseWriter> | |
1862 | + | |
1863 | + <!-- | |
1864 | + Custom response writers can be declared as needed... | |
1865 | + --> | |
1866 | + <queryResponseWriter name="velocity" class="solr.VelocityResponseWriter" startup="lazy"/> | |
1867 | + | |
1868 | + | |
1869 | + <!-- XSLT response writer transforms the XML output by any xslt file found | |
1870 | + in Solr's conf/xslt directory. Changes to xslt files are checked for | |
1871 | + every xsltCacheLifetimeSeconds. | |
1872 | + --> | |
1873 | + <queryResponseWriter name="xslt" class="solr.XSLTResponseWriter"> | |
1874 | + <int name="xsltCacheLifetimeSeconds">5</int> | |
1875 | + </queryResponseWriter> | |
1876 | + | |
1877 | + <!-- Query Parsers | |
1878 | + | |
1879 | + http://wiki.apache.org/solr/SolrQuerySyntax | |
1880 | + | |
1881 | + Multiple QParserPlugins can be registered by name, and then | |
1882 | + used in either the "defType" param for the QueryComponent (used | |
1883 | + by SearchHandler) or in LocalParams | |
1884 | + --> | |
1885 | + <!-- example of registering a query parser --> | |
1886 | + <!-- | |
1887 | + <queryParser name="myparser" class="com.mycompany.MyQParserPlugin"/> | |
1888 | + --> | |
1889 | + | |
1890 | + <!-- Function Parsers | |
1891 | + | |
1892 | + http://wiki.apache.org/solr/FunctionQuery | |
1893 | + | |
1894 | + Multiple ValueSourceParsers can be registered by name, and then | |
1895 | + used as function names when using the "func" QParser. | |
1896 | + --> | |
1897 | + <!-- example of registering a custom function parser --> | |
1898 | + <!-- | |
1899 | + <valueSourceParser name="myfunc" | |
1900 | + class="com.mycompany.MyValueSourceParser" /> | |
1901 | + --> | |
1902 | + <valueSourceParser name="wordfreq" | |
1903 | + class="jp.sf.fess.solr.plugin.search.WordFreqValueSourceParser" /> | |
1904 | + | |
1905 | + | |
1906 | + <!-- Document Transformers | |
1907 | + http://wiki.apache.org/solr/DocTransformers | |
1908 | + --> | |
1909 | + <!-- | |
1910 | + Could be something like: | |
1911 | + <transformer name="db" class="com.mycompany.LoadFromDatabaseTransformer" > | |
1912 | + <int name="connection">jdbc://....</int> | |
1913 | + </transformer> | |
1914 | + | |
1915 | + To add a constant value to all docs, use: | |
1916 | + <transformer name="mytrans2" class="org.apache.solr.response.transform.ValueAugmenterFactory" > | |
1917 | + <int name="value">5</int> | |
1918 | + </transformer> | |
1919 | + | |
1920 | + If you want the user to still be able to change it with _value:something_ use this: | |
1921 | + <transformer name="mytrans3" class="org.apache.solr.response.transform.ValueAugmenterFactory" > | |
1922 | + <double name="defaultValue">5</double> | |
1923 | + </transformer> | |
1924 | + | |
1925 | + If you are using the QueryElevationComponent, you may wish to mark documents that get boosted. The | |
1926 | + EditorialMarkerFactory will do exactly that: | |
1927 | + <transformer name="qecBooster" class="org.apache.solr.response.transform.EditorialMarkerFactory" /> | |
1928 | + --> | |
1929 | + | |
1930 | + | |
1931 | + <!-- Legacy config for the admin interface --> | |
1932 | + <admin> | |
1933 | + <defaultQuery>*:*</defaultQuery> | |
1934 | + </admin> | |
1935 | + | |
1936 | +</config> |
@@ -0,0 +1,44 @@ | ||
1 | +<?xml version="1.0" encoding="UTF-8" ?> | |
2 | +<!-- | |
3 | + Licensed to the Apache Software Foundation (ASF) under one or more | |
4 | + contributor license agreements. See the NOTICE file distributed with | |
5 | + this work for additional information regarding copyright ownership. | |
6 | + The ASF licenses this file to You under the Apache License, Version 2.0 | |
7 | + (the "License"); you may not use this file except in compliance with | |
8 | + the License. You may obtain a copy of the License at | |
9 | + | |
10 | + http://www.apache.org/licenses/LICENSE-2.0 | |
11 | + | |
12 | + Unless required by applicable law or agreed to in writing, software | |
13 | + distributed under the License is distributed on an "AS IS" BASIS, | |
14 | + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
15 | + See the License for the specific language governing permissions and | |
16 | + limitations under the License. | |
17 | +--> | |
18 | + | |
19 | +<!-- | |
20 | + This is an example of a simple "solr.xml" file for configuring one or | |
21 | + more Solr Cores, as well as allowing Cores to be added, removed, and | |
22 | + reloaded via HTTP requests. | |
23 | + | |
24 | + More information about options available in this configuration file, | |
25 | + and Solr Core administration can be found online: | |
26 | + http://wiki.apache.org/solr/CoreAdmin | |
27 | +--> | |
28 | + | |
29 | +<solr> | |
30 | + | |
31 | +<!-- | |
32 | + <solrcloud> | |
33 | + <int name="zkClientTimeout">${zkClientTimeout:15000}</int> | |
34 | + <bool name="genericCoreNodeNames">${genericCoreNodeNames:true}</bool> | |
35 | + </solrcloud> | |
36 | +--> | |
37 | + | |
38 | + <shardHandlerFactory name="shardHandlerFactory" | |
39 | + class="HttpShardHandlerFactory"> | |
40 | + <int name="socketTimeout">${socketTimeout:0}</int> | |
41 | + <int name="connTimeout">${connTimeout:0}</int> | |
42 | + </shardHandlerFactory> | |
43 | + | |
44 | +</solr> |
@@ -0,0 +1,189 @@ | ||
1 | +<?xml version="1.0" encoding="UTF-8"?> | |
2 | +<!-- | |
3 | + Licensed to the Apache Software Foundation (ASF) under one or more | |
4 | + contributor license agreements. See the NOTICE file distributed with | |
5 | + this work for additional information regarding copyright ownership. | |
6 | + The ASF licenses this file to You under the Apache License, Version 2.0 | |
7 | + (the "License"); you may not use this file except in compliance with | |
8 | + the License. You may obtain a copy of the License at | |
9 | + | |
10 | + http://www.apache.org/licenses/LICENSE-2.0 | |
11 | + | |
12 | + Unless required by applicable law or agreed to in writing, software | |
13 | + distributed under the License is distributed on an "AS IS" BASIS, | |
14 | + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
15 | + See the License for the specific language governing permissions and | |
16 | + limitations under the License. | |
17 | +--> | |
18 | +<web-app xmlns="http://java.sun.com/xml/ns/javaee" | |
19 | + xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" | |
20 | + xsi:schemaLocation="http://java.sun.com/xml/ns/javaee http://java.sun.com/xml/ns/javaee/web-app_2_5.xsd" | |
21 | + version="2.5" | |
22 | + metadata-complete="true" | |
23 | +> | |
24 | + | |
25 | + | |
26 | + <!-- Uncomment if you are trying to use a Resin version before 3.0.19. | |
27 | + Their XML implementation isn't entirely compatible with Xerces. | |
28 | + Below are the implementations to use with Sun's JVM. | |
29 | + <system-property javax.xml.xpath.XPathFactory= | |
30 | + "com.sun.org.apache.xpath.internal.jaxp.XPathFactoryImpl"/> | |
31 | + <system-property javax.xml.parsers.DocumentBuilderFactory= | |
32 | + "com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl"/> | |
33 | + <system-property javax.xml.parsers.SAXParserFactory= | |
34 | + "com.sun.org.apache.xerces.internal.jaxp.SAXParserFactoryImpl"/> | |
35 | + --> | |
36 | + | |
37 | + <!-- People who want to hardcode their "Solr Home" directly into the | |
38 | + WAR File can set the JNDI property here... | |
39 | + --> | |
40 | + <!-- | |
41 | + <env-entry> | |
42 | + <env-entry-name>solr/home</env-entry-name> | |
43 | + <env-entry-value>/put/your/solr/home/here</env-entry-value> | |
44 | + <env-entry-type>java.lang.String</env-entry-type> | |
45 | + </env-entry> | |
46 | + --> | |
47 | + | |
48 | + <!-- Any path (name) registered in solrconfig.xml will be sent to that filter --> | |
49 | + <filter> | |
50 | + <filter-name>SolrRequestFilter</filter-name> | |
