kreuzberg 4.0.0.pre.rc.13 → 4.0.0.pre.rc.15

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 (369) hide show
  1. checksums.yaml +4 -4
  2. data/.gitignore +14 -14
  3. data/.rspec +3 -3
  4. data/.rubocop.yaml +1 -1
  5. data/.rubocop.yml +538 -538
  6. data/Gemfile +8 -8
  7. data/Gemfile.lock +104 -2
  8. data/README.md +454 -454
  9. data/Rakefile +33 -25
  10. data/Steepfile +47 -47
  11. data/examples/async_patterns.rb +341 -341
  12. data/ext/kreuzberg_rb/extconf.rb +45 -45
  13. data/ext/kreuzberg_rb/native/.cargo/config.toml +2 -2
  14. data/ext/kreuzberg_rb/native/Cargo.lock +6750 -6941
  15. data/ext/kreuzberg_rb/native/Cargo.toml +53 -54
  16. data/ext/kreuzberg_rb/native/README.md +425 -425
  17. data/ext/kreuzberg_rb/native/build.rs +52 -15
  18. data/ext/kreuzberg_rb/native/include/ieeefp.h +11 -11
  19. data/ext/kreuzberg_rb/native/include/msvc_compat/strings.h +14 -14
  20. data/ext/kreuzberg_rb/native/include/strings.h +20 -20
  21. data/ext/kreuzberg_rb/native/include/unistd.h +47 -47
  22. data/ext/kreuzberg_rb/native/src/lib.rs +3158 -3158
  23. data/extconf.rb +28 -28
  24. data/kreuzberg.gemspec +214 -214
  25. data/lib/kreuzberg/api_proxy.rb +142 -142
  26. data/lib/kreuzberg/cache_api.rb +81 -81
  27. data/lib/kreuzberg/cli.rb +55 -55
  28. data/lib/kreuzberg/cli_proxy.rb +127 -127
  29. data/lib/kreuzberg/config.rb +724 -724
  30. data/lib/kreuzberg/error_context.rb +80 -80
  31. data/lib/kreuzberg/errors.rb +118 -118
  32. data/lib/kreuzberg/extraction_api.rb +340 -340
  33. data/lib/kreuzberg/mcp_proxy.rb +186 -186
  34. data/lib/kreuzberg/ocr_backend_protocol.rb +113 -113
  35. data/lib/kreuzberg/post_processor_protocol.rb +86 -86
  36. data/lib/kreuzberg/result.rb +279 -279
  37. data/lib/kreuzberg/setup_lib_path.rb +80 -80
  38. data/lib/kreuzberg/validator_protocol.rb +89 -89
  39. data/lib/kreuzberg/version.rb +5 -5
  40. data/lib/kreuzberg.rb +109 -109
  41. data/lib/{pdfium.dll → libpdfium.so} +0 -0
  42. data/sig/kreuzberg/internal.rbs +184 -184
  43. data/sig/kreuzberg.rbs +546 -546
  44. data/spec/binding/cache_spec.rb +227 -227
  45. data/spec/binding/cli_proxy_spec.rb +85 -85
  46. data/spec/binding/cli_spec.rb +55 -55
  47. data/spec/binding/config_spec.rb +345 -345
  48. data/spec/binding/config_validation_spec.rb +283 -283
  49. data/spec/binding/error_handling_spec.rb +213 -213
  50. data/spec/binding/errors_spec.rb +66 -66
  51. data/spec/binding/plugins/ocr_backend_spec.rb +307 -307
  52. data/spec/binding/plugins/postprocessor_spec.rb +269 -269
  53. data/spec/binding/plugins/validator_spec.rb +274 -274
  54. data/spec/fixtures/config.toml +39 -39
  55. data/spec/fixtures/config.yaml +41 -41
  56. data/spec/fixtures/invalid_config.toml +4 -4
  57. data/spec/smoke/package_spec.rb +178 -178
  58. data/spec/spec_helper.rb +42 -42
  59. data/vendor/Cargo.toml +2 -2
  60. data/vendor/kreuzberg/Cargo.toml +5 -5
  61. data/vendor/kreuzberg/README.md +230 -230
  62. data/vendor/kreuzberg/benches/otel_overhead.rs +48 -48
  63. data/vendor/kreuzberg/build.rs +887 -843
  64. data/vendor/kreuzberg/src/api/error.rs +81 -81
  65. data/vendor/kreuzberg/src/api/handlers.rs +199 -199
  66. data/vendor/kreuzberg/src/api/mod.rs +87 -79
