iiif_print 1.0.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.coveralls.yml +2 -0
- data/.env +5 -0
- data/.fcrepo_wrapper +4 -0
- data/.github/release.yml +20 -0
- data/.github/workflows/branches.yml +24 -0
- data/.github/workflows/build-lint-test-action.yaml +33 -0
- data/.github/workflows/release_labels.yml +25 -0
- data/.gitignore +52 -0
- data/.rubocop.yml +177 -0
- data/.solr_wrapper +8 -0
- data/.travis.yml +49 -0
- data/CONTRIBUTING.md +181 -0
- data/Dockerfile +15 -0
- data/Gemfile +52 -0
- data/LICENSE +203 -0
- data/README.md +203 -0
- data/Rakefile +38 -0
- data/app/actors/iiif_print/actors/file_set_actor_decorator.rb +56 -0
- data/app/assets/config/iiif_print_manifest.js +2 -0
- data/app/assets/images/iiif_print/.keep +0 -0
- data/app/assets/javascripts/iiif_print/autocomplete_fix.js +33 -0
- data/app/assets/javascripts/iiif_print/ocr_search.js.erb +6 -0
- data/app/assets/javascripts/iiif_print.js +3 -0
- data/app/assets/stylesheets/iiif_print/_iiif_print.scss +4 -0
- data/app/assets/stylesheets/iiif_print/_issue_search.scss +13 -0
- data/app/assets/stylesheets/iiif_print/_issues_calendar.scss +18 -0
- data/app/assets/stylesheets/iiif_print/_newspapers_search.scss +38 -0
- data/app/assets/stylesheets/iiif_print/_search_results.scss +6 -0
- data/app/helpers/hyrax/iiif_helper.rb +22 -0
- data/app/helpers/iiif_print/application_helper.rb +5 -0
- data/app/helpers/iiif_print_helper.rb +64 -0
- data/app/indexers/concerns/iiif_print/child_indexer.rb +34 -0
- data/app/indexers/concerns/iiif_print/file_set_indexer.rb +29 -0
- data/app/mailers/iiif_print/application_mailer.rb +8 -0
- data/app/models/concerns/iiif_print/set_child_flag.rb +29 -0
- data/app/models/concerns/iiif_print/solr/document.rb +47 -0
- data/app/models/iiif_print/application_record.rb +6 -0
- data/app/models/iiif_print/derivative_attachment.rb +8 -0
- data/app/models/iiif_print/iiif_search_response_decorator.rb +17 -0
- data/app/models/iiif_print/ingest_file_relation.rb +14 -0
- data/app/models/iiif_print/pending_relationship.rb +7 -0
- data/app/presenters/iiif_print/iiif_manifest_presenter_behavior.rb +10 -0
- data/app/presenters/iiif_print/iiif_manifest_presenter_factory_behavior.rb +33 -0
- data/app/presenters/iiif_print/work_show_presenter_decorator.rb +29 -0
- data/app/renderers/hyrax/renderers/faceted_attribute_renderer_decorator.rb +18 -0
- data/app/search_builders/concerns/iiif_print/exclude_models.rb +17 -0
- data/app/search_builders/concerns/iiif_print/highlight_search_params.rb +14 -0
- data/app/services/iiif_print/manifest_builder_service_behavior.rb +97 -0
- data/app/services/iiif_print/pluggable_derivative_service.rb +120 -0
- data/app/views/catalog/_snippets_more.html.erb +16 -0
- data/app/views/hyrax/base/_representative_media.html.erb +9 -0
- data/app/views/hyrax/base/iiif_viewers/_universal_viewer.html.erb +8 -0
- data/app/views/hyrax/file_sets/_actions.html.erb +45 -0
- data/bin/rails +13 -0
- data/config/fcrepo_wrapper_test.yml +5 -0
- data/config/initializers/assets.rb +2 -0
- data/config/locales/iiif_print.de.yml +148 -0
- data/config/locales/iiif_print.en.yml +119 -0
- data/config/locales/iiif_print.es.yml +148 -0
- data/config/locales/iiif_print.fr.yml +149 -0
- data/config/locales/iiif_print.it.yml +142 -0
- data/config/locales/iiif_print.pt-BR.yml +148 -0
- data/config/locales/iiif_print.zh.yml +142 -0
- data/config/solr_wrapper_test.yml +9 -0
- data/config/test-fixture/solr-config/_rest_managed.json +3 -0
- data/config/test-fixture/solr-config/admin-extra.html +31 -0
- data/config/test-fixture/solr-config/elevate.xml +36 -0
- data/config/test-fixture/solr-config/mapping-ISOLatin1Accent.txt +246 -0
- data/config/test-fixture/solr-config/protwords.txt +21 -0
- data/config/test-fixture/solr-config/schema.xml +366 -0
- data/config/test-fixture/solr-config/scripts.conf +24 -0
- data/config/test-fixture/solr-config/solrconfig.xml +322 -0
- data/config/test-fixture/solr-config/spellings.txt +2 -0
- data/config/test-fixture/solr-config/stopwords.txt +58 -0
- data/config/test-fixture/solr-config/stopwords_en.txt +58 -0
- data/config/test-fixture/solr-config/synonyms.txt +31 -0
- data/config/test-fixture/solr-config/xslt/example.xsl +132 -0
- data/config/test-fixture/solr-config/xslt/example_atom.xsl +67 -0
- data/config/test-fixture/solr-config/xslt/example_rss.xsl +66 -0
- data/config/test-fixture/solr-config/xslt/luke.xsl +337 -0
- data/config/vendor/fits.xml +55 -0
- data/config/vendor/imagemagick-6-policy.xml +76 -0
- data/db/migrate/20181214181358_create_iiif_print_derivative_attachments.rb +12 -0
- data/db/migrate/20190107165909_create_iiif_print_ingest_file_relations.rb +11 -0