51 | + <filter-class>org.apache.solr.servlet.SolrDispatchFilter</filter-class> | |
52 | + <!-- If you are wiring Solr into a larger web application which controls | |
53 | + the web context root, you will probably want to mount Solr under | |
54 | + a path prefix (app.war with /app/solr mounted into it, for example). | |
55 | + You will need to put this prefix in front of the SolrDispatchFilter | |
56 | + url-pattern mapping too (/solr/*), and also on any paths for | |
57 | + legacy Solr servlet mappings you may be using. | |
58 | + For the Admin UI to work properly in a path-prefixed configuration, | |
59 | + the admin folder containing the resources needs to be under the app context root | |
60 | + named to match the path-prefix. For example: | |
61 | + | |
62 | + .war | |
63 | + xxx | |
64 | + js | |
65 | + main.js | |
66 | + --> | |
67 | + <!-- | |
68 | + <init-param> | |
69 | + <param-name>path-prefix</param-name> | |
70 | + <param-value>/xxx</param-value> | |
71 | + </init-param> | |
72 | + --> | |
73 | + </filter> | |
74 | + | |
75 | + <filter-mapping> | |
76 | + <!-- | |
77 | + NOTE: When using multicore, /admin JSP URLs with a core specified | |
78 | + such as /solr/coreName/admin/stats.jsp get forwarded by a | |
79 | + RequestDispatcher to /solr/admin/stats.jsp with the specified core | |
80 | + put into request scope keyed as "org.apache.solr.SolrCore". | |
81 | + | |
82 | + It is unnecessary, and potentially problematic, to have the SolrDispatchFilter | |
83 | + configured to also filter on forwards. Do not configure | |
84 | + this dispatcher as <dispatcher>FORWARD</dispatcher>. | |
85 | + --> | |
86 | + <filter-name>SolrRequestFilter</filter-name> | |
87 | + <url-pattern>/*</url-pattern> | |
88 | + </filter-mapping> | |
89 | + | |
90 | + <servlet> | |
91 | + <servlet-name>Zookeeper</servlet-name> | |
92 | + <servlet-class>org.apache.solr.servlet.ZookeeperInfoServlet</servlet-class> | |
93 | + </servlet> | |
94 | + | |
95 | + <servlet> | |
96 | + <servlet-name>LoadAdminUI</servlet-name> | |
97 | + <servlet-class>org.apache.solr.servlet.LoadAdminUiServlet</servlet-class> | |
98 | + </servlet> | |
99 | + | |
100 | + <!-- Remove in Solr 5.0 --> | |
101 | + <!-- This sends SC_MOVED_PERMANENTLY (301) for resources that changed in 4.0 --> | |
102 | + <servlet> | |
103 | + <servlet-name>RedirectOldAdminUI</servlet-name> | |
104 | + <servlet-class>org.apache.solr.servlet.RedirectServlet</servlet-class> | |
105 | + <init-param> | |
106 | + <param-name>destination</param-name> | |
107 | + <param-value>${context}/#/</param-value> | |
108 | + </init-param> | |
109 | + </servlet> | |
110 | + | |
111 | + <servlet> | |
112 | + <servlet-name>RedirectOldZookeeper</servlet-name> | |
113 | + <servlet-class>org.apache.solr.servlet.RedirectServlet</servlet-class> | |
114 | + <init-param> | |
115 | + <param-name>destination</param-name> | |
116 | + <param-value>${context}/zookeeper</param-value> | |
117 | + </init-param> | |
118 | + </servlet> | |
119 | + | |
120 | + <servlet> | |
121 | + <servlet-name>RedirectLogging</servlet-name> | |
122 | + <servlet-class>org.apache.solr.servlet.RedirectServlet</servlet-class> | |
123 | + <init-param> | |
124 | + <param-name>destination</param-name> | |
125 | + <param-value>${context}/#/~logging</param-value> | |
126 | + </init-param> | |
127 | + </servlet> | |
128 | + | |
129 | + <servlet-mapping> | |
130 | + <servlet-name>RedirectOldAdminUI</servlet-name> | |
131 | + <url-pattern>/admin/</url-pattern> | |
132 | + </servlet-mapping> | |
133 | + <servlet-mapping> | |
134 | + <servlet-name>RedirectOldAdminUI</servlet-name> | |
135 | + <url-pattern>/admin</url-pattern> | |
136 | + </servlet-mapping> | |
137 | + <servlet-mapping> | |
138 | + <servlet-name>RedirectOldZookeeper</servlet-name> | |
139 | + <url-pattern>/zookeeper.jsp</url-pattern> | |
140 | + </servlet-mapping> | |
141 | + <servlet-mapping> | |
142 | + <servlet-name>RedirectLogging</servlet-name> | |
143 | + <url-pattern>/logging</url-pattern> | |
144 | + </servlet-mapping> | |
145 | + | |
146 | + <!-- Servlet Mapping --> | |
147 | + <servlet-mapping> | |
148 | + <servlet-name>Zookeeper</servlet-name> | |
149 | + <url-pattern>/zookeeper</url-pattern> | |
150 | + </servlet-mapping> | |
151 | + | |
152 | + <servlet-mapping> | |
153 | + <servlet-name>LoadAdminUI</servlet-name> | |
154 | + <url-pattern>/admin.html</url-pattern> | |
155 | + </servlet-mapping> | |
156 | + | |
157 | + <mime-mapping> | |
158 | + <extension>.xsl</extension> | |
159 | + <!-- per http://www.w3.org/TR/2006/PR-xslt20-20061121/ --> | |
160 | + <mime-type>application/xslt+xml</mime-type> | |
161 | + </mime-mapping> | |
162 | + | |
163 | + <welcome-file-list> | |
164 | + <welcome-file>admin.html</welcome-file> | |
165 | + </welcome-file-list> | |
166 | + | |
167 | + <security-constraint> | |
168 | + <web-resource-collection> | |
169 | + <web-resource-name>Protected Solr</web-resource-name> | |
170 | + <url-pattern>/*</url-pattern> | |
171 | + <http-method>GET</http-method> | |
172 | + <http-method>POST</http-method> | |
173 | + </web-resource-collection> | |
174 | + <auth-constraint> | |
175 | + <role-name>solr</role-name> | |
176 | + </auth-constraint> | |
177 | + </security-constraint> | |
178 | + | |
179 | + <login-config> | |
180 | + <auth-method>BASIC</auth-method> | |
181 | + <realm-name>Basic Authentication</realm-name> | |
182 | + </login-config> | |
183 | + | |
184 | + <security-role> | |
185 | + <description>Role for Accessing Solr</description> | |
186 | + <role-name>solr</role-name> | |
187 | + </security-role> | |
188 | + | |
189 | +</web-app> |
@@ -0,0 +1,32 @@ | ||
1 | +<?xml version='1.0' encoding='utf-8'?> | |
2 | +<!-- | |
3 | + Licensed to the Apache Software Foundation (ASF) under one or more | |
4 | + contributor license agreements. See the NOTICE file distributed with | |
5 | + this work for additional information regarding copyright ownership. | |
6 | + The ASF licenses this file to You under the Apache License, Version 2.0 | |
7 | + (the "License"); you may not use this file except in compliance with | |
8 | + the License. You may obtain a copy of the License at | |
9 | + | |
10 | + http://www.apache.org/licenses/LICENSE-2.0 | |
11 | + | |
12 | + Unless required by applicable law or agreed to in writing, software | |
13 | + distributed under the License is distributed on an "AS IS" BASIS, | |
14 | + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
15 | + See the License for the specific language governing permissions and | |
16 | + limitations under the License. | |
17 | +--> | |
18 | +<tomcat-users> | |
19 | + <role rolename="fess"/> | |
20 | + <role rolename="solr"/> | |
21 | + <role rolename="manager-script"/> | |
22 | + <user username="admin" password="admin" roles="fess"/> | |
23 | + <user username="solradmin" password="solradmin" roles="solr"/> | |
24 | + <user username="manager" password="manager" roles="manager-script"/> | |
25 | +<!-- | |
26 | + <role rolename="tomcat"/> | |
27 | + <role rolename="role1"/> | |
28 | + <user username="tomcat" password="tomcat" roles="tomcat"/> | |
29 | + <user username="both" password="tomcat" roles="tomcat,role1"/> | |
30 | + <user username="role1" password="tomcat" roles="role1"/> | |
31 | +--> | |
32 | +</tomcat-users> |
@@ -0,0 +1,142 @@ | ||
1 | +<?xml version='1.0' encoding='utf-8'?> | |
2 | +<!-- | |
3 | + Licensed to the Apache Software Foundation (ASF) under one or more | |
4 | + contributor license agreements. See the NOTICE file distributed with | |
5 | + this work for additional information regarding copyright ownership. | |
6 | + The ASF licenses this file to You under the Apache License, Version 2.0 | |
7 | + (the "License"); you may not use this file except in compliance with | |
8 | + the License. You may obtain a copy of the License at | |
