kreuzberg 4.0.0.rc1 → 4.0.0.rc2

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (342) hide show
  1. checksums.yaml +4 -4
  2. data/.gitignore +14 -8
  3. data/.rspec +3 -3
  4. data/.rubocop.yaml +1 -534
  5. data/.rubocop.yml +538 -0
  6. data/Gemfile +8 -9
  7. data/Gemfile.lock +9 -109
  8. data/README.md +426 -421
  9. data/Rakefile +25 -25
  10. data/Steepfile +47 -47
  11. data/examples/async_patterns.rb +341 -340
  12. data/ext/kreuzberg_rb/extconf.rb +45 -35
  13. data/ext/kreuzberg_rb/native/Cargo.lock +6535 -0
  14. data/ext/kreuzberg_rb/native/Cargo.toml +44 -36
  15. data/ext/kreuzberg_rb/native/README.md +425 -425
  16. data/ext/kreuzberg_rb/native/build.rs +15 -17
  17. data/ext/kreuzberg_rb/native/include/ieeefp.h +11 -11
  18. data/ext/kreuzberg_rb/native/include/msvc_compat/strings.h +14 -14
  19. data/ext/kreuzberg_rb/native/include/strings.h +20 -20
  20. data/ext/kreuzberg_rb/native/include/unistd.h +47 -47
  21. data/ext/kreuzberg_rb/native/src/lib.rs +2998 -2939
  22. data/extconf.rb +28 -28
  23. data/kreuzberg.gemspec +148 -105
  24. data/lib/kreuzberg/api_proxy.rb +142 -142
  25. data/lib/kreuzberg/cache_api.rb +46 -45
  26. data/lib/kreuzberg/cli.rb +55 -55
  27. data/lib/kreuzberg/cli_proxy.rb +127 -127
  28. data/lib/kreuzberg/config.rb +691 -684
  29. data/lib/kreuzberg/error_context.rb +32 -0
  30. data/lib/kreuzberg/errors.rb +118 -50
  31. data/lib/kreuzberg/extraction_api.rb +85 -84
  32. data/lib/kreuzberg/mcp_proxy.rb +186 -186
  33. data/lib/kreuzberg/ocr_backend_protocol.rb +113 -113
  34. data/lib/kreuzberg/post_processor_protocol.rb +86 -86
  35. data/lib/kreuzberg/result.rb +216 -216
  36. data/lib/kreuzberg/setup_lib_path.rb +80 -79
  37. data/lib/kreuzberg/validator_protocol.rb +89 -89
  38. data/lib/kreuzberg/version.rb +5 -5
  39. data/lib/kreuzberg.rb +103 -82
  40. data/sig/kreuzberg/internal.rbs +184 -184
  41. data/sig/kreuzberg.rbs +520 -468
  42. data/spec/binding/cache_spec.rb +227 -227
  43. data/spec/binding/cli_proxy_spec.rb +85 -87
  44. data/spec/binding/cli_spec.rb +55 -54
  45. data/spec/binding/config_spec.rb +345 -345
  46. data/spec/binding/config_validation_spec.rb +283 -283
  47. data/spec/binding/error_handling_spec.rb +213 -213
  48. data/spec/binding/errors_spec.rb +66 -66
  49. data/spec/binding/plugins/ocr_backend_spec.rb +307 -307
  50. data/spec/binding/plugins/postprocessor_spec.rb +269 -269
  51. data/spec/binding/plugins/validator_spec.rb +274 -274
  52. data/spec/fixtures/config.toml +39 -39
  53. data/spec/fixtures/config.yaml +41 -42
  54. data/spec/fixtures/invalid_config.toml +4 -4
  55. data/spec/smoke/package_spec.rb +178 -178
  56. data/spec/spec_helper.rb +42 -42
  57. data/vendor/kreuzberg/Cargo.toml +204 -134
  58. data/vendor/kreuzberg/README.md +175 -175
  59. data/vendor/kreuzberg/benches/otel_overhead.rs +48 -0
  60. data/vendor/kreuzberg/build.rs +474 -460
  61. data/vendor/kreuzberg/src/api/error.rs +81 -81
  62. data/vendor/kreuzberg/src/api/handlers.rs +199 -199
  63. data/vendor/kreuzberg/src/api/mod.rs +79 -79
  64. data/vendor/kreuzberg/src/api/server.rs +353 -353
  65. data/vendor/kreuzberg/src/api/types.rs +170 -170
  66. data/vendor/kreuzberg/src/cache/mod.rs +1167 -1143
  67. data/vendor/kreuzberg/src/chunking/mod.rs +677 -677
  68. data/vendor/kreuzberg/src/core/batch_mode.rs +95 -35
  69. data/vendor/kreuzberg/src/core/config.rs +1032 -1032
  70. data/vendor/kreuzberg/src/core/extractor.rs +1024 -903
  71. data/vendor/kreuzberg/src/core/io.rs +329 -327
  72. data/vendor/kreuzberg/src/core/mime.rs +605 -615
  73. data/vendor/kreuzberg/src/core/mod.rs +45 -42
  74. data/vendor/kreuzberg/src/core/pipeline.rs +984 -906
  75. data/vendor/kreuzberg/src/embeddings.rs +432 -323
  76. data/vendor/kreuzberg/src/error.rs +431 -431
  77. data/vendor/kreuzberg/src/extraction/archive.rs +954 -954
  78. data/vendor/kreuzberg/src/extraction/docx.rs +40 -40
  79. data/vendor/kreuzberg/src/extraction/email.rs +854 -854
  80. data/vendor/kreuzberg/src/extraction/excel.rs +688 -688
  81. data/vendor/kreuzberg/src/extraction/html.rs +553 -553
  82. data/vendor/kreuzberg/src/extraction/image.rs +368 -368
  83. data/vendor/kreuzberg/src/extraction/libreoffice.rs +563 -564
  84. data/vendor/kreuzberg/src/extraction/markdown.rs +213 -0
  85. data/vendor/kreuzberg/src/extraction/mod.rs +81 -77
  86. data/vendor/kreuzberg/src/extraction/office_metadata/app_properties.rs +398 -398
  87. data/vendor/kreuzberg/src/extraction/office_metadata/core_properties.rs +247 -247
  88. data/vendor/kreuzberg/src/extraction/office_metadata/custom_properties.rs +240 -240
  89. data/vendor/kreuzberg/src/extraction/office_metadata/mod.rs +130 -128
  90. data/vendor/kreuzberg/src/extraction/office_metadata/odt_properties.rs +287 -0
  91. data/vendor/kreuzberg/src/extraction/pptx.rs +3000 -3000
  92. data/vendor/kreuzberg/src/extraction/structured.rs +490 -490
  93. data/vendor/kreuzberg/src/extraction/table.rs +328 -328
  94. data/vendor/kreuzberg/src/extraction/text.rs +269 -269
  95. data/vendor/kreuzberg/src/extraction/xml.rs +333 -333
  96. data/vendor/kreuzberg/src/extractors/archive.rs +446 -425
  97. data/vendor/kreuzberg/src/extractors/bibtex.rs +469 -0
  98. data/vendor/kreuzberg/src/extractors/docbook.rs +502 -0
  99. data/vendor/kreuzberg/src/extractors/docx.rs +367 -479
  100. data/vendor/kreuzberg/src/extractors/email.rs +143 -129
  101. data/vendor/kreuzberg/src/extractors/epub.rs +707 -0
  102. data/vendor/kreuzberg/src/extractors/excel.rs +343 -344
  103. data/vendor/kreuzberg/src/extractors/fictionbook.rs +491 -0
  104. data/vendor/kreuzberg/src/extractors/fictionbook.rs.backup2 +738 -0
  105. data/vendor/kreuzberg/src/extractors/html.rs +393 -410
  106. data/vendor/kreuzberg/src/extractors/image.rs +198 -195
  107. data/vendor/kreuzberg/src/extractors/jats.rs +1051 -0
  108. data/vendor/kreuzberg/src/extractors/jupyter.rs +367 -0
  109. data/vendor/kreuzberg/src/extractors/latex.rs +652 -0
  110. data/vendor/kreuzberg/src/extractors/markdown.rs +700 -0
  111. data/vendor/kreuzberg/src/extractors/mod.rs +365 -268
  112. data/vendor/kreuzberg/src/extractors/odt.rs +628 -0
  113. data/vendor/kreuzberg/src/extractors/opml.rs +634 -0
  114. data/vendor/kreuzberg/src/extractors/orgmode.rs +528 -0
  115. data/vendor/kreuzberg/src/extractors/pdf.rs +493 -496
  116. data/vendor/kreuzberg/src/extractors/pptx.rs +248 -234
  117. data/vendor/kreuzberg/src/extractors/rst.rs +576 -0
  118. data/vendor/kreuzberg/src/extractors/rtf.rs +810 -0
  119. data/vendor/kreuzberg/src/extractors/security.rs +484 -0
  120. data/vendor/kreuzberg/src/extractors/security_tests.rs +367 -0
  121. data/vendor/kreuzberg/src/extractors/structured.rs +140 -126
  122. data/vendor/kreuzberg/src/extractors/text.rs +260 -242
  123. data/vendor/kreuzberg/src/extractors/typst.rs +650 -0
  124. data/vendor/kreuzberg/src/extractors/xml.rs +135 -128
  125. data/vendor/kreuzberg/src/image/dpi.rs +164 -164
  126. data/vendor/kreuzberg/src/image/mod.rs +6 -6
  127. data/vendor/kreuzberg/src/image/preprocessing.rs +417 -417
  128. data/vendor/kreuzberg/src/image/resize.rs +89 -89
  129. data/vendor/kreuzberg/src/keywords/config.rs +154 -154
  130. data/vendor/kreuzberg/src/keywords/mod.rs +237 -237
  131. data/vendor/kreuzberg/src/keywords/processor.rs +267 -267
  132. data/vendor/kreuzberg/src/keywords/rake.rs +293 -294
  133. data/vendor/kreuzberg/src/keywords/types.rs +68 -68
  134. data/vendor/kreuzberg/src/keywords/yake.rs +163 -163
  135. data/vendor/kreuzberg/src/language_detection/mod.rs +942 -942
  136. data/vendor/kreuzberg/src/lib.rs +105 -102
  137. data/vendor/kreuzberg/src/mcp/mod.rs +32 -32
  138. data/vendor/kreuzberg/src/mcp/server.rs +1968 -1966
  139. data/vendor/kreuzberg/src/ocr/cache.rs +469 -469
  140. data/vendor/kreuzberg/src/ocr/error.rs +37 -37
  141. data/vendor/kreuzberg/src/ocr/hocr.rs +216 -216
  142. data/vendor/kreuzberg/src/ocr/mod.rs +58 -58
  143. data/vendor/kreuzberg/src/ocr/processor.rs +863 -847
  144. data/vendor/kreuzberg/src/ocr/table/mod.rs +4 -4
  145. data/vendor/kreuzberg/src/ocr/table/tsv_parser.rs +144 -144
  146. data/vendor/kreuzberg/src/ocr/tesseract_backend.rs +450 -450
  147. data/vendor/kreuzberg/src/ocr/types.rs +393 -393
  148. data/vendor/kreuzberg/src/ocr/utils.rs +47 -47
  149. data/vendor/kreuzberg/src/ocr/validation.rs +206 -206
  150. data/vendor/kreuzberg/src/panic_context.rs +154 -0
  151. data/vendor/kreuzberg/src/pdf/error.rs +122 -122
  152. data/vendor/kreuzberg/src/pdf/images.rs +139 -139
  153. data/vendor/kreuzberg/src/pdf/metadata.rs +346 -346
  154. data/vendor/kreuzberg/src/pdf/mod.rs +50 -50
  155. data/vendor/kreuzberg/src/pdf/rendering.rs +369 -369
  156. data/vendor/kreuzberg/src/pdf/table.rs +393 -420
  157. data/vendor/kreuzberg/src/pdf/text.rs +158 -161
  158. data/vendor/kreuzberg/src/plugins/extractor.rs +1013 -1010
  159. data/vendor/kreuzberg/src/plugins/mod.rs +209 -209
  160. data/vendor/kreuzberg/src/plugins/ocr.rs +620 -629
  161. data/vendor/kreuzberg/src/plugins/processor.rs +642 -641
  162. data/vendor/kreuzberg/src/plugins/registry.rs +1337 -1324
  163. data/vendor/kreuzberg/src/plugins/traits.rs +258 -258