  67. data/vendor/kreuzberg/src/api/server.rs +353 -353
  68. data/vendor/kreuzberg/src/api/types.rs +170 -170
  69. data/vendor/kreuzberg/src/cache/mod.rs +1167 -1167
  70. data/vendor/kreuzberg/src/chunking/mod.rs +1877 -1877
  71. data/vendor/kreuzberg/src/chunking/processor.rs +220 -220
  72. data/vendor/kreuzberg/src/core/batch_mode.rs +95 -95
  73. data/vendor/kreuzberg/src/core/config.rs +1080 -1080
  74. data/vendor/kreuzberg/src/core/extractor.rs +1156 -1156
  75. data/vendor/kreuzberg/src/core/io.rs +329 -329
  76. data/vendor/kreuzberg/src/core/mime.rs +605 -605
  77. data/vendor/kreuzberg/src/core/mod.rs +47 -47
  78. data/vendor/kreuzberg/src/core/pipeline.rs +1184 -1184
  79. data/vendor/kreuzberg/src/embeddings.rs +500 -500
  80. data/vendor/kreuzberg/src/error.rs +431 -431
  81. data/vendor/kreuzberg/src/extraction/archive.rs +954 -954
  82. data/vendor/kreuzberg/src/extraction/docx.rs +398 -398
  83. data/vendor/kreuzberg/src/extraction/email.rs +854 -854
  84. data/vendor/kreuzberg/src/extraction/excel.rs +688 -688
  85. data/vendor/kreuzberg/src/extraction/html.rs +634 -601
  86. data/vendor/kreuzberg/src/extraction/image.rs +491 -491
  87. data/vendor/kreuzberg/src/extraction/libreoffice.rs +574 -574
  88. data/vendor/kreuzberg/src/extraction/markdown.rs +213 -213
  89. data/vendor/kreuzberg/src/extraction/mod.rs +81 -81
  90. data/vendor/kreuzberg/src/extraction/office_metadata/app_properties.rs +398 -398
  91. data/vendor/kreuzberg/src/extraction/office_metadata/core_properties.rs +247 -247
  92. data/vendor/kreuzberg/src/extraction/office_metadata/custom_properties.rs +240 -240
  93. data/vendor/kreuzberg/src/extraction/office_metadata/mod.rs +130 -130
  94. data/vendor/kreuzberg/src/extraction/office_metadata/odt_properties.rs +284 -284
  95. data/vendor/kreuzberg/src/extraction/pptx.rs +3100 -3100
  96. data/vendor/kreuzberg/src/extraction/structured.rs +490 -490
  97. data/vendor/kreuzberg/src/extraction/table.rs +328 -328
  98. data/vendor/kreuzberg/src/extraction/text.rs +269 -269
  99. data/vendor/kreuzberg/src/extraction/xml.rs +333 -333
  100. data/vendor/kreuzberg/src/extractors/archive.rs +447 -447
  101. data/vendor/kreuzberg/src/extractors/bibtex.rs +470 -470
  102. data/vendor/kreuzberg/src/extractors/docbook.rs +504 -504
  103. data/vendor/kreuzberg/src/extractors/docx.rs +400 -400
  104. data/vendor/kreuzberg/src/extractors/email.rs +157 -157
  105. data/vendor/kreuzberg/src/extractors/epub.rs +708 -708
  106. data/vendor/kreuzberg/src/extractors/excel.rs +345 -345
  107. data/vendor/kreuzberg/src/extractors/fictionbook.rs +492 -492
  108. data/vendor/kreuzberg/src/extractors/html.rs +407 -407
  109. data/vendor/kreuzberg/src/extractors/image.rs +219 -219
  110. data/vendor/kreuzberg/src/extractors/jats.rs +1054 -1054
  111. data/vendor/kreuzberg/src/extractors/jupyter.rs +368 -368
  112. data/vendor/kreuzberg/src/extractors/latex.rs +653 -653
  113. data/vendor/kreuzberg/src/extractors/markdown.rs +701 -701
  114. data/vendor/kreuzberg/src/extractors/mod.rs +429 -429
  115. data/vendor/kreuzberg/src/extractors/odt.rs +628 -628
  116. data/vendor/kreuzberg/src/extractors/opml.rs +635 -635