9 | + | |
10 | + http://www.apache.org/licenses/LICENSE-2.0 | |
11 | + | |
12 | + Unless required by applicable law or agreed to in writing, software | |
13 | + distributed under the License is distributed on an "AS IS" BASIS, | |
14 | + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
15 | + See the License for the specific language governing permissions and | |
16 | + limitations under the License. | |
17 | +--> | |
18 | +<!-- Note: A "Server" is not itself a "Container", so you may not | |
19 | + define subcomponents such as "Valves" at this level. | |
20 | + Documentation at /docs/config/server.html | |
21 | + --> | |
22 | +<Server port="8005" shutdown="SHUTDOWN"> | |
23 | + <!-- Security listener. Documentation at /docs/config/listeners.html | |
24 | + <Listener className="org.apache.catalina.security.SecurityListener" /> | |
25 | + --> | |
26 | + <!--APR library loader. Documentation at /docs/apr.html --> | |
27 | + <Listener className="org.apache.catalina.core.AprLifecycleListener" SSLEngine="on" /> | |
28 | + <!--Initialize Jasper prior to webapps are loaded. Documentation at /docs/jasper-howto.html --> | |
29 | + <Listener className="org.apache.catalina.core.JasperListener" /> | |
30 | + <!-- Prevent memory leaks due to use of particular java/javax APIs--> | |
31 | + <Listener className="org.apache.catalina.core.JreMemoryLeakPreventionListener" /> | |
32 | + <Listener className="org.apache.catalina.mbeans.GlobalResourcesLifecycleListener" /> | |
33 | + <Listener className="org.apache.catalina.core.ThreadLocalLeakPreventionListener" /> | |
34 | + | |
35 | + <!-- Global JNDI resources | |
36 | + Documentation at /docs/jndi-resources-howto.html | |
37 | + --> | |
38 | + <GlobalNamingResources> | |
39 | + <!-- Editable user database that can also be used by | |
40 | + UserDatabaseRealm to authenticate users | |
41 | + --> | |
42 | + <Resource name="UserDatabase" auth="Container" | |
43 | + type="org.apache.catalina.UserDatabase" | |
44 | + description="User database that can be updated and saved" | |
45 | + factory="org.apache.catalina.users.MemoryUserDatabaseFactory" | |
46 | + pathname="conf/tomcat-users.xml" /> | |
47 | + </GlobalNamingResources> | |
48 | + | |
49 | + <!-- A "Service" is a collection of one or more "Connectors" that share | |
50 | + a single "Container" Note: A "Service" is not itself a "Container", | |
51 | + so you may not define subcomponents such as "Valves" at this level. | |
52 | + Documentation at /docs/config/service.html | |
53 | + --> | |
54 | + <Service name="Catalina"> | |
55 | + | |
56 | + <!--The connectors can use a shared executor, you can define one or more named thread pools--> | |
57 | + <!-- | |
58 | + <Executor name="tomcatThreadPool" namePrefix="catalina-exec-" | |
59 | + maxThreads="150" minSpareThreads="4"/> | |
60 | + --> | |
61 | + | |
62 | + | |
63 | + <!-- A "Connector" represents an endpoint by which requests are received | |
64 | + and responses are returned. Documentation at : | |
65 | + Java HTTP Connector: /docs/config/http.html (blocking & non-blocking) | |
66 | + Java AJP Connector: /docs/config/ajp.html | |
67 | + APR (HTTP/AJP) Connector: /docs/apr.html | |
68 | + Define a non-SSL HTTP/1.1 Connector on port 8080 | |
69 | + --> | |
70 | + <Connector port="8080" protocol="HTTP/1.1" | |
71 | + connectionTimeout="20000" | |
72 | + redirectPort="8443" URIEncoding="UTF-8" /> | |
73 | + <!-- A "Connector" using the shared thread pool--> | |
74 | + <!-- | |
75 | + <Connector executor="tomcatThreadPool" | |
76 | + port="8080" protocol="HTTP/1.1" | |
77 | + connectionTimeout="20000" | |
78 | + redirectPort="8443" /> | |
79 | + --> | |
80 | + <!-- Define a SSL HTTP/1.1 Connector on port 8443 | |
81 | + This connector uses the JSSE configuration, when using APR, the | |
82 | + connector should be using the OpenSSL style configuration | |
83 | + described in the APR documentation --> | |
84 | + <!-- | |
85 | + <Connector port="8443" protocol="HTTP/1.1" SSLEnabled="true" | |
86 | + maxThreads="150" scheme="https" secure="true" | |
87 | + clientAuth="false" sslProtocol="TLS" /> | |
88 | + --> | |
89 | + | |
90 | + <!-- Define an AJP 1.3 Connector on port 8009 --> | |
91 | + <Connector port="8009" protocol="AJP/1.3" redirectPort="8443" URIEncoding="UTF-8" /> | |
92 | + | |
93 | + | |
94 | + <!-- An Engine represents the entry point (within Catalina) that processes | |
95 | + every request. The Engine implementation for Tomcat stand alone | |
96 | + analyzes the HTTP headers included with the request, and passes them | |
97 | + on to the appropriate Host (virtual host). | |
98 | + Documentation at /docs/config/engine.html --> | |
99 | + | |
100 | + <!-- You should set jvmRoute to support load-balancing via AJP ie : | |
101 | + <Engine name="Catalina" defaultHost="localhost" jvmRoute="jvm1"> | |
102 | + --> | |
103 | + <Engine name="Catalina" defaultHost="localhost"> | |
104 | + | |
105 | + <!--For clustering, please take a look at documentation at: | |
106 | + /docs/cluster-howto.html (simple how to) | |
107 | + /docs/config/cluster.html (reference documentation) --> | |
108 | + <!-- | |
109 | + <Cluster className="org.apache.catalina.ha.tcp.SimpleTcpCluster"/> | |
110 | + --> | |
111 | + | |
112 | + <!-- Use the LockOutRealm to prevent attempts to guess user passwords | |
113 | + via a brute-force attack --> | |
114 | + <Realm className="org.apache.catalina.realm.LockOutRealm"> | |
115 | + <!-- This Realm uses the UserDatabase configured in the global JNDI | |
116 | + resources under the key "UserDatabase". Any edits | |
117 | + that are performed against this UserDatabase are immediately | |
118 | + available for use by the Realm. --> | |
119 | + <Realm className="org.apache.catalina.realm.UserDatabaseRealm" | |
120 | + resourceName="UserDatabase"/> | |
121 | + </Realm> | |
122 | + | |
123 | + <Host name="localhost" appBase="webapps" | |
124 | + unpackWARs="true" autoDeploy="true"> | |
125 | + | |
126 | + <!-- SingleSignOn valve, share authentication between web applications | |
127 | + Documentation at: /docs/config/valve.html --> | |
128 | + <!-- | |
129 | + <Valve className="org.apache.catalina.authenticator.SingleSignOn" /> | |
130 | + --> | |
131 | + | |
132 | + <!-- Access log processes all example. | |
133 | + Documentation at: /docs/config/valve.html | |
134 | + Note: The pattern used is equivalent to using pattern="common" --> | |
135 | + <Valve className="org.apache.catalina.valves.AccessLogValve" directory="logs" | |
136 | + prefix="localhost_access_log." suffix=".txt" | |
137 | + pattern="%h %l %u %t "%r" %s %b" resolveHosts="false"/> | |
138 | + | |
139 | + </Host> | |
140 | + </Engine> | |
141 | + </Service> | |
142 | +</Server> |
@@ -0,0 +1,166 @@ | ||
1 | +@echo off | |
2 | +rem Licensed to the Apache Software Foundation (ASF) under one or more | |
3 | +rem contributor license agreements. See the NOTICE file distributed with | |
4 | +rem this work for additional information regarding copyright ownership. | |
5 | +rem The ASF licenses this file to You under the Apache License, Version 2.0 | |
6 | +rem (the "License"); you may not use this file except in compliance with | |
7 | +rem the License. You may obtain a copy of the License at | |
8 | +rem | |
9 | +rem http://www.apache.org/licenses/LICENSE-2.0 | |
10 | +rem | |
11 | +rem Unless required by applicable law or agreed to in writing, software | |
12 | +rem distributed under the License is distributed on an "AS IS" BASIS, | |
13 | +rem WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
14 | +rem See the License for the specific language governing permissions and | |
15 | +rem limitations under the License. | |
16 | + | |
17 | +if "%OS%" == "Windows_NT" setlocal | |
