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
@@ -0,0 +1,496 @@
1
+ //! Pandoc server mode for persistent process conversions
2
+ //!
3
+ //! Manages a long-running pandoc-server process to amortize startup costs
4
+ //! across multiple document conversions. Suitable for high-throughput scenarios.
5
+ //!
6
+ //! # Overview
7
+ //!
8
+ //! This module provides low-level control over pandoc-server lifecycle. For most use cases,
9
+ //! prefer [`BatchExtractor`](super::batch::BatchExtractor) which automatically handles
10
+ //! server detection, startup, and fallback.
11
+ //!
12
+ //! # Server Activation
13
+ //!
14
+ //! Pandoc only runs in server mode when invoked as "pandoc-server" (checks argv[0]).
15
+ //! This module handles activation automatically:
16
+ //!
17
+ //! 1. Detects pandoc binary location via `which pandoc`
18
+ //! 2. Creates temporary symlink: `/tmp/pandoc-server-{pid}`
19
+ //! 3. Invokes symlink with `--port` and `--timeout` flags
20
+ //!
21
+ //! # Example
22
+ //!
23
+ //! ```no_run
24
+ //! use kreuzberg::extraction::pandoc::server::PandocServer;
25
+ //!
26
+ //! #[tokio::main]
27
+ //! async fn main() -> kreuzberg::Result<()> {
28
+ //! // Create and start server
29
+ //! let server = PandocServer::new(Some(3030), Some(120)).await?;
30
+ //! server.start().await?;
31
+ //!
32
+ //! // Health check
33
+ //! let version = server.health_check().await?;
34
+ //! println!("Pandoc version: {}", version);
35
+ //!
36
+ //! // Convert document
37
+ //! let markdown = "# Hello World\n\nTest content.";
38
+ //! let html = server.convert(markdown, "markdown", "html").await?;
39
+ //!
40
+ //! // Extract with metadata
41
+ //! let (content, metadata) = server.extract_with_server(markdown, "markdown").await?;
42
+ //!
43
+ //! // Cleanup
44
+ //! server.stop().await?;
45
+ //!
46
+ //! Ok(())
47
+ //! }
48
+ //! ```
49
+ //!
50
+ //! # Thread Safety
51
+ //!
52
+ //! `PandocServer` uses Arc<RwLock<Option<Child>>> for thread-safe process management.
53
+ //! It can be safely shared across async tasks.
54
+
55
+ use crate::error::{KreuzbergError, Result};
56
+ use serde::{Deserialize, Serialize};
57
+ use serde_json::Value;
58
+ use std::collections::HashMap;
59
+ use std::path::{Path, PathBuf};
60
+ use std::process::Stdio;
61
+ use std::sync::Arc;
62
+ use tokio::io::{AsyncBufReadExt, BufReader};
63
+ use tokio::process::{Child, Command};
64
+ use tokio::sync::RwLock;
65
+ use tokio::time::{Duration, sleep, timeout};
66
+
67
+ /// Default port for pandoc-server
68
+ const DEFAULT_PORT: u16 = 3030;
69
+
70
+ /// Default timeout for conversions (seconds)
71
+ const DEFAULT_TIMEOUT: u64 = 120;
72
+
73
+ /// Maximum retries for failed requests
74
+ const MAX_RETRIES: usize = 3;
75
+
76
+ /// Pandoc server request payload
77
+ #[derive(Debug, Serialize)]
78
+ struct PandocRequest {
79
+ text: String,
80
+ #[serde(skip_serializing_if = "Option::is_none")]
81
+ from: Option<String>,
82
+ #[serde(skip_serializing_if = "Option::is_none")]
83
+ to: Option<String>,
84
+ #[serde(skip_serializing_if = "Option::is_none")]
85
+ standalone: Option<bool>,
86
+ }
87
+
88
+ /// Pandoc server response payload
89
+ #[derive(Debug, Deserialize)]
90
+ #[serde(untagged)]
91
+ enum PandocResponse {
92
+ Success {
93
+ output: String,
94
+ #[serde(default)]
95
+ base64: bool,
96
+ },
97
+ Error {
98