  164. data/vendor/kreuzberg/src/plugins/validator.rs +956 -955
  165. data/vendor/kreuzberg/src/stopwords/mod.rs +1470 -1470
  166. data/vendor/kreuzberg/src/text/mod.rs +19 -19
  167. data/vendor/kreuzberg/src/text/quality.rs +697 -697
  168. data/vendor/kreuzberg/src/text/string_utils.rs +217 -217
  169. data/vendor/kreuzberg/src/text/token_reduction/cjk_utils.rs +164 -164
  170. data/vendor/kreuzberg/src/text/token_reduction/config.rs +100 -100
  171. data/vendor/kreuzberg/src/text/token_reduction/core.rs +796 -796
  172. data/vendor/kreuzberg/src/text/token_reduction/filters.rs +902 -902
  173. data/vendor/kreuzberg/src/text/token_reduction/mod.rs +160 -160
  174. data/vendor/kreuzberg/src/text/token_reduction/semantic.rs +619 -619
  175. data/vendor/kreuzberg/src/text/token_reduction/simd_text.rs +147 -147
  176. data/vendor/kreuzberg/src/types.rs +903 -873
  177. data/vendor/kreuzberg/src/utils/mod.rs +17 -17
  178. data/vendor/kreuzberg/src/utils/quality.rs +959 -959
  179. data/vendor/kreuzberg/src/utils/string_utils.rs +381 -381
  180. data/vendor/kreuzberg/stopwords/af_stopwords.json +53 -53
  181. data/vendor/kreuzberg/stopwords/ar_stopwords.json +482 -482
  182. data/vendor/kreuzberg/stopwords/bg_stopwords.json +261 -261
  183. data/vendor/kreuzberg/stopwords/bn_stopwords.json +400 -400
  184. data/vendor/kreuzberg/stopwords/br_stopwords.json +1205 -1205
  185. data/vendor/kreuzberg/stopwords/ca_stopwords.json +280 -280
  186. data/vendor/kreuzberg/stopwords/cs_stopwords.json +425 -425
  187. data/vendor/kreuzberg/stopwords/da_stopwords.json +172 -172
  188. data/vendor/kreuzberg/stopwords/de_stopwords.json +622 -622
  189. data/vendor/kreuzberg/stopwords/el_stopwords.json +849 -849
  190. data/vendor/kreuzberg/stopwords/en_stopwords.json +1300 -1300
  191. data/vendor/kreuzberg/stopwords/eo_stopwords.json +175 -175
  192. data/vendor/kreuzberg/stopwords/es_stopwords.json +734 -734
  193. data/vendor/kreuzberg/stopwords/et_stopwords.json +37 -37
  194. data/vendor/kreuzberg/stopwords/eu_stopwords.json +100 -100
  195. data/vendor/kreuzberg/stopwords/fa_stopwords.json +801 -801
  196. data/vendor/kreuzberg/stopwords/fi_stopwords.json +849 -849
  197. data/vendor/kreuzberg/stopwords/fr_stopwords.json +693 -693
  198. data/vendor/kreuzberg/stopwords/ga_stopwords.json +111 -111
  199. data/vendor/kreuzberg/stopwords/gl_stopwords.json +162 -162
  200. data/vendor/kreuzberg/stopwords/gu_stopwords.json +226 -226
  201. data/vendor/kreuzberg/stopwords/ha_stopwords.json +41 -41
  202. data/vendor/kreuzberg/stopwords/he_stopwords.json +196 -196
  203. data/vendor/kreuzberg/stopwords/hi_stopwords.json +227 -227
  204. data/vendor/kreuzberg/stopwords/hr_stopwords.json +181 -181
  205. data/vendor/kreuzberg/stopwords/hu_stopwords.json +791 -791
  206. data/vendor/kreuzberg/stopwords/hy_stopwords.json +47 -47
  207. data/vendor/kreuzberg/stopwords/id_stopwords.json +760 -760
  208. data/vendor/kreuzberg/stopwords/it_stopwords.json +634 -634
  209. data/vendor/kreuzberg/stopwords/ja_stopwords.json +136 -136
  210. data/vendor/kreuzberg/stopwords/kn_stopwords.json +84 -84
  211. data/vendor/kreuzberg/stopwords/ko_stopwords.json +681 -681
  212. data/vendor/kreuzberg/stopwords/ku_stopwords.json +64 -64
  213. data/vendor/kreuzberg/stopwords/la_stopwords.json +51 -51
  214. data/vendor/kreuzberg/stopwords/lt_stopwords.json +476 -476
  215. data/vendor/kreuzberg/stopwords/lv_stopwords.json +163 -163
  216. data/vendor/kreuzberg/stopwords/ml_stopwords.json +1 -1
  217. data/vendor/kreuzberg/stopwords/mr_stopwords.json +101 -101
  218. data/vendor/kreuzberg/stopwords/ms_stopwords.json +477 -477
  219. data/vendor/kreuzberg/stopwords/ne_stopwords.json +490 -490
  220. data/vendor/kreuzberg/stopwords/nl_stopwords.json +415 -415
  221. data/vendor/kreuzberg/stopwords/no_stopwords.json +223 -223
  222. data/vendor/kreuzberg/stopwords/pl_stopwords.json +331 -331
  223. data/vendor/kreuzberg/stopwords/pt_stopwords.json +562 -562
  224. data/vendor/kreuzberg/stopwords/ro_stopwords.json +436 -436
  225. data/vendor/kreuzberg/stopwords/ru_stopwords.json +561 -561
  226. data/vendor/kreuzberg/stopwords/si_stopwords.json +193 -193
  227. data/vendor/kreuzberg/stopwords/sk_stopwords.json +420 -420
  228. data/vendor/kreuzberg/stopwords/sl_stopwords.json +448 -448
  229. data/vendor/kreuzberg/stopwords/so_stopwords.json +32 -32
  230. data/vendor/kreuzberg/stopwords/st_stopwords.json +33 -33
  231. data/vendor/kreuzberg/stopwords/sv_stopwords.json +420 -420
  232. data/vendor/kreuzberg/stopwords/sw_stopwords.json +76 -76
  233. data/vendor/kreuzberg/stopwords/ta_stopwords.json +129 -129
  234. data/vendor/kreuzberg/stopwords/te_stopwords.json +54 -54
  235. data/vendor/kreuzberg/stopwords/th_stopwords.json +118 -118
  236. data/vendor/kreuzberg/stopwords/tl_stopwords.json +149 -149
  237. data/vendor/kreuzberg/stopwords/tr_stopwords.json +506 -506
  238. data/vendor/kreuzberg/stopwords/uk_stopwords.json +75 -75
  239. data/vendor/kreuzberg/stopwords/ur_stopwords.json +519 -519
  240. data/vendor/kreuzberg/stopwords/vi_stopwords.json +647 -647
  241. data/vendor/kreuzberg/stopwords/yo_stopwords.json +62 -62
  242. data/vendor/kreuzberg/stopwords/zh_stopwords.json +796 -796
  243. data/vendor/kreuzberg/stopwords/zu_stopwords.json +31 -31
  244. data/vendor/kreuzberg/tests/api_extract_multipart.rs +52 -0
  245. data/vendor/kreuzberg/tests/api_tests.rs +966 -966
  246. data/vendor/kreuzberg/tests/archive_integration.rs +543 -543
  247. data/vendor/kreuzberg/tests/batch_orchestration.rs +556 -542
  248. data/vendor/kreuzberg/tests/batch_processing.rs +316 -304
  249. data/vendor/kreuzberg/tests/bibtex_parity_test.rs +421 -0
  250. data/vendor/kreuzberg/tests/concurrency_stress.rs +525 -509
  251. data/vendor/kreuzberg/tests/config_features.rs +598 -580
  252. data/vendor/kreuzberg/tests/config_loading_tests.rs +415 -439
  253. data/vendor/kreuzberg/tests/core_integration.rs +510 -493
  254. data/vendor/kreuzberg/tests/csv_integration.rs +414 -424
  255. data/vendor/kreuzberg/tests/docbook_extractor_tests.rs +498 -0
  256. data/vendor/kreuzberg/tests/docx_metadata_extraction_test.rs +122 -124
  257. data/vendor/kreuzberg/tests/docx_vs_pandoc_comparison.rs +370 -0
  258. data/vendor/kreuzberg/tests/email_integration.rs +325 -325
  259. data/vendor/kreuzberg/tests/epub_native_extractor_tests.rs +275 -0
  260. data/vendor/kreuzberg/tests/error_handling.rs +393 -393
  261. data/vendor/kreuzberg/tests/fictionbook_extractor_tests.rs +228 -0
  262. data/vendor/kreuzberg/tests/format_integration.rs +159 -159
  263. data/vendor/kreuzberg/tests/helpers/mod.rs +142 -142
  264. data/vendor/kreuzberg/tests/html_table_test.rs +551 -0
  265. data/vendor/kreuzberg/tests/image_integration.rs +253 -253
  266. data/vendor/kreuzberg/tests/instrumentation_test.rs +139 -0
  267. data/vendor/kreuzberg/tests/jats_extractor_tests.rs +639 -0
  268. data/vendor/kreuzberg/tests/jupyter_extractor_tests.rs +704 -0
  269. data/vendor/kreuzberg/tests/keywords_integration.rs +479 -479
  270. data/vendor/kreuzberg/tests/keywords_quality.rs +509 -509
  271. data/vendor/kreuzberg/tests/latex_extractor_tests.rs +496 -0
  272. data/vendor/kreuzberg/tests/markdown_extractor_tests.rs +490 -0
  273. data/vendor/kreuzberg/tests/mime_detection.rs +428 -428
  274. data/vendor/kreuzberg/tests/ocr_configuration.rs +510 -510
  275. data/vendor/kreuzberg/tests/ocr_errors.rs +676 -676
  276. data/vendor/kreuzberg/tests/ocr_quality.rs +627 -627
  277. data/vendor/kreuzberg/tests/ocr_stress.rs +469 -469
  278. data/vendor/kreuzberg/tests/odt_extractor_tests.rs +695 -0
  279. data/vendor/kreuzberg/tests/opml_extractor_tests.rs +616 -0
  280. data/vendor/kreuzberg/tests/orgmode_extractor_tests.rs +822 -0
  281. data/vendor/kreuzberg/tests/pdf_integration.rs +43 -43
  282. data/vendor/kreuzberg/tests/pipeline_integration.rs +1411 -1412
  283. data/vendor/kreuzberg/tests/plugin_ocr_backend_test.rs +771 -771
  284. data/vendor/kreuzberg/tests/plugin_postprocessor_test.rs +560 -561
  285. data/vendor/kreuzberg/tests/plugin_system.rs +921 -921
  286. data/vendor/kreuzberg/tests/plugin_validator_test.rs +783 -783
  287. data/vendor/kreuzberg/tests/registry_integration_tests.rs +586 -607
  288. data/vendor/kreuzberg/tests/rst_extractor_tests.rs +692 -0
  289. data/vendor/kreuzberg/tests/rtf_extractor_tests.rs +776 -0
  290. data/vendor/kreuzberg/tests/security_validation.rs +415 -404
  291. data/vendor/kreuzberg/tests/stopwords_integration_test.rs +888 -888
  292. data/vendor/kreuzberg/tests/test_fastembed.rs +609 -609
  293. data/vendor/kreuzberg/tests/typst_behavioral_tests.rs +1259 -0
  294. data/vendor/kreuzberg/tests/typst_extractor_tests.rs +647 -0
  295. data/vendor/kreuzberg/tests/xlsx_metadata_extraction_test.rs +87 -87
  296. data/vendor/rb-sys/.cargo-ok +1 -0
  297. data/vendor/rb-sys/.cargo_vcs_info.json +6 -0
  298. data/vendor/rb-sys/Cargo.lock +393 -0
  299. data/vendor/rb-sys/Cargo.toml +70 -0
  300. data/vendor/rb-sys/Cargo.toml.orig +57 -0
  301. data/vendor/rb-sys/LICENSE-APACHE +190 -0
  302. data/vendor/rb-sys/LICENSE-MIT +21 -0
  303. data/vendor/rb-sys/bin/release.sh +21 -0
  304. data/vendor/rb-sys/build/features.rs +108 -0