18 | +rem --------------------------------------------------------------------------- | |
19 | +rem NT Service Install/Uninstall script | |
20 | +rem | |
21 | +rem Options | |
22 | +rem install Install the service using Tomcat7 as service name. | |
23 | +rem Service is installed using default settings. | |
24 | +rem remove Remove the service from the System. | |
25 | +rem | |
26 | +rem name (optional) If the second argument is present it is considered | |
27 | +rem to be new service name | |
28 | +rem | |
29 | +rem $Id: service.bat 1000718 2010-09-24 06:00:00Z mturk $ | |
30 | +rem --------------------------------------------------------------------------- | |
31 | + | |
32 | +set "SELF=%~dp0%service.bat" | |
33 | +rem Guess CATALINA_HOME if not defined | |
34 | +set "CURRENT_DIR=%cd%" | |
35 | +if not "%CATALINA_HOME%" == "" goto gotHome | |
36 | +set "CATALINA_HOME=%cd%" | |
37 | +if exist "%CATALINA_HOME%\bin\tomcat7.exe" goto okHome | |
38 | +rem CD to the upper dir | |
39 | +cd .. | |
40 | +set "CATALINA_HOME=%cd%" | |
41 | +:gotHome | |
42 | +if exist "%CATALINA_HOME%\bin\tomcat7.exe" goto okHome | |
43 | +echo The tomcat.exe was not found... | |
44 | +echo The CATALINA_HOME environment variable is not defined correctly. | |
45 | +echo This environment variable is needed to run this program | |
46 | +goto end | |
47 | +:okHome | |
48 | +rem Make sure prerequisite environment variables are set | |
49 | +if not "%JAVA_HOME%" == "" goto gotJdkHome | |
50 | +if not "%JRE_HOME%" == "" goto gotJreHome | |
51 | +echo Neither the JAVA_HOME nor the JRE_HOME environment variable is defined | |
52 | +echo Service will try to guess them from the registry. | |
53 | +goto okJavaHome | |
54 | +:gotJreHome | |
55 | +if not exist "%JRE_HOME%\bin\java.exe" goto noJavaHome | |
56 | +if not exist "%JRE_HOME%\bin\javaw.exe" goto noJavaHome | |
57 | +goto okJavaHome | |
58 | +:gotJdkHome | |
59 | +if not exist "%JAVA_HOME%\jre\bin\java.exe" goto noJavaHome | |
60 | +if not exist "%JAVA_HOME%\jre\bin\javaw.exe" goto noJavaHome | |
61 | +if not exist "%JAVA_HOME%\bin\javac.exe" goto noJavaHome | |
62 | +if not "%JRE_HOME%" == "" goto okJavaHome | |
63 | +set "JRE_HOME=%JAVA_HOME%\jre" | |
64 | +goto okJavaHome | |
65 | +:noJavaHome | |
66 | +echo The JAVA_HOME environment variable is not defined correctly | |
67 | +echo This environment variable is needed to run this program | |
68 | +echo NB: JAVA_HOME should point to a JDK not a JRE | |
69 | +goto end | |
70 | +:okJavaHome | |
71 | +if not "%CATALINA_BASE%" == "" goto gotBase | |
72 | +set "CATALINA_BASE=%CATALINA_HOME%" | |
73 | +:gotBase | |
74 | + | |
75 | +set "EXECUTABLE=%CATALINA_HOME%\bin\tomcat7.exe" | |
76 | + | |
77 | +rem Set default Service name | |
78 | +set SERVICE_NAME=Tomcat7 | |
79 | +set PR_DISPLAYNAME=Apache Tomcat 7 | |
80 | + | |
81 | +if "x%1x" == "xx" goto displayUsage | |
82 | +set SERVICE_CMD=%1 | |
83 | +shift | |
84 | +if "x%1x" == "xx" goto checkServiceCmd | |
85 | +:checkUser | |
86 | +if "x%1x" == "x/userx" goto runAsUser | |
87 | +if "x%1x" == "x--userx" goto runAsUser | |
88 | +set SERVICE_NAME=%1 | |
89 | +set PR_DISPLAYNAME=Apache Tomcat %1 | |
90 | +shift | |
91 | +if "x%1x" == "xx" goto checkServiceCmd | |
92 | +goto checkUser | |
93 | +:runAsUser | |
94 | +shift | |
95 | +if "x%1x" == "xx" goto displayUsage | |
96 | +set SERVICE_USER=%1 | |
97 | +shift | |
98 | +runas /env /savecred /user:%SERVICE_USER% "%COMSPEC% /K \"%SELF%\" %SERVICE_CMD% %SERVICE_NAME%" | |
99 | +goto end | |
100 | +:checkServiceCmd | |
101 | +if /i %SERVICE_CMD% == install goto doInstall | |
102 | +if /i %SERVICE_CMD% == remove goto doRemove | |
103 | +if /i %SERVICE_CMD% == uninstall goto doRemove | |
104 | +echo Unknown parameter "%1" | |
105 | +:displayUsage | |
106 | +echo. | |
107 | +echo Usage: service.bat install/remove [service_name] [/user username] | |
108 | +goto end | |
109 | + | |
110 | +:doRemove | |
111 | +rem Remove the service | |
112 | +"%EXECUTABLE%" //DS//%SERVICE_NAME% | |
113 | +if not errorlevel 1 goto removed | |
114 | +echo Failed removing '%SERVICE_NAME%' service | |
115 | +goto end | |
116 | +:removed | |
117 | +echo The service '%SERVICE_NAME%' has been removed | |
118 | +goto end | |
119 | + | |
120 | +:doInstall | |
121 | +rem Install the service | |
122 | +echo Installing the service '%SERVICE_NAME%' ... | |
123 | +echo Using CATALINA_HOME: "%CATALINA_HOME%" | |
124 | +echo Using CATALINA_BASE: "%CATALINA_BASE%" | |
125 | +echo Using JAVA_HOME: "%JAVA_HOME%" | |
126 | +echo Using JRE_HOME: "%JRE_HOME%" | |
127 | + | |
128 | +rem Use the environment variables as an example | |
129 | +rem Each command line option is prefixed with PR_ | |
130 | + | |
131 | +set PR_DESCRIPTION=Apache Tomcat 7.0.27 Server - http://tomcat.apache.org/ | |
132 | +set "PR_INSTALL=%EXECUTABLE%" | |
133 | +set "PR_LOGPATH=%CATALINA_BASE%\logs" | |
134 | +set "PR_CLASSPATH=%CATALINA_HOME%\bin\bootstrap.jar;%CATALINA_BASE%\bin\tomcat-juli.jar;%CATALINA_HOME%\bin\tomcat-juli.jar" | |
135 | +rem Set the server jvm from JAVA_HOME | |
136 | +set "PR_JVM=%JRE_HOME%\bin\server\jvm.dll" | |
137 | +if exist "%PR_JVM%" goto foundJvm | |
138 | +rem Set the client jvm from JAVA_HOME | |
139 | +set "PR_JVM=%JRE_HOME%\bin\client\jvm.dll" | |
140 | +if exist "%PR_JVM%" goto foundJvm | |
141 | +set PR_JVM=auto | |
142 | +:foundJvm | |
143 | +echo Using JVM: "%PR_JVM%" | |
144 | +"%EXECUTABLE%" //IS//%SERVICE_NAME% --StartClass org.apache.catalina.startup.Bootstrap --StopClass org.apache.catalina.startup.Bootstrap --StartParams start --StopParams stop | |
145 | +if not errorlevel 1 goto installed | |
146 | +echo Failed installing '%SERVICE_NAME%' service | |
147 | +goto end | |
148 | +:installed | |
149 | +rem Clear the environment variables. They are not needed any more. | |
150 | +set PR_DISPLAYNAME= | |
151 | +set PR_DESCRIPTION= | |
152 | +set PR_INSTALL= | |
153 | +set PR_LOGPATH= | |
154 | +set PR_CLASSPATH= | |
155 | +set PR_JVM= | |
156 | +rem Set extra parameters | |
157 | +"%EXECUTABLE%" //US//%SERVICE_NAME% --JvmOptions "-Dcatalina.base=%CATALINA_BASE%;-Dcatalina.home=%CATALINA_HOME%;-Djava.endorsed.dirs=%CATALINA_HOME%\endorsed" --StartMode jvm --StopMode jvm | |
158 | +rem More extra parameters | |
159 | +set "PR_LOGPATH=%CATALINA_BASE%\logs" | |
160 | +set PR_STDOUTPUT=auto | |
161 | +set PR_STDERROR=auto | |
162 | +"%EXECUTABLE%" //US//%SERVICE_NAME% ++JvmOptions "-Djava.io.tmpdir=%CATALINA_BASE%\temp;-Djava.util.logging.manager=org.apache.juli.ClassLoaderLogManager;-Djava.util.logging.config.file=%CATALINA_BASE%\conf\logging.properties;-Dsolr.solr.home=%CATALINA_HOME%\solr;-Dsolr.data.dir=%CATALINA_HOME%\solr\core1\data;-Dfess.log.file=%CATALINA_HOME%\webapps\fess\WEB-INF\logs\fess.out;-Djava.awt.headless=true;-XX:+UseGCOverheadLimit;-XX:+UseConcMarkSweepGC;-XX:+CMSIncrementalMode;-XX:+UseTLAB;-XX:MaxPermSize=128m" --JvmMs 128 --JvmMx 512 | |
163 | +echo The service '%SERVICE_NAME%' has been installed. | |
164 | + | |
165 | +:end | |
166 | +cd "%CURRENT_DIR%" |
@@ -0,0 +1,26 @@ | ||
1 | +#!/bin/sh | |
2 | + | |
3 | +CONTEXT_NAME=fess | |
4 | + | |
5 | +cd `dirname $0` | |
6 | +cd .. | |
7 | +TOMCAT_HOME=`pwd` | |
8 | +cd webapps/$CONTEXT_NAME | |
9 | +BASE_DIR=`pwd` | |
10 | + | |
11 | +CP_PATH=WEB-INF/cmd/resources:WEB-INF/classes | |
12 | +# added WEB-INF/lib/*.jar | |
13 | +for cp_path in `ls WEB-INF/lib/*.jar` ; do | |
14 | + CP_PATH="$CP_PATH:$cp_path" | |
15 | +done | |
16 | +# added WEB-INF/cmd/lib/*.jar | |
17 | +for cp_path in `ls WEB-INF/cmd/lib/*.jar` ; do | |
18 | + CP_PATH="$CP_PATH:$cp_path" | |
19 | +done | |
20 | + | |
21 | +SOLR_OPTS="$SOLR_OPTS -Dsolr.solr.home=$TOMCAT_HOME/solr -Dsolr.data.dir=$TOMCAT_HOME/solr/core1/data -Dfess.log.file=$BASE_DIR/WEB-INF/logs/fess_crawler.out" | |