- data/db/migrate/20230109000000_create_iiif_print_pending_relationships.rb +11 -0
- data/docker-compose.yml +129 -0
- data/iiif_print.gemspec +43 -0
- data/lib/generators/iiif_print/assets_generator.rb +29 -0
- data/lib/generators/iiif_print/catalog_controller_generator.rb +32 -0
- data/lib/generators/iiif_print/install_generator.rb +52 -0
- data/lib/generators/iiif_print/templates/config/initializers/iiif_print.rb +22 -0
- data/lib/generators/iiif_print/templates/iiif_print.scss +1 -0
- data/lib/iiif_print/base_derivative_service.rb +113 -0
- data/lib/iiif_print/blacklight_iiif_search/annotation_decorator.rb +84 -0
- data/lib/iiif_print/catalog_search_builder.rb +31 -0
- data/lib/iiif_print/configuration.rb +99 -0
- data/lib/iiif_print/data/fileset_helper.rb +25 -0
- data/lib/iiif_print/data/path_helper.rb +40 -0
- data/lib/iiif_print/data/work_derivatives.rb +323 -0
- data/lib/iiif_print/data/work_file.rb +92 -0
- data/lib/iiif_print/data/work_files.rb +199 -0
- data/lib/iiif_print/data.rb +35 -0
- data/lib/iiif_print/engine.rb +77 -0
- data/lib/iiif_print/errors.rb +9 -0
- data/lib/iiif_print/image_tool.rb +119 -0
- data/lib/iiif_print/jobs/application_job.rb +8 -0
- data/lib/iiif_print/jobs/child_works_from_pdf_job.rb +107 -0
- data/lib/iiif_print/jobs/create_relationships_job.rb +78 -0
- data/lib/iiif_print/jp2_derivative_service.rb +118 -0
- data/lib/iiif_print/jp2_image_metadata.rb +81 -0
- data/lib/iiif_print/lineage_service.rb +41 -0
- data/lib/iiif_print/metadata.rb +125 -0
- data/lib/iiif_print/pdf_derivative_service.rb +42 -0
- data/lib/iiif_print/split_pdfs/child_work_creation_from_pdf_service.rb +75 -0
- data/lib/iiif_print/split_pdfs/pages_into_images_service.rb +130 -0
- data/lib/iiif_print/split_pdfs/pdf_image_extraction_service.rb +85 -0
- data/lib/iiif_print/text_extraction/alto_reader.rb +123 -0
- data/lib/iiif_print/text_extraction/hocr_reader.rb +172 -0
- data/lib/iiif_print/text_extraction/page_ocr.rb +87 -0
- data/lib/iiif_print/text_extraction/render_alto.rb +84 -0
- data/lib/iiif_print/text_extraction/word_coords_builder.rb +38 -0
- data/lib/iiif_print/text_extraction.rb +11 -0
- data/lib/iiif_print/text_extraction_derivative_service.rb +47 -0
- data/lib/iiif_print/text_formats_from_alto_service.rb +77 -0
- data/lib/iiif_print/tiff_derivative_service.rb +50 -0
- data/lib/iiif_print/version.rb +3 -0
- data/lib/iiif_print/works_controller_behavior.rb +9 -0
- data/lib/iiif_print.rb +136 -0
- data/lib/tasks/set_child_works.rake +22 -0
- data/spec/.keep.txt +1 -0
- data/spec/factories/ability.rb +6 -0
- data/spec/factories/newspaper_issue.rb +7 -0
- data/spec/factories/newspaper_page.rb +7 -0
- data/spec/factories/newspaper_page_solr_document.rb +12 -0
- data/spec/factories/newspaper_title.rb +8 -0
- data/spec/factories/uploaded_pdf_file.rb +9 -0
- data/spec/factories/uploaded_txt_file.rb +9 -0
- data/spec/factories/user.rb +13 -0
- data/spec/fixtures/files/4.1.07.jp2 +0 -0
- data/spec/fixtures/files/4.1.07.tiff +0 -0
- data/spec/fixtures/files/README.md +7 -0
- data/spec/fixtures/files/alto-2-0.xsd +714 -0
- data/spec/fixtures/files/broken-truncated.pdf +0 -0
- data/spec/fixtures/files/credits.md +16 -0
- data/spec/fixtures/files/lowres-gray-via-ndnp-sample.tiff +0 -0
- data/spec/fixtures/files/minimal-1-page.pdf +0 -0
- data/spec/fixtures/files/minimal-2-page.pdf +0 -0
- data/spec/fixtures/files/minimal-alto.xml +31 -0
- data/spec/fixtures/files/ndnp-alto-sample.xml +24 -0
- data/spec/fixtures/files/ndnp-sample1-json.json +1 -0
- data/spec/fixtures/files/ndnp-sample1-txt.txt +1 -0
- data/spec/fixtures/files/ndnp-sample1.pdf +0 -0
- data/spec/fixtures/files/ocr_alto.xml +202 -0
- data/spec/fixtures/files/ocr_alto_scaled_4pts_per_px.xml +202 -0
- data/spec/fixtures/files/ocr_color.tiff +0 -0
- data/spec/fixtures/files/ocr_gray.jp2 +0 -0
- data/spec/fixtures/files/ocr_gray.tiff +0 -0
- data/spec/fixtures/files/ocr_mono.tiff +0 -0
- data/spec/fixtures/files/ocr_mono_text_hocr.html +78 -0
- data/spec/fixtures/files/page1.tiff +0 -0
- data/spec/fixtures/files/sample-4page-issue.pdf +0 -0
- data/spec/fixtures/files/sample-color-newsletter.pdf +0 -0
- data/spec/fixtures/files/thumbnail.jpg +0 -0
- data/spec/helpers/hyrax/iiif_helper_spec.rb +65 -0
- data/spec/helpers/iiif_print_helper_spec.rb +43 -0
- data/spec/iiif_print/base_derivative_service_spec.rb +11 -0
- data/spec/iiif_print/blacklight_iiif_search/annotation_decorator_spec.rb +51 -0
- data/spec/iiif_print/catalog_search_builder_spec.rb +60 -0
- data/spec/iiif_print/configuration_spec.rb +67 -0
- data/spec/iiif_print/data/work_derivatives_spec.rb +245 -0
- data/spec/iiif_print/data/work_file_spec.rb +99 -0