  305. data/vendor/rb-sys/build/main.rs +246 -0
  306. data/vendor/rb-sys/build/stable_api_config.rs +153 -0
  307. data/vendor/rb-sys/build/version.rs +48 -0
  308. data/vendor/rb-sys/readme.md +36 -0
  309. data/vendor/rb-sys/src/bindings.rs +21 -0
  310. data/vendor/rb-sys/src/hidden.rs +11 -0
  311. data/vendor/rb-sys/src/lib.rs +34 -0
  312. data/vendor/rb-sys/src/macros.rs +371 -0
  313. data/vendor/rb-sys/src/memory.rs +53 -0
  314. data/vendor/rb-sys/src/ruby_abi_version.rs +38 -0
  315. data/vendor/rb-sys/src/special_consts.rs +31 -0
  316. data/vendor/rb-sys/src/stable_api/compiled.c +179 -0
  317. data/vendor/rb-sys/src/stable_api/compiled.rs +257 -0
  318. data/vendor/rb-sys/src/stable_api/ruby_2_6.rs +316 -0
  319. data/vendor/rb-sys/src/stable_api/ruby_2_7.rs +316 -0
  320. data/vendor/rb-sys/src/stable_api/ruby_3_0.rs +324 -0
  321. data/vendor/rb-sys/src/stable_api/ruby_3_1.rs +317 -0
  322. data/vendor/rb-sys/src/stable_api/ruby_3_2.rs +315 -0
  323. data/vendor/rb-sys/src/stable_api/ruby_3_3.rs +326 -0
  324. data/vendor/rb-sys/src/stable_api/ruby_3_4.rs +327 -0
  325. data/vendor/rb-sys/src/stable_api.rs +261 -0
  326. data/vendor/rb-sys/src/symbol.rs +31 -0
  327. data/vendor/rb-sys/src/tracking_allocator.rs +332 -0
  328. data/vendor/rb-sys/src/utils.rs +89 -0
  329. data/vendor/rb-sys/src/value_type.rs +7 -0
  330. metadata +90 -95
  331. data/pkg/kreuzberg-4.0.0.rc1.gem +0 -0
  332. data/spec/examples.txt +0 -104
  333. data/vendor/kreuzberg/src/bin/profile_extract.rs +0 -455
  334. data/vendor/kreuzberg/src/extraction/pandoc/batch.rs +0 -275
  335. data/vendor/kreuzberg/src/extraction/pandoc/mime_types.rs +0 -178
  336. data/vendor/kreuzberg/src/extraction/pandoc/mod.rs +0 -491
  337. data/vendor/kreuzberg/src/extraction/pandoc/server.rs +0 -496
  338. data/vendor/kreuzberg/src/extraction/pandoc/subprocess.rs +0 -1188
  339. data/vendor/kreuzberg/src/extraction/pandoc/version.rs +0 -162
  340. data/vendor/kreuzberg/src/extractors/pandoc.rs +0 -201
  341. data/vendor/kreuzberg/tests/chunking_offset_demo.rs +0 -92
  342. data/vendor/kreuzberg/tests/pandoc_integration.rs +0 -503
@@ -1,955 +1,956 @@
1
- //! Validator plugin trait.
2
- //!
3
- //! This module defines the trait for implementing custom validation logic.
4
-
5
- use crate::Result;
6
- use crate::core::config::ExtractionConfig;
7
- use crate::plugins::Plugin;
8
- use crate::types::ExtractionResult;
9
- use async_trait::async_trait;
10
- use std::sync::Arc;
11
-
12
- /// Trait for validator plugins.
13
- ///
14
- /// Validators check extraction results for quality, completeness, or correctness.
15
- /// Unlike post-processors, validator errors **fail fast** - if a validator returns
16
- /// an error, the extraction fails immediately.
17
- ///
18
- /// # Use Cases
19
- ///
20
- /// - **Quality Gates**: Ensure extracted content meets minimum quality standards
21
- /// - **Compliance**: Verify content meets regulatory requirements
22
- /// - **Content Filtering**: Reject documents containing unwanted content
23
- /// - **Format Validation**: Verify extracted content structure
24
- /// - **Security Checks**: Scan for malicious content
25
- ///
26
- /// # Error Handling
27
- ///
28
- /// Validator errors are **fatal** - they cause the extraction to fail and bubble up
29
- /// to the caller. Use validators for hard requirements that must be met.
30
- ///
31
- /// For non-fatal checks, use post-processors instead.
32
- ///
33
- /// # Thread Safety
34
- ///
35
- /// Validators must be thread-safe (`Send + Sync`).
36
- ///
37
- /// # Example
38
- ///
39
- /// ```rust
40
- /// use kreuzberg::plugins::{Plugin, Validator};
41
- /// use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
42
- /// use async_trait::async_trait;
43
- ///
44
- /// /// Validate that extracted content has minimum length
45
- /// struct MinimumLengthValidator {
46
- /// min_length: usize,
47
- /// }
48
- ///
49
- /// impl Plugin for MinimumLengthValidator {
50
- /// fn name(&self) -> &str { "min-length-validator" }
51
- /// fn version(&self) -> String { "1.0.0".to_string() }
52
- /// fn initialize(&self) -> Result<()> { Ok(()) }
53
- /// fn shutdown(&self) -> Result<()> { Ok(()) }
54
- /// }
55
- ///
56
- /// #[async_trait]
57
- /// impl Validator for MinimumLengthValidator {
58
- /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
59
- /// -> Result<()> {
60
- /// if result.content.len() < self.min_length {
61
- /// return Err(KreuzbergError::validation(format!(
62
- /// "Content too short: {} < {} characters",
63
- /// result.content.len(),
64
- /// self.min_length
65
- /// )));
66
- /// }
67
- /// Ok(())
68
- /// }
69
- /// }
70
- /// ```
71
- #[async_trait]
72
- pub trait Validator: Plugin {
73
- /// Validate an extraction result.
74
- ///
75
- /// Check the extraction result and return `Ok(())` if valid, or an error
76
- /// if validation fails.
77
- ///
78
- /// # Arguments
79
- ///
80
- /// * `result` - The extraction result to validate
81
- /// * `config` - Extraction configuration
82
- ///
83
- /// # Returns
84
- ///
85
- /// - `Ok(())` if validation passes
86
- /// - `Err(...)` if validation fails (extraction will fail)
87
- ///
88
- /// # Errors
89
- ///
90
- /// - `KreuzbergError::Validation` - Validation failed
91
- /// - Any other error type appropriate for the failure
92
- ///
93
- /// # Example - Content Length Validation
94
- ///
95
- /// ```rust
96
- /// # use kreuzberg::plugins::{Plugin, Validator};
97
- /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
98
- /// # use async_trait::async_trait;
99
- /// # struct ContentLengthValidator { min: usize, max: usize }
100
- /// # impl Plugin for ContentLengthValidator {
101
- /// # fn name(&self) -> &str { "length-validator" }
102
- /// # fn version(&self) -> String { "1.0.0".to_string() }
103
- /// # fn initialize(&self) -> Result<()> { Ok(()) }
104
- /// # fn shutdown(&self) -> Result<()> { Ok(()) }
105
- /// # }
106
- /// # #[async_trait]
107
- /// # impl Validator for ContentLengthValidator {
108
- /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
109
- /// -> Result<()> {
110
- /// let length = result.content.len();
111
- ///
112
- /// if length < self.min {
113
- /// return Err(KreuzbergError::validation(format!(
114
- /// "Content too short: {} < {} characters",
115
- /// length, self.min
116
- /// )));
117
- /// }
118
- ///
119
- /// if length > self.max {
120
- /// return Err(KreuzbergError::validation(format!(
121
- /// "Content too long: {} > {} characters",
122
- /// length, self.max
123
- /// )));
124
- /// }
125
- ///
126
- /// Ok(())
127
- /// }
128
- /// # }
129
- /// ```
130
- ///
131
- /// # Example - Quality Score Validation
132
- ///
133
- /// ```rust
134
- /// # use kreuzberg::plugins::{Plugin, Validator};
135
- /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
136
- /// # use async_trait::async_trait;
137
- /// # struct QualityValidator { min_score: f64 }
138
- /// # impl Plugin for QualityValidator {
139
- /// # fn name(&self) -> &str { "quality-validator" }
140
- /// # fn version(&self) -> String { "1.0.0".to_string() }
141
- /// # fn initialize(&self) -> Result<()> { Ok(()) }
142
- /// # fn shutdown(&self) -> Result<()> { Ok(()) }
143
- /// # }
144
- /// # #[async_trait]
145
- /// # impl Validator for QualityValidator {
146
- /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
147
- /// -> Result<()> {
148
- /// // Check if quality_score exists in metadata
149
- /// let score = result.metadata
150
- /// .additional
151
- /// .get("quality_score")
152
- /// .and_then(|v| v.as_f64())
153
- /// .unwrap_or(0.0);
154
- ///
155
- /// if score < self.min_score {
156
- /// return Err(KreuzbergError::validation(format!(
157
- /// "Quality score too low: {} < {}",
158
- /// score, self.min_score
159
- /// )));
160
- /// }
161
- ///
162
- /// Ok(())
163
- /// }
164
- /// # }
165
- /// ```
166
- ///
167
- /// # Example - Security Validation
168
- ///
169
- /// ```rust
170
- /// # use kreuzberg::plugins::{Plugin, Validator};
171
- /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
172
- /// # use async_trait::async_trait;
173
- /// # struct SecurityValidator { blocked_patterns: Vec<String> }
174
- /// # impl Plugin for SecurityValidator {
175
- /// # fn name(&self) -> &str { "security-validator" }
176
- /// # fn version(&self) -> String { "1.0.0".to_string() }
177
- /// # fn initialize(&self) -> Result<()> { Ok(()) }
178
- /// # fn shutdown(&self) -> Result<()> { Ok(()) }
179
- /// # }
180
- /// # #[async_trait]
181
- /// # impl Validator for SecurityValidator {
182
- /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
183
- /// -> Result<()> {
184
- /// // Check for blocked patterns
185
- /// for pattern in &self.blocked_patterns {
186
- /// if result.content.contains(pattern) {
187
- /// return Err(KreuzbergError::validation(format!(
188
- /// "Content contains blocked pattern: {}",
189
- /// pattern
190
- /// )));
191
- /// }
192
- /// }
193
- ///
194
- /// Ok(())
195
- /// }
196
- /// # }
197
- /// ```
198
- async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig) -> Result<()>;
199
-
200
- /// Optional: Check if this validator should run for a given result.
