kreuzberg 4.0.0.pre.rc.29 → 4.0.0.rc1

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 (321) hide show
  1. checksums.yaml +4 -4
  2. data/.gitignore +0 -6
  3. data/.rubocop.yaml +534 -1
  4. data/Gemfile +2 -1
  5. data/Gemfile.lock +28 -116
  6. data/README.md +269 -629
  7. data/Rakefile +0 -9
  8. data/Steepfile +4 -8
  9. data/examples/async_patterns.rb +58 -1
  10. data/ext/kreuzberg_rb/extconf.rb +5 -35
  11. data/ext/kreuzberg_rb/native/Cargo.toml +16 -55
  12. data/ext/kreuzberg_rb/native/build.rs +14 -12
  13. data/ext/kreuzberg_rb/native/include/ieeefp.h +1 -1
  14. data/ext/kreuzberg_rb/native/include/msvc_compat/strings.h +1 -1
  15. data/ext/kreuzberg_rb/native/include/strings.h +2 -2
  16. data/ext/kreuzberg_rb/native/include/unistd.h +1 -1
  17. data/ext/kreuzberg_rb/native/src/lib.rs +34 -897
  18. data/extconf.rb +6 -38
  19. data/kreuzberg.gemspec +20 -114
  20. data/lib/kreuzberg/api_proxy.rb +18 -2
  21. data/lib/kreuzberg/cache_api.rb +0 -22
  22. data/lib/kreuzberg/cli.rb +10 -2
  23. data/lib/kreuzberg/cli_proxy.rb +10 -0
  24. data/lib/kreuzberg/config.rb +22 -274
  25. data/lib/kreuzberg/errors.rb +7 -73
  26. data/lib/kreuzberg/extraction_api.rb +8 -237
  27. data/lib/kreuzberg/mcp_proxy.rb +11 -2
  28. data/lib/kreuzberg/ocr_backend_protocol.rb +73 -0
  29. data/lib/kreuzberg/post_processor_protocol.rb +71 -0
  30. data/lib/kreuzberg/result.rb +33 -151
  31. data/lib/kreuzberg/setup_lib_path.rb +2 -22
  32. data/lib/kreuzberg/validator_protocol.rb +73 -0
  33. data/lib/kreuzberg/version.rb +1 -1
  34. data/lib/kreuzberg.rb +13 -27
  35. data/pkg/kreuzberg-4.0.0.rc1.gem +0 -0
  36. data/sig/kreuzberg.rbs +12 -105
  37. data/spec/binding/cache_spec.rb +22 -22
  38. data/spec/binding/cli_proxy_spec.rb +4 -2
  39. data/spec/binding/cli_spec.rb +11 -12
  40. data/spec/binding/config_spec.rb +0 -74
  41. data/spec/binding/config_validation_spec.rb +6 -100
  42. data/spec/binding/error_handling_spec.rb +97 -283
  43. data/spec/binding/plugins/ocr_backend_spec.rb +8 -8
  44. data/spec/binding/plugins/postprocessor_spec.rb +11 -11
  45. data/spec/binding/plugins/validator_spec.rb +13 -12
  46. data/spec/examples.txt +104 -0
  47. data/spec/fixtures/config.toml +1 -0
  48. data/spec/fixtures/config.yaml +1 -0
  49. data/spec/fixtures/invalid_config.toml +1 -0
  50. data/spec/smoke/package_spec.rb +3 -2
  51. data/spec/spec_helper.rb +3 -1
  52. data/vendor/kreuzberg/Cargo.toml +67 -192
  53. data/vendor/kreuzberg/README.md +9 -97
  54. data/vendor/kreuzberg/build.rs +194 -516
  55. data/vendor/kreuzberg/src/api/handlers.rs +9 -130
  56. data/vendor/kreuzberg/src/api/mod.rs +3 -18
  57. data/vendor/kreuzberg/src/api/server.rs +71 -236
  58. data/vendor/kreuzberg/src/api/types.rs +7 -43
  59. data/vendor/kreuzberg/src/bin/profile_extract.rs +455 -0
  60. data/vendor/kreuzberg/src/cache/mod.rs +3 -27
  61. data/vendor/kreuzberg/src/chunking/mod.rs +79 -1705
  62. data/vendor/kreuzberg/src/core/batch_mode.rs +0 -60
  63. data/vendor/kreuzberg/src/core/config.rs +23 -905
  64. data/vendor/kreuzberg/src/core/extractor.rs +106 -403
  65. data/vendor/kreuzberg/src/core/io.rs +2 -4
  66. data/vendor/kreuzberg/src/core/mime.rs +12 -2
  67. data/vendor/kreuzberg/src/core/mod.rs +3 -22
  68. data/vendor/kreuzberg/src/core/pipeline.rs +78 -395
  69. data/vendor/kreuzberg/src/embeddings.rs +21 -169
  70. data/vendor/kreuzberg/src/error.rs +2 -2
  71. data/vendor/kreuzberg/src/extraction/archive.rs +31 -36
  72. data/vendor/kreuzberg/src/extraction/docx.rs +1 -365
  73. data/vendor/kreuzberg/src/extraction/email.rs +11 -12
  74. data/vendor/kreuzberg/src/extraction/excel.rs +129 -138
  75. data/vendor/kreuzberg/src/extraction/html.rs +170 -1447
  76. data/vendor/kreuzberg/src/extraction/image.rs +14 -138
  77. data/vendor/kreuzberg/src/extraction/libreoffice.rs +3 -13
  78. data/vendor/kreuzberg/src/extraction/mod.rs +5 -21
  79. data/vendor/kreuzberg/src/extraction/office_metadata/mod.rs +0 -2
  80. data/vendor/kreuzberg/src/extraction/pandoc/batch.rs +275 -0
  81. data/vendor/kreuzberg/src/extraction/pandoc/mime_types.rs +178 -0
  82. data/vendor/kreuzberg/src/extraction/pandoc/mod.rs +491 -0
  83. data/vendor/kreuzberg/src/extraction/pandoc/server.rs +496 -0
  84. data/vendor/kreuzberg/src/extraction/pandoc/subprocess.rs +1188 -0
  85. data/vendor/kreuzberg/src/extraction/pandoc/version.rs +162 -0
  86. data/vendor/kreuzberg/src/extraction/pptx.rs +94 -196
  87. data/vendor/kreuzberg/src/extraction/structured.rs +4 -5
  88. data/vendor/kreuzberg/src/extraction/table.rs +1 -2
  89. data/vendor/kreuzberg/src/extraction/text.rs +10 -18
  90. data/vendor/kreuzberg/src/extractors/archive.rs +0 -22
  91. data/vendor/kreuzberg/src/extractors/docx.rs +148 -69