- data/spec/iiif_print/data/work_files_spec.rb +237 -0
- data/spec/iiif_print/image_tool_spec.rb +109 -0
- data/spec/iiif_print/jobs/child_works_from_pdf_job_spec.rb +30 -0
- data/spec/iiif_print/jobs/create_relationships_job_spec.rb +17 -0
- data/spec/iiif_print/jp2_image_metadata_spec.rb +37 -0
- data/spec/iiif_print/lineage_service_spec.rb +13 -0
- data/spec/iiif_print/metadata_spec.rb +115 -0
- data/spec/iiif_print/split_pdfs/pages_into_images_service_spec.rb +6 -0
- data/spec/iiif_print/text_extraction/alto_reader_spec.rb +49 -0
- data/spec/iiif_print/text_extraction/hocr_reader_spec.rb +45 -0
- data/spec/iiif_print/text_extraction/page_ocr_spec.rb +84 -0
- data/spec/iiif_print/text_extraction/render_alto_spec.rb +54 -0
- data/spec/iiif_print/text_extraction/word_coords_builder_spec.rb +44 -0
- data/spec/iiif_print_spec.rb +51 -0
- data/spec/misc_shared.rb +111 -0
- data/spec/models/iiif_print/derivative_attachment_spec.rb +37 -0
- data/spec/models/iiif_print/ingest_file_relation_spec.rb +56 -0
- data/spec/models/solr_document_spec.rb +14 -0
- data/spec/presenters/iiif_print/iiif_manifest_presenter_behavior_spec.rb +19 -0
- data/spec/presenters/iiif_print/iiif_manifest_presenter_factory_behavior_spec.rb +49 -0
- data/spec/services/iiif_print/jp2_derivative_service_spec.rb +59 -0
- data/spec/services/iiif_print/pdf_derivative_service_spec.rb +66 -0
- data/spec/services/iiif_print/pluggable_derivative_service_spec.rb +178 -0
- data/spec/services/iiif_print/text_extraction_derivative_service_spec.rb +82 -0
- data/spec/services/iiif_print/text_formats_from_alto_service_spec.rb +127 -0
- data/spec/services/iiif_print/tiff_derivative_service_spec.rb +65 -0
- data/spec/spec_helper.rb +181 -0
- data/spec/support/controller_level_helpers.rb +28 -0
- data/spec/support/iiif_print_models.rb +127 -0
- data/spec/test_app_templates/blacklight.yml +9 -0
- data/spec/test_app_templates/fedora.yml +15 -0
- data/spec/test_app_templates/lib/generators/test_app_generator.rb +40 -0
- data/spec/test_app_templates/redis.yml +9 -0
- data/spec/test_app_templates/solr/conf/schema.xml +362 -0
- data/spec/test_app_templates/solr/conf/solrconfig.xml +322 -0
- data/spec/test_app_templates/solr.yml +7 -0
- data/tasks/iiif_print_dev.rake +34 -0
- data/tmp/.keep +0 -0
- metadata +605 -0
@@ -0,0 +1,246 @@
|
|
1
|
+
# The ASF licenses this file to You under the Apache License, Version 2.0
|
2
|
+
# (the "License"); you may not use this file except in compliance with
|
3
|
+
# the License. You may obtain a copy of the License at
|
4
|
+
#
|
5
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
6
|
+
#
|
7
|
+
# Unless required by applicable law or agreed to in writing, software
|
8
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
9
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
10
|
+
# See the License for the specific language governing permissions and
|
11
|
+
# limitations under the License.
|
12
|
+
|
13
|
+
# Syntax:
|
14
|
+
# "source" => "target"
|
15
|
+
# "source".length() > 0 (source cannot be empty.)
|
16
|
+
# "target".length() >= 0 (target can be empty.)
|
17
|
+
|
18
|
+
# example:
|
19
|
+
# "??" => "A"
|
20
|
+
# "\u00C0" => "A"
|
21
|
+
# "\u00C0" => "\u0041"
|
22
|
+
# "??" => "ss"
|
23
|
+
# "\t" => " "
|
24
|
+
# "\n" => ""
|
25
|
+
|
26
|
+
# ?? => A
|
27
|
+
"\u00C0" => "A"
|
28
|
+
|
29
|
+
# ?? => A
|
30
|
+
"\u00C1" => "A"
|
31
|
+
|
32
|
+
# ?? => A
|
33
|
+
"\u00C2" => "A"
|
34
|
+
|
35
|
+
# ?? => A
|
36
|
+
"\u00C3" => "A"
|
37
|
+
|
38
|
+
# ?? => A
|
39
|
+
"\u00C4" => "A"
|
40
|
+
|
41
|
+
# ?? => A
|
42
|
+
"\u00C5" => "A"
|
43
|
+
|
44
|
+
# ?? => AE
|
45
|
+
"\u00C6" => "AE"
|
46
|
+
|
47
|
+
# ?? => C
|
48
|
+
"\u00C7" => "C"
|
49
|
+
|
50
|
+
# ?? => E
|
51
|
+
"\u00C8" => "E"
|
52
|
+
|
53
|
+
# ?? => E
|
54
|
+
"\u00C9" => "E"
|
55
|
+
|
56
|
+
# ?? => E
|
57
|
+
"\u00CA" => "E"
|
58
|
+
|
59
|
+
# ?? => E
|
60
|
+
"\u00CB" => "E"
|
61
|
+
|
62
|
+
# ?? => I
|
63
|
+
"\u00CC" => "I"
|
64
|
+
|
65
|
+
# ?? => I
|
66
|
+
"\u00CD" => "I"
|
67
|
+
|
68
|
+
# ?? => I
|
69
|
+
"\u00CE" => "I"
|
70
|
+
|
71
|
+
# ?? => I
|
72
|
+
"\u00CF" => "I"
|
73
|
+
|
74
|
+
# ?? => IJ
|
75
|
+
"\u0132" => "IJ"
|
76
|
+
|
77
|
+
# ?? => D
|
78
|
+
"\u00D0" => "D"
|
79
|
+
|
80
|
+
# ?? => N
|
81
|
+
"\u00D1" => "N"
|
82
|
+
|
83
|
+
# ?? => O
|
84
|
+
"\u00D2" => "O"
|
85
|
+
|
86
|
+
# ?? => O
|
87
|
+
"\u00D3" => "O"
|
88
|
+
|
89
|
+
# ?? => O
|
90
|
+
"\u00D4" => "O"
|
91
|
+
|
92
|
+
# ?? => O
|
93
|
+
"\u00D5" => "O"
|
94
|
+
|
95
|
+
# ?? => O
|
96
|
+
"\u00D6" => "O"
|
97
|
+
|
98
|
+
# ?? => O
|
99
|
+
"\u00D8" => "O"
|
100
|
+
|
101
|
+
# ?? => OE
|
102
|
+
"\u0152" => "OE"