  117. data/vendor/kreuzberg/src/extractors/orgmode.rs +529 -529
  118. data/vendor/kreuzberg/src/extractors/pdf.rs +749 -749
  119. data/vendor/kreuzberg/src/extractors/pptx.rs +267 -267
  120. data/vendor/kreuzberg/src/extractors/rst.rs +577 -577
  121. data/vendor/kreuzberg/src/extractors/rtf.rs +809 -809
  122. data/vendor/kreuzberg/src/extractors/security.rs +484 -484
  123. data/vendor/kreuzberg/src/extractors/security_tests.rs +367 -367
  124. data/vendor/kreuzberg/src/extractors/structured.rs +142 -142
  125. data/vendor/kreuzberg/src/extractors/text.rs +265 -265
  126. data/vendor/kreuzberg/src/extractors/typst.rs +651 -651
  127. data/vendor/kreuzberg/src/extractors/xml.rs +147 -147
  128. data/vendor/kreuzberg/src/image/dpi.rs +164 -164
  129. data/vendor/kreuzberg/src/image/mod.rs +6 -6
  130. data/vendor/kreuzberg/src/image/preprocessing.rs +417 -417
  131. data/vendor/kreuzberg/src/image/resize.rs +89 -89
  132. data/vendor/kreuzberg/src/keywords/config.rs +154 -154
  133. data/vendor/kreuzberg/src/keywords/mod.rs +237 -237
  134. data/vendor/kreuzberg/src/keywords/processor.rs +275 -275
  135. data/vendor/kreuzberg/src/keywords/rake.rs +293 -293
  136. data/vendor/kreuzberg/src/keywords/types.rs +68 -68
  137. data/vendor/kreuzberg/src/keywords/yake.rs +163 -163
  138. data/vendor/kreuzberg/src/language_detection/mod.rs +985 -985
  139. data/vendor/kreuzberg/src/language_detection/processor.rs +219 -219
  140. data/vendor/kreuzberg/src/lib.rs +113 -113
  141. data/vendor/kreuzberg/src/mcp/mod.rs +35 -35
  142. data/vendor/kreuzberg/src/mcp/server.rs +2076 -2076
  143. data/vendor/kreuzberg/src/ocr/cache.rs +469 -469
  144. data/vendor/kreuzberg/src/ocr/error.rs +37 -37
  145. data/vendor/kreuzberg/src/ocr/hocr.rs +216 -216
  146. data/vendor/kreuzberg/src/ocr/mod.rs +58 -58
  147. data/vendor/kreuzberg/src/ocr/processor.rs +863 -863
  148. data/vendor/kreuzberg/src/ocr/table/mod.rs +4 -4
  149. data/vendor/kreuzberg/src/ocr/table/tsv_parser.rs +144 -144
  150. data/vendor/kreuzberg/src/ocr/tesseract_backend.rs +452 -452
  151. data/vendor/kreuzberg/src/ocr/types.rs +393 -393
  152. data/vendor/kreuzberg/src/ocr/utils.rs +47 -47
  153. data/vendor/kreuzberg/src/ocr/validation.rs +206 -206
  154. data/vendor/kreuzberg/src/panic_context.rs +154 -154
  155. data/vendor/kreuzberg/src/pdf/bindings.rs +44 -44
  156. data/vendor/kreuzberg/src/pdf/bundled.rs +452 -346
  157. data/vendor/kreuzberg/src/pdf/error.rs +130 -130
  158. data/vendor/kreuzberg/src/pdf/images.rs +139 -139
  159. data/vendor/kreuzberg/src/pdf/metadata.rs +489 -489
  160. data/vendor/kreuzberg/src/pdf/mod.rs +68 -68
  161. data/vendor/kreuzberg/src/pdf/rendering.rs +368 -368
  162. data/vendor/kreuzberg/src/pdf/table.rs +420 -420
  163. data/vendor/kreuzberg/src/pdf/text.rs +240 -240
  164. data/vendor/kreuzberg/src/plugins/extractor.rs +1044 -1044
  165. data/vendor/kreuzberg/src/plugins/mod.rs +212 -212
  166. data/vendor/kreuzberg/src/plugins/ocr.rs +639 -639
  167. data/vendor/kreuzberg/src/plugins/processor.rs +650 -650
  168. data/vendor/kreuzberg/src/plugins/registry.rs +1339 -1339
  169. data/vendor/kreuzberg/src/plugins/traits.rs +258 -258
  170. data/vendor/kreuzberg/src/plugins/validator.rs +967 -967