201
- ///
202
- /// Allows conditional validation based on MIME type, metadata, or content.
203
- /// Defaults to `true` (always run).
204
- ///
205
- /// # Arguments
206
- ///
207
- /// * `result` - The extraction result to check
208
- /// * `config` - Extraction configuration
209
- ///
210
- /// # Returns
211
- ///
212
- /// `true` if the validator should run, `false` to skip.
213
- ///
214
- /// # Example
215
- ///
216
- /// ```rust
217
- /// # use kreuzberg::plugins::{Plugin, Validator};
218
- /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig};
219
- /// # use async_trait::async_trait;
220
- /// # struct PdfValidator;
221
- /// # impl Plugin for PdfValidator {
222
- /// # fn name(&self) -> &str { "pdf-validator" }
223
- /// # fn version(&self) -> String { "1.0.0".to_string() }
224
- /// # fn initialize(&self) -> Result<()> { Ok(()) }
225
- /// # fn shutdown(&self) -> Result<()> { Ok(()) }
226
- /// # }
227
- /// # #[async_trait]
228
- /// # impl Validator for PdfValidator {
229
- /// # async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> { Ok(()) }
230
- /// /// Only validate PDF documents
231
- /// fn should_validate(&self, result: &ExtractionResult, config: &ExtractionConfig) -> bool {
232
- /// result.mime_type == "application/pdf"
233
- /// }
234
- /// # }
235
- /// ```
236
- fn should_validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> bool {
237
- true
238
- }
239
-
240
- /// Optional: Get the validation priority.
241
- ///
242
- /// Higher priority validators run first. Useful for ordering validation checks
243
- /// (e.g., run cheap validations before expensive ones).
244
- ///
245
- /// Default priority is 50.
246
- ///
247
- /// # Returns
248
- ///
249
- /// Priority value (higher = runs earlier).
250
- ///
251
- /// # Example
252
- ///
253
- /// ```rust
254
- /// # use kreuzberg::plugins::{Plugin, Validator};
255
- /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig};
256
- /// # use async_trait::async_trait;
257
- /// # struct FastValidator;
258
- /// # impl Plugin for FastValidator {
259
- /// # fn name(&self) -> &str { "fast-validator" }
260
- /// # fn version(&self) -> String { "1.0.0".to_string() }
261
- /// # fn initialize(&self) -> Result<()> { Ok(()) }
262
- /// # fn shutdown(&self) -> Result<()> { Ok(()) }
263
- /// # }
264
- /// # #[async_trait]
265
- /// # impl Validator for FastValidator {
266
- /// # async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> { Ok(()) }
267
- /// /// Run this validator first (it's fast)
268
- /// fn priority(&self) -> i32 {
269
- /// 100
270
- /// }
271
- /// # }
272
- /// ```
273
- fn priority(&self) -> i32 {
274
- 50
275
- }
276
- }
277
-
278
- // Public registration APIs
279
-
280
- /// Register a validator with the global registry.
281
- ///
282
- /// The validator will be registered with its default priority and will be called
283
- /// during extraction validation. The validator's `name()` method is used as the
284
- /// registration name.
285
- ///
286
- /// # Arguments
287
- ///
288
- /// * `validator` - The validator implementation wrapped in Arc
289
- ///
290
- /// # Returns
291
- ///
292
- /// - `Ok(())` if registration succeeded
293
- /// - `Err(...)` if validation failed or initialization failed
294
- ///
295
- /// # Errors
296
- ///
297
- /// - `KreuzbergError::Validation` - Invalid validator name (empty or contains whitespace)
298
- /// - Any error from the validator's `initialize()` method
299
- ///
300
- /// # Example
301
- ///
302
- /// ```rust
303
- /// use kreuzberg::plugins::{Plugin, Validator, register_validator};
304
- /// use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
305
- /// use async_trait::async_trait;
306
- /// use std::sync::Arc;
307
- ///
308
- /// struct MinLengthValidator { min_length: usize }
309
- ///
310
- /// impl Plugin for MinLengthValidator {
311
- /// fn name(&self) -> &str { "min-length" }
312
- /// fn version(&self) -> String { "1.0.0".to_string() }
313
- /// fn initialize(&self) -> Result<()> { Ok(()) }
314
- /// fn shutdown(&self) -> Result<()> { Ok(()) }
315
- /// }
316
- ///
317
- /// #[async_trait]
318
- /// impl Validator for MinLengthValidator {
319
- /// async fn validate(&self, result: &ExtractionResult, _: &ExtractionConfig) -> Result<()> {
320
- /// if result.content.len() < self.min_length {
321
- /// return Err(KreuzbergError::validation(
322
- /// format!("Content too short: {} < {}", result.content.len(), self.min_length)
323
- /// ));
324
- /// }
325
- /// Ok(())
326
- /// }
327
- /// }
328
- ///
329
- /// # tokio_test::block_on(async {
330
- /// let validator = Arc::new(MinLengthValidator { min_length: 10 });
331
- /// register_validator(validator)?;
332
- /// # Ok::<(), KreuzbergError>(())
333
- /// # });
334
- /// ```
335
- pub fn register_validator(validator: Arc<dyn Validator>) -> crate::Result<()> {
336
- use crate::plugins::registry::get_validator_registry;
337
-
338
- let registry = get_validator_registry();
339
- let mut registry = registry
340
- .write()
341
- .expect("~keep Failed to acquire write lock on validator registry"); // ~keep
342
-
343
- registry.register(validator)
344
- }
345
-
346
- /// Unregister a validator by name.
347
- ///
348
- /// Removes the validator from the global registry and calls its `shutdown()` method.
349
- ///
350
- /// # Arguments
351
- ///
352
- /// * `name` - Name of the validator to unregister
353
- ///
354
- /// # Returns
355
- ///
356
- /// - `Ok(())` if the validator was unregistered or didn't exist
357
- /// - `Err(...)` if the shutdown method failed
358
- ///
359
- /// # Example
360
- ///
361
- /// ```rust
362
- /// use kreuzberg::plugins::unregister_validator;
363
- ///
364
- /// # tokio_test::block_on(async {
365
- /// unregister_validator("min-length")?;
366
- /// # Ok::<(), kreuzberg::KreuzbergError>(())
367
- /// # });
368
- /// ```
369
- pub fn unregister_validator(name: &str) -> crate::Result<()> {
370
- use crate::plugins::registry::get_validator_registry;
371
-
372
- let registry = get_validator_registry();
373
- let mut registry = registry
374
- .write()
375
- .expect("~keep Failed to acquire write lock on validator registry"); // ~keep
376
-
377
- registry.remove(name)
378
- }
379
-
380
- /// List all registered validators.
381
- ///
382
- /// Returns the names of all validators currently registered in the global registry.
383
- ///
384
- /// # Returns
385
- ///
386
- /// A vector of validator names.
387
- ///
388
- /// # Example
389
- ///
390
- /// ```rust
391
- /// use kreuzberg::plugins::list_validators;
392
- ///
393
- /// # tokio_test::block_on(async {
394
- /// let validators = list_validators()?;
395
- /// for name in validators {
396
- /// println!("Registered validator: {}", name);
397
- /// }
398
- /// # Ok::<(), kreuzberg::KreuzbergError>(())
399
- /// # });
400
- /// ```
401
- pub fn list_validators() -> crate::Result<Vec<String>> {
402
- use crate::plugins::registry::get_validator_registry;
403
-
404
- let registry = get_validator_registry();
405
- let registry = registry
406
- .read()
407
- .expect("~keep Failed to acquire read lock on validator registry"); // ~keep
408
-
409
- Ok(registry.list())
410
- }
411
-
412
- /// Clear all validators from the global registry.
413
- ///
414
- /// Removes all validators and calls their `shutdown()` methods.
415
- ///
416
- /// # Returns
417
- ///
418
- /// - `Ok(())` if all validators were cleared successfully
419
- /// - `Err(...)` if any shutdown method failed
420
- ///
421
- /// # Example
422
- ///
423
- /// ```rust
424
- /// use kreuzberg::plugins::clear_validators;
425
- ///
426
- /// # tokio_test::block_on(async {
427
- /// clear_validators()?;
428
- /// # Ok::<(), kreuzberg::KreuzbergError>(())
429
- /// # });
430
- /// ```
431
- pub fn clear_validators() -> crate::Result<()> {
432
- use crate::plugins::registry::get_validator_registry;
433
-
434
- let registry = get_validator_registry();
435
- let mut registry = registry
436
- .write()
437
- .expect("~keep Failed to acquire write lock on validator registry"); // ~keep
438
-
439
- registry.shutdown_all()
440
- }
441
-
442
- #[cfg(test)]
443
- mod tests {
444
- use super::*;
445
- use crate::KreuzbergError;
446
- use std::collections::HashMap;
447
-
448
- struct MockValidator {
449
- should_fail: bool,
450
- }
451
-
452
- impl Plugin for MockValidator {
453
- fn name(&self) -> &str {
454
- "mock-validator"
455
- }
456
-
457
- fn version(&self) -> String {
458
- "1.0.0".to_string()
459
- }
460
-
461
- fn initialize(&self) -> Result<()> {
462
- Ok(())
463
- }
464
-
465
- fn shutdown(&self) -> Result<()> {
466
- Ok(())
467
- }
468
- }
469
-
470
- #[async_trait]
471
- impl Validator for MockValidator {
472
- async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
473
- if self.should_fail {
474
- Err(KreuzbergError::validation("Validation failed".to_string()))
475
- } else {
476
- Ok(())
477
- }
478
- }
479
- }
480
-
481
- #[tokio::test]
482
- async fn test_validator_success() {
483
- let validator = MockValidator { should_fail: false };
484
-
485
- let result = ExtractionResult {
486
- content: "test content".to_string(),
487
- mime_type: "text/plain".to_string(),
488
- metadata: crate::types::Metadata::default(),
489
- tables: vec![],
490
- detected_languages: None,
491
- chunks: None,
492
- images: None,
493
- };
494
-
495
- let config = ExtractionConfig::default();
496
- assert!(validator.validate(&result, &config).await.is_ok());
497
- }
498
-
499
- #[tokio::test]
500
- async fn test_validator_failure() {
501
- let validator = MockValidator { should_fail: true };
502
-
503
- let result = ExtractionResult {
504
- content: "test content".to_string(),
505
- mime_type: "text/plain".to_string(),
506
- metadata: crate::types::Metadata::default(),
507
- tables: vec![],
508
- detected_languages: None,
509
- chunks: None,
510
- images: None,
511
- };
512
-
513
- let config = ExtractionConfig::default();
514
- let validation_result = validator.validate(&result, &config).await;
515
-
516
- assert!(matches!(validation_result, Err(KreuzbergError::Validation { .. })));
517
- }
518
-
519
- #[test]
520
- fn test_validator_should_validate_default() {
521
- let validator = MockValidator { should_fail: false };
522
-
523
- let result = ExtractionResult {
524
- content: "test".to_string(),
525
- mime_type: "text/plain".to_string(),
526
- metadata: crate::types::Metadata::default(),
527
- tables: vec![],
528
- detected_languages: None,
529
- chunks: None,
530
- images: None,
531
- };
532
-
533
- let config = ExtractionConfig::default();
534
-
535
- assert!(validator.should_validate(&result, &config));
536
- }
537
-
538
- #[test]
539
- fn test_validator_priority_default() {
540
- let validator = MockValidator { should_fail: false };
541
- assert_eq!(validator.priority(), 50);
542
- }
543
-
544
- #[tokio::test]
545
- async fn test_validator_plugin_interface() {
546
- let validator = MockValidator { should_fail: false };
547
-
548
- assert_eq!(validator.name(), "mock-validator");
549
- assert_eq!(validator.version(), "1.0.0");
550
- assert!(validator.initialize().is_ok());
551
- assert!(validator.shutdown().is_ok());
552
- }
553