  92. data/vendor/kreuzberg/src/extractors/email.rs +9 -37
  93. data/vendor/kreuzberg/src/extractors/excel.rs +40 -81
  94. data/vendor/kreuzberg/src/extractors/html.rs +173 -182
  95. data/vendor/kreuzberg/src/extractors/image.rs +8 -32
  96. data/vendor/kreuzberg/src/extractors/mod.rs +10 -171
  97. data/vendor/kreuzberg/src/extractors/pandoc.rs +201 -0
  98. data/vendor/kreuzberg/src/extractors/pdf.rs +64 -329
  99. data/vendor/kreuzberg/src/extractors/pptx.rs +34 -79
  100. data/vendor/kreuzberg/src/extractors/structured.rs +0 -16
  101. data/vendor/kreuzberg/src/extractors/text.rs +7 -30
  102. data/vendor/kreuzberg/src/extractors/xml.rs +8 -27
  103. data/vendor/kreuzberg/src/keywords/processor.rs +1 -9
  104. data/vendor/kreuzberg/src/keywords/rake.rs +1 -0
  105. data/vendor/kreuzberg/src/language_detection/mod.rs +51 -94
  106. data/vendor/kreuzberg/src/lib.rs +5 -17
  107. data/vendor/kreuzberg/src/mcp/mod.rs +1 -4
  108. data/vendor/kreuzberg/src/mcp/server.rs +21 -145
  109. data/vendor/kreuzberg/src/ocr/mod.rs +0 -2
  110. data/vendor/kreuzberg/src/ocr/processor.rs +8 -19
  111. data/vendor/kreuzberg/src/ocr/tesseract_backend.rs +0 -2
  112. data/vendor/kreuzberg/src/pdf/error.rs +1 -93
  113. data/vendor/kreuzberg/src/pdf/metadata.rs +100 -263
  114. data/vendor/kreuzberg/src/pdf/mod.rs +2 -33
  115. data/vendor/kreuzberg/src/pdf/rendering.rs +12 -12
  116. data/vendor/kreuzberg/src/pdf/table.rs +64 -61
  117. data/vendor/kreuzberg/src/pdf/text.rs +24 -416
  118. data/vendor/kreuzberg/src/plugins/extractor.rs +8 -40
  119. data/vendor/kreuzberg/src/plugins/mod.rs +0 -3
  120. data/vendor/kreuzberg/src/plugins/ocr.rs +14 -22
  121. data/vendor/kreuzberg/src/plugins/processor.rs +1 -10
  122. data/vendor/kreuzberg/src/plugins/registry.rs +0 -15
  123. data/vendor/kreuzberg/src/plugins/validator.rs +8 -20
  124. data/vendor/kreuzberg/src/stopwords/mod.rs +2 -2
  125. data/vendor/kreuzberg/src/text/mod.rs +0 -8
  126. data/vendor/kreuzberg/src/text/quality.rs +15 -28
  127. data/vendor/kreuzberg/src/text/string_utils.rs +10 -22
  128. data/vendor/kreuzberg/src/text/token_reduction/core.rs +50 -86
  129. data/vendor/kreuzberg/src/text/token_reduction/filters.rs +16 -37
  130. data/vendor/kreuzberg/src/text/token_reduction/simd_text.rs +1 -2
  131. data/vendor/kreuzberg/src/types.rs +67 -907
  132. data/vendor/kreuzberg/src/utils/mod.rs +0 -14
  133. data/vendor/kreuzberg/src/utils/quality.rs +3 -12
  134. data/vendor/kreuzberg/tests/api_tests.rs +0 -506
  135. data/vendor/kreuzberg/tests/archive_integration.rs +0 -2
  136. data/vendor/kreuzberg/tests/batch_orchestration.rs +12 -57
  137. data/vendor/kreuzberg/tests/batch_processing.rs +8 -32
  138. data/vendor/kreuzberg/tests/chunking_offset_demo.rs +92 -0
  139. data/vendor/kreuzberg/tests/concurrency_stress.rs +8 -40
  140. data/vendor/kreuzberg/tests/config_features.rs +1 -33
  141. data/vendor/kreuzberg/tests/config_loading_tests.rs +39 -16
  142. data/vendor/kreuzberg/tests/core_integration.rs +9 -35
  143. data/vendor/kreuzberg/tests/csv_integration.rs +81 -71
  144. data/vendor/kreuzberg/tests/docx_metadata_extraction_test.rs +25 -23
  145. data/vendor/kreuzberg/tests/email_integration.rs +1 -3
  146. data/vendor/kreuzberg/tests/error_handling.rs +34 -43
  147. data/vendor/kreuzberg/tests/format_integration.rs +1 -7
  148. data/vendor/kreuzberg/tests/helpers/mod.rs +0 -60
  149. data/vendor/kreuzberg/tests/image_integration.rs +0 -2
  150. data/vendor/kreuzberg/tests/mime_detection.rs +16 -17
  151. data/vendor/kreuzberg/tests/ocr_configuration.rs +0 -4
  152. data/vendor/kreuzberg/tests/ocr_errors.rs +0 -22
  153. data/vendor/kreuzberg/tests/ocr_quality.rs +0 -2
  154. data/vendor/kreuzberg/tests/pandoc_integration.rs +503 -0
  155. data/vendor/kreuzberg/tests/pdf_integration.rs +0 -2
  156. data/vendor/kreuzberg/tests/pipeline_integration.rs +2 -36
  157. data/vendor/kreuzberg/tests/plugin_ocr_backend_test.rs +0 -5
  158. data/vendor/kreuzberg/tests/plugin_postprocessor_test.rs +1 -17
  159. data/vendor/kreuzberg/tests/plugin_system.rs +0 -6
  160. data/vendor/kreuzberg/tests/registry_integration_tests.rs +22 -2
  161. data/vendor/kreuzberg/tests/security_validation.rs +1 -13
  162. data/vendor/kreuzberg/tests/test_fastembed.rs +23 -45
  163. metadata +25 -171
  164. data/.rubocop.yml +0 -543
  165. data/ext/kreuzberg_rb/native/.cargo/config.toml +0 -23
  166. data/ext/kreuzberg_rb/native/Cargo.lock +0 -7619
  167. data/lib/kreuzberg/error_context.rb +0 -136
  168. data/lib/kreuzberg/types.rb +0 -170
  169. data/lib/libpdfium.so +0 -0
  170. data/spec/binding/async_operations_spec.rb +0 -473
  171. data/spec/binding/batch_operations_spec.rb +0 -595