22 | +FESS_OPTS="$FESS_OPTS -Dfess.crawler.process=true -Djava.awt.headless=true -server -XX:+UseGCOverheadLimit -XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode -XX:+UseTLAB -Xmx512m -XX:MaxPermSize=128m" | |
23 | + | |
24 | +java -cp $CP_PATH $SOLR_OPTS $FESS_OPTS jp.sf.fess.exec.Crawler $@ | |
25 | +# args: --sessionId 20100518061741 | |
26 | + |
@@ -0,0 +1,14 @@ | ||
1 | +#!/bin/bash | |
2 | + | |
3 | +SCREEN_SHOT_CMD="wkhtmltoimage-amd64 --use-xserver " | |
4 | +CONVERT_CMD=/usr/bin/convert | |
5 | + | |
6 | +SIZE=300x | |
7 | +URL=$1 | |
8 | +OUTPUT_FILE=$2 | |
9 | +TMP_FILE=/tmp/$$_`basename $OUTPUT_FILE` | |
10 | + | |
11 | +$SCREEN_SHOT_CMD $URL $TMP_FILE | |
12 | +$CONVERT_CMD -resize $SIZE $TMP_FILE $OUTPUT_FILE | |
13 | +rm $TMP_FILE | |
14 | + |
@@ -0,0 +1,5 @@ | ||
1 | +set JAVA_OPTS=%JAVA_OPTS% -Dsolr.solr.home="%CATALINA_HOME%\solr" -Dfess.solr.data.dir="%CATALINA_HOME%\solr\core1\data" -Dfess.log.file="%CATALINA_HOME%\webapps\fess\WEB-INF\logs\fess.out" -Djava.awt.headless=true -server -Xmx1g -XX:MaxPermSize=128m -XX:-UseGCOverheadLimit -XX:+UseConcMarkSweepGC -XX:CMSInitiatingOccupancyFraction=75 -XX:+CMSIncrementalMode -XX:+CMSIncrementalPacing -XX:CMSIncrementalDutyCycleMin=0 -XX:+UseParNewGC -XX:+UseStringCache -XX:+UseTLAB -XX:+DisableExplicitGC | |
2 | +rem -XX:+UseCompressedOops | |
3 | +rem -XX:+UseCompressedStrings if java6u20 or above | |
4 | +rem -XX:+OptimizeStringConcat if java6u21 or above | |
5 | + |
@@ -0,0 +1,11 @@ | ||
1 | +#!/bin/sh | |
2 | + | |
3 | +export JAVA_OPTS="$JAVA_OPTS -Dsolr.solr.home=$CATALINA_HOME/solr -Dfess.solr.data.dir=$CATALINA_HOME/solr/core1/data -Dfess.log.file=$CATALINA_HOME/webapps/fess/WEB-INF/logs/fess.out -Djava.awt.headless=true -server -Xmx1g -XX:MaxPermSize=128m -XX:-UseGCOverheadLimit -XX:+UseConcMarkSweepGC -XX:CMSInitiatingOccupancyFraction=75 -XX:+CMSIncrementalMode -XX:+CMSIncrementalPacing -XX:CMSIncrementalDutyCycleMin=0 -XX:+UseParNewGC -XX:+UseStringCache -XX:+UseTLAB -XX:+DisableExplicitGC" | |
4 | +# -XX:+UseCompressedOops for 64bit OS | |
5 | +# -XX:+UseCompressedStrings if java6u20 or above | |
6 | +# -XX:+OptimizeStringConcat if java6u21 or above | |
7 | + | |
8 | +# if Java heap size is over 32GB heap, remove -XX:+UseCompressedOops | |
9 | +# if your machine has 2 or more cpu, enabled -XX:+CMSParallelRemarkEnabled by -XX:+UseConcMarkSweepGC | |
10 | + | |
11 | +export CATALINA_OUT=/dev/null |
@@ -0,0 +1,58 @@ | ||
1 | +#!/bin/bash | |
2 | + | |
3 | +CONFIG_FILE=$1 | |
4 | +cd `dirname $0` | |
5 | +BASE_DIR=`pwd` | |
6 | + | |
7 | +PRODUCT_NAME=fess | |
8 | + | |
9 | +if [ x$CONFIG_FILE != "x" -a -f "$CONFIG_FILE" ] ; then | |
10 | + . $CONFIG_FILE | |
11 | +else | |
12 | + echo -n "Please enter a new DB password(search server)> " | |
13 | + read -s FESS_PASSWORD | |
14 | + echo | |
15 | + | |
16 | + echo -n "Please enter a new DB password(crawler)> " | |
17 | + read -s ROBOT_PASSWORD | |
18 | + echo | |
19 | + | |
20 | + echo -n "MySQL(root) Password: " | |
21 | + read -s MYSQL_PASSWORD | |
22 | + echo | |
23 | +fi | |
24 | + | |
25 | +CREATE_SQL_FILE=/tmp/fess_db.sql.$$ | |
26 | +echo "create database fess_db;" >> $CREATE_SQL_FILE | |
27 | +echo "GRANT ALL PRIVILEGES ON fess_db.* TO fess_user@localhost IDENTIFIED BY '$FESS_PASSWORD';" >> $CREATE_SQL_FILE | |
28 | +echo "create database fess_robot;" >> $CREATE_SQL_FILE | |
29 | +echo "GRANT ALL PRIVILEGES ON fess_robot.* TO s2robot@localhost IDENTIFIED BY '$ROBOT_PASSWORD';" >> $CREATE_SQL_FILE | |
30 | +echo "FLUSH PRIVILEGES;" >> $CREATE_SQL_FILE | |
31 | +echo -n "Creating DB on MySQL..." | |
32 | +mysql -u root --password=$MYSQL_PASSWORD < $CREATE_SQL_FILE | |
33 | +ret=`echo $?` | |
34 | +if [ $ret == 0 ] ; then | |
35 | + echo "Created." | |
36 | +fi | |
37 | +rm $CREATE_SQL_FILE | |
38 | + | |
39 | +echo -n "Creating tables..." | |
40 | +mysql -u fess_user --password=$FESS_PASSWORD fess_db < $BASE_DIR/fess.ddl | |
41 | +ret1=`echo $?` | |
42 | +mysql -u s2robot --password=$ROBOT_PASSWORD fess_robot < $BASE_DIR/robot.ddl | |
43 | +ret2=`echo $?` | |
44 | +if [ $ret1 == 0 -a $ret2 == 0 ] ; then | |
45 | + echo " Created." | |
46 | +fi | |
47 | + | |
48 | +cd ../.. | |
49 | + | |
50 | +TMP_FILE=/tmp/fess_dicon.$$ | |
51 | +sed -e "s/password\">[^<]*</password\">\"$FESS_PASSWORD\"</" \ | |
52 | + ./webapps/$PRODUCT_NAME/WEB-INF/classes/jdbc.dicon > $TMP_FILE | |
53 | +cp $TMP_FILE ./webapps/$PRODUCT_NAME/WEB-INF/classes/jdbc.dicon | |
54 | +sed -e "s/password\">[^<]*</password\">\"$ROBOT_PASSWORD\"</" \ | |
55 | + ./webapps/$PRODUCT_NAME/WEB-INF/classes/s2robot_jdbc.dicon > $TMP_FILE | |
56 | +cp $TMP_FILE ./webapps/$PRODUCT_NAME/WEB-INF/classes/s2robot_jdbc.dicon | |
57 | +rm $TMP_FILE | |
58 | + |
@@ -0,0 +1,26 @@ | ||
1 | +#!/bin/bash | |
2 | + | |
3 | +CONFIG_FILE=$1 | |
4 | +cd `dirname $0` | |
5 | +BASE_DIR=`pwd` | |
6 | + | |
7 | +if [ x$CONFIG_FILE != "x" -a -f "$CONFIG_FILE" ] ; then | |
8 | + . $CONFIG_FILE | |
9 | +else | |
10 | + echo -n "MySQL(root) Password: " | |
11 | + read -s MYSQL_PASSWORD | |
12 | + echo | |
13 | +fi | |
14 | + | |
15 | +DROP_SQL_FILE=/tmp/fess_db.sql.$$ | |
16 | +echo "drop database fess_db;" >> $DROP_SQL_FILE | |
17 | +echo "drop database fess_robot;" >> $DROP_SQL_FILE | |
18 | +echo "FLUSH PRIVILEGES;" >> $DROP_SQL_FILE | |
19 | +echo -n "Dropping DB on MySQL..." | |
20 | +mysql -u root --password=$MYSQL_PASSWORD < $DROP_SQL_FILE | |
21 | +ret=`echo $?` | |
22 | +if [ $ret == 0 ] ; then | |
23 | + echo "Dropped." | |
24 | +fi | |
25 | +rm $DROP_SQL_FILE | |
26 | + |
@@ -0,0 +1,145 @@ | ||
1 | +%define __jar_repack 0 | |
2 | + | |
3 | +%global dbtype __DBTYPE__ | |
4 | +%global fess_version __FESS_VERSION__ | |
5 | +%global fess_pkg_version __FESS_PKG_VERSION__ | |
6 | +%global fess_release __FESS_REVISION__ | |
7 | +%global product_root __PRODUCT_ROOT__ | |
8 | +%global product_name __PRODUCT_NAME__ | |
9 | +%global user_name __USER_NAME__ | |
10 | + | |
11 | +%if %{dbtype} | |
12 | +Name: %{product_name}-server-%{dbtype} | |
13 | +%else | |
14 | +Name: %{product_name}-server-h2 | |
15 | +%endif | |
16 | +Version: %{fess_version} | |
17 | +Release: %{fess_release} | |
18 | +Epoch: 1 | |
19 | +Summary: Fess Server | |
20 | +Group: Applications/Multimedia | |
21 | +BuildArch: noarch | |
22 | +License: Apache License, Version 2.0 | |
23 | +URL: http://fess.sourceforge.jp/ | |
24 | +BuildRoot: %{_tmpdir}/%{name}-%{version}-%{release} | |
25 | + | |
26 | +%if %{dbtype} | |
27 | +Source0: fess-server-%{dbtype}-%{fess_pkg_version}.zip | |
28 | +%else | |
29 | +Source0: fess-server-%{fess_pkg_version}.zip | |
30 | +%endif | |
31 | + | |
32 | +BuildRequires: unzip | |
33 | + | |
34 | +Requires: jdk >= 1.6.0 | |
35 | +Requires(pre): user(%{user_name}) | |
36 | +Requires(pre): group(%{user_name}) | |
37 | +Requires(pre): shadow-utils | |
38 | +Provides: user(%{user_name}) | |
39 | +Provides: group(%{user_name}) | |
40 | + | |
41 | +Requires(post): chkconfig | |
42 | +Requires(preun): chkconfig | |
43 | +Requires(preun): initscripts | |
44 | +Requires(postun): initscripts | |
45 | + | |
46 | +%if "%{dbtype}" == "mysql" | |
47 | +Requires: mysql-server | |
48 | +%endif | |
49 | + | |
50 | +%description | |
51 | +Fess is Java based full text search server provided as OSS product. | |
52 | + | |
53 | +%prep | |
54 | +rm -rf "${RPM_BUILD_ROOT}" | |
55 | +%setup -q -n fess-server-%{fess_pkg_version} | |
56 | + | |
57 | +%build | |
58 | + | |
59 | +%install | |
60 | +mkdir -p $RPM_BUILD_ROOT%{product_root} | |
61 | +cp -r $RPM_BUILD_DIR/fess-server-%{fess_pkg_version} $RPM_BUILD_ROOT%{product_root}/%{product_name} | |
62 | +%if "%{product_name}" != "fess" | |
63 | + mv $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/fess $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/%{product_name} | |