+ error: String,
99
+ },
100
+ }
101
+
102
+ /// Pandoc server instance manager
103
+ pub struct PandocServer {
104
+ port: u16,
105
+ timeout_secs: u64,
106
+ process: Arc<RwLock<Option<Child>>>,
107
+ pandoc_path: PathBuf,
108
+ server_symlink: Option<PathBuf>,
109
+ }
110
+
111
+ impl PandocServer {
112
+ /// Create a new Pandoc server manager
113
+ ///
114
+ /// # Arguments
115
+ /// * `port` - HTTP port for the server (default: 3030)
116
+ /// * `timeout_secs` - Conversion timeout in seconds (default: 120)
117
+ pub async fn new(port: Option<u16>, timeout_secs: Option<u64>) -> Result<Self> {
118
+ let port = port.unwrap_or(DEFAULT_PORT);
119
+ let timeout_secs = timeout_secs.unwrap_or(DEFAULT_TIMEOUT);
120
+
121
+ let pandoc_path = Self::find_pandoc().await?;
122
+
123
+ let server_symlink = if pandoc_path.file_name() == Some(std::ffi::OsStr::new("pandoc")) {
124
+ Some(Self::create_server_symlink(&pandoc_path).await?)
125
+ } else {
126
+ None
127
+ };
128
+
129
+ Ok(Self {
130
+ port,
131
+ timeout_secs,
132
+ process: Arc::new(RwLock::new(None)),
133
+ pandoc_path,
134
+ server_symlink,
135
+ })
136
+ }
137
+
138
+ /// Check if pandoc-server is available (pandoc 3.8+)
139
+ pub async fn is_server_available() -> bool {
140
+ if let Ok(output) = Command::new("which").arg("pandoc-server").output().await
141
+ && output.status.success()
142
+ {
143
+ tracing::debug!("Found pandoc-server binary in PATH");
144
+ return true;
145
+ }
146
+
147
+ if let Ok(output) = Command::new("pandoc").arg("--version").output().await
148
+ && let Ok(version_str) = String::from_utf8(output.stdout)
149
+ && let Some(version_line) = version_str.lines().next()
150
+ && let Some(version) = version_line.split_whitespace().nth(1)
151
+ && let Some(major_minor) = version.split('.').take(2).collect::<Vec<_>>().get(0..2)
152
+ && let (Ok(major), Ok(minor)) = (major_minor[0].parse::<u32>(), major_minor[1].parse::<u32>())
153
+ {
154
+ let supports_server = major > 3 || (major == 3 && minor >= 8);
155
+ if supports_server {
156
+ tracing::debug!("Pandoc version {}.{} supports server mode", major, minor);
157
+ } else {
158
+ tracing::debug!(
159
+ "Pandoc version {}.{} does not support server mode (requires 3.8+)",
160
+ major,
161
+ minor
162
+ );
163
+ }
164
+ return supports_server;
165
+ }
166
+
167
+ tracing::debug!("Could not determine pandoc server availability");
168
+ false
169
+ }
170
+
171
+ /// Find pandoc binary in PATH
172
+ async fn find_pandoc() -> Result<PathBuf> {
173
+ let output = Command::new("which")
174
+ .arg("pandoc")
175
+ .output()
176
+ .await
177
+ .map_err(|e| KreuzbergError::MissingDependency(format!("Failed to locate pandoc: {}", e)))?;
178
+
179
+ if !output.status.success() {
180
+ return Err(KreuzbergError::MissingDependency(
181
+ "pandoc not found in PATH".to_string(),
182
+ ));
183
+ }
184
+
185
+ let path_str = String::from_utf8_lossy(&output.stdout);
186
+ let path = PathBuf::from(path_str.trim());
187
+
188
+ if !path.exists() {
189
+ return Err(KreuzbergError::MissingDependency(
190
+ "pandoc binary does not exist".to_string(),
191
+ ));
192
+ }
193
+
194
+ Ok(path)
195
+ }
196
+
197
+ /// Create a temporary symlink to pandoc named "pandoc-server"
198
+ ///
199
+ /// This is required because pandoc only runs in server mode when invoked
200
+ /// as "pandoc-server" (checks argv[0]).