  172. data/spec/binding/batch_spec.rb +0 -359
  173. data/spec/binding/config_result_spec.rb +0 -377
  174. data/spec/binding/embeddings_spec.rb +0 -816
  175. data/spec/binding/error_recovery_spec.rb +0 -488
  176. data/spec/binding/font_config_spec.rb +0 -220
  177. data/spec/binding/images_spec.rb +0 -738
  178. data/spec/binding/keywords_extraction_spec.rb +0 -600
  179. data/spec/binding/metadata_types_spec.rb +0 -1228
  180. data/spec/binding/pages_extraction_spec.rb +0 -471
  181. data/spec/binding/tables_spec.rb +0 -641
  182. data/spec/unit/config/chunking_config_spec.rb +0 -213
  183. data/spec/unit/config/embedding_config_spec.rb +0 -343
  184. data/spec/unit/config/extraction_config_spec.rb +0 -438
  185. data/spec/unit/config/font_config_spec.rb +0 -285
  186. data/spec/unit/config/hierarchy_config_spec.rb +0 -314
  187. data/spec/unit/config/image_extraction_config_spec.rb +0 -209
  188. data/spec/unit/config/image_preprocessing_config_spec.rb +0 -249
  189. data/spec/unit/config/keyword_config_spec.rb +0 -229
  190. data/spec/unit/config/language_detection_config_spec.rb +0 -258
  191. data/spec/unit/config/ocr_config_spec.rb +0 -171
  192. data/spec/unit/config/page_config_spec.rb +0 -221
  193. data/spec/unit/config/pdf_config_spec.rb +0 -267
  194. data/spec/unit/config/postprocessor_config_spec.rb +0 -290
  195. data/spec/unit/config/tesseract_config_spec.rb +0 -181
  196. data/spec/unit/config/token_reduction_config_spec.rb +0 -251
  197. data/test/metadata_types_test.rb +0 -959
  198. data/vendor/Cargo.toml +0 -61
  199. data/vendor/kreuzberg/examples/bench_fixes.rs +0 -71
  200. data/vendor/kreuzberg/examples/test_pdfium_fork.rs +0 -62
  201. data/vendor/kreuzberg/src/chunking/processor.rs +0 -219
  202. data/vendor/kreuzberg/src/core/batch_optimizations.rs +0 -385
  203. data/vendor/kreuzberg/src/core/config_validation.rs +0 -949
  204. data/vendor/kreuzberg/src/core/formats.rs +0 -235
  205. data/vendor/kreuzberg/src/core/server_config.rs +0 -1220
  206. data/vendor/kreuzberg/src/extraction/capacity.rs +0 -263
  207. data/vendor/kreuzberg/src/extraction/markdown.rs +0 -216
  208. data/vendor/kreuzberg/src/extraction/office_metadata/odt_properties.rs +0 -284
  209. data/vendor/kreuzberg/src/extractors/bibtex.rs +0 -470
  210. data/vendor/kreuzberg/src/extractors/docbook.rs +0 -504
  211. data/vendor/kreuzberg/src/extractors/epub.rs +0 -696
  212. data/vendor/kreuzberg/src/extractors/fictionbook.rs +0 -492
  213. data/vendor/kreuzberg/src/extractors/jats.rs +0 -1054
  214. data/vendor/kreuzberg/src/extractors/jupyter.rs +0 -368
  215. data/vendor/kreuzberg/src/extractors/latex.rs +0 -653
  216. data/vendor/kreuzberg/src/extractors/markdown.rs +0 -701
  217. data/vendor/kreuzberg/src/extractors/odt.rs +0 -628
  218. data/vendor/kreuzberg/src/extractors/opml.rs +0 -635
  219. data/vendor/kreuzberg/src/extractors/orgmode.rs +0 -529
  220. data/vendor/kreuzberg/src/extractors/rst.rs +0 -577
  221. data/vendor/kreuzberg/src/extractors/rtf.rs +0 -809
  222. data/vendor/kreuzberg/src/extractors/security.rs +0 -484
  223. data/vendor/kreuzberg/src/extractors/security_tests.rs +0 -367
  224. data/vendor/kreuzberg/src/extractors/typst.rs +0 -651
  225. data/vendor/kreuzberg/src/language_detection/processor.rs +0 -218
  226. data/vendor/kreuzberg/src/ocr/language_registry.rs +0 -520
  227. data/vendor/kreuzberg/src/panic_context.rs +0 -154
  228. data/vendor/kreuzberg/src/pdf/bindings.rs +0 -306
  229. data/vendor/kreuzberg/src/pdf/bundled.rs +0 -408
  230. data/vendor/kreuzberg/src/pdf/fonts.rs +0 -358
  231. data/vendor/kreuzberg/src/pdf/hierarchy.rs +0 -903
  232. data/vendor/kreuzberg/src/text/quality_processor.rs +0 -231
  233. data/vendor/kreuzberg/src/text/utf8_validation.rs +0 -193
  234. data/vendor/kreuzberg/src/utils/pool.rs +0 -503
  235. data/vendor/kreuzberg/src/utils/pool_sizing.rs +0 -364
  236. data/vendor/kreuzberg/src/utils/string_pool.rs +0 -761
  237. data/vendor/kreuzberg/tests/api_embed.rs +0 -360
  238. data/vendor/kreuzberg/tests/api_extract_multipart.rs +0 -52
  239. data/vendor/kreuzberg/tests/api_large_pdf_extraction.rs +0 -471
  240. data/vendor/kreuzberg/tests/api_large_pdf_extraction_diagnostics.rs +0 -289
  241. data/vendor/kreuzberg/tests/batch_pooling_benchmark.rs +0 -154
  242. data/vendor/kreuzberg/tests/bibtex_parity_test.rs +0 -421
  243. data/vendor/kreuzberg/tests/config_integration_test.rs +0 -753
  244. data/vendor/kreuzberg/tests/data/hierarchy_ground_truth.json +0 -294
  245. data/vendor/kreuzberg/tests/docbook_extractor_tests.rs +0 -500
  246. data/vendor/kreuzberg/tests/docx_vs_pandoc_comparison.rs +0 -370
  247. data/vendor/kreuzberg/tests/epub_native_extractor_tests.rs +0 -275