  171. data/vendor/kreuzberg/src/stopwords/mod.rs +1470 -1470
  172. data/vendor/kreuzberg/src/text/mod.rs +25 -25
  173. data/vendor/kreuzberg/src/text/quality.rs +697 -697
  174. data/vendor/kreuzberg/src/text/quality_processor.rs +219 -219
  175. data/vendor/kreuzberg/src/text/string_utils.rs +217 -217
  176. data/vendor/kreuzberg/src/text/token_reduction/cjk_utils.rs +164 -164
  177. data/vendor/kreuzberg/src/text/token_reduction/config.rs +100 -100
  178. data/vendor/kreuzberg/src/text/token_reduction/core.rs +796 -796
  179. data/vendor/kreuzberg/src/text/token_reduction/filters.rs +902 -902
  180. data/vendor/kreuzberg/src/text/token_reduction/mod.rs +160 -160
  181. data/vendor/kreuzberg/src/text/token_reduction/semantic.rs +619 -619
  182. data/vendor/kreuzberg/src/text/token_reduction/simd_text.rs +147 -147
  183. data/vendor/kreuzberg/src/types.rs +1055 -1055
  184. data/vendor/kreuzberg/src/utils/mod.rs +17 -17
  185. data/vendor/kreuzberg/src/utils/quality.rs +959 -959
  186. data/vendor/kreuzberg/src/utils/string_utils.rs +381 -381
  187. data/vendor/kreuzberg/stopwords/af_stopwords.json +53 -53
  188. data/vendor/kreuzberg/stopwords/ar_stopwords.json +482 -482
  189. data/vendor/kreuzberg/stopwords/bg_stopwords.json +261 -261
  190. data/vendor/kreuzberg/stopwords/bn_stopwords.json +400 -400
  191. data/vendor/kreuzberg/stopwords/br_stopwords.json +1205 -1205
  192. data/vendor/kreuzberg/stopwords/ca_stopwords.json +280 -280
  193. data/vendor/kreuzberg/stopwords/cs_stopwords.json +425 -425
  194. data/vendor/kreuzberg/stopwords/da_stopwords.json +172 -172
  195. data/vendor/kreuzberg/stopwords/de_stopwords.json +622 -622
  196. data/vendor/kreuzberg/stopwords/el_stopwords.json +849 -849
  197. data/vendor/kreuzberg/stopwords/en_stopwords.json +1300 -1300
  198. data/vendor/kreuzberg/stopwords/eo_stopwords.json +175 -175
  199. data/vendor/kreuzberg/stopwords/es_stopwords.json +734 -734
  200. data/vendor/kreuzberg/stopwords/et_stopwords.json +37 -37
  201. data/vendor/kreuzberg/stopwords/eu_stopwords.json +100 -100
  202. data/vendor/kreuzberg/stopwords/fa_stopwords.json +801 -801
  203. data/vendor/kreuzberg/stopwords/fi_stopwords.json +849 -849
  204. data/vendor/kreuzberg/stopwords/fr_stopwords.json +693 -693
  205. data/vendor/kreuzberg/stopwords/ga_stopwords.json +111 -111
  206. data/vendor/kreuzberg/stopwords/gl_stopwords.json +162 -162
  207. data/vendor/kreuzberg/stopwords/gu_stopwords.json +226 -226
  208. data/vendor/kreuzberg/stopwords/ha_stopwords.json +41 -41
  209. data/vendor/kreuzberg/stopwords/he_stopwords.json +196 -196
  210. data/vendor/kreuzberg/stopwords/hi_stopwords.json +227 -227
  211. data/vendor/kreuzberg/stopwords/hr_stopwords.json +181 -181
  212. data/vendor/kreuzberg/stopwords/hu_stopwords.json +791 -791
  213. data/vendor/kreuzberg/stopwords/hy_stopwords.json +47 -47
  214. data/vendor/kreuzberg/stopwords/id_stopwords.json +760 -760
  215. data/vendor/kreuzberg/stopwords/it_stopwords.json +634 -634
  216. data/vendor/kreuzberg/stopwords/ja_stopwords.json +136 -136
  217. data/vendor/kreuzberg/stopwords/kn_stopwords.json +84 -84
  218. data/vendor/kreuzberg/stopwords/ko_stopwords.json +681 -681