201
+ async fn create_server_symlink(pandoc_path: &Path) -> Result<PathBuf> {
202
+ #[cfg(unix)]
203
+ {
204
+ let temp_dir = std::env::temp_dir();
205
+ let symlink_path = temp_dir.join(format!("pandoc-server-{}", std::process::id()));
206
+
207
+ let _ = tokio::fs::remove_file(&symlink_path).await;
208
+
209
+ tokio::fs::symlink(pandoc_path, &symlink_path).await.map_err(|e| {
210
+ KreuzbergError::Io(std::io::Error::other(format!(
211
+ "Failed to create pandoc-server symlink: {}",
212
+ e
213
+ )))
214
+ })?;
215
+
216
+ Ok(symlink_path)
217
+ }
218
+
219
+ #[cfg(not(unix))]
220
+ {
221
+ let _ = pandoc_path;
222
+ Err(KreuzbergError::validation(
223
+ "Pandoc server mode requires Unix-like system for symlinks",
224
+ ))
225
+ }
226
+ }
227
+
228
+ /// Start the pandoc server process
229
+ pub async fn start(&self) -> Result<()> {
230
+ let mut process_lock = self.process.write().await;
231
+
232
+ if let Some(mut child) = process_lock.take() {
233
+ let _ = child.kill().await;
234
+ }
235
+
236
+ let command_path = self.server_symlink.as_ref().unwrap_or(&self.pandoc_path);
237
+
238
+ let mut child = Command::new(command_path)
239
+ .arg("--port")
240
+ .arg(self.port.to_string())
241
+ .arg("--timeout")
242
+ .arg(self.timeout_secs.to_string())
243
+ .stdout(Stdio::piped())
244
+ .stderr(Stdio::piped())
245
+ .spawn()
246
+ .map_err(|e| KreuzbergError::Io(std::io::Error::other(format!("Failed to start pandoc-server: {}", e))))?;
247
+
248
+ let stdout = child
249
+ .stdout
250
+ .take()
251
+ .ok_or_else(|| KreuzbergError::Io(std::io::Error::other("Failed to capture server stdout")))?;
252
+
253
+ let mut reader = BufReader::new(stdout).lines();
254
+
255
+ match timeout(Duration::from_secs(5), reader.next_line()).await {
256
+ Ok(Ok(Some(line))) if line.contains("Starting server") => {
257
+ *process_lock = Some(child);
258
+ Ok(())
259
+ }
260
+ _ => {
261
+ let _ = child.kill().await;
262
+ Err(KreuzbergError::Io(std::io::Error::other(
263
+ "Pandoc server failed to start within 5 seconds",
264
+ )))
265
+ }
266
+ }
267
+ }
268
+
269
+ /// Stop the pandoc server process
270
+ pub async fn stop(&self) -> Result<()> {
271
+ let mut process_lock = self.process.write().await;
272
+
273
+ if let Some(mut child) = process_lock.take() {
274
+ child.kill().await.map_err(|e| {
275
+ KreuzbergError::Io(std::io::Error::other(format!("Failed to kill pandoc-server: {}", e)))
276
+ })?;
277
+ }
278
+
279
+ if let Some(symlink) = &self.server_symlink {
280
+ let _ = tokio::fs::remove_file(symlink).await;
281
+ }
282
+
283
+ Ok(())
284
+ }
285
+
286
+ /// Check if server is running
287
+ pub async fn is_running(&self) -> bool {
288
+ let process_lock = self.process.read().await;
289
+ process_lock.is_some()
290
+ }
291
+
292
+ /// Perform health check by querying /version endpoint
293
+ pub async fn health_check(&self) -> Result<String> {
294
+ let url = format!("http://localhost:{}/version", self.port);
295
+
296
+ let client = reqwest::Client::builder()
297
+ .timeout(Duration::from_secs(2))
298
+ .build()
299
+ .map_err(|e| KreuzbergError::Io(std::io::Error::other(format!("Failed to create HTTP client: {}", e))))?;
300
+
301
+ let response = client
302
+ .get(&url)
303
+ .send()
304
+ .await
305
+ .map_err(|e| KreuzbergError::Io(std::io::Error::other(format!("Health check failed: {}", e))))?;