  248. data/vendor/kreuzberg/tests/fictionbook_extractor_tests.rs +0 -228
  249. data/vendor/kreuzberg/tests/html_table_test.rs +0 -551
  250. data/vendor/kreuzberg/tests/instrumentation_test.rs +0 -139
  251. data/vendor/kreuzberg/tests/jats_extractor_tests.rs +0 -639
  252. data/vendor/kreuzberg/tests/jupyter_extractor_tests.rs +0 -704
  253. data/vendor/kreuzberg/tests/latex_extractor_tests.rs +0 -496
  254. data/vendor/kreuzberg/tests/markdown_extractor_tests.rs +0 -490
  255. data/vendor/kreuzberg/tests/ocr_language_registry.rs +0 -191
  256. data/vendor/kreuzberg/tests/odt_extractor_tests.rs +0 -674
  257. data/vendor/kreuzberg/tests/opml_extractor_tests.rs +0 -616
  258. data/vendor/kreuzberg/tests/orgmode_extractor_tests.rs +0 -822
  259. data/vendor/kreuzberg/tests/page_markers.rs +0 -297
  260. data/vendor/kreuzberg/tests/pdf_hierarchy_detection.rs +0 -301
  261. data/vendor/kreuzberg/tests/pdf_hierarchy_quality.rs +0 -589
  262. data/vendor/kreuzberg/tests/pdf_ocr_triggering.rs +0 -301
  263. data/vendor/kreuzberg/tests/pdf_text_merging.rs +0 -475
  264. data/vendor/kreuzberg/tests/pdfium_linking.rs +0 -340
  265. data/vendor/kreuzberg/tests/rst_extractor_tests.rs +0 -694
  266. data/vendor/kreuzberg/tests/rtf_extractor_tests.rs +0 -775
  267. data/vendor/kreuzberg/tests/typst_behavioral_tests.rs +0 -1260
  268. data/vendor/kreuzberg/tests/typst_extractor_tests.rs +0 -648
  269. data/vendor/kreuzberg-ffi/Cargo.toml +0 -67
  270. data/vendor/kreuzberg-ffi/README.md +0 -851
  271. data/vendor/kreuzberg-ffi/benches/result_view_benchmark.rs +0 -227
  272. data/vendor/kreuzberg-ffi/build.rs +0 -168
  273. data/vendor/kreuzberg-ffi/cbindgen.toml +0 -37
  274. data/vendor/kreuzberg-ffi/kreuzberg-ffi.pc.in +0 -12
  275. data/vendor/kreuzberg-ffi/kreuzberg.h +0 -3012
  276. data/vendor/kreuzberg-ffi/src/batch_streaming.rs +0 -588
  277. data/vendor/kreuzberg-ffi/src/config.rs +0 -1341
  278. data/vendor/kreuzberg-ffi/src/error.rs +0 -901
  279. data/vendor/kreuzberg-ffi/src/extraction.rs +0 -555
  280. data/vendor/kreuzberg-ffi/src/helpers.rs +0 -879
  281. data/vendor/kreuzberg-ffi/src/lib.rs +0 -977
  282. data/vendor/kreuzberg-ffi/src/memory.rs +0 -493
  283. data/vendor/kreuzberg-ffi/src/mime.rs +0 -329
  284. data/vendor/kreuzberg-ffi/src/panic_shield.rs +0 -265
  285. data/vendor/kreuzberg-ffi/src/plugins/document_extractor.rs +0 -442
  286. data/vendor/kreuzberg-ffi/src/plugins/mod.rs +0 -14
  287. data/vendor/kreuzberg-ffi/src/plugins/ocr_backend.rs +0 -628
  288. data/vendor/kreuzberg-ffi/src/plugins/post_processor.rs +0 -438
  289. data/vendor/kreuzberg-ffi/src/plugins/validator.rs +0 -329
  290. data/vendor/kreuzberg-ffi/src/result.rs +0 -510
  291. data/vendor/kreuzberg-ffi/src/result_pool.rs +0 -639
  292. data/vendor/kreuzberg-ffi/src/result_view.rs +0 -773
  293. data/vendor/kreuzberg-ffi/src/string_intern.rs +0 -568
  294. data/vendor/kreuzberg-ffi/src/types.rs +0 -363
  295. data/vendor/kreuzberg-ffi/src/util.rs +0 -210
  296. data/vendor/kreuzberg-ffi/src/validation.rs +0 -848
  297. data/vendor/kreuzberg-ffi/tests.disabled/README.md +0 -48
  298. data/vendor/kreuzberg-ffi/tests.disabled/config_loading_tests.rs +0 -299
  299. data/vendor/kreuzberg-ffi/tests.disabled/config_tests.rs +0 -346
  300. data/vendor/kreuzberg-ffi/tests.disabled/extractor_tests.rs +0 -232
  301. data/vendor/kreuzberg-ffi/tests.disabled/plugin_registration_tests.rs +0 -470
  302. data/vendor/kreuzberg-tesseract/.commitlintrc.json +0 -13
  303. data/vendor/kreuzberg-tesseract/.crate-ignore +0 -2
  304. data/vendor/kreuzberg-tesseract/Cargo.lock +0 -2933
  305. data/vendor/kreuzberg-tesseract/Cargo.toml +0 -57
  306. data/vendor/kreuzberg-tesseract/LICENSE +0 -22
  307. data/vendor/kreuzberg-tesseract/README.md +0 -399
  308. data/vendor/kreuzberg-tesseract/build.rs +0 -1127
  309. data/vendor/kreuzberg-tesseract/patches/README.md +0 -71
  310. data/vendor/kreuzberg-tesseract/patches/tesseract.diff +0 -199
  311. data/vendor/kreuzberg-tesseract/src/api.rs +0 -1371
  312. data/vendor/kreuzberg-tesseract/src/choice_iterator.rs +0 -77
  313. data/vendor/kreuzberg-tesseract/src/enums.rs +0 -297
  314. data/vendor/kreuzberg-tesseract/src/error.rs +0 -81
  315. data/vendor/kreuzberg-tesseract/src/lib.rs +0 -145
  316. data/vendor/kreuzberg-tesseract/src/monitor.rs +0 -57
  317. data/vendor/kreuzberg-tesseract/src/mutable_iterator.rs +0 -197
  318. data/vendor/kreuzberg-tesseract/src/page_iterator.rs +0 -253
  319. data/vendor/kreuzberg-tesseract/src/result_iterator.rs +0 -286
  320. data/vendor/kreuzberg-tesseract/src/result_renderer.rs +0 -183
  321. data/vendor/kreuzberg-tesseract/tests/integration_test.rs +0 -211
@@ -1,628 +0,0 @@
1
- //! OCR Backend Plugin FFI
2
- //!