|
103
|
+
|
104
|
+
# ??
|
105
|
+
"\u00DE" => "TH"
|
106
|
+
|
107
|
+
# ?? => U
|
108
|
+
"\u00D9" => "U"
|
109
|
+
|
110
|
+
# ?? => U
|
111
|
+
"\u00DA" => "U"
|
112
|
+
|
113
|
+
# ?? => U
|
114
|
+
"\u00DB" => "U"
|
115
|
+
|
116
|
+
# ?? => U
|
117
|
+
"\u00DC" => "U"
|
118
|
+
|
119
|
+
# ?? => Y
|
120
|
+
"\u00DD" => "Y"
|
121
|
+
|
122
|
+
# ?? => Y
|
123
|
+
"\u0178" => "Y"
|
124
|
+
|
125
|
+
# ?? => a
|
126
|
+
"\u00E0" => "a"
|
127
|
+
|
128
|
+
# ?? => a
|
129
|
+
"\u00E1" => "a"
|
130
|
+
|
131
|
+
# ?? => a
|
132
|
+
"\u00E2" => "a"
|
133
|
+
|
134
|
+
# ?? => a
|
135
|
+
"\u00E3" => "a"
|
136
|
+
|
137
|
+
# ?? => a
|
138
|
+
"\u00E4" => "a"
|
139
|
+
|
140
|
+
# ?? => a
|
141
|
+
"\u00E5" => "a"
|
142
|
+
|
143
|
+
# ?? => ae
|
144
|
+
"\u00E6" => "ae"
|
145
|
+
|
146
|
+
# ?? => c
|
147
|
+
"\u00E7" => "c"
|
148
|
+
|
149
|
+
# ?? => e
|
150
|
+
"\u00E8" => "e"
|
151
|
+
|
152
|
+
# ?? => e
|
153
|
+
"\u00E9" => "e"
|
154
|
+
|
155
|
+
# ?? => e
|
156
|
+
"\u00EA" => "e"
|
157
|
+
|
158
|
+
# ?? => e
|
159
|
+
"\u00EB" => "e"
|
160
|
+
|
161
|
+
# ?? => i
|
162
|
+
"\u00EC" => "i"
|
163
|
+
|
164
|
+
# ?? => i
|
165
|
+
"\u00ED" => "i"
|
166
|
+
|
167
|
+
# ?? => i
|
168
|
+
"\u00EE" => "i"
|
169
|
+
|
170
|
+
# ?? => i
|
171
|
+
"\u00EF" => "i"
|
172
|
+
|
173
|
+
# ?? => ij
|
174
|
+
"\u0133" => "ij"
|
175
|
+
|
176
|
+
# ?? => d
|
177
|
+
"\u00F0" => "d"
|
178
|
+
|
179
|
+
# ?? => n
|
180
|
+
"\u00F1" => "n"
|
181
|
+
|
182
|
+
# ?? => o
|
183
|
+
"\u00F2" => "o"
|
184
|
+
|
185
|
+
# ?? => o
|
186
|
+
"\u00F3" => "o"
|
187
|
+
|
188
|
+
# ?? => o
|
189
|
+
"\u00F4" => "o"
|
190
|
+
|
191
|
+
# ?? => o
|
192
|
+
"\u00F5" => "o"
|
193
|
+
|
194
|
+
# ?? => o
|
195
|
+
"\u00F6" => "o"
|
196
|
+
|
197
|
+
# ?? => o
|
198
|
+
"\u00F8" => "o"
|
199
|
+
|
200
|
+
# ?? => oe
|
201
|
+
"\u0153" => "oe"
|
202
|
+
|
203
|
+
# ?? => ss
|
204
|
+
"\u00DF" => "ss"
|
205
|
+
|
206
|
+
# ?? => th
|
207
|
+
"\u00FE" => "th"
|
208
|
+
|
209
|
+
# ?? => u
|
210
|
+
"\u00F9" => "u"
|
211
|
+
|
212
|
+
# ?? => u
|
213
|
+
"\u00FA" => "u"
|
214
|
+
|
215
|
+
# ?? => u
|
216
|
+
"\u00FB" => "u"
|
217
|
+
|
218
|
+
# ?? => u
|
219
|
+
"\u00FC" => "u"
|
220
|
+
|
221
|
+
# ?? => y
|
222
|
+
"\u00FD" => "y"
|
223
|
+
|
224
|
+
# ?? => y
|
225
|
+
"\u00FF" => "y"
|
226
|
+
|
227
|
+
# ??? => ff
|
228
|
+
"\uFB00" => "ff"
|
229
|
+
|
230
|
+
# ??? => fi
|
231
|
+
"\uFB01" => "fi"
|
232
|
+
|
233
|
+
# ??? => fl
|
234
|
+
"\uFB02" => "fl"
|
235
|
+
|
236
|
+
# ??? => ffi
|
237
|
+
"\uFB03" => "ffi"
|
238
|
+
|
239
|
+
# ??? => ffl
|
240
|
+
"\uFB04" => "ffl"
|
241
|
+
|
242
|
+
# ??? => ft
|
243
|
+
"\uFB05" => "ft"
|
244
|
+
|
245
|
+
# ??? => st
|
246
|
+
"\uFB06" => "st"
|
@@ -0,0 +1,21 @@
|
|
1
|
+
# The ASF licenses this file to You under the Apache License, Version 2.0
|
2
|
+
# (the "License"); you may not use this file except in compliance with
|
3
|
+
# the License. You may obtain a copy of the License at
|
4
|
+
#
|
5
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
6
|
+
#
|
7
|
+
# Unless required by applicable law or agreed to in writing, software
|
8
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
9
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
10
|
+
# See the License for the specific language governing permissions and
|
11
|
+
# limitations under the License.
|
12
|
+
|
13
|
+
#-----------------------------------------------------------------------
|
14
|
+
# Use a protected word file to protect against the stemmer reducing two
|
15
|
+
# unrelated words to the same base word.
|
16
|
+
|
17
|
+
# Some non-words that normally won't be encountered,
|
18
|
+
# just to test that they won't be stemmed.
|
19
|
+
dontstems
|
20
|
+
zwhacky
|
21
|
+
|
@@ -0,0 +1,366 @@
|
|
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="Hydra Demo Index" version="1.5">
|
49
|
+
<!-- attribute "name" is the name of this schema and is only used for display purposes.
|
50
|
+
Applications should change this to reflect the nature of the search collection.
|
51
|
+
version="1.4" is Solr's version number for the schema syntax and semantics. It should
|
52
|
+
not normally be changed by applications.
|
53
|
+
1.0: multiValued attribute did not exist, all fields are multiValued by nature
|
54
|
+
1.1: multiValued attribute introduced, false by default
|
55
|
+
1.2: omitTermFreqAndPositions attribute introduced, true by default except for text fields.