-
554
- #[tokio::test]
555
- async fn test_validator_empty_content() {
556
- let validator = MockValidator { should_fail: false };
557
-
558
- let result = ExtractionResult {
559
- content: String::new(),
560
- mime_type: "text/plain".to_string(),
561
- metadata: crate::types::Metadata::default(),
562
- tables: vec![],
563
- detected_languages: None,
564
- chunks: None,
565
- images: None,
566
- };
567
-
568
- let config = ExtractionConfig::default();
569
- assert!(validator.validate(&result, &config).await.is_ok());
570
- }
571
-
572
- #[test]
573
- fn test_validator_should_validate_conditional() {
574
- struct PdfOnlyValidator;
575
-
576
- impl Plugin for PdfOnlyValidator {
577
- fn name(&self) -> &str {
578
- "pdf-only"
579
- }
580
- fn version(&self) -> String {
581
- "1.0.0".to_string()
582
- }
583
- fn initialize(&self) -> Result<()> {
584
- Ok(())
585
- }
586
- fn shutdown(&self) -> Result<()> {
587
- Ok(())
588
- }
589
- }
590
-
591
- #[async_trait]
592
- impl Validator for PdfOnlyValidator {
593
- async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
594
- Ok(())
595
- }
596
-
597
- fn should_validate(&self, result: &ExtractionResult, _config: &ExtractionConfig) -> bool {
598
- result.mime_type == "application/pdf"
599
- }
600
- }
601
-
602
- let validator = PdfOnlyValidator;
603
- let config = ExtractionConfig::default();
604
-
605
- let pdf_result = ExtractionResult {
606
- content: "test".to_string(),
607
- mime_type: "application/pdf".to_string(),
608
- metadata: crate::types::Metadata::default(),
609
- tables: vec![],
610
- detected_languages: None,
611
- chunks: None,
612
- images: None,
613
- };
614
-
615
- let txt_result = ExtractionResult {
616
- content: "test".to_string(),
617
- mime_type: "text/plain".to_string(),
618
- metadata: crate::types::Metadata::default(),
619
- tables: vec![],
620
- detected_languages: None,
621
- chunks: None,
622
- images: None,
623
- };
624
-
625
- assert!(validator.should_validate(&pdf_result, &config));
626
- assert!(!validator.should_validate(&txt_result, &config));
627
- }
628
-
629
- #[test]
630
- fn test_validator_priority_ranges() {
631
- struct HighPriorityValidator;
632
- struct LowPriorityValidator;
633
-
634
- impl Plugin for HighPriorityValidator {
635
- fn name(&self) -> &str {
636
- "high-priority"
637
- }
638
- fn version(&self) -> String {
639
- "1.0.0".to_string()
640
- }
641
- fn initialize(&self) -> Result<()> {
642
- Ok(())
643
- }
644
- fn shutdown(&self) -> Result<()> {
645
- Ok(())
646
- }
647
- }
648
-
649
- impl Plugin for LowPriorityValidator {
650
- fn name(&self) -> &str {
651
- "low-priority"
652
- }
653
- fn version(&self) -> String {
654
- "1.0.0".to_string()
655
- }
656
- fn initialize(&self) -> Result<()> {
657
- Ok(())
658
- }
659
- fn shutdown(&self) -> Result<()> {
660
- Ok(())
661
- }
662
- }
663
-
664
- #[async_trait]
665
- impl Validator for HighPriorityValidator {
666
- async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
667
- Ok(())
668
- }
669
-
670
- fn priority(&self) -> i32 {
671
- 100
672
- }
673
- }
674
-
675
- #[async_trait]
676
- impl Validator for LowPriorityValidator {
677
- async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
678
- Ok(())
679
- }
680
-
681
- fn priority(&self) -> i32 {
682
- 10
683
- }
684
- }
685
-
686
- let high = HighPriorityValidator;
687
- let low = LowPriorityValidator;
688
-
689
- assert_eq!(high.priority(), 100);
690
- assert_eq!(low.priority(), 10);
691
- assert!(high.priority() > low.priority());
692
- }
693
-
694
- #[tokio::test]
695
- async fn test_validator_error_message() {
696
- let validator = MockValidator { should_fail: true };
697
-
698
- let result = ExtractionResult {
699
- content: "test".to_string(),
700
- mime_type: "text/plain".to_string(),
701
- metadata: crate::types::Metadata::default(),
702
- tables: vec![],
703
- detected_languages: None,
704
- chunks: None,
705
- images: None,
706
- };
707
-
708
- let config = ExtractionConfig::default();
709
- let err = validator.validate(&result, &config).await.unwrap_err();
710
-
711
- match err {
712
- KreuzbergError::Validation { message: msg, .. } => {
713
- assert_eq!(msg, "Validation failed");
714
- }
715
- _ => panic!("Expected Validation error"),
716
- }
717
- }
718
-
719
- #[tokio::test]
720
- async fn test_validator_with_metadata() {
721
- let validator = MockValidator { should_fail: false };
722
-
723
- let mut additional = HashMap::new();
724
- additional.insert("quality_score".to_string(), serde_json::json!(0.95));
725
-
726
- let result = ExtractionResult {
727
- content: "test".to_string(),
728
- mime_type: "text/plain".to_string(),
729
- metadata: crate::types::Metadata {
730
- additional,
731
- ..Default::default()
732
- },
733
- tables: vec![],
734
- detected_languages: None,
735
- chunks: None,
736
- images: None,
737
- };
738
-
739
- let config = ExtractionConfig::default();
740
- assert!(validator.validate(&result, &config).await.is_ok());
741
- }
742
-
743
- #[tokio::test]
744
- async fn test_validator_with_tables() {
745
- use crate::types::Table;
746
-
747
- let validator = MockValidator { should_fail: false };
748
-
749
- let table = Table {
750
- cells: vec![vec!["A".to_string(), "B".to_string()]],
751
- markdown: "| A | B |".to_string(),
752
- page_number: 0,
753
- };
754
-
755
- let result = ExtractionResult {
756
- content: "test".to_string(),
757
- mime_type: "text/plain".to_string(),
758
- metadata: crate::types::Metadata::default(),
759
- tables: vec![table],
760
- detected_languages: None,
761
- chunks: None,
762
- images: None,
763
- };
764
-
765
- let config = ExtractionConfig::default();
766
- assert!(validator.validate(&result, &config).await.is_ok());
767
- }
768
-
769
- #[tokio::test]
770
- async fn test_validator_different_mime_types() {
771
- let validator = MockValidator { should_fail: false };
772
- let config = ExtractionConfig::default();
773
-
774
- let mime_types = vec![
775
- "text/plain",
776
- "application/pdf",
777
- "application/json",
778
- "text/html",
779
- "image/png",
780
- ];
781
-
782
- for mime_type in mime_types {
783
- let result = ExtractionResult {
784
- content: "test".to_string(),
785
- mime_type: mime_type.to_string(),
786
- metadata: crate::types::Metadata::default(),
787
- tables: vec![],
788
- detected_languages: None,
789
- chunks: None,
790
- images: None,
791
- };
792
-
793
- assert!(validator.validate(&result, &config).await.is_ok());
794
- }
795
- }
796
-
797
- #[tokio::test]
798
- async fn test_validator_long_content() {
799
- let validator = MockValidator { should_fail: false };
800
-
801
- let result = ExtractionResult {
802
- content: "test content ".repeat(10000),
803
- mime_type: "text/plain".to_string(),
804
- metadata: crate::types::Metadata::default(),
805
- tables: vec![],
806
- detected_languages: None,
807
- chunks: None,
808
- images: None,
809
- };
810
-
811
- let config = ExtractionConfig::default();
812
- assert!(validator.validate(&result, &config).await.is_ok());
813
- }
814
-
815
- // Tests for public registration APIs
816
-
817
- #[test]
818
- fn test_register_validator() {
819
- use std::sync::Arc;
820
-
821
- let validator = Arc::new(MockValidator { should_fail: false });
822
- let result = super::register_validator(validator);
823
- assert!(result.is_ok());
824
-
825
- let _ = super::unregister_validator("mock-validator");
826
- }
827
-
828
- #[test]
829
- fn test_unregister_validator() {
830
- use std::sync::Arc;
831
-
832
- let validator = Arc::new(MockValidator { should_fail: false });
833
- super::register_validator(validator).unwrap();
834
-
835
- let result = super::unregister_validator("mock-validator");
836
- assert!(result.is_ok());
837
- }
838
-
839
- #[test]
840
- fn test_unregister_nonexistent_validator() {
841
- let result = super::unregister_validator("nonexistent-validator-xyz");
842
- assert!(result.is_ok());
843
- }
844
-
845
- #[test]
846
- fn test_list_validators() {
847
- use std::sync::Arc;
848
-
849
- super::clear_validators().unwrap();
850
-
851
- let validator1 = Arc::new(MockValidator { should_fail: false });
852
- // Both validators have the same name, so only one will be registered
853
- let validator2 = Arc::new(MockValidator { should_fail: false });
854
-
855
- let list_before = super::list_validators().unwrap();
856
- assert_eq!(list_before.len(), 0);
857
-
858
- super::register_validator(validator1).unwrap();
859
- super::register_validator(validator2).unwrap();
860
-
861
- let list = super::list_validators().unwrap();
862
- // Only 1 validator registered since they have the same name
863
- assert_eq!(list.len(), 1);
864
- assert!(list.contains(&"mock-validator".to_string()));
865
-
866
- super::unregister_validator("mock-validator").unwrap();
867
- }
868
-
869
- #[test]
870
- fn test_clear_validators() {
871
- use std::sync::Arc;
872
-
873
- super::clear_validators().unwrap();
874
-
875
- let validator1 = Arc::new(MockValidator { should_fail: false });
876
- let validator2 = Arc::new(MockValidator { should_fail: false });
877
-
878
- super::register_validator(validator1).unwrap();
879
- super::register_validator(validator2).unwrap();
880
-
881
- // Verify at least one validator is registered
882
- let list_before = super::list_validators().unwrap();
883
- assert!(!list_before.is_empty());
884
-
885
- let result = super::clear_validators();
886
- assert!(result.is_ok());
887
-
888
- let list = super::list_validators().unwrap();
889
- assert_eq!(list.len(), 0);
890
- }
891
-
892
- #[test]
893
- fn test_register_validator_with_invalid_name() {
894
- use std::sync::Arc;
895
-
896
- struct InvalidNameValidator;
897
- impl Plugin for InvalidNameValidator {
898
- fn name(&self) -> &str {
899
- "invalid name with spaces"
900
- }
901
- fn version(&self) -> String {
902
- "1.0.0".to_string()
903
- }
904
- fn initialize(&self) -> Result<()> {
905
- Ok(())
906
- }
907
- fn shutdown(&self) -> Result<()> {
908
- Ok(())
909
- }
910
- }
911
-
912
- #[async_trait]
913
- impl Validator for InvalidNameValidator {
914
- async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> {
915
- Ok(())
916
- }
917
- }
918
-
919
- let validator = Arc::new(InvalidNameValidator);
920
- let result = super::register_validator(validator);
921
- assert!(matches!(result, Err(KreuzbergError::Validation { .. })));
922
- }
923
-
924
- #[test]
925
- fn test_register_validator_with_empty_name() {
926
- use std::sync::Arc;
927
-
928
- struct EmptyNameValidator;
929
- impl Plugin for EmptyNameValidator {
930
- fn name(&self) -> &str {
931
- ""
932
- }
933
- fn version(&self) -> String {
934
- "1.0.0".to_string()
935
- }
936
- fn initialize(&self) -> Result<()> {
937
- Ok(())
938
- }
939
- fn shutdown(&self) -> Result<()> {
940
- Ok(())
941
- }
942
- }
943
-
944
- #[async_trait]
945
- impl Validator for EmptyNameValidator {
946
- async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> {
947
- Ok(())
948
- }
949
- }
950
-
951
- let validator = Arc::new(EmptyNameValidator);
952
- let result = super::register_validator(validator);
953
- assert!(matches!(result, Err(KreuzbergError::Validation { .. })));
954
- }
955
- }
1
+ //! Validator plugin trait.