3
- //! Provides FFI bindings for registering and managing custom OCR backends.
4
-
5
- use std::ffi::{CStr, CString};
6
- use std::os::raw::c_char;
7
- use std::ptr;
8
- use std::sync::Arc;
9
-
10
- use async_trait::async_trait;
11
- use kreuzberg::core::config::OcrConfig;
12
- use kreuzberg::plugins::registry::get_ocr_backend_registry;
13
- use kreuzberg::plugins::{OcrBackend, Plugin};
14
- use kreuzberg::types::ExtractionResult;
15
- use kreuzberg::{KreuzbergError, Result};
16
-
17
- use crate::helpers::{clear_last_error, set_last_error};
18
- use crate::memory::kreuzberg_free_string;
19
- // Macros are exported at the crate root due to #[macro_export]
20
- use crate::{ffi_panic_guard, ffi_panic_guard_bool, ffi_panic_guard_i32};
21
-
22
- /// Type alias for the OCR backend callback function.
23
- ///
24
- /// # Parameters
25
- ///
26
- /// - `image_bytes`: Raw image bytes
27
- /// - `image_length`: Length of image data in bytes
28
- /// - `config_json`: JSON-encoded OcrConfig (null-terminated string)
29
- ///
30
- /// # Returns
31
- ///
32
- /// Null-terminated string containing extracted text (must be freed by Rust via kreuzberg_free_string),
33
- /// or NULL on error.
34
- ///
35
- /// # Safety
36
- ///
37
- /// The callback must:
38
- /// - Not store the image_bytes pointer (it's only valid for the duration of the call)
39
- /// - Return a valid null-terminated UTF-8 string allocated by the caller
40
- /// - Return NULL on error (error message should be retrievable separately)
41
- type OcrBackendCallback =
42
- unsafe extern "C" fn(image_bytes: *const u8, image_length: usize, config_json: *const c_char) -> *mut c_char;
43
-
44
- fn parse_languages_from_json(languages_json: *const c_char) -> std::result::Result<Option<Vec<String>>, String> {
45
- if languages_json.is_null() {
46
- return Ok(None);
47
- }
48
-
49
- let raw = unsafe { CStr::from_ptr(languages_json) }
50
- .to_str()
51
- .map_err(|e| format!("Invalid UTF-8 in languages JSON: {}", e))?;
52
-
53
- if raw.trim().is_empty() {
54
- return Ok(None);
55
- }
56
-
57
- let langs: Vec<String> = serde_json::from_str(raw).map_err(|e| format!("Failed to parse languages JSON: {}", e))?;
58
-
59
- if langs.is_empty() {
60
- return Ok(None);
61
- }
62
-
63
- let normalized = langs
64
- .into_iter()
65
- .map(|l| l.trim().to_string())
66
- .filter(|l| !l.is_empty())
67
- .collect::<Vec<_>>();
68
-
69
- if normalized.is_empty() {
70
- return Ok(None);
71
- }
72
-
73
- Ok(Some(normalized))
74
- }
75
-
76
- /// FFI wrapper for custom OCR backends registered from Java/C.
77
- ///
78
- /// This struct wraps a C function pointer and implements the OcrBackend trait,
79
- /// allowing custom OCR implementations from FFI languages to be registered
80
- /// and used within the Rust extraction pipeline.
81
- struct FfiOcrBackend {
82
- name: String,
83
- callback: OcrBackendCallback,
84
- supported_languages: Option<Vec<String>>,
85
- }
86
-
87
- impl FfiOcrBackend {
88
- fn new(name: String, callback: OcrBackendCallback, supported_languages: Option<Vec<String>>) -> Self {
89
- Self {
90
- name,
91
- callback,
92
- supported_languages,
93
- }
94
- }
95
- }
96
-
97
- impl Plugin for FfiOcrBackend {
98
- fn name(&self) -> &str {
99
- &self.name
100
- }
101
-
102
- fn version(&self) -> String {
103
- "ffi-1.0.0".to_string()
104
- }
105
-
106
- fn initialize(&self) -> Result<()> {
107
- Ok(())
108
- }
109
-
110
- fn shutdown(&self) -> Result<()> {
111
- Ok(())
112
- }
113
- }
114
-
115
- #[async_trait]
116
- impl OcrBackend for FfiOcrBackend {
117
- async fn process_image(&self, image_bytes: &[u8], config: &OcrConfig) -> Result<ExtractionResult> {
118
- let config_json = serde_json::to_string(config).map_err(|e| KreuzbergError::Validation {
119
- message: format!("Failed to serialize OCR config: {}", e),
120
- source: Some(Box::new(e)),
121
- })?;
122
-
123
- let callback = self.callback;
124
- let image_data = image_bytes.to_vec();
125
- let config_json_owned = config_json.clone();
126
-
127
- let result_text = tokio::task::spawn_blocking(move || {
128
- let config_cstring = CString::new(config_json_owned).map_err(|e| KreuzbergError::Validation {
129
- message: format!("Failed to create C string from config JSON: {}", e),
130
- source: Some(Box::new(e)),
131
- })?;
132
-
133
- let result_ptr = unsafe { callback(image_data.as_ptr(), image_data.len(), config_cstring.as_ptr()) };
134
-
135
- if result_ptr.is_null() {
136
- return Err(KreuzbergError::Ocr {
137
- message: "OCR backend returned NULL (operation failed)".to_string(),
138
- source: None,
139
- });
140
- }
141
-
142
- let result_cstr = unsafe { CStr::from_ptr(result_ptr) };
143
- let text = result_cstr
144
- .to_str()
145
- .map_err(|e| KreuzbergError::Ocr {
146
- message: format!("OCR backend returned invalid UTF-8: {}", e),
147
- source: Some(Box::new(e)),
148
- })?