64 | +%endif | |
65 | +rm -f $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/*.war | |
66 | + | |
67 | +TMP_FILE=/tmp/build.$$ | |
68 | + | |
69 | +sed -e "s/\(echo \"Using.*\)/\1 > \/dev\/null/" \ | |
70 | + $RPM_BUILD_ROOT%{product_root}/%{product_name}/bin/catalina.sh > $TMP_FILE | |
71 | +cp $TMP_FILE $RPM_BUILD_ROOT%{product_root}/%{product_name}/bin/catalina.sh | |
72 | + | |
73 | +sed -e "s/\/fess\//\/%{product_name}\//g" \ | |
74 | + $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/%{product_name}/WEB-INF/web.xml > $TMP_FILE | |
75 | +cp $TMP_FILE $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/%{product_name}/WEB-INF/web.xml | |
76 | + | |
77 | +sed -e "s/\/fess\//\/%{product_name}\//g" \ | |
78 | + $RPM_BUILD_ROOT%{product_root}/%{product_name}/bin/setenv.sh > $TMP_FILE | |
79 | +cp $TMP_FILE $RPM_BUILD_ROOT%{product_root}/%{product_name}/bin/setenv.sh | |
80 | + | |
81 | +sed -e "s/PRODUCT_NAME=.*/PRODUCT_NAME=%{product_name}/g" \ | |
82 | + $RPM_BUILD_ROOT%{product_root}/%{product_name}/extension/mysql/install.sh > $TMP_FILE | |
83 | +cp $TMP_FILE $RPM_BUILD_ROOT%{product_root}/%{product_name}/extension/mysql/install.sh | |
84 | + | |
85 | +rm $TMP_FILE | |
86 | + | |
87 | +mkdir -p $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/%{product_name}/WEB-INF/conf | |
88 | +touch $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/%{product_name}/WEB-INF/conf/solrserver.properties | |
89 | +touch $RPM_BUILD_ROOT%{product_root}/%{product_name}/webapps/%{product_name}/WEB-INF/conf/crawler.properties | |
90 | + | |
91 | +chmod +x $RPM_BUILD_ROOT%{product_root}/%{product_name}/bin/*.sh | |
92 | +chmod +x $RPM_BUILD_ROOT%{product_root}/%{product_name}/extension/mysql/*.sh | |
93 | + | |
94 | +mkdir -p $RPM_BUILD_ROOT/etc/init.d | |
95 | +sed -e "s/FESS_PROG=fess/FESS_PROG=%{product_name}/" \ | |
96 | + -e "s/FESS_USER=\"fess\"/FESS_USER=\"%{user_name}\"/" \ | |
97 | + $RPM_BUILD_DIR/fess-server-%{fess_pkg_version}/extension/scripts/fess > $RPM_BUILD_ROOT/etc/init.d/%{product_name} | |
98 | + | |
99 | +%clean | |
100 | +rm -rf "${RPM_BUILD_ROOT}" | |
101 | + | |
102 | +%pre | |
103 | +getent group %{user_name} >/dev/null || groupadd -r %{user_name} | |
104 | +getent passwd %{user_name} >/dev/null || \ | |
105 | +useradd -d %{product_root}/%{product_name} -g %{user_name} -M -r %{user_name} | |
106 | +exit 0 | |
107 | + | |
108 | +%post | |
109 | +/sbin/chkconfig --add %{product_name} | |
110 | + | |
111 | +%if "%{dbtype}" == "mysql" | |
112 | + echo "Please run the following script to configure database." | |
113 | + echo " %{product_root}/%{product_name}/extension/mysql/install.sh" | |
114 | +%endif | |
115 | + | |
116 | +%preun | |
117 | +# only delete user on removal, not upgrade | |
118 | +if [ $1 = 0 ]; then | |
119 | + /sbin/service %{product_name} stop > /dev/null 2>&1 | |
120 | + /sbin/chkconfig --del %{product_name} | |
121 | + pkill -U %{user_name} | |
122 | + userdel %{user_name} | |
123 | + rm -rf %{product_root}/%{product_name}/logs/* | |
124 | + rm -rf %{product_root}/%{product_name}/conf/Catalina | |
125 | + rm -rf %{product_root}/%{product_name}/webapps/%{product_name}/WEB-INF/logs/* | |
126 | +else | |
127 | + /sbin/service %{product_name} stop > /dev/null 2>&1 | |
128 | +fi | |
129 | +rm -rf %{product_root}/%{product_name}/webapps/%{product_name}/jar | |
130 | +rm -rf %{product_root}/%{product_name}/work/* | |
131 | + | |
132 | +%postun | |
133 | +if [ $1 -ge 1 ]; then | |
134 | + /sbin/service %{product_name} start > /dev/null 2>&1 | |
135 | +fi | |
136 | + | |
137 | +%files | |
138 | +%defattr(-,%{user_name},%{user_name},0755) | |
139 | +%{product_root}/%{product_name} | |
140 | + | |
141 | +%attr(755,root,root) /etc/init.d/%{product_name} | |
142 | + | |
143 | +%changelog | |
144 | + | |
145 | + |
@@ -0,0 +1,454 @@ | ||
1 | +<?xml version="1.0" encoding="UTF-8"?> | |
2 | + | |
3 | +<project name="fess" default="all" basedir="."> | |
4 | + | |
5 | + <property file="override.properties" /> | |
6 | + <property file="build.properties" /> | |
7 | + | |
8 | + <taskdef resource="net/sf/antcontrib/antlib.xml"> | |
9 | + <classpath> | |
10 | + <pathelement location="${basedir}/lib/ant-contrib-1.0b2.jar" /> | |
11 | + </classpath> | |
12 | + </taskdef> | |
13 | + | |
14 | + <condition property="profile.mysql"> | |
15 | + <equals arg1="${build.fess.profile}" arg2="mysql"/> | |
16 | + </condition> | |
17 | + | |
18 | + <condition property="profile.oracle"> | |
19 | + <equals arg1="${build.fess.profile}" arg2="oracle"/> | |
20 | + </condition> | |
21 | + | |
22 | + <target name="setup.shell" | |
23 | + description="--> Setup shell parameters"> | |
24 | + <condition property="shell.cmd" value="sh"> | |
25 | + <os family="unix" /> | |
26 | + </condition> | |
27 | + <condition property="shell.options" value="-c"> | |
28 | + <os family="unix" /> | |
29 | + </condition> | |
30 | + <condition property="shell.cmd" value="cmd.exe"> | |
31 | + <os family="windows" /> | |
32 | + </condition> | |
33 | + <condition property="shell.options" value="/c"> | |
34 | + <os family="windows" /> | |
35 | + </condition> | |
36 | + </target> | |
37 | + | |
38 | + <!-- =================================================================== --> | |
39 | + <!-- DISPLAY LOGO --> | |
40 | + <!-- =================================================================== --> | |
41 | + | |
42 | + <target name="display.logo" | |
43 | + description="--> Display Logo"> | |
44 | + <echo> | |
45 | +-------------------------------- | |
46 | + ~ Full tExt Search System ~ | |
47 | +-------------------------------- | |
48 | + </echo> | |
49 | + </target> | |
50 | + | |
51 | + <!-- =================================================================== --> | |
52 | + <!-- CLEAN --> | |
53 | + <!-- =================================================================== --> | |
54 | + | |
55 | + <target name="clean" depends="display.logo" | |
56 | + description="--> Clean All"> | |
57 | + <delete dir="${build.dir}"/> | |
58 | + </target> | |
59 | + | |
60 | + <!-- =================================================================== --> | |
61 | + <!-- BUILD --> | |
62 | + <!-- =================================================================== --> | |
63 | + | |
64 | + <target name="build" depends="display.logo" | |
65 | + description="--> Build Fess"> | |
66 | + <antcall target="build.fess"/> | |
67 | + <antcall target="build.lucene-gosen"/> | |
68 | + <antcall target="setup.tomcat"/> | |
69 | + <antcall target="setup.solr"/> | |
70 | + <antcall target="setup.server"/> | |
71 | + <antcall target="package"/> | |
72 | + </target> | |
73 | + | |
74 | + <!-- =================================================================== --> | |
75 | + <!-- BUILD FESS APPLICATION --> | |
76 | + <!-- =================================================================== --> | |
77 | + | |
78 | + <target name="build.fess" | |
79 | + description="--> Build Fess Application"> | |
80 | + <mkdir dir="${build.dir}"/> | |
81 | + | |
82 | + <antcall target="checkout.fess"/> | |
83 | + <antcall target="build.fess.app"/> | |
84 | + | |
85 | + <antcall target="checkout.fess.solr.plugin"/> | |
86 | + <antcall target="build.fess.solr.plugin"/> | |
87 | + </target> | |
88 | + | |
89 | + <target name="checkout.fess" | |
90 | + description="--> Checkout Fess Application"> | |
91 | + <exec | |
92 | + executable="svn" | |
93 | + dir="${build.dir}" | |
94 | + > | |
95 | + <arg line="checkout ${fess.app.svn.path} --revision ${fess.app.svn.revision} ${fess.app.svn.module}"/> | |
96 | + </exec> | |
97 | + </target> | |
98 | + | |
99 | + <target name="build.fess.app.mysql" if="profile.mysql" | |
100 | + description="--> Build Fess app for mysql"> | |
101 | + <!-- build --> | |