2
+ //!
3
+ //! This module defines the trait for implementing custom validation logic.
4
+
5
+ use crate::Result;
6
+ use crate::core::config::ExtractionConfig;
7
+ use crate::plugins::Plugin;
8
+ use crate::types::ExtractionResult;
9
+ use async_trait::async_trait;
10
+ use std::sync::Arc;
11
+
12
+ /// Trait for validator plugins.
13
+ ///
14
+ /// Validators check extraction results for quality, completeness, or correctness.
15
+ /// Unlike post-processors, validator errors **fail fast** - if a validator returns
16
+ /// an error, the extraction fails immediately.
17
+ ///
18
+ /// # Use Cases
19
+ ///
20
+ /// - **Quality Gates**: Ensure extracted content meets minimum quality standards
21
+ /// - **Compliance**: Verify content meets regulatory requirements
22
+ /// - **Content Filtering**: Reject documents containing unwanted content
23
+ /// - **Format Validation**: Verify extracted content structure
24
+ /// - **Security Checks**: Scan for malicious content
25
+ ///
26
+ /// # Error Handling
27
+ ///
28
+ /// Validator errors are **fatal** - they cause the extraction to fail and bubble up
29
+ /// to the caller. Use validators for hard requirements that must be met.
30
+ ///
31
+ /// For non-fatal checks, use post-processors instead.
32
+ ///
33
+ /// # Thread Safety
34
+ ///
35
+ /// Validators must be thread-safe (`Send + Sync`).
36
+ ///
37
+ /// # Example
38
+ ///
39
+ /// ```rust
40
+ /// use kreuzberg::plugins::{Plugin, Validator};
41
+ /// use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
42
+ /// use async_trait::async_trait;
43
+ ///
44
+ /// /// Validate that extracted content has minimum length
45
+ /// struct MinimumLengthValidator {
46
+ /// min_length: usize,
47
+ /// }
48
+ ///
49
+ /// impl Plugin for MinimumLengthValidator {
50
+ /// fn name(&self) -> &str { "min-length-validator" }
51
+ /// fn version(&self) -> String { "1.0.0".to_string() }
52
+ /// fn initialize(&self) -> Result<()> { Ok(()) }
53
+ /// fn shutdown(&self) -> Result<()> { Ok(()) }
54
+ /// }
55
+ ///
56
+ /// #[async_trait]
57
+ /// impl Validator for MinimumLengthValidator {
58
+ /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
59
+ /// -> Result<()> {
60
+ /// if result.content.len() < self.min_length {
61
+ /// return Err(KreuzbergError::validation(format!(
62
+ /// "Content too short: {} < {} characters",
63
+ /// result.content.len(),
64
+ /// self.min_length
65
+ /// )));
66
+ /// }
67
+ /// Ok(())
68
+ /// }
69
+ /// }
70
+ /// ```
71
+ #[cfg_attr(not(target_arch = "wasm32"), async_trait)]
72
+ #[cfg_attr(target_arch = "wasm32", async_trait(?Send))]
73
+ pub trait Validator: Plugin {
74
+ /// Validate an extraction result.
75
+ ///
76
+ /// Check the extraction result and return `Ok(())` if valid, or an error
77
+ /// if validation fails.
78
+ ///
79
+ /// # Arguments
80
+ ///
81
+ /// * `result` - The extraction result to validate
82
+ /// * `config` - Extraction configuration
83
+ ///
84
+ /// # Returns
85
+ ///
86
+ /// - `Ok(())` if validation passes
87
+ /// - `Err(...)` if validation fails (extraction will fail)
88
+ ///
89
+ /// # Errors
90
+ ///
91
+ /// - `KreuzbergError::Validation` - Validation failed
92
+ /// - Any other error type appropriate for the failure
93
+ ///
94
+ /// # Example - Content Length Validation
95
+ ///
96
+ /// ```rust
97
+ /// # use kreuzberg::plugins::{Plugin, Validator};
98
+ /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
99
+ /// # use async_trait::async_trait;
100
+ /// # struct ContentLengthValidator { min: usize, max: usize }
101
+ /// # impl Plugin for ContentLengthValidator {
102
+ /// # fn name(&self) -> &str { "length-validator" }
103
+ /// # fn version(&self) -> String { "1.0.0".to_string() }
104
+ /// # fn initialize(&self) -> Result<()> { Ok(()) }
105
+ /// # fn shutdown(&self) -> Result<()> { Ok(()) }
106
+ /// # }
107
+ /// # #[async_trait]
108
+ /// # impl Validator for ContentLengthValidator {
109
+ /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
110
+ /// -> Result<()> {
111
+ /// let length = result.content.len();
112
+ ///
113
+ /// if length < self.min {
114
+ /// return Err(KreuzbergError::validation(format!(
115
+ /// "Content too short: {} < {} characters",
116
+ /// length, self.min
117
+ /// )));
118
+ /// }
119
+ ///
120
+ /// if length > self.max {
121
+ /// return Err(KreuzbergError::validation(format!(
122
+ /// "Content too long: {} > {} characters",
123
+ /// length, self.max
124
+ /// )));
125
+ /// }
126
+ ///
127
+ /// Ok(())
128
+ /// }
129
+ /// # }
130
+ /// ```
131
+ ///
132
+ /// # Example - Quality Score Validation
133
+ ///
134
+ /// ```rust
135
+ /// # use kreuzberg::plugins::{Plugin, Validator};
136
+ /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
137
+ /// # use async_trait::async_trait;
138
+ /// # struct QualityValidator { min_score: f64 }
139
+ /// # impl Plugin for QualityValidator {
140
+ /// # fn name(&self) -> &str { "quality-validator" }
141
+ /// # fn version(&self) -> String { "1.0.0".to_string() }
142
+ /// # fn initialize(&self) -> Result<()> { Ok(()) }
143
+ /// # fn shutdown(&self) -> Result<()> { Ok(()) }
144
+ /// # }
145
+ /// # #[async_trait]
146
+ /// # impl Validator for QualityValidator {
147
+ /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
148
+ /// -> Result<()> {
149
+ /// // Check if quality_score exists in metadata
150
+ /// let score = result.metadata
151
+ /// .additional
152
+ /// .get("quality_score")
153
+ /// .and_then(|v| v.as_f64())
154
+ /// .unwrap_or(0.0);
155
+ ///
156
+ /// if score < self.min_score {
157
+ /// return Err(KreuzbergError::validation(format!(
158
+ /// "Quality score too low: {} < {}",
159
+ /// score, self.min_score
160
+ /// )));
161
+ /// }
162
+ ///
163
+ /// Ok(())
164
+ /// }
165
+ /// # }
166
+ /// ```
167
+ ///
168
+ /// # Example - Security Validation
169
+ ///
170
+ /// ```rust
171
+ /// # use kreuzberg::plugins::{Plugin, Validator};
172
+ /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
173
+ /// # use async_trait::async_trait;
174
+ /// # struct SecurityValidator { blocked_patterns: Vec<String> }
175
+ /// # impl Plugin for SecurityValidator {
176
+ /// # fn name(&self) -> &str { "security-validator" }
177
+ /// # fn version(&self) -> String { "1.0.0".to_string() }
178
+ /// # fn initialize(&self) -> Result<()> { Ok(()) }
179
+ /// # fn shutdown(&self) -> Result<()> { Ok(()) }
180
+ /// # }
181
+ /// # #[async_trait]
182
+ /// # impl Validator for SecurityValidator {
183
+ /// async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig)
184
+ /// -> Result<()> {
185
+ /// // Check for blocked patterns
186
+ /// for pattern in &self.blocked_patterns {
187
+ /// if result.content.contains(pattern) {
188
+ /// return Err(KreuzbergError::validation(format!(
189
+ /// "Content contains blocked pattern: {}",
190
+ /// pattern
191
+ /// )));
192
+ /// }
193
+ /// }
194
+ ///
195
+ /// Ok(())
196
+ /// }
197
+ /// # }
198
+ /// ```
199
+ async fn validate(&self, result: &ExtractionResult, config: &ExtractionConfig) -> Result<()>;
200
+
201
+ /// Optional: Check if this validator should run for a given result.
202
+ ///
203
+ /// Allows conditional validation based on MIME type, metadata, or content.
204
+ /// Defaults to `true` (always run).
205
+ ///
206
+ /// # Arguments
207
+ ///
208
+ /// * `result` - The extraction result to check
209
+ /// * `config` - Extraction configuration
210
+ ///
211
+ /// # Returns
212
+ ///
213
+ /// `true` if the validator should run, `false` to skip.
214
+ ///
215
+ /// # Example
216
+ ///
217
+ /// ```rust
218
+ /// # use kreuzberg::plugins::{Plugin, Validator};
219
+ /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig};
220
+ /// # use async_trait::async_trait;
221
+ /// # struct PdfValidator;
222
+ /// # impl Plugin for PdfValidator {
223
+ /// # fn name(&self) -> &str { "pdf-validator" }
224
+ /// # fn version(&self) -> String { "1.0.0".to_string() }
225
+ /// # fn initialize(&self) -> Result<()> { Ok(()) }
226
+ /// # fn shutdown(&self) -> Result<()> { Ok(()) }
227
+ /// # }
228
+ /// # #[async_trait]
229
+ /// # impl Validator for PdfValidator {
230
+ /// # async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> { Ok(()) }
231
+ /// /// Only validate PDF documents
232
+ /// fn should_validate(&self, result: &ExtractionResult, config: &ExtractionConfig) -> bool {
233
+ /// result.mime_type == "application/pdf"
234
+ /// }
235
+ /// # }
236
+ /// ```
237
+ fn should_validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> bool {
238
+ true
239
+ }
240
+
241
+ /// Optional: Get the validation priority.
242
+ ///
243
+ /// Higher priority validators run first. Useful for ordering validation checks
244
+ /// (e.g., run cheap validations before expensive ones).
245
+ ///
246
+ /// Default priority is 50.
247
+ ///
248
+ /// # Returns
249
+ ///
250
+ /// Priority value (higher = runs earlier).
251
+ ///
252
+ /// # Example
253
+ ///
254
+ /// ```rust
255
+ /// # use kreuzberg::plugins::{Plugin, Validator};
256
+ /// # use kreuzberg::{Result, ExtractionResult, ExtractionConfig};
257
+ /// # use async_trait::async_trait;
258
+ /// # struct FastValidator;
259
+ /// # impl Plugin for FastValidator {
260
+ /// # fn name(&self) -> &str { "fast-validator" }
261
+ /// # fn version(&self) -> String { "1.0.0".to_string() }
262
+ /// # fn initialize(&self) -> Result<()> { Ok(()) }
263
+ /// # fn shutdown(&self) -> Result<()> { Ok(()) }
264
+ /// # }
265
+ /// # #[async_trait]
266
+ /// # impl Validator for FastValidator {
267
+ /// # async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> { Ok(()) }
268
+ /// /// Run this validator first (it's fast)
269
+ /// fn priority(&self) -> i32 {
270
+ /// 100
271
+ /// }
272
+ /// # }
273
+ /// ```
274
+ fn priority(&self) -> i32 {
275
+ 50
276
+ }
277
+ }
278
+
279
+ /// Register a validator with the global registry.