149
- .to_string();
150
-
151
- unsafe { kreuzberg_free_string(result_ptr) };
152
-
153
- Ok(text)
154
- })
155
- .await
156
- .map_err(|e| KreuzbergError::Ocr {
157
- message: format!("OCR backend task panicked: {}", e),
158
- source: Some(Box::new(e)),
159
- })??;
160
-
161
- Ok(ExtractionResult {
162
- content: result_text,
163
- mime_type: "text/plain".to_string(),
164
- metadata: kreuzberg::types::Metadata::default(),
165
- tables: vec![],
166
- detected_languages: None,
167
- chunks: None,
168
- images: None,
169
- pages: None,
170
- })
171
- }
172
-
173
- fn supports_language(&self, _lang: &str) -> bool {
174
- match &self.supported_languages {
175
- Some(langs) => langs.iter().any(|candidate| candidate.eq_ignore_ascii_case(_lang)),
176
- None => true,
177
- }
178
- }
179
-
180
- fn backend_type(&self) -> kreuzberg::plugins::OcrBackendType {
181
- kreuzberg::plugins::OcrBackendType::Custom
182
- }
183
- }
184
-
185
- /// Register a custom OCR backend via FFI callback.
186
- ///
187
- /// # Safety
188
- ///
189
- /// - `name` must be a valid null-terminated C string
190
- /// - `callback` must be a valid function pointer that:
191
- /// - Does not store the image_bytes pointer
192
- /// - Returns a null-terminated UTF-8 string or NULL on error
193
- /// - The returned string must be freeable by kreuzberg_free_string
194
- /// - Returns true on success, false on error (check kreuzberg_last_error)
195
- ///
196
- /// # Example (C)
197
- ///
198
- /// ```c
199
- /// char* my_ocr_backend(const uint8_t* image_bytes, size_t image_length, const char* config_json) {
200
- /// // Implement OCR logic here
201
- /// // Return allocated string with result, or NULL on error
202
- /// return strdup("Extracted text");
203
- /// }
204
- ///
205
- /// bool success = kreuzberg_register_ocr_backend("my-ocr", my_ocr_backend);
206
- /// if (!success) {
207
- /// const char* error = kreuzberg_last_error();
208
- /// printf("Failed to register: %s\n", error);
209
- /// }
210
- /// ```
211
- #[unsafe(no_mangle)]
212
- pub unsafe extern "C" fn kreuzberg_register_ocr_backend(name: *const c_char, callback: OcrBackendCallback) -> bool {
213
- ffi_panic_guard_bool!("kreuzberg_register_ocr_backend", {
214
- clear_last_error();
215
-
216
- if name.is_null() {
217
- set_last_error("Backend name cannot be NULL".to_string());
218
- return false;
219
- }
220
-
221
- let name_str = match unsafe { CStr::from_ptr(name) }.to_str() {
222
- Ok(s) => s,
223
- Err(e) => {
224
- set_last_error(format!("Invalid UTF-8 in backend name: {}", e));
225
- return false;
226
- }
227
- };
228
-
229
- if name_str.is_empty() {
230
- set_last_error("Plugin name cannot be empty".to_string());
231
- return false;
232
- }
233
-
234
- if name_str.chars().any(|c| c.is_whitespace()) {
235
- set_last_error("Plugin name cannot contain whitespace".to_string());
236
- return false;
237
- }
238
-
239
- let backend = Arc::new(FfiOcrBackend::new(name_str.to_string(), callback, None));
240
-
241
- let registry = get_ocr_backend_registry();
242
- let mut registry_guard = match registry.write() {
243
- Ok(guard) => guard,
244
- Err(e) => {
245
- // ~keep: Lock poisoning indicates a panic in another thread holding the lock.
246
- set_last_error(format!("Failed to acquire registry write lock: {}", e));
247
- return false;
248
- }
249
- };
250
-
251
- match registry_guard.register(backend) {
252
- Ok(()) => true,
253
- Err(e) => {
254
- set_last_error(format!("Failed to register OCR backend: {}", e));
255
- false
256
- }
257
- }
258
- })
259
- }
260
-
261
- /// Register a custom OCR backend with explicit language support via FFI callback.
262
- ///
263
- /// # Safety
264
- ///
265
- /// - `languages_json` must be a null-terminated JSON array of language codes or NULL
266
- /// - See `kreuzberg_register_ocr_backend` for additional safety notes.
267
- #[unsafe(no_mangle)]
268
- pub unsafe extern "C" fn kreuzberg_register_ocr_backend_with_languages(
269
- name: *const c_char,
270
- callback: OcrBackendCallback,
271
- languages_json: *const c_char,
272
- ) -> bool {
273
- ffi_panic_guard_bool!("kreuzberg_register_ocr_backend_with_languages", {
274
- clear_last_error();
275
-
276
- if name.is_null() {
277
- set_last_error("Backend name cannot be NULL".to_string());
278
- return false;
279
- }
280
-
281
- let name_str = match unsafe { CStr::from_ptr(name) }.to_str() {
282
- Ok(s) => s,
283
- Err(e) => {
284
- set_last_error(format!("Invalid UTF-8 in backend name: {}", e));
285
- return false;
286
- }
287
- };
288
-
289
- if name_str.is_empty() {
290
- set_last_error("Plugin name cannot be empty".to_string());
291
- return false;
292
- }
293
-
294
- if name_str.chars().any(|c| c.is_whitespace()) {
295
- set_last_error("Plugin name cannot contain whitespace".to_string());
296
- return false;
297
- }
298
-
299
- let supported_languages = match parse_languages_from_json(languages_json) {
300
- Ok(langs) => langs,
301
- Err(e) => {
302
- set_last_error(e);
303
- return false;
304
- }
305
- };
306
-
307
- let backend = Arc::new(FfiOcrBackend::new(name_str.to_string(), callback, supported_languages));
308
-
309
- let registry = get_ocr_backend_registry();
310
- let mut registry_guard = match registry.write() {
311
- Ok(guard) => guard,
312
- Err(e) => {
313
- // ~keep: Lock poisoning indicates a panic in another thread holding the lock.