  219. data/vendor/kreuzberg/stopwords/ku_stopwords.json +64 -64
  220. data/vendor/kreuzberg/stopwords/la_stopwords.json +51 -51
  221. data/vendor/kreuzberg/stopwords/lt_stopwords.json +476 -476
  222. data/vendor/kreuzberg/stopwords/lv_stopwords.json +163 -163
  223. data/vendor/kreuzberg/stopwords/ml_stopwords.json +1 -1
  224. data/vendor/kreuzberg/stopwords/mr_stopwords.json +101 -101
  225. data/vendor/kreuzberg/stopwords/ms_stopwords.json +477 -477
  226. data/vendor/kreuzberg/stopwords/ne_stopwords.json +490 -490
  227. data/vendor/kreuzberg/stopwords/nl_stopwords.json +415 -415
  228. data/vendor/kreuzberg/stopwords/no_stopwords.json +223 -223
  229. data/vendor/kreuzberg/stopwords/pl_stopwords.json +331 -331
  230. data/vendor/kreuzberg/stopwords/pt_stopwords.json +562 -562
  231. data/vendor/kreuzberg/stopwords/ro_stopwords.json +436 -436
  232. data/vendor/kreuzberg/stopwords/ru_stopwords.json +561 -561
  233. data/vendor/kreuzberg/stopwords/si_stopwords.json +193 -193
  234. data/vendor/kreuzberg/stopwords/sk_stopwords.json +420 -420
  235. data/vendor/kreuzberg/stopwords/sl_stopwords.json +448 -448
  236. data/vendor/kreuzberg/stopwords/so_stopwords.json +32 -32
  237. data/vendor/kreuzberg/stopwords/st_stopwords.json +33 -33
  238. data/vendor/kreuzberg/stopwords/sv_stopwords.json +420 -420
  239. data/vendor/kreuzberg/stopwords/sw_stopwords.json +76 -76
  240. data/vendor/kreuzberg/stopwords/ta_stopwords.json +129 -129
  241. data/vendor/kreuzberg/stopwords/te_stopwords.json +54 -54
  242. data/vendor/kreuzberg/stopwords/th_stopwords.json +118 -118
  243. data/vendor/kreuzberg/stopwords/tl_stopwords.json +149 -149
  244. data/vendor/kreuzberg/stopwords/tr_stopwords.json +506 -506
  245. data/vendor/kreuzberg/stopwords/uk_stopwords.json +75 -75
  246. data/vendor/kreuzberg/stopwords/ur_stopwords.json +519 -519
  247. data/vendor/kreuzberg/stopwords/vi_stopwords.json +647 -647
  248. data/vendor/kreuzberg/stopwords/yo_stopwords.json +62 -62
  249. data/vendor/kreuzberg/stopwords/zh_stopwords.json +796 -796
  250. data/vendor/kreuzberg/stopwords/zu_stopwords.json +31 -31
  251. data/vendor/kreuzberg/tests/api_extract_multipart.rs +52 -52
  252. data/vendor/kreuzberg/tests/api_tests.rs +966 -966
  253. data/vendor/kreuzberg/tests/archive_integration.rs +545 -545
  254. data/vendor/kreuzberg/tests/batch_orchestration.rs +556 -556
  255. data/vendor/kreuzberg/tests/batch_processing.rs +318 -318
  256. data/vendor/kreuzberg/tests/bibtex_parity_test.rs +421 -421
  257. data/vendor/kreuzberg/tests/concurrency_stress.rs +533 -533
  258. data/vendor/kreuzberg/tests/config_features.rs +612 -612
  259. data/vendor/kreuzberg/tests/config_loading_tests.rs +416 -416
  260. data/vendor/kreuzberg/tests/core_integration.rs +510 -510
  261. data/vendor/kreuzberg/tests/csv_integration.rs +414 -414
  262. data/vendor/kreuzberg/tests/docbook_extractor_tests.rs +500 -500
  263. data/vendor/kreuzberg/tests/docx_metadata_extraction_test.rs +122 -122
  264. data/vendor/kreuzberg/tests/docx_vs_pandoc_comparison.rs +370 -370
  265. data/vendor/kreuzberg/tests/email_integration.rs +327 -327
  266. data/vendor/kreuzberg/tests/epub_native_extractor_tests.rs +275 -275