102 | + <delete file="${build.fess.app.dir}/src/main/java/jp/sf/fess/helper/impl/H2DatabaseHelperImpl.java"/> | |
103 | + <delete file="${build.fess.app.dir}/src/main/java/jp/sf/fess/servlet/H2ConfigServlet.java"/> | |
104 | + | |
105 | + <exec | |
106 | + executable="${shell.cmd}" | |
107 | + failonerror="true" | |
108 | + dir="${build.fess.app.dir}" | |
109 | + > | |
110 | + <arg line="${shell.options} 'mvn dbflute:download'"/> | |
111 | + </exec> | |
112 | + <exec | |
113 | + executable="${shell.cmd}" | |
114 | + failonerror="true" | |
115 | + dir="${build.fess.app.dir}" | |
116 | + > | |
117 | + <arg line="${shell.options} 'mvn dbflute:jdbc -P ${build.fess.profile}'"/> | |
118 | + </exec> | |
119 | + <exec | |
120 | + executable="${shell.cmd}" | |
121 | + failonerror="true" | |
122 | + dir="${build.fess.app.dir}" | |
123 | + > | |
124 | + <arg line="${shell.options} 'mvn dbflute:generate -P ${build.fess.profile}'"/> | |
125 | + </exec> | |
126 | + <exec | |
127 | + executable="${shell.cmd}" | |
128 | + failonerror="true" | |
129 | + dir="${build.fess.app.dir}" | |
130 | + > | |
131 | + <arg line="${shell.options} 'mvn dbflute:sql2entity -P ${build.fess.profile}'"/> | |
132 | + </exec> | |
133 | + | |
134 | + <!-- workaround for mysql --> | |
135 | + <replace dir="${build.fess.app.dir}/src/main/java/jp/sf/fess/db" token="// MySQL: " value=""> | |
136 | + <include name="**/*.java" /> | |
137 | + </replace> | |
138 | + </target> | |
139 | + | |
140 | + <target name="build.fess.app.oracle" if="profile.oracle" | |
141 | + description="--> Build Fess app for oracle"> | |
142 | + <!-- build --> | |
143 | + <delete file="${build.fess.app.dir}/src/main/java/jp/sf/fess/helper/impl/H2DatabaseHelperImpl.java"/> | |
144 | + <delete file="${build.fess.app.dir}/src/main/java/jp/sf/fess/servlet/H2ConfigServlet.java"/> | |
145 | + | |
146 | + <exec | |
147 | + executable="${shell.cmd}" | |
148 | + failonerror="true" | |
149 | + dir="${build.fess.app.dir}" | |
150 | + > | |
151 | + <arg line="${shell.options} 'mvn dbflute:download'"/> | |
152 | + </exec> | |
153 | + <copy file="${driver.oracle.file}" | |
154 | + todir="${build.fess.mydbflute.dir}/lib" overwrite="true"/> | |
155 | + <exec | |
156 | + executable="${shell.cmd}" | |
157 | + failonerror="true" | |
158 | + dir="${build.fess.app.dir}" | |
159 | + > | |
160 | + <arg line="${shell.options} 'mvn dbflute:jdbc -P ${build.fess.profile}'"/> | |
161 | + </exec> | |
162 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
163 | + token="name="DATA_CONFIG_TO_B_TYPE"" | |
164 | + value="javaName="DataConfigToBrowserTypeMapping" name="DATA_CONFIG_TO_B_TYPE"" | |
165 | + /> | |
166 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
167 | + token="name="DATA_CONFIG_TO_LABEL_TYPE"" | |
168 | + value="javaName="DataConfigToLabelTypeMapping" name="DATA_CONFIG_TO_LABEL_TYPE"" | |
169 | + /> | |
170 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
171 | + token="name="DATA_CONFIG_TO_ROLE_TYPE"" | |
172 | + value="javaName="DataConfigToRoleTypeMapping" name="DATA_CONFIG_TO_ROLE_TYPE"" | |
173 | + /> | |
174 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
175 | + token="name="FILE_CONFIG_TO_B_TYPE"" | |
176 | + value="javaName="FileConfigToBrowserTypeMapping" name="FILE_CONFIG_TO_B_TYPE"" | |
177 | + /> | |
178 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
179 | + token="name="FILE_CONFIG_TO_LABEL_TYPE"" | |
180 | + value="javaName="FileConfigToLabelTypeMapping" name="FILE_CONFIG_TO_LABEL_TYPE"" | |
181 | + /> | |
182 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
183 | + token="name="FILE_CONFIG_TO_ROLE_TYPE"" | |
184 | + value="javaName="FileConfigToRoleTypeMapping" name="FILE_CONFIG_TO_ROLE_TYPE"" | |
185 | + /> | |
186 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
187 | + token="name="LABEL_TYPE_TO_ROLE_TYPE"" | |
188 | + value="javaName="LabelTypeToRoleTypeMapping" name="LABEL_TYPE_TO_ROLE_TYPE"" | |
189 | + /> | |
190 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
191 | + token="name="WEB_CONFIG_TO_B_TYPE"" | |
192 | + value="javaName="WebConfigToBrowserTypeMapping" name="WEB_CONFIG_TO_B_TYPE"" | |
193 | + /> | |
194 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
195 | + token="name="WEB_CONFIG_TO_LABEL_TYPE"" | |
196 | + value="javaName="WebConfigToLabelTypeMapping" name="WEB_CONFIG_TO_LABEL_TYPE"" | |
197 | + /> | |
198 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
199 | + token="name="WEB_CONFIG_TO_ROLE_TYPE"" | |
200 | + value="javaName="WebConfigToRoleTypeMapping" name="WEB_CONFIG_TO_ROLE_TYPE"" | |
201 | + /> | |
202 | + <replace file="${build.fess.app.dir}/dbflute_oracle/schema/project-schema-fess.xml" | |
203 | + token="dbType="FLOAT" javaType="java.math.BigDecimal"" | |
204 | + value="dbType="FLOAT" javaType="Float"" | |
205 | + /> | |
206 | + <exec | |
207 | + executable="${shell.cmd}" | |
208 | + failonerror="true" | |
209 | + dir="${build.fess.app.dir}" | |
210 | + > | |
211 | + <arg line="${shell.options} 'mvn dbflute:generate -P ${build.fess.profile}'"/> | |
212 | + </exec> | |
213 | + <exec | |
214 | + executable="${shell.cmd}" | |
215 | + failonerror="true" | |
216 | + dir="${build.fess.app.dir}" | |
217 | + > | |
218 | + <arg line="${shell.options} 'mvn dbflute:sql2entity -P ${build.fess.profile}'"/> | |
219 | + </exec> | |
220 | + | |
221 | + </target> | |
222 | + | |
223 | + <target name="build.fess.app" depends="setup.shell,build.fess.app.mysql,build.fess.app.oracle" | |
224 | + description="--> Build Fess app by Maven 2"> | |
225 | + <!-- build --> | |
226 | + <exec | |
227 | + executable="${shell.cmd}" | |
228 | + failonerror="true" | |
229 | + dir="${build.fess.app.dir}" | |
230 | + > | |
231 | + <arg line="${shell.options} 'mvn install -P ${build.fess.profile} -DskipTests=true'"/> | |
232 | + </exec> | |
233 | + </target> | |
234 | + | |
235 | + <!-- =================================================================== --> | |
236 | + <!-- BUILD lucene-gosen --> | |
237 | + <!-- =================================================================== --> | |
238 | + | |
239 | + <target name="build.lucene-gosen" | |
240 | + description="--> Build lucene-gosen"> | |
241 | + <mkdir dir="${build.dir}"/> | |
242 | + | |
243 | + <antcall target="checkout.lucene-gosen"/> | |
244 | + <antcall target="build.lucene-gosen.app"/> | |
245 | + </target> | |
246 | + | |
247 | + <target name="checkout.lucene-gosen" | |
248 | + description="--> Checkout lucene-gosen"> | |
249 | + <exec | |
250 | + executable="svn" | |
251 | + dir="${build.dir}" | |
252 | + > | |
253 | + <arg line="checkout ${lucene-gosen.svn.path} --revision ${lucene-gosen.svn.revision} ${lucene-gosen.svn.module}"/> | |
254 | + </exec> | |
255 | + </target> | |
256 | + | |
257 | + <target name="build.lucene-gosen.app" depends="setup.shell" | |
258 | + description="--> Build lucene-gosen by Ant"> | |
259 | + <!-- build --> | |
260 | + <exec | |
261 | + executable="${shell.cmd}" | |
262 | + failonerror="true" | |
263 | + dir="${build.lucene-gosen.dir}" | |
264 | + > | |
265 | + <arg line="${shell.options} 'ant nodic-jar'"/> | |
266 | + </exec> | |
267 | + <!-- build dic --> | |
268 | + <exec | |
269 | + executable="${shell.cmd}" | |
270 | + failonerror="true" | |
271 | + dir="${build.lucene-gosen.dir}" | |
272 | + > | |
273 | + <arg line="${shell.options} 'ant build-dic-naist-chasen'"/> | |
274 | + </exec> | |
275 | + </target> | |
276 | + | |
277 | + <!-- =================================================================== --> | |
278 | + <!-- BUILD FESS SOLR PLUGIN --> | |
279 | + <!-- =================================================================== --> | |
280 | + | |
281 | + <target name="checkout.fess.solr.plugin" | |
282 | + description="--> Checkout Fess Solr Plugin"> | |
283 | + <exec | |