314
- set_last_error(format!("Failed to acquire registry write lock: {}", e));
315
- return false;
316
- }
317
- };
318
-
319
- match registry_guard.register(backend) {
320
- Ok(()) => true,
321
- Err(e) => {
322
- set_last_error(format!("Failed to register OCR backend: {}", e));
323
- false
324
- }
325
- }
326
- })
327
- }
328
-
329
- /// Unregister an OCR backend by name.
330
- ///
331
- /// # Safety
332
- ///
333
- /// - `name` must be a valid null-terminated C string
334
- /// - Returns true on success, false on error (check kreuzberg_last_error)
335
- ///
336
- /// # Example (C)
337
- ///
338
- /// ```c
339
- /// bool success = kreuzberg_unregister_ocr_backend("custom-ocr");
340
- /// if (!success) {
341
- /// const char* error = kreuzberg_last_error();
342
- /// printf("Failed to unregister: %s\n", error);
343
- /// }
344
- /// ```
345
- #[unsafe(no_mangle)]
346
- pub unsafe extern "C" fn kreuzberg_unregister_ocr_backend(name: *const c_char) -> bool {
347
- ffi_panic_guard_bool!("kreuzberg_unregister_ocr_backend", {
348
- clear_last_error();
349
-
350
- if name.is_null() {
351
- set_last_error("OCR backend name cannot be NULL".to_string());
352
- return false;
353
- }
354
-
355
- let name_str = match unsafe { CStr::from_ptr(name) }.to_str() {
356
- Ok(s) => s,
357
- Err(e) => {
358
- set_last_error(format!("Invalid UTF-8 in OCR backend name: {}", e));
359
- return false;
360
- }
361
- };
362
-
363
- if name_str.is_empty() {
364
- set_last_error("OCR backend name cannot be empty".to_string());
365
- return false;
366
- }
367
-
368
- if name_str.chars().any(|c| c.is_whitespace()) {
369
- set_last_error("OCR backend name cannot contain whitespace".to_string());
370
- return false;
371
- }
372
-
373
- match kreuzberg::plugins::unregister_ocr_backend(name_str) {
374
- Ok(()) => true,
375
- Err(e) => {
376
- set_last_error(e.to_string());
377
- false
378
- }
379
- }
380
- })
381
- }
382
-
383
- /// List all registered OCR backends as a JSON array of names.
384
- ///
385
- /// # Safety
386
- ///
387
- /// - Returned string must be freed with `kreuzberg_free_string`.
388
- /// - Returns NULL on error (check `kreuzberg_last_error`).
389
- ///
390
- /// # Example (C)
391
- ///
392
- /// ```c
393
- /// char* backends = kreuzberg_list_ocr_backends();
394
- /// if (backends == NULL) {
395
- /// const char* error = kreuzberg_last_error();
396
- /// printf("Failed to list backends: %s\n", error);
397
- /// } else {
398
- /// printf("OCR backends: %s\n", backends);
399
- /// kreuzberg_free_string(backends);
400
- /// }
401
- /// ```
402
- #[unsafe(no_mangle)]
403
- pub unsafe extern "C" fn kreuzberg_list_ocr_backends() -> *mut c_char {
404
- ffi_panic_guard!("kreuzberg_list_ocr_backends", {
405
- clear_last_error();
406
-
407
- match kreuzberg::plugins::list_ocr_backends() {
408
- Ok(backends) => match serde_json::to_string(&backends) {
409
- Ok(json) => match CString::new(json) {
410
- Ok(cstr) => cstr.into_raw(),
411
- Err(e) => {
412
- set_last_error(format!("Failed to create C string: {}", e));
413
- ptr::null_mut()
414
- }
415
- },
416
- Err(e) => {
417
- set_last_error(format!("Failed to serialize OCR backend list: {}", e));
418
- ptr::null_mut()
419
- }
420
- },
421
- Err(e) => {
422
- set_last_error(e.to_string());
423
- ptr::null_mut()
424
- }
425
- }
426
- })
427
- }
428
-
429
- /// Clear all registered OCR backends.
430
- ///
431
- /// # Safety
432
- ///
433
- /// - Removes all registered OCR backends. Subsequent extractions will use only built-in backends.
434
- /// - Returns true on success, false on error.
435
- ///
436
- /// # Example (C)
437
- ///
438
- /// ```c
439
- /// bool success = kreuzberg_clear_ocr_backends();
440
- /// if (!success) {
441
- /// const char* error = kreuzberg_last_error();
442
- /// printf("Failed to clear OCR backends: %s\n", error);
443
- /// }
444
- /// ```
445
- #[unsafe(no_mangle)]
446
- pub unsafe extern "C" fn kreuzberg_clear_ocr_backends() -> bool {
447
- ffi_panic_guard_bool!("kreuzberg_clear_ocr_backends", {
448
- clear_last_error();
449
-
450
- match kreuzberg::plugins::clear_ocr_backends() {
451
- Ok(()) => true,
452
- Err(e) => {
453
- set_last_error(e.to_string());
454
- false
455
- }
456
- }
457
- })
458
- }
459
-
460
- /// Get supported languages for an OCR backend.
461
- ///
462
- /// Returns a JSON array of supported language codes for the given backend.