306
+
307
+ if !response.status().is_success() {
308
+ return Err(KreuzbergError::Io(std::io::Error::other(format!(
309
+ "Health check returned status: {}",
310
+ response.status()
311
+ ))));
312
+ }
313
+
314
+ let version = response.text().await.map_err(|e| {
315
+ KreuzbergError::Io(std::io::Error::other(format!(
316
+ "Failed to read health check response: {}",
317
+ e
318
+ )))
319
+ })?;
320
+
321
+ Ok(version.trim().to_string())
322
+ }
323
+
324
+ /// Convert document using the pandoc server
325
+ ///
326
+ /// # Arguments
327
+ /// * `content` - Document content as string
328
+ /// * `from_format` - Input format (e.g., "markdown", "docx", "rst")
329
+ /// * `to_format` - Output format (e.g., "json", "markdown", "html")
330
+ pub async fn convert(&self, content: &str, from_format: &str, to_format: &str) -> Result<String> {
331
+ let url = format!("http://localhost:{}/", self.port);
332
+
333
+ let request_body = PandocRequest {
334
+ text: content.to_string(),
335
+ from: Some(from_format.to_string()),
336
+ to: Some(to_format.to_string()),
337
+ standalone: Some(true),
338
+ };
339
+
340
+ let client = reqwest::Client::builder()
341
+ .timeout(Duration::from_secs(self.timeout_secs))
342
+ .build()
343
+ .map_err(|e| KreuzbergError::Io(std::io::Error::other(format!("Failed to create HTTP client: {}", e))))?;
344
+
345
+ for attempt in 0..MAX_RETRIES {
346
+ let response = client
347
+ .post(&url)
348
+ .json(&request_body)
349
+ .header("Accept", "application/json")
350
+ .send()
351
+ .await;
352
+
353
+ match response {
354
+ Ok(resp) => {
355
+ if !resp.status().is_success() {
356
+ if attempt < MAX_RETRIES - 1 {
357
+ sleep(Duration::from_millis(100 * (attempt as u64 + 1))).await;
358
+ continue;
359
+ }
360
+ return Err(KreuzbergError::parsing(format!(
361
+ "Pandoc server returned status: {}",
362
+ resp.status()
363
+ )));
364
+ }
365
+
366
+ let pandoc_response: PandocResponse = resp
367
+ .json()
368
+ .await
369
+ .map_err(|e| KreuzbergError::parsing(format!("Failed to parse server response: {}", e)))?;
370
+
371
+ match pandoc_response {
372
+ PandocResponse::Success { output, base64 } => {
373
+ if base64 {
374
+ let decoded = base64_simd::STANDARD.decode_to_vec(output.as_bytes()).map_err(|e| {
375
+ KreuzbergError::parsing(format!("Failed to decode base64 output: {}", e))
376
+ })?;
377
+ return String::from_utf8(decoded).map_err(|e| {
378
+ KreuzbergError::parsing(format!("Failed to decode UTF-8 output: {}", e))
379
+ });
380
+ }
381
+ return Ok(output);
382
+ }
383
+ PandocResponse::Error { error } => {
384
+ return Err(KreuzbergError::parsing(format!("Pandoc server error: {}", error)));
385
+ }
386
+ }
387
+ }
388
+ Err(e) => {
389
+ if attempt < MAX_RETRIES - 1 {
390
+ sleep(Duration::from_millis(100 * (attempt as u64 + 1))).await;
391
+ continue;
392
+ }
393
+ return Err(KreuzbergError::Io(std::io::Error::other(format!(
394
+ "HTTP request failed after {} retries: {}",
395
+ MAX_RETRIES, e
396
+ ))));
397
+ }
398
+ }
399
+ }
400
+
401
+ Err(KreuzbergError::parsing(
402
+ "Max retries exceeded for pandoc server conversion".to_string(),
403
+ ))
404
+ }
405
+
406
+ /// Extract both content and metadata using server mode
407
+ ///
408
+ /// This is the optimized version that matches our combined extraction approach.