|
56
|
+
1.3: removed optional field compress feature
|
57
|
+
1.4: default auto-phrase (QueryParser feature) to off
|
58
|
+
-->
|
59
|
+
|
60
|
+
<types>
|
61
|
+
<fieldType name="string" class="solr.StrField" sortMissingLast="true" />
|
62
|
+
<fieldType name="boolean" class="solr.BoolField" sortMissingLast="true"/>
|
63
|
+
<fieldType name="rand" class="solr.RandomSortField" omitNorms="true"/>
|
64
|
+
|
65
|
+
<!-- Default numeric field types. -->
|
66
|
+
<fieldType name="int" class="solr.TrieIntField" precisionStep="0" positionIncrementGap="0"/>
|
67
|
+
<fieldType name="float" class="solr.TrieFloatField" precisionStep="0" positionIncrementGap="0"/>
|
68
|
+
<fieldType name="long" class="solr.TrieLongField" precisionStep="0" positionIncrementGap="0"/>
|
69
|
+
<fieldType name="double" class="solr.TrieDoubleField" precisionStep="0" positionIncrementGap="0"/>
|
70
|
+
|
71
|
+
<!-- trie numeric field types for faster range queries -->
|
72
|
+
<fieldType name="tint" class="solr.TrieIntField" precisionStep="8" positionIncrementGap="0"/>
|
73
|
+
<fieldType name="tfloat" class="solr.TrieFloatField" precisionStep="8" positionIncrementGap="0"/>
|
74
|
+
<fieldType name="tlong" class="solr.TrieLongField" precisionStep="8" positionIncrementGap="0"/>
|
75
|
+
<fieldType name="tdouble" class="solr.TrieDoubleField" precisionStep="8" positionIncrementGap="0"/>
|
76
|
+
|
77
|
+
<!-- The format for this date field is of the form 1995-12-31T23:59:59Z
|
78
|
+
Optional fractional seconds are allowed: 1995-12-31T23:59:59.999Z
|
79
|
+
-->
|
80
|
+
<fieldType name="date" class="solr.TrieDateField" precisionStep="0" positionIncrementGap="0"/>
|
81
|
+
<!-- A Trie based date field for faster date range queries and date faceting. -->
|
82
|
+
<fieldType name="tdate" class="solr.TrieDateField" precisionStep="6" positionIncrementGap="0"/>
|
83
|
+
|
84
|
+
|
85
|
+
<!-- This point type indexes the coordinates as separate fields (subFields)
|
86
|
+
If subFieldType is defined, it references a type, and a dynamic field
|
87
|
+
definition is created matching *___<typename>. Alternately, if
|
88
|
+
subFieldSuffix is defined, that is used to create the subFields.
|
89
|
+
Example: if subFieldType="double", then the coordinates would be
|
90
|
+
indexed in fields myloc_0___double,myloc_1___double.
|
91
|
+
Example: if subFieldSuffix="_d" then the coordinates would be indexed
|
92
|
+
in fields myloc_0_d,myloc_1_d
|
93
|
+
The subFields are an implementation detail of the fieldType, and end
|
94
|
+
users normally should not need to know about them.
|
95
|
+
-->
|
96
|
+
<fieldType name="point" class="solr.PointType" dimension="2" subFieldSuffix="_d"/>
|
97
|
+
|
98
|
+
<!-- A specialized field for geospatial search. If indexed, this fieldType must not be multivalued. -->
|
99
|
+
<fieldType name="location" class="solr.LatLonType" subFieldSuffix="_coordinate"/>
|
100
|
+
|
101
|
+
<!-- An alternative geospatial field type new to Solr 4. It supports multiValued and polygon shapes.
|
102
|
+
For more information about this and other Spatial fields new to Solr 4, see:
|
103
|
+
http://wiki.apache.org/solr/SolrAdaptersForLuceneSpatial4
|
104
|
+
-->
|
105
|
+
<fieldType name="location_rpt" class="solr.SpatialRecursivePrefixTreeFieldType"
|
106
|
+
geo="true" distErrPct="0.025" maxDistErr="0.000009" distanceUnits="degrees" />
|
107
|
+
|
108
|
+
<fieldType name="text" class="solr.TextField" omitNorms="false">
|
109
|
+
<analyzer>
|
110
|
+
<tokenizer class="solr.ICUTokenizerFactory"/>
|
111
|
+
<filter class="solr.ICUFoldingFilterFactory"/> <!-- NFKC, case folding, diacritics removed -->
|
112
|
+
<filter class="solr.TrimFilterFactory"/>
|
113
|
+
</analyzer>
|
114
|
+
</fieldType>
|
115
|
+
|
116
|
+
<!-- A text field that only splits on whitespace for exact matching of words -->
|
117
|
+
<fieldType name="text_ws" class="solr.TextField" positionIncrementGap="100">
|
118
|
+
<analyzer>
|
119
|
+
<tokenizer class="solr.WhitespaceTokenizerFactory"/>
|
120
|
+
<filter class="solr.TrimFilterFactory"/>
|
121
|
+
</analyzer>
|
122
|
+
</fieldType>
|
123
|
+
|
124
|
+
<!-- single token analyzed text, for sorting. Punctuation is significant. -->
|
125
|
+
<fieldtype name="alphaSort" class="solr.TextField" sortMissingLast="true" omitNorms="true">
|
126
|
+
<analyzer>
|
127
|
+
<tokenizer class="solr.KeywordTokenizerFactory" />
|
128
|
+
<filter class="solr.ICUFoldingFilterFactory"/>
|
129
|
+
<filter class="solr.TrimFilterFactory" />
|
130
|
+
</analyzer>
|
131
|
+
</fieldtype>
|
132
|
+
|
133
|
+
<!-- A text field with defaults appropriate for English -->
|
134
|
+
<fieldType name="text_en" class="solr.TextField" positionIncrementGap="100">
|
135
|
+
<analyzer>
|
136
|
+
<tokenizer class="solr.ICUTokenizerFactory"/>
|
137
|
+
<filter class="solr.ICUFoldingFilterFactory"/> <!-- NFKC, case folding, diacritics removed -->
|
138
|
+
<filter class="solr.EnglishPossessiveFilterFactory"/>
|
139
|
+
<!-- EnglishMinimalStemFilterFactory is less aggressive than PorterStemFilterFactory: -->
|
140
|
+
<filter class="solr.EnglishMinimalStemFilterFactory"/>
|
141
|
+
<!--
|
142
|
+
<filter class="solr.PorterStemFilterFactory"/>
|
143
|
+
-->
|
144
|
+
<filter class="solr.TrimFilterFactory"/>
|
145
|
+
</analyzer>
|
146
|
+
</fieldType>
|
147
|
+
|
148
|
+
<!-- queries for paths match documents at that path, or in descendent paths -->
|
149
|
+
<fieldType name="descendent_path" class="solr.TextField">
|
150
|
+
<analyzer type="index">
|
151
|
+