  267. data/vendor/kreuzberg/tests/error_handling.rs +402 -402
  268. data/vendor/kreuzberg/tests/fictionbook_extractor_tests.rs +228 -228
  269. data/vendor/kreuzberg/tests/format_integration.rs +165 -164
  270. data/vendor/kreuzberg/tests/helpers/mod.rs +142 -142
  271. data/vendor/kreuzberg/tests/html_table_test.rs +551 -551
  272. data/vendor/kreuzberg/tests/image_integration.rs +255 -255
  273. data/vendor/kreuzberg/tests/instrumentation_test.rs +139 -139
  274. data/vendor/kreuzberg/tests/jats_extractor_tests.rs +639 -639
  275. data/vendor/kreuzberg/tests/jupyter_extractor_tests.rs +704 -704
  276. data/vendor/kreuzberg/tests/keywords_integration.rs +479 -479
  277. data/vendor/kreuzberg/tests/keywords_quality.rs +509 -509
  278. data/vendor/kreuzberg/tests/latex_extractor_tests.rs +496 -496
  279. data/vendor/kreuzberg/tests/markdown_extractor_tests.rs +490 -490
  280. data/vendor/kreuzberg/tests/mime_detection.rs +429 -429
  281. data/vendor/kreuzberg/tests/ocr_configuration.rs +514 -514
  282. data/vendor/kreuzberg/tests/ocr_errors.rs +698 -698
  283. data/vendor/kreuzberg/tests/ocr_quality.rs +629 -629
  284. data/vendor/kreuzberg/tests/ocr_stress.rs +469 -469
  285. data/vendor/kreuzberg/tests/odt_extractor_tests.rs +674 -674
  286. data/vendor/kreuzberg/tests/opml_extractor_tests.rs +616 -616
  287. data/vendor/kreuzberg/tests/orgmode_extractor_tests.rs +822 -822
  288. data/vendor/kreuzberg/tests/pdf_integration.rs +45 -45
  289. data/vendor/kreuzberg/tests/pdfium_linking.rs +374 -374
  290. data/vendor/kreuzberg/tests/pipeline_integration.rs +1436 -1436
  291. data/vendor/kreuzberg/tests/plugin_ocr_backend_test.rs +776 -776
  292. data/vendor/kreuzberg/tests/plugin_postprocessor_test.rs +560 -560
  293. data/vendor/kreuzberg/tests/plugin_system.rs +927 -927
  294. data/vendor/kreuzberg/tests/plugin_validator_test.rs +783 -783
  295. data/vendor/kreuzberg/tests/registry_integration_tests.rs +587 -587
  296. data/vendor/kreuzberg/tests/rst_extractor_tests.rs +694 -694
  297. data/vendor/kreuzberg/tests/rtf_extractor_tests.rs +775 -775
  298. data/vendor/kreuzberg/tests/security_validation.rs +416 -416
  299. data/vendor/kreuzberg/tests/stopwords_integration_test.rs +888 -888
  300. data/vendor/kreuzberg/tests/test_fastembed.rs +631 -631
  301. data/vendor/kreuzberg/tests/typst_behavioral_tests.rs +1260 -1260
  302. data/vendor/kreuzberg/tests/typst_extractor_tests.rs +648 -648
  303. data/vendor/kreuzberg/tests/xlsx_metadata_extraction_test.rs +87 -87
  304. data/vendor/kreuzberg-tesseract/.commitlintrc.json +13 -13
  305. data/vendor/kreuzberg-tesseract/.crate-ignore +2 -2
  306. data/vendor/kreuzberg-tesseract/Cargo.lock +2933 -2933
  307. data/vendor/kreuzberg-tesseract/Cargo.toml +2 -2
  308. data/vendor/kreuzberg-tesseract/LICENSE +22 -22
  309. data/vendor/kreuzberg-tesseract/README.md +399 -399
  310. data/vendor/kreuzberg-tesseract/build.rs +1354 -1354
  311. data/vendor/kreuzberg-tesseract/patches/README.md +71 -71
  312. data/vendor/kreuzberg-tesseract/patches/tesseract.diff +199 -199
  313. data/vendor/kreuzberg-tesseract/src/api.rs +1371 -1371
  314. data/vendor/kreuzberg-tesseract/src/choice_iterator.rs +77 -77
  315. data/vendor/kreuzberg-tesseract/src/enums.rs +297 -297