409
+ pub async fn extract_with_server(
410
+ &self,
411
+ content: &str,
412
+ from_format: &str,
413
+ ) -> Result<(String, HashMap<String, Value>)> {
414
+ let json_output = self.convert(content, from_format, "json").await?;
415
+
416
+ let json_data: Value = serde_json::from_str(&json_output)
417
+ .map_err(|e| KreuzbergError::parsing(format!("Failed to parse JSON from server: {}", e)))?;
418
+
419
+ let content = super::subprocess::extract_content_from_json(&json_data)?;
420
+ let metadata = super::subprocess::extract_metadata_from_json(&json_data)?;
421
+
422
+ Ok((content, metadata))
423
+ }
424
+ }
425
+
426
+ impl Drop for PandocServer {
427
+ fn drop(&mut self) {
428
+ if let Some(symlink) = &self.server_symlink {
429
+ let _ = std::fs::remove_file(symlink);
430
+ }
431
+ }
432
+ }
433
+
434
+ #[cfg(test)]
435
+ mod tests {
436
+ use super::*;
437
+
438
+ #[tokio::test]
439
+ async fn test_find_pandoc() {
440
+ let result = PandocServer::find_pandoc().await;
441
+ if result.is_ok() {
442
+ let path = result.unwrap();
443
+ assert!(path.exists());
444
+ assert!(path.ends_with("pandoc"));
445
+ }
446
+ }
447
+
448
+ #[tokio::test]
449
+ async fn test_server_lifecycle() {
450
+ let server = PandocServer::new(Some(3032), Some(120)).await;
451
+ if server.is_err() {
452
+ return;
453
+ }
454
+
455
+ let server = server.unwrap();
456
+
457
+ let start_result = server.start().await;
458
+ if start_result.is_err() {
459
+ return;
460
+ }
461
+
462
+ assert!(server.is_running().await);
463
+
464
+ let version = server.health_check().await;
465
+ assert!(version.is_ok());
466
+
467
+ let stop_result = server.stop().await;
468
+ assert!(stop_result.is_ok());
469
+ assert!(!server.is_running().await);
470
+ }
471
+
472
+ #[tokio::test]
473
+ async fn test_server_conversion() {
474
+ let server = PandocServer::new(Some(3033), Some(120)).await;
475
+ if server.is_err() {
476
+ return;
477
+ }
478
+
479
+ let server = server.unwrap();
480
+
481
+ if server.start().await.is_err() {
482
+ return;
483
+ }
484
+
485
+ let result = server
486
+ .convert("# Hello World\n\nTest paragraph.", "markdown", "html")
487
+ .await;
488
+
489
+ let _ = server.stop().await;
490
+
491
+ if let Ok(html) = result {
492
+ assert!(html.contains("Hello World"));
493
+ assert!(html.contains("Test paragraph"));
494
+ }
495
+ }
496
+ }