284 | + executable="svn" | |
285 | + dir="${build.dir}" | |
286 | + > | |
287 | + <arg line="checkout ${fess.solr.plugin.svn.path} --revision ${fess.solr.plugin.svn.revision} ${fess.solr.plugin.svn.module}"/> | |
288 | + </exec> | |
289 | + </target> | |
290 | + | |
291 | + <target name="build.fess.solr.plugin" depends="setup.shell" | |
292 | + description="--> Build Fess Solr Plugin by Maven 2"> | |
293 | + <!-- build --> | |
294 | + <exec | |
295 | + executable="${shell.cmd}" | |
296 | + failonerror="true" | |
297 | + dir="${build.fess.solr.plugin.dir}" | |
298 | + > | |
299 | + <arg line="${shell.options} 'mvn install dependency:copy-dependencies -DskipTests=true'"/> | |
300 | + </exec> | |
301 | + </target> | |
302 | + | |
303 | + <!-- =================================================================== --> | |
304 | + <!-- SETUP TOMCAT --> | |
305 | + <!-- =================================================================== --> | |
306 | + | |
307 | + <target name="setup.tomcat" | |
308 | + description="--> Setup Tomcat"> | |
309 | + <mkdir dir="${build.dir}"/> | |
310 | + <get usetimestamp="true" dest="${build.tomcat.file}" src="${tomcat.url}"/> | |
311 | + <unzip src="${build.tomcat.file}" dest="${build.dir}"/> | |
312 | + | |
313 | + <!-- Clean webapps --> | |
314 | + <delete dir="${build.tomcat.dir}/webapps/ROOT"/> | |
315 | + <delete dir="${build.tomcat.dir}/webapps/docs"/> | |
316 | + <delete dir="${build.tomcat.dir}/webapps/examples"/> | |
317 | + <delete dir="${build.tomcat.dir}/webapps/host-manager"/> | |
318 | + | |
319 | + <move todir="${package.build.dir}"> | |
320 | + <fileset dir="${build.tomcat.dir}"/> | |
321 | + </move> | |
322 | + </target> | |
323 | + | |
324 | + <!-- =================================================================== --> | |
325 | + <!-- SETUP SOLR --> | |
326 | + <!-- =================================================================== --> | |
327 | + | |
328 | + <target name="setup.solr" | |
329 | + description="--> Setup Solr"> | |
330 | + <mkdir dir="${build.dir}"/> | |
331 | + <get usetimestamp="true" dest="${build.solr.file}" src="${solr.url}"/> | |
332 | + <unzip src="${build.solr.file}" dest="${build.dir}"/> | |
333 | + </target> | |
334 | + | |
335 | + <!-- =================================================================== --> | |
336 | + <!-- SETUP SERVER --> | |
337 | + <!-- =================================================================== --> | |
338 | + | |
339 | + <target name="setup.server" | |
340 | + description="--> Setup Server"> | |
341 | + <mkdir dir="${build.dir}"/> | |
342 | + <!-- copy fess.war --> | |
343 | + <unwar src="${build.fess.app.war.file}" | |
344 | + dest="${build.fess.app.deployed.war}"/> | |
345 | + <!-- copy solr.war --> | |
346 | + <unwar src="${build.solr.war.file}" | |
347 | + dest="${build.solr.deployed.war}"/> | |
348 | + <!-- copy solr example --> | |
349 | + <copy todir="${package.build.dir}/solr/core1" overwrite="true"> | |
350 | + <fileset dir="${build.solr.example.dir}/solr"> | |
351 | + <include name="**"/> | |
352 | + </fileset> | |
353 | + </copy> | |
354 | + <!-- Solr4 --> | |
355 | + <copy todir="${package.build.dir}/solr/core1/conf" overwrite="true"> | |
356 | + <fileset dir="${build.solr.example.dir}/solr/collection1/conf"> | |
357 | + <include name="**"/> | |
358 | + </fileset> | |
359 | + </copy> | |
360 | + <delete dir="${package.build.dir}/solr/core1/collection1"/> | |
361 | + <delete file="${package.build.dir}/solr/core1/README.txt"/> | |
362 | + <delete file="${package.build.dir}/solr/core1/solr.xml"/> | |
363 | + <move file="${package.build.dir}/solr/core1/zoo.cfg" | |
364 | + tofile="${package.build.dir}/solr/zoo.cfg"/> | |
365 | + <!-- Logging --> | |
366 | + <copy todir="${package.build.dir}/webapps/solr/WEB-INF/lib" overwrite="true"> | |
367 | + <fileset dir="${build.solr.example.dir}/lib/ext"> | |
368 | + <include name="*.jar"/> | |
369 | + </fileset> | |
370 | + </copy> | |
371 | + <mkdir dir="${package.build.dir}/webapps/solr/WEB-INF/classes"/> | |
372 | + <copy file="${build.solr.example.dir}/resources/log4j.properties" | |
373 | + todir="${package.build.dir}/webapps/solr/WEB-INF/classes" overwrite="true"/> | |
374 | + <!-- copy solr dist --> | |
375 | + <mkdir dir="${package.build.dir}/solr/dist"/> | |
376 | + <copy todir="${package.build.dir}/solr/dist" overwrite="true"> | |
377 | + <fileset dir="${build.solr.dir}/dist"> | |
378 | + <include name="**"/> | |
379 | + </fileset> | |
380 | + </copy> | |
381 | + <!-- copy solr contrib --> | |
382 | + <mkdir dir="${package.build.dir}/solr/contrib"/> | |
383 | + <copy todir="${package.build.dir}/solr/contrib" overwrite="true"> | |
384 | + <fileset dir="${build.solr.dir}/contrib"> | |
385 | + <include name="**"/> | |
386 | + </fileset> | |
387 | + </copy> | |
388 | + <!-- copy solr plugin --> | |
389 | + <mkdir dir="${build.fess.solr.plugin.deployed.jar.dir}"/> | |
390 | + <copy file="${build.fess.solr.plugin.jar.file}" | |
391 | + tofile="${build.fess.solr.plugin.deployed.jar.file}" overwrite="true"/> | |
392 | + <copy file="${build.fess.suggest.plugin.jar.file}" | |
393 | + tofile="${build.fess.suggest.plugin.deployed.jar.file}" overwrite="true"/> | |
394 | + <copy todir="${package.build.dir}/solr/lib" overwrite="true"> | |
395 | + <fileset dir="${build.fess.solr.plugin.dir}/target/dependency"> | |
396 | + <include name="icu4j-*.jar"/> | |
397 | + </fileset> | |
398 | + </copy> | |
399 | + <copy file="${package.build.dir}/solr/core1/conf/lang/userdict_ja.txt" | |
400 | + tofile="${package.build.dir}/solr/core1/conf/lang/userdict_suggest_ja.txt" | |
401 | + overwrite="true"/> | |
402 | + <!-- copy lucene-gosen --> | |
403 | + <mkdir dir="${package.build.dir}/solr/lib"/> | |
404 | + <copy todir="${package.build.dir}/solr/lib" overwrite="true"> | |
405 | + <fileset dir="${build.lucene-gosen.dir}/dist"> | |
406 | + <include name="*.jar"/> | |
407 | + </fileset> | |
408 | + </copy> | |
409 | + <mkdir dir="${package.build.dir}/solr/core1/conf/dictionary/naist-chasen"/> | |
410 | + <copy todir="${package.build.dir}/solr/core1/conf/dictionary/naist-chasen" | |
411 | + overwrite="true"> | |
412 | + <fileset dir="${build.lucene-gosen.dir}/dictionary/naist-chasen"> | |
413 | + <include name="*.sen"/> | |
414 | + </fileset> | |
415 | + </copy> | |
416 | + <copy todir="${package.build.dir}/solr/core1/conf" overwrite="true"> | |
417 | + <fileset dir="${build.lucene-gosen.dir}/example"> | |
418 | + <include name="*_ja.txt"/> | |
419 | + </fileset> | |
420 | + </copy> | |
421 | + <!-- copy tomcat --> | |
422 | + <copy todir="${package.build.dir}" overwrite="true"> | |
423 | + <fileset dir="${src.tomcat.dir}"> | |
424 | + <include name="**"/> | |
425 | + </fileset> | |
426 | + </copy> | |
427 | + <concat destfile="${package.build.dir}/NOTICE" append="true" | |
428 | + encoding="UTF-8" outputencoding="UTF-8"> | |
429 | + <filelist dir="${basedir}" files="NOTICE"/> | |
430 | + </concat> | |
431 | + </target> | |
432 | + | |
433 | + <target name="package" | |
434 | + description="--> Package Fess"> | |
435 | + <!-- zip --> | |
436 | + <zip file="${package.build.file}"> | |
437 | + <zipfileset dir="${build.dir}"> | |
438 | + <include name="${package.build.name}/**"/> | |
439 | + </zipfileset> | |
440 | + </zip> | |
441 | + </target> | |
442 | + | |
443 | + <!-- =================================================================== --> | |
444 | + <!-- ALL BUILD --> | |
445 | + <!-- =================================================================== --> | |
446 | + | |
447 | + <target name="all" depends="display.logo" | |
448 | + description="--> Build all"> | |
449 | + <antcall target="clean"/> | |
450 | + <antcall target="build"/> | |
451 | + </target> | |
452 | + | |
453 | +</project> | |
454 | + |