  316. data/vendor/kreuzberg-tesseract/src/error.rs +81 -81
  317. data/vendor/kreuzberg-tesseract/src/lib.rs +145 -145
  318. data/vendor/kreuzberg-tesseract/src/monitor.rs +57 -57
  319. data/vendor/kreuzberg-tesseract/src/mutable_iterator.rs +197 -197
  320. data/vendor/kreuzberg-tesseract/src/page_iterator.rs +253 -253
  321. data/vendor/kreuzberg-tesseract/src/result_iterator.rs +286 -286
  322. data/vendor/kreuzberg-tesseract/src/result_renderer.rs +183 -183
  323. data/vendor/kreuzberg-tesseract/tests/integration_test.rs +211 -211
  324. data/vendor/rb-sys/.cargo_vcs_info.json +5 -5
  325. data/vendor/rb-sys/Cargo.lock +393 -393
  326. data/vendor/rb-sys/Cargo.toml +70 -70
  327. data/vendor/rb-sys/Cargo.toml.orig +57 -57
  328. data/vendor/rb-sys/LICENSE-APACHE +190 -190
  329. data/vendor/rb-sys/LICENSE-MIT +21 -21
  330. data/vendor/rb-sys/build/features.rs +111 -111
  331. data/vendor/rb-sys/build/main.rs +286 -286
  332. data/vendor/rb-sys/build/stable_api_config.rs +155 -155
  333. data/vendor/rb-sys/build/version.rs +50 -50
  334. data/vendor/rb-sys/readme.md +36 -36
  335. data/vendor/rb-sys/src/bindings.rs +21 -21
  336. data/vendor/rb-sys/src/hidden.rs +11 -11
  337. data/vendor/rb-sys/src/lib.rs +35 -35
  338. data/vendor/rb-sys/src/macros.rs +371 -371
  339. data/vendor/rb-sys/src/memory.rs +53 -53
  340. data/vendor/rb-sys/src/ruby_abi_version.rs +38 -38
  341. data/vendor/rb-sys/src/special_consts.rs +31 -31
  342. data/vendor/rb-sys/src/stable_api/compiled.c +179 -179
  343. data/vendor/rb-sys/src/stable_api/compiled.rs +257 -257
  344. data/vendor/rb-sys/src/stable_api/ruby_2_7.rs +324 -324
  345. data/vendor/rb-sys/src/stable_api/ruby_3_0.rs +332 -332
  346. data/vendor/rb-sys/src/stable_api/ruby_3_1.rs +325 -325
  347. data/vendor/rb-sys/src/stable_api/ruby_3_2.rs +323 -323
  348. data/vendor/rb-sys/src/stable_api/ruby_3_3.rs +339 -339
  349. data/vendor/rb-sys/src/stable_api/ruby_3_4.rs +339 -339
  350. data/vendor/rb-sys/src/stable_api.rs +260 -260
  351. data/vendor/rb-sys/src/symbol.rs +31 -31
  352. data/vendor/rb-sys/src/tracking_allocator.rs +330 -330
  353. data/vendor/rb-sys/src/utils.rs +89 -89
  354. data/vendor/rb-sys/src/value_type.rs +7 -7
  355. metadata +81 -22
  356. data/vendor/kreuzberg-ffi/Cargo.toml +0 -63
  357. data/vendor/kreuzberg-ffi/README.md +0 -851
  358. data/vendor/kreuzberg-ffi/build.rs +0 -176
  359. data/vendor/kreuzberg-ffi/cbindgen.toml +0 -27
  360. data/vendor/kreuzberg-ffi/kreuzberg-ffi-install.pc +0 -12
  361. data/vendor/kreuzberg-ffi/kreuzberg-ffi.pc.in +0 -12
  362. data/vendor/kreuzberg-ffi/kreuzberg.h +0 -1087
  363. data/vendor/kreuzberg-ffi/src/lib.rs +0 -3616
  364. data/vendor/kreuzberg-ffi/src/panic_shield.rs +0 -247
  365. data/vendor/kreuzberg-ffi/tests.disabled/README.md +0 -48
  366. data/vendor/kreuzberg-ffi/tests.disabled/config_loading_tests.rs +0 -299
  367. data/vendor/kreuzberg-ffi/tests.disabled/config_tests.rs +0 -346
  368. data/vendor/kreuzberg-ffi/tests.disabled/extractor_tests.rs +0 -232
  369. data/vendor/kreuzberg-ffi/tests.disabled/plugin_registration_tests.rs +0 -470
@@ -1,53 +1,53 @@
1
- /// Prevents premature destruction of local objects.