280
+ ///
281
+ /// The validator will be registered with its default priority and will be called
282
+ /// during extraction validation. The validator's `name()` method is used as the
283
+ /// registration name.
284
+ ///
285
+ /// # Arguments
286
+ ///
287
+ /// * `validator` - The validator implementation wrapped in Arc
288
+ ///
289
+ /// # Returns
290
+ ///
291
+ /// - `Ok(())` if registration succeeded
292
+ /// - `Err(...)` if validation failed or initialization failed
293
+ ///
294
+ /// # Errors
295
+ ///
296
+ /// - `KreuzbergError::Validation` - Invalid validator name (empty or contains whitespace)
297
+ /// - Any error from the validator's `initialize()` method
298
+ ///
299
+ /// # Example
300
+ ///
301
+ /// ```rust
302
+ /// use kreuzberg::plugins::{Plugin, Validator, register_validator};
303
+ /// use kreuzberg::{Result, ExtractionResult, ExtractionConfig, KreuzbergError};
304
+ /// use async_trait::async_trait;
305
+ /// use std::sync::Arc;
306
+ ///
307
+ /// struct MinLengthValidator { min_length: usize }
308
+ ///
309
+ /// impl Plugin for MinLengthValidator {
310
+ /// fn name(&self) -> &str { "min-length" }
311
+ /// fn version(&self) -> String { "1.0.0".to_string() }
312
+ /// fn initialize(&self) -> Result<()> { Ok(()) }
313
+ /// fn shutdown(&self) -> Result<()> { Ok(()) }
314
+ /// }
315
+ ///
316
+ /// #[async_trait]
317
+ /// impl Validator for MinLengthValidator {
318
+ /// async fn validate(&self, result: &ExtractionResult, _: &ExtractionConfig) -> Result<()> {
319
+ /// if result.content.len() < self.min_length {
320
+ /// return Err(KreuzbergError::validation(
321
+ /// format!("Content too short: {} < {}", result.content.len(), self.min_length)
322
+ /// ));
323
+ /// }
324
+ /// Ok(())
325
+ /// }
326
+ /// }
327
+ ///
328
+ /// # tokio_test::block_on(async {
329
+ /// let validator = Arc::new(MinLengthValidator { min_length: 10 });
330
+ /// register_validator(validator)?;
331
+ /// # Ok::<(), KreuzbergError>(())
332
+ /// # });
333
+ /// ```
334
+ pub fn register_validator(validator: Arc<dyn Validator>) -> crate::Result<()> {
335
+ use crate::plugins::registry::get_validator_registry;
336
+
337
+ let registry = get_validator_registry();
338
+ let mut registry = registry
339
+ .write()
340
+ .expect("~keep Failed to acquire write lock on validator registry"); // ~keep
341
+
342
+ registry.register(validator)
343
+ }
344
+
345
+ /// Unregister a validator by name.
346
+ ///
347
+ /// Removes the validator from the global registry and calls its `shutdown()` method.
348
+ ///
349
+ /// # Arguments
350
+ ///
351
+ /// * `name` - Name of the validator to unregister
352
+ ///
353
+ /// # Returns
354
+ ///
355
+ /// - `Ok(())` if the validator was unregistered or didn't exist
356
+ /// - `Err(...)` if the shutdown method failed
357
+ ///
358
+ /// # Example
359
+ ///
360
+ /// ```rust
361
+ /// use kreuzberg::plugins::unregister_validator;
362
+ ///
363
+ /// # tokio_test::block_on(async {
364
+ /// unregister_validator("min-length")?;
365
+ /// # Ok::<(), kreuzberg::KreuzbergError>(())
366
+ /// # });
367
+ /// ```
368
+ pub fn unregister_validator(name: &str) -> crate::Result<()> {
369
+ use crate::plugins::registry::get_validator_registry;
370
+
371
+ let registry = get_validator_registry();
372
+ let mut registry = registry
373
+ .write()
374
+ .expect("~keep Failed to acquire write lock on validator registry"); // ~keep
375
+
376
+ registry.remove(name)
377
+ }
378
+
379
+ /// List all registered validators.
380
+ ///
381
+ /// Returns the names of all validators currently registered in the global registry.
382
+ ///
383
+ /// # Returns
384
+ ///
385
+ /// A vector of validator names.
386
+ ///
387
+ /// # Example
388
+ ///
389
+ /// ```rust
390
+ /// use kreuzberg::plugins::list_validators;
391
+ ///
392
+ /// # tokio_test::block_on(async {
393
+ /// let validators = list_validators()?;
394
+ /// for name in validators {
395
+ /// println!("Registered validator: {}", name);
396
+ /// }
397
+ /// # Ok::<(), kreuzberg::KreuzbergError>(())
398
+ /// # });
399
+ /// ```
400
+ pub fn list_validators() -> crate::Result<Vec<String>> {
401
+ use crate::plugins::registry::get_validator_registry;
402
+
403
+ let registry = get_validator_registry();
404
+ let registry = registry
405
+ .read()
406
+ .expect("~keep Failed to acquire read lock on validator registry"); // ~keep
407
+
408
+ Ok(registry.list())
409
+ }
410
+
411
+ /// Clear all validators from the global registry.
412
+ ///
413
+ /// Removes all validators and calls their `shutdown()` methods.
414
+ ///
415
+ /// # Returns
416
+ ///
417
+ /// - `Ok(())` if all validators were cleared successfully
418
+ /// - `Err(...)` if any shutdown method failed
419
+ ///
420
+ /// # Example
421
+ ///
422
+ /// ```rust
423
+ /// use kreuzberg::plugins::clear_validators;
424
+ ///
425
+ /// # tokio_test::block_on(async {
426
+ /// clear_validators()?;
427
+ /// # Ok::<(), kreuzberg::KreuzbergError>(())
428
+ /// # });
429
+ /// ```
430
+ pub fn clear_validators() -> crate::Result<()> {
431
+ use crate::plugins::registry::get_validator_registry;
432
+
433
+ let registry = get_validator_registry();
434
+ let mut registry = registry
435
+ .write()
436
+ .expect("~keep Failed to acquire write lock on validator registry"); // ~keep
437
+
438
+ registry.shutdown_all()
439
+ }
440
+
441
+ #[cfg(test)]
442
+ mod tests {
443
+ use super::*;
444
+ use crate::KreuzbergError;
445
+ use std::collections::HashMap;
446
+
447
+ struct MockValidator {
448
+ should_fail: bool,
449
+ }
450
+
451
+ impl Plugin for MockValidator {
452
+ fn name(&self) -> &str {
453
+ "mock-validator"
454
+ }
455
+
456
+ fn version(&self) -> String {
457
+ "1.0.0".to_string()
458
+ }
459
+
460
+ fn initialize(&self) -> Result<()> {
461
+ Ok(())
462
+ }
463
+
464
+ fn shutdown(&self) -> Result<()> {
465
+ Ok(())
466
+ }
467
+ }
468
+
469
+ #[async_trait]
470
+ impl Validator for MockValidator {
471
+ async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
472
+ if self.should_fail {
473
+ Err(KreuzbergError::validation("Validation failed".to_string()))
474
+ } else {
475
+ Ok(())
476
+ }
477
+ }
478
+ }
479
+
480
+ #[tokio::test]
481
+ async fn test_validator_success() {
482
+ let validator = MockValidator { should_fail: false };
483
+
484
+ let result = ExtractionResult {
485
+ content: "test content".to_string(),
486
+ mime_type: "text/plain".to_string(),
487
+ metadata: crate::types::Metadata::default(),
488
+ tables: vec![],
489
+ detected_languages: None,
490
+ chunks: None,
491
+ images: None,
492
+ };
493
+
494
+ let config = ExtractionConfig::default();
495
+ assert!(validator.validate(&result, &config).await.is_ok());
496
+ }
497
+
498
+ #[tokio::test]
499
+ async fn test_validator_failure() {
500
+ let validator = MockValidator { should_fail: true };
501
+
502
+ let result = ExtractionResult {
503
+ content: "test content".to_string(),
504
+ mime_type: "text/plain".to_string(),
505
+ metadata: crate::types::Metadata::default(),
506
+ tables: vec![],
507
+ detected_languages: None,
508
+ chunks: None,
509
+ images: None,
510
+ };
511
+
512
+ let config = ExtractionConfig::default();
513
+ let validation_result = validator.validate(&result, &config).await;
514
+
515
+ assert!(matches!(validation_result, Err(KreuzbergError::Validation { .. })));
516
+ }
517
+
518
+ #[test]
519
+ fn test_validator_should_validate_default() {
520
+ let validator = MockValidator { should_fail: false };
521
+
522
+ let result = ExtractionResult {
523
+ content: "test".to_string(),
524
+ mime_type: "text/plain".to_string(),
525
+ metadata: crate::types::Metadata::default(),
526
+ tables: vec![],
527
+ detected_languages: None,
528
+ chunks: None,
529
+ images: None,
530
+ };
531
+
532
+ let config = ExtractionConfig::default();
533
+
534
+ assert!(validator.should_validate(&result, &config));
535
+ }
536
+
537
+ #[test]
538
+ fn test_validator_priority_default() {
539
+ let validator = MockValidator { should_fail: false };
540
+ assert_eq!(validator.priority(), 50);
541
+ }
542
+
543
+ #[tokio::test]
544
+ async fn test_validator_plugin_interface() {
545
+ let validator = MockValidator { should_fail: false };
546
+
547
+ assert_eq!(validator.name(), "mock-validator");
548
+ assert_eq!(validator.version(), "1.0.0");
549
+ assert!(validator.initialize().is_ok());
550
+ assert!(validator.shutdown().is_ok());
551
+ }
552
+
553
+ #[tokio::test]
554
+ async fn test_validator_empty_content() {
555
+ let validator = MockValidator { should_fail: false };
556
+
557
+ let result = ExtractionResult {
558
+ content: String::new(),
559
+ mime_type: "text/plain".to_string(),
560
+ metadata: crate::types::Metadata::default(),
561
+ tables: vec![],
562
+ detected_languages: None,
563
+ chunks: None,
564
+ images: None,
565
+ };
566
+
567
+ let config = ExtractionConfig::default();
568
+ assert!(validator.validate(&result, &config).await.is_ok());
569
+ }
570
+
571
+ #[test]
572
+ fn test_validator_should_validate_conditional() {
573
+ struct PdfOnlyValidator;
574
+
575
+ impl Plugin for PdfOnlyValidator {
576
+ fn name(&self) -> &str {
577
+ "pdf-only"
578
+ }
579
+ fn version(&self) -> String {
580
+ "1.0.0".to_string()
581
+ }
582
+ fn initialize(&self) -> Result<()> {
583
+ Ok(())
584
+ }
585
+ fn shutdown(&self) -> Result<()> {
586
+ Ok(())
587
+ }
588
+ }
589
+
590
+ #[async_trait]
591
+ impl Validator for PdfOnlyValidator {
592
+ async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
593
+ Ok(())
594
+ }
595
+
596
+ fn should_validate(&self, result: &ExtractionResult, _config: &ExtractionConfig) -> bool {
597
+ result.mime_type == "application/pdf"
598
+ }
599
+ }
600
+
601
+ let validator = PdfOnlyValidator;
602
+ let config = ExtractionConfig::default();
603
+
604