463
- /// Supported backends: "easyocr", "paddleocr", "tesseract"
464
- ///
465
- /// # Safety
466
- ///
467
- /// - The returned string must be freed with `kreuzberg_free_string`
468
- /// - Returns NULL if backend not found or on error (check `kreuzberg_last_error`)
469
- ///
470
- /// # Example (C)
471
- ///
472
- /// ```c
473
- /// char* languages = kreuzberg_get_ocr_languages("easyocr");
474
- /// if (languages != NULL) {
475
- /// printf("EasyOCR languages: %s\n", languages);
476
- /// kreuzberg_free_string(languages);
477
- /// }
478
- /// ```
479
- #[unsafe(no_mangle)]
480
- pub unsafe extern "C" fn kreuzberg_get_ocr_languages(backend: *const c_char) -> *mut c_char {
481
- ffi_panic_guard!("kreuzberg_get_ocr_languages", {
482
- clear_last_error();
483
-
484
- if backend.is_null() {
485
- set_last_error("Backend name cannot be NULL".to_string());
486
- return ptr::null_mut();
487
- }
488
-
489
- let backend_str = match unsafe { CStr::from_ptr(backend) }.to_str() {
490
- Ok(s) => s,
491
- Err(e) => {
492
- set_last_error(format!("Invalid UTF-8 in backend name: {}", e));
493
- return ptr::null_mut();
494
- }
495
- };
496
-
497
- use kreuzberg::ocr::LanguageRegistry;
498
- let registry = LanguageRegistry::global();
499
-
500
- match registry.get_supported_languages(backend_str) {
501
- Some(languages) => match serde_json::to_string(languages) {
502
- Ok(json) => match CString::new(json) {
503
- Ok(cstr) => cstr.into_raw(),
504
- Err(e) => {
505
- set_last_error(format!("Failed to serialize language list: {}", e));
506
- ptr::null_mut()
507
- }
508
- },
509
- Err(e) => {
510
- set_last_error(format!("Failed to serialize language list: {}", e));
511
- ptr::null_mut()
512
- }
513
- },
514
- None => {
515
- set_last_error(format!("Unknown OCR backend: '{}'", backend_str));
516
- ptr::null_mut()
517
- }
518
- }
519
- })
520
- }
521
-
522
- /// Check if a language is supported by an OCR backend.
523
- ///
524
- /// Returns 1 (true) if the language is supported, 0 (false) otherwise.
525
- ///
526
- /// # Arguments
527
- ///
528
- /// * `backend` - Backend name (e.g., "easyocr", "paddleocr", "tesseract")
529
- /// * `language` - Language code to check
530
- ///
531
- /// # Returns
532
- ///
533
- /// 1 if supported, 0 if not supported or backend not found.
534
- ///
535
- /// # Example (C)
536
- ///
537
- /// ```c
538
- /// int is_supported = kreuzberg_is_language_supported("easyocr", "en");
539
- /// if (is_supported) {
540
- /// printf("English is supported by EasyOCR\n");
541
- /// }
542
- /// ```
543
- ///
544
- /// # Safety
545
- ///
546
- /// - `backend` and `language` must be valid pointers to valid UTF-8 C strings.
547
- /// - Both pointers can be checked for NULL; returns 0 if either is NULL.
548
- /// - The C strings must remain valid for the duration of the function call.
549
- #[unsafe(no_mangle)]
550
- pub unsafe extern "C" fn kreuzberg_is_language_supported(backend: *const c_char, language: *const c_char) -> i32 {
551
- ffi_panic_guard_i32!("kreuzberg_is_language_supported", {
552
- clear_last_error();
553
-
554
- if backend.is_null() || language.is_null() {
555
- set_last_error("Backend and language parameters cannot be NULL".to_string());
556
- return 0;
557
- }
558
-
559
- let backend_str = match unsafe { CStr::from_ptr(backend) }.to_str() {
560
- Ok(s) => s,
561
- Err(e) => {
562
- set_last_error(format!("Invalid UTF-8 in backend name: {}", e));
563
- return 0;
564
- }
565
- };
566
-
567
- let language_str = match unsafe { CStr::from_ptr(language) }.to_str() {
568
- Ok(s) => s,
569
- Err(e) => {
570
- set_last_error(format!("Invalid UTF-8 in language code: {}", e));
571
- return 0;
572
- }
573
- };
574
-
575
- use kreuzberg::ocr::LanguageRegistry;
576
- let registry = LanguageRegistry::global();
577
-
578
- if registry.is_language_supported(backend_str, language_str) {
579
- 1
580
- } else {
581
- 0
582
- }
583
- })
584
- }
585
-
586
- /// Get list of all registered OCR backends with language support.
587
- ///
588
- /// Returns a JSON object mapping backend names to language counts.
589
- /// Example: `{"easyocr": 80, "paddleocr": 14, "tesseract": 100}`
590
- ///
591
- /// # Safety
592
- ///
593
- /// - The returned string must be freed with `kreuzberg_free_string`
594
- /// - Returns NULL on error (check `kreuzberg_last_error`)
595
- ///
596
- /// # Example (C)
597
- ///
598
- /// ```c
599
- /// char* backends = kreuzberg_list_ocr_backends_with_languages();
600
- /// if (backends != NULL) {
601
- /// printf("Available backends: %s\n", backends);
602
- /// kreuzberg_free_string(backends);
603
- /// }
604
- /// ```
605
- #[unsafe(no_mangle)]
606
- pub unsafe extern "C" fn kreuzberg_list_ocr_backends_with_languages() -> *mut c_char {
607
- ffi_panic_guard!("kreuzberg_list_ocr_backends_with_languages", {
608
- clear_last_error();
609
-
610
- use kreuzberg::ocr::LanguageRegistry;
611
- let registry = LanguageRegistry::global();
612
- let backends = registry.get_backends();
613
-
614
- let mut backend_map = serde_json::json!({});
615
- for backend in backends {
616
- let count = registry.get_language_count(&backend);
617
- backend_map[&backend] = serde_json::json!(count);
618
- }
619
-
620
- match CString::new(backend_map.to_string()) {
621
- Ok(cstr) => cstr.into_raw(),
622
- Err(e) => {
623
- set_last_error(format!("Failed to serialize backend list: {}", e));
624
- ptr::null_mut()
625
- }
626
- }
627
- })
628
- }