<tokenizer class="solr.PathHierarchyTokenizerFactory" delimiter="/" />
|
152
|
+
</analyzer>
|
153
|
+
<analyzer type="query">
|
154
|
+
<tokenizer class="solr.KeywordTokenizerFactory" />
|
155
|
+
</analyzer>
|
156
|
+
</fieldType>
|
157
|
+
|
158
|
+
<!-- queries for paths match documents at that path, or in ancestor paths -->
|
159
|
+
<fieldType name="ancestor_path" class="solr.TextField">
|
160
|
+
<analyzer type="index">
|
161
|
+
<tokenizer class="solr.KeywordTokenizerFactory" />
|
162
|
+
</analyzer>
|
163
|
+
<analyzer type="query">
|
164
|
+
<tokenizer class="solr.PathHierarchyTokenizerFactory" delimiter="/" />
|
165
|
+
</analyzer>
|
166
|
+
</fieldType>
|
167
|
+
|
168
|
+
<fieldType class="solr.TextField" name="textSuggest" positionIncrementGap="100">
|
169
|
+
<analyzer>
|
170
|
+
<tokenizer class="solr.KeywordTokenizerFactory"/>
|
171
|
+
<filter class="solr.StandardFilterFactory"/>
|
172
|
+
<filter class="solr.LowerCaseFilterFactory"/>
|
173
|
+
<filter class="solr.RemoveDuplicatesTokenFilterFactory"/>
|
174
|
+
</analyzer>
|
175
|
+
</fieldType>
|
176
|
+
</types>
|
177
|
+
|
178
|
+
|
179
|
+
<fields>
|
180
|
+
<!-- If you remove this field, you must _also_ disable the update log in solrconfig.xml
|
181
|
+
or Solr won't start. _version_ and update log are required for SolrCloud
|
182
|
+
-->
|
183
|
+
<field name="_version_" type="long" indexed="true" stored="true"/>
|
184
|
+
|
185
|
+
<field name="id" type="string" stored="true" indexed="true" multiValued="false" required="true"/>
|
186
|
+
<field name="timestamp" type="date" indexed="true" stored="true" default="NOW" multiValued="false"/>
|
187
|
+
|
188
|
+
<field name="lat" type="tdouble" stored="true" indexed="true" multiValued="false"/>
|
189
|
+
<field name="lng" type="tdouble" stored="true" indexed="true" multiValued="false"/>
|
190
|
+
|
191
|
+
<!-- NOTE: not all possible Solr field types are represented in the dynamic fields -->
|
192
|
+
|
193
|
+
<!-- text (_t...) -->
|
194
|
+
<dynamicField name="*_ti" type="text" stored="false" indexed="true" multiValued="false"/>
|
195
|
+
<dynamicField name="*_tim" type="text" stored="false" indexed="true" multiValued="true"/>
|
196
|
+
<dynamicField name="*_ts" type="text" stored="true" indexed="false" multiValued="false"/>
|
197
|
+
<dynamicField name="*_tsm" type="text" stored="true" indexed="false" multiValued="true"/>
|
198
|
+
<dynamicField name="*_tsi" type="text" stored="true" indexed="true" multiValued="false"/>
|
199
|
+
<dynamicField name="*_tsim" type="text" stored="true" indexed="true" multiValued="true"/>
|
200
|
+
<dynamicField name="*_tiv" type="text" stored="false" indexed="true" multiValued="false" termVectors="true" termPositions="true" termOffsets="true"/>
|
201
|
+
<dynamicField name="*_timv" type="text" stored="false" indexed="true" multiValued="true" termVectors="true" termPositions="true" termOffsets="true"/>
|
202
|
+
<dynamicField name="*_tsiv" type="text" stored="true" indexed="true" multiValued="false" termVectors="true" termPositions="true" termOffsets="true"/>
|
203
|
+
<dynamicField name="*_tsimv" type="text" stored="true" indexed="true" multiValued="true" termVectors="true" termPositions="true" termOffsets="true"/>
|
204
|
+
|
205
|
+
<!-- English text (_te...) -->
|
206
|
+
<dynamicField name="*_tei" type="text_en" stored="false" indexed="true" multiValued="false"/>
|
207
|
+
<dynamicField name="*_teim" type="text_en" stored="false" indexed="true" multiValued="true"/>
|
208
|
+
<dynamicField name="*_tes" type="text_en" stored="true" indexed="false" multiValued="false"/>
|
209
|
+
<dynamicField name="*_tesm" type="text_en" stored="true" indexed="false" multiValued="true"/>
|
210
|
+
<dynamicField name="*_tesi" type="text_en" stored="true" indexed="true" multiValued="false"/>
|
211
|
+
<dynamicField name="*_tesim" type="text_en" stored="true" indexed="true" multiValued="true"/>
|
212
|
+
<dynamicField name="*_teiv" type="text_en" stored="false" indexed="true" multiValued="false" termVectors="true" termPositions="true" termOffsets="true"/>
|
213
|
+
<dynamicField name="*_teimv" type="text_en" stored="false" indexed="true" multiValued="true" termVectors="true" termPositions="true" termOffsets="true"/>
|
214
|
+
<dynamicField name="*_tesiv" type="text_en" stored="true" indexed="true" multiValued="false" termVectors="true" termPositions="true" termOffsets="true"/>
|
215
|
+
<dynamicField name="*_tesimv" type="text_en" stored="true" indexed="true" multiValued="true" termVectors="true" termPositions="true" termOffsets="true"/>
|
216
|
+
|
217
|
+
<!-- string (_s...) -->
|
218
|
+
<dynamicField name="*_si" type="string" stored="false" indexed="true" multiValued="false"/>
|
219
|
+
<dynamicField name="*_sim" type="string" stored="false" indexed="true" multiValued="true"/>
|
220
|
+
<dynamicField name="*_ss" type="string" stored="true" indexed="false" multiValued="false"/>
|
221
|
+
<dynamicField name="*_ssm" type="string" stored="true" indexed="false" multiValued="true"/>
|
222
|
+
<dynamicField name="*_ssi" type="string" stored="true" indexed="true" multiValued="false"/>
|
223
|
+
<dynamicField name="*_ssim" type="string" stored="true" indexed="true" multiValued="true"/>
|
224
|
+
<dynamicField name="*_ssort" type="alphaSort" stored="false" indexed="true" multiValued="false"/>
|
225
|
+
|
226
|
+
<!-- integer (_i...) -->
|
227
|
+
<dynamicField name="*_ii" type="int" stored="false" indexed="true" multiValued="false"/>
|
228
|
+
<dynamicField name="*_iim" type="int" stored="false" indexed="true" multiValued="true"/>
|