+ let pdf_result = ExtractionResult {
605
+ content: "test".to_string(),
606
+ mime_type: "application/pdf".to_string(),
607
+ metadata: crate::types::Metadata::default(),
608
+ tables: vec![],
609
+ detected_languages: None,
610
+ chunks: None,
611
+ images: None,
612
+ };
613
+
614
+ let txt_result = ExtractionResult {
615
+ content: "test".to_string(),
616
+ mime_type: "text/plain".to_string(),
617
+ metadata: crate::types::Metadata::default(),
618
+ tables: vec![],
619
+ detected_languages: None,
620
+ chunks: None,
621
+ images: None,
622
+ };
623
+
624
+ assert!(validator.should_validate(&pdf_result, &config));
625
+ assert!(!validator.should_validate(&txt_result, &config));
626
+ }
627
+
628
+ #[test]
629
+ fn test_validator_priority_ranges() {
630
+ struct HighPriorityValidator;
631
+ struct LowPriorityValidator;
632
+
633
+ impl Plugin for HighPriorityValidator {
634
+ fn name(&self) -> &str {
635
+ "high-priority"
636
+ }
637
+ fn version(&self) -> String {
638
+ "1.0.0".to_string()
639
+ }
640
+ fn initialize(&self) -> Result<()> {
641
+ Ok(())
642
+ }
643
+ fn shutdown(&self) -> Result<()> {
644
+ Ok(())
645
+ }
646
+ }
647
+
648
+ impl Plugin for LowPriorityValidator {
649
+ fn name(&self) -> &str {
650
+ "low-priority"
651
+ }
652
+ fn version(&self) -> String {
653
+ "1.0.0".to_string()
654
+ }
655
+ fn initialize(&self) -> Result<()> {
656
+ Ok(())
657
+ }
658
+ fn shutdown(&self) -> Result<()> {
659
+ Ok(())
660
+ }
661
+ }
662
+
663
+ #[async_trait]
664
+ impl Validator for HighPriorityValidator {
665
+ async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
666
+ Ok(())
667
+ }
668
+
669
+ fn priority(&self) -> i32 {
670
+ 100
671
+ }
672
+ }
673
+
674
+ #[async_trait]
675
+ impl Validator for LowPriorityValidator {
676
+ async fn validate(&self, _result: &ExtractionResult, _config: &ExtractionConfig) -> Result<()> {
677
+ Ok(())
678
+ }
679
+
680
+ fn priority(&self) -> i32 {
681
+ 10
682
+ }
683
+ }
684
+
685
+ let high = HighPriorityValidator;
686
+ let low = LowPriorityValidator;
687
+
688
+ assert_eq!(high.priority(), 100);
689
+ assert_eq!(low.priority(), 10);
690
+ assert!(high.priority() > low.priority());
691
+ }
692
+
693
+ #[tokio::test]
694
+ async fn test_validator_error_message() {
695
+ let validator = MockValidator { should_fail: true };
696
+
697
+ let result = ExtractionResult {
698
+ content: "test".to_string(),
699
+ mime_type: "text/plain".to_string(),
700
+ metadata: crate::types::Metadata::default(),
701
+ tables: vec![],
702
+ detected_languages: None,
703
+ chunks: None,
704
+ images: None,
705
+ };
706
+
707
+ let config = ExtractionConfig::default();
708
+ let err = validator.validate(&result, &config).await.unwrap_err();
709
+
710
+ match err {
711
+ KreuzbergError::Validation { message: msg, .. } => {
712
+ assert_eq!(msg, "Validation failed");
713
+ }
714
+ _ => panic!("Expected Validation error"),
715
+ }
716
+ }
717
+
718
+ #[tokio::test]
719
+ async fn test_validator_with_metadata() {
720
+ let validator = MockValidator { should_fail: false };
721
+
722
+ let mut additional = HashMap::new();
723
+ additional.insert("quality_score".to_string(), serde_json::json!(0.95));
724
+
725
+ let result = ExtractionResult {
726
+ content: "test".to_string(),
727
+ mime_type: "text/plain".to_string(),
728
+ metadata: crate::types::Metadata {
729
+ additional,
730
+ ..Default::default()
731
+ },
732
+ tables: vec![],
733
+ detected_languages: None,
734
+ chunks: None,
735
+ images: None,
736
+ };
737
+
738
+ let config = ExtractionConfig::default();
739
+ assert!(validator.validate(&result, &config).await.is_ok());
740
+ }
741
+
742
+ #[tokio::test]
743
+ async fn test_validator_with_tables() {
744
+ use crate::types::Table;
745
+
746
+ let validator = MockValidator { should_fail: false };
747
+
748
+ let table = Table {
749
+ cells: vec![vec!["A".to_string(), "B".to_string()]],
750
+ markdown: "| A | B |".to_string(),
751
+ page_number: 0,
752
+ };
753
+
754
+ let result = ExtractionResult {
755
+ content: "test".to_string(),
756
+ mime_type: "text/plain".to_string(),
757
+ metadata: crate::types::Metadata::default(),
758
+ tables: vec![table],
759
+ detected_languages: None,
760
+ chunks: None,
761
+ images: None,
762
+ };
763
+
764
+ let config = ExtractionConfig::default();
765
+ assert!(validator.validate(&result, &config).await.is_ok());
766
+ }
767
+
768
+ #[tokio::test]
769
+ async fn test_validator_different_mime_types() {
770
+ let validator = MockValidator { should_fail: false };
771
+ let config = ExtractionConfig::default();
772
+
773
+ let mime_types = vec![
774
+ "text/plain",
775
+ "application/pdf",
776
+ "application/json",
777
+ "text/html",
778
+ "image/png",
779
+ ];
780
+
781
+ for mime_type in mime_types {
782
+ let result = ExtractionResult {
783
+ content: "test".to_string(),
784
+ mime_type: mime_type.to_string(),
785
+ metadata: crate::types::Metadata::default(),
786
+ tables: vec![],
787
+ detected_languages: None,
788
+ chunks: None,
789
+ images: None,
790
+ };
791
+
792
+ assert!(validator.validate(&result, &config).await.is_ok());
793
+ }
794
+ }
795
+
796
+ #[tokio::test]
797
+ async fn test_validator_long_content() {
798
+ let validator = MockValidator { should_fail: false };
799
+
800
+ let result = ExtractionResult {
801
+ content: "test content ".repeat(10000),
802
+ mime_type: "text/plain".to_string(),
803
+ metadata: crate::types::Metadata::default(),
804
+ tables: vec![],
805
+ detected_languages: None,
806
+ chunks: None,
807
+ images: None,
808
+ };
809
+
810
+ let config = ExtractionConfig::default();
811
+ assert!(validator.validate(&result, &config).await.is_ok());
812
+ }
813
+
814
+ #[test]
815
+ #[serial_test::serial]
816
+ fn test_register_validator() {
817
+ use std::sync::Arc;
818
+
819
+ let validator = Arc::new(MockValidator { should_fail: false });
820
+ let result = super::register_validator(validator);
821
+ assert!(result.is_ok());
822
+
823
+ let _ = super::unregister_validator("mock-validator");
824
+ }
825
+
826
+ #[test]
827
+ #[serial_test::serial]
828
+ fn test_unregister_validator() {
829
+ use std::sync::Arc;
830
+
831
+ let validator = Arc::new(MockValidator { should_fail: false });
832
+ super::register_validator(validator).unwrap();
833
+
834
+ let result = super::unregister_validator("mock-validator");
835
+ assert!(result.is_ok());
836
+ }
837
+
838
+ #[test]
839
+ #[serial_test::serial]
840
+ fn test_unregister_nonexistent_validator() {
841
+ let result = super::unregister_validator("nonexistent-validator-xyz");
842
+ assert!(result.is_ok());
843
+ }
844
+
845
+ #[test]
846
+ #[serial_test::serial]
847
+ fn test_list_validators() {
848
+ use std::sync::Arc;
849
+
850
+ super::clear_validators().unwrap();
851
+
852
+ let validator1 = Arc::new(MockValidator { should_fail: false });
853
+ let validator2 = Arc::new(MockValidator { should_fail: false });
854
+
855
+ let list_before = super::list_validators().unwrap();
856
+ assert_eq!(list_before.len(), 0);
857
+
858
+ super::register_validator(validator1).unwrap();
859
+ super::register_validator(validator2).unwrap();
860
+
861
+ let list = super::list_validators().unwrap();
862
+ assert_eq!(list.len(), 1);
863
+ assert!(list.contains(&"mock-validator".to_string()));
864
+
865
+ super::unregister_validator("mock-validator").unwrap();
866
+ }
867
+
868
+ #[test]
869
+ #[serial_test::serial]
870
+ fn test_clear_validators() {
871
+ use std::sync::Arc;
872
+
873
+ super::clear_validators().unwrap();
874
+
875
+ let validator1 = Arc::new(MockValidator { should_fail: false });
876
+ let validator2 = Arc::new(MockValidator { should_fail: false });
877
+
878
+ super::register_validator(validator1).unwrap();
879
+ super::register_validator(validator2).unwrap();
880
+
881
+ let list_before = super::list_validators().unwrap();
882
+ assert!(!list_before.is_empty());
883
+
884
+ let result = super::clear_validators();
885
+ assert!(result.is_ok());
886
+
887
+ let list = super::list_validators().unwrap();
888
+ assert_eq!(list.len(), 0);
889
+ }
890
+
891
+ #[test]
892
+ #[serial_test::serial]
893
+ fn test_register_validator_with_invalid_name() {
894
+ use std::sync::Arc;
895
+
896
+ struct InvalidNameValidator;
897
+ impl Plugin for InvalidNameValidator {
898
+ fn name(&self) -> &str {
899
+ "invalid name with spaces"
900
+ }
901
+ fn version(&self) -> String {
902
+ "1.0.0".to_string()
903
+ }
904
+ fn initialize(&self) -> Result<()> {
905
+ Ok(())
906
+ }
907
+ fn shutdown(&self) -> Result<()> {
908
+ Ok(())
909
+ }
910
+ }
911
+
912
+ #[async_trait]
913
+ impl Validator for InvalidNameValidator {
914
+ async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> {
915
+ Ok(())
916
+ }
917
+ }
918
+
919
+ let validator = Arc::new(InvalidNameValidator);
920
+ let result = super::register_validator(validator);
921
+ assert!(matches!(result, Err(KreuzbergError::Validation { .. })));
922
+ }
923
+
924
+ #[test]
925
+ #[serial_test::serial]
926
+ fn test_register_validator_with_empty_name() {
927
+ use std::sync::Arc;
928
+
929
+ struct EmptyNameValidator;
930
+ impl Plugin for EmptyNameValidator {
931
+ fn name(&self) -> &str {
932
+ ""
933
+ }
934
+ fn version(&self) -> String {
935
+ "1.0.0".to_string()
936
+ }
937
+ fn initialize(&self) -> Result<()> {
938
+ Ok(())
939
+ }
940
+ fn shutdown(&self) -> Result<()> {
941
+ Ok(())
942
+ }
943
+ }
944
+
945
+ #[async_trait]
946
+ impl Validator for EmptyNameValidator {
947
+ async fn validate(&self, _: &ExtractionResult, _: &ExtractionConfig) -> Result<()> {
948
+ Ok(())
949
+ }
950
+ }
951
+
952
+ let validator = Arc::new(EmptyNameValidator);
953
+ let result = super::register_validator(validator);
954
+ assert!(matches!(result, Err(KreuzbergError::Validation { .. })));
955
+ }
956
+ }