229
|
+
<dynamicField name="*_is" type="int" stored="true" indexed="false" multiValued="false"/>
|
230
|
+
<dynamicField name="*_ism" type="int" stored="true" indexed="false" multiValued="true"/>
|
231
|
+
<dynamicField name="*_isi" type="int" stored="true" indexed="true" multiValued="false"/>
|
232
|
+
<dynamicField name="*_isim" type="int" stored="true" indexed="true" multiValued="true"/>
|
233
|
+
|
234
|
+
<!-- trie integer (_it...) (for faster range queries) -->
|
235
|
+
<dynamicField name="*_iti" type="tint" stored="false" indexed="true" multiValued="false"/>
|
236
|
+
<dynamicField name="*_itim" type="tint" stored="false" indexed="true" multiValued="true"/>
|
237
|
+
<dynamicField name="*_its" type="tint" stored="true" indexed="false" multiValued="false"/>
|
238
|
+
<dynamicField name="*_itsm" type="tint" stored="true" indexed="false" multiValued="true"/>
|
239
|
+
<dynamicField name="*_itsi" type="tint" stored="true" indexed="true" multiValued="false"/>
|
240
|
+
<dynamicField name="*_itsim" type="tint" stored="true" indexed="true" multiValued="true"/>
|
241
|
+
|
242
|
+
<!-- date (_dt...) -->
|
243
|
+
<!-- The format for this date field is of the form 1995-12-31T23:59:59Z
|
244
|
+
Optional fractional seconds are allowed: 1995-12-31T23:59:59.999Z -->
|
245
|
+
<dynamicField name="*_dti" type="date" stored="false" indexed="true" multiValued="false"/>
|
246
|
+
<dynamicField name="*_dtim" type="date" stored="false" indexed="true" multiValued="true"/>
|
247
|
+
<dynamicField name="*_dts" type="date" stored="true" indexed="false" multiValued="false"/>
|
248
|
+
<dynamicField name="*_dtsm" type="date" stored="true" indexed="false" multiValued="true"/>
|
249
|
+
<dynamicField name="*_dtsi" type="date" stored="true" indexed="true" multiValued="false"/>
|
250
|
+
<dynamicField name="*_dtsim" type="date" stored="true" indexed="true" multiValued="true"/>
|
251
|
+
|
252
|
+
<!-- trie date (_dtt...) (for faster range queries) -->
|
253
|
+
<dynamicField name="*_dtti" type="tdate" stored="false" indexed="true" multiValued="false"/>
|
254
|
+
<dynamicField name="*_dttim" type="tdate" stored="false" indexed="true" multiValued="true"/>
|
255
|
+
<dynamicField name="*_dtts" type="tdate" stored="true" indexed="false" multiValued="false"/>
|
256
|
+
<dynamicField name="*_dttsm" type="tdate" stored="true" indexed="false" multiValued="true"/>
|
257
|
+
<dynamicField name="*_dttsi" type="tdate" stored="true" indexed="true" multiValued="false"/>
|
258
|
+
<dynamicField name="*_dttsim" type="tdate" stored="true" indexed="true" multiValued="true"/>
|
259
|
+
|
260
|
+
<!-- long (_l...) -->
|
261
|
+
<dynamicField name="*_li" type="long" stored="false" indexed="true" multiValued="false"/>
|
262
|
+
<dynamicField name="*_lim" type="long" stored="false" indexed="true" multiValued="true"/>
|
263
|
+
<dynamicField name="*_ls" type="long" stored="true" indexed="false" multiValued="false"/>
|
264
|
+
<dynamicField name="*_lsm" type="long" stored="true" indexed="false" multiValued="true"/>
|
265
|
+
<dynamicField name="*_lsi" type="long" stored="true" indexed="true" multiValued="false"/>
|
266
|
+
<dynamicField name="*_lsim" type="long" stored="true" indexed="true" multiValued="true"/>
|
267
|
+
|
268
|
+
<!-- trie long (_lt...) (for faster range queries) -->
|
269
|
+
<dynamicField name="*_lti" type="tlong" stored="false" indexed="true" multiValued="false"/>
|
270
|
+
<dynamicField name="*_ltim" type="tlong" stored="false" indexed="true" multiValued="true"/>
|
271
|
+
<dynamicField name="*_lts" type="tlong" stored="true" indexed="false" multiValued="false"/>
|
272
|
+
<dynamicField name="*_ltsm" type="tlong" stored="true" indexed="false" multiValued="true"/>
|
273
|
+
<dynamicField name="*_ltsi" type="tlong" stored="true" indexed="true" multiValued="false"/>
|
274
|
+
<dynamicField name="*_ltsim" type="tlong" stored="true" indexed="true" multiValued="true"/>
|
275
|
+
|
276
|
+
<!-- double (_db...) -->
|
277
|
+
<dynamicField name="*_dbi" type="double" stored="false" indexed="true" multiValued="false"/>
|
278
|
+
<dynamicField name="*_dbim" type="double" stored="false" indexed="true" multiValued="true"/>
|
279
|
+
<dynamicField name="*_dbs" type="double" stored="true" indexed="false" multiValued="false"/>
|
280
|
+
<dynamicField name="*_dbsm" type="double" stored="true" indexed="false" multiValued="true"/>
|
281
|
+
<dynamicField name="*_dbsi" type="double" stored="true" indexed="true" multiValued="false"/>
|
282
|
+
<dynamicField name="*_dbsim" type="double" stored="true" indexed="true" multiValued="true"/>
|
283
|
+
|
284
|
+
<!-- trie double (_dbt...) (for faster range queries) -->
|
285
|
+
<dynamicField name="*_dbti" type="tdouble" stored="false" indexed="true" multiValued="false"/>
|
286
|
+
<dynamicField name="*_dbtim" type="tdouble" stored="false" indexed="true" multiValued="true"/>
|
287
|
+
<dynamicField name="*_dbts" type="tdouble" stored="true" indexed="false" multiValued="false"/>
|
288
|
+
<dynamicField name="*_dbtsm" type="tdouble" stored="true" indexed="false" multiValued="true"/>
|
289
|
+
<dynamicField name="*_dbtsi" type="tdouble" stored="true" indexed="true" multiValued="false"/>
|
290
|
+
<dynamicField name="*_dbtsim" type="tdouble" stored="true" indexed="true" multiValued="true"/>
|
291
|
+
|
292
|
+
<!-- float (_f...) -->
|
293
|
+
<dynamicField name="*_fi" type="float" stored="false" indexed="true" multiValued="false"/>
|
294
|
+
<dynamicField name="*_fim" type="float" stored="false" indexed="true" multiValued="true"/>
|
295
|
+
<dynamicField name="*_fs" type="float" stored="true" indexed="false" multiValued="false"/>
|
296
|
+
<dynamicField name="*_fsm" type="float" stored="true" indexed="false" multiValued="true"/>
|
297
|
+
<dynamicField name="*_fsi" type="float" stored="true" indexed="true" multiValued="false"/>
|
298
|
+
<dynamicField name="*_fsim" type="float" stored="true" indexed="true" multiValued="true"/>
|
299
|
+
|
300
|
+
<!-- trie float (_ft...) (for faster range queries) -->
|
301
|
+
<dynamicField name="*_fti" type="tfloat" stored="false" indexed="true" multiValued="false"/>
|
302
|
+
<dynamicField name="*_ftim" type="tfloat" stored="false" indexed="true" multiValued="true"/>
|
303
|
+
<dynamicField name="*_fts" type="tfloat" stored="true" indexed="false" multiValued="false"/>
|
304
|
+
<dynamicField name="*_ftsm" type="tfloat" stored="true" indexed="false" multiValued="true"/>
|
305
|
+
<dynamicField name="*_ftsi" type="tfloat" stored="true" indexed="true" multiValued="false"/>
|
306
|
+
<dynamicField name="*_ftsim" type="tfloat" stored="true" indexed="true" multiValued="true"/>
|
307
|
+
|
308
|
+
<!-- boolean (_b...) -->
|
309
|
+
<dynamicField name="*_bi" type="boolean" stored="false" indexed="true" multiValued="false"/>
|
310
|
+
<dynamicField name="*_bs" type="boolean" stored="true" indexed="false" multiValued="false"/>
|
311
|
+
<dynamicField name="*_bsi" type="boolean" stored="true" indexed="true" multiValued="false"/>
|
312
|
+
|
313
|
+
<!-- Type used to index the lat and lon components for the "location" FieldType -->
|
314
|
+
<dynamicField name="*_coordinate" type="tdouble" indexed="true" stored="false" />
|
315
|
+
|
316
|
+
<!-- location (_ll...) -->
|
317
|
+
<dynamicField name="*_lli" type="location" stored="false" indexed="true" multiValued="false"/>
|
318
|
+
<dynamicField name="*_llim" type="location" stored="false" indexed="true" multiValued="true"/>
|
319
|
+
<dynamicField name="*_lls" type="location" stored="true" indexed="false" multiValued="false"/>
|
320
|
+
<dynamicField name="*_llsm" type="location" stored="true" indexed="false" multiValued="true"/>
|
321
|
+
<dynamicField name="*_llsi" type="location" stored="true" indexed="true" multiValued="false"/>
|
322
|
+
<dynamicField name="*_llsim" type="location" stored="true" indexed="true" multiValued="true"/>
|
323
|
+
|
324
|
+
<dynamicField name="*suggest" type="textSuggest" indexed="true" stored="false" multiValued="true" />
|
325
|
+
|
326
|
+
<!-- you must define copyField source and dest fields explicity or schemaBrowser doesn't work -->
|
327
|
+
<field name="all_text_timv" type="text" stored="false" indexed="true" multiValued="true" termVectors="true" termPositions="true" termOffsets="true"/>
|
328
|
+
|
329
|
+
|
330
|
+
</fields>
|
331
|
+
|
332
|
+
<!-- Field to use to determine and enforce document uniqueness.
|
333
|
+
Unless this field is marked with required="false", it will be a required field
|
334
|
+
-->
|
335
|
+
<uniqueKey>id</uniqueKey>
|
336
|
+
|
337
|
+
<!-- copyField commands copy one field to another at the time a document
|
338
|
+
is added to the index. It's used either to index the same field differently,
|
339
|
+
or to add multiple fields to the same field for easier/faster searching. -->
|
340
|
+
<!-- Copy Fields -->
|
341
|
+
|
342
|
+
<!-- Above, multiple source fields are copied to the [text] field.
|
343
|
+
Another way to map multiple source fields to the same
|
344
|
+
destination field is to use the dynamic field syntax.
|
345
|
+
copyField also supports a maxChars to copy setting. -->
|
346
|
+
|
347
|
+
<!-- <copyField source="*_tesim" dest="all_text_timv" maxChars="3000"/> -->
|
348
|
+
<!-- for suggestions -->
|
349
|
+
<copyField source="*_tesim" dest="suggest"/>
|
350
|
+
<copyField source="*_ssim" dest="suggest"/>
|
351
|
+
|
352
|
+
<!-- Similarity is the scoring routine for each document vs. a query.
|
353
|
+
A custom similarity may be specified here, but the default is fine
|
354
|
+
for most applications. -->
|
355
|
+
<!-- <similarity class="org.apache.lucene.search.DefaultSimilarity"/> -->
|
356
|
+
<!-- ... OR ...
|
357
|
+
Specify a SimilarityFactory class name implementation
|
358
|
+
allowing parameters to be used.
|
359
|
+
-->
|
360
|
+
<!--
|
361
|
+
<similarity class="com.example.solr.CustomSimilarityFactory">
|
362
|
+
<str name="paramkey">param value</str>
|
363
|
+
</similarity>
|
364
|
+
-->
|
365
|
+
|
366
|
+
</schema>
|
@@ -0,0 +1,24 @@
|
|
1
|
+
# Licensed to the Apache Software Foundation (ASF) under one or more
|
2
|
+
# contributor license agreements. See the NOTICE file distributed with
|
3
|
+
# this work for additional information regarding copyright ownership.
|
4
|
+
# The ASF licenses this file to You under the Apache License, Version 2.0
|
5
|
+
# (the "License"); you may not use this file except in compliance with
|
6
|
+
# the License. You may obtain a copy of the License at
|
7
|
+
#
|
8
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
9
|
+
#
|
10
|
+
# Unless required by applicable law or agreed to in writing, software
|
11
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
12
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
13
|
+
# See the License for the specific language governing permissions and
|
14
|
+
# limitations under the License.
|
15
|
+
|
16
|
+
user=
|
17
|
+
solr_hostname=localhost
|
18
|
+
solr_port=8983
|
19
|
+
rsyncd_port=18983
|
20
|
+
data_dir=
|
21
|
+
webapp_name=solr
|
22
|
+
master_host=
|
23
|
+
master_data_dir=
|
24
|
+
master_status_dir=
|