2
- ///
3
- /// Ruby's garbage collector is conservative; it scans the C level machine stack as well.
4
- /// Possible in-use Ruby objects must remain visible on stack, to be properly marked as such.
5
- /// However, Rust's compiler optimizations might remove the references to these objects from
6
- /// the stack when they are not being used directly.
7
- ///
8
- /// Consider the following example:
9
- ///
10
- /// ```ignore
11
- /// use rb_sys::{rb_str_new_cstr, rb_str_cat_cstr, RSTRING_PTR, rb_gc_guard};
12
- ///
13
- /// unsafe {
14
- /// let s = rb_str_new_cstr(" world\0".as_ptr() as _);
15
- /// let sptr = RSTRING_PTR(s);
16
- /// let t = rb_str_new_cstr("hello,\0".as_ptr() as _); // Possible GC invocation
17
- /// let u = rb_str_cat_cstr(t, sptr);
18
- /// rb_gc_guard!(s); // ensure `s` (and thus `sptr`) do not get GC-ed
19
- /// }
20
- /// ```
21
- ///
22
- /// In this example, without the `rb_gc_guard!`, the last use of `s` is before the last use
23
- /// of `sptr`. Compilers could think `s` and `t` are allowed to overlap. That would
24
- /// eliminate `s` from the stack, while `sptr` is still in use. If our GC runs at that
25
- /// very moment, `s` gets swept out, which also destroys `sptr`.
26
- ///
27
- /// In order to prevent this scenario, `rb_gc_guard!` must be placed after the last use
28
- /// of `sptr`. Placing `rb_gc_guard!` before dereferencing `sptr` would be of no use.
29
- ///
30
- /// Using the `rb_gc_guard!` macro has the following advantages:
31
- ///
32
- /// - the intent of the macro use is clear.
33
- ///
34
- /// - `rb_gc_guard!` only affects its call site, without negatively affecting other systems.
35
- ///
36
- /// # Example
37
- /// ```no_run
38
- /// use rb_sys::{rb_utf8_str_new_cstr, rb_gc_guard};
39
- ///
40
- /// let my_string = unsafe { rb_utf8_str_new_cstr("hello world\0".as_ptr() as _) };
41
- /// let _ = rb_gc_guard!(my_string);
42
- /// ```
43
- #[macro_export]
44
- macro_rules! rb_gc_guard {
45
- ($v:expr) => {{
46
- unsafe {
47
- let val: $crate::VALUE = $v;
48
- let rb_gc_guarded_ptr = std::ptr::read_volatile(&&val);
49
- std::arch::asm!("/* {0} */", in(reg) rb_gc_guarded_ptr);
50
- *rb_gc_guarded_ptr
51
- }
52
- }};
53
- }
1
+ /// Prevents premature destruction of local objects.
2
+ ///
3
+ /// Ruby's garbage collector is conservative; it scans the C level machine stack as well.
4
+ /// Possible in-use Ruby objects must remain visible on stack, to be properly marked as such.
5
+ /// However, Rust's compiler optimizations might remove the references to these objects from
6
+ /// the stack when they are not being used directly.
7
+ ///
8
+ /// Consider the following example:
9
+ ///
10
+ /// ```ignore
11
+ /// use rb_sys::{rb_str_new_cstr, rb_str_cat_cstr, RSTRING_PTR, rb_gc_guard};
12
+ ///
13
+ /// unsafe {
14
+ /// let s = rb_str_new_cstr(" world\0".as_ptr() as _);
15
+ /// let sptr = RSTRING_PTR(s);
16
+ /// let t = rb_str_new_cstr("hello,\0".as_ptr() as _); // Possible GC invocation
17
+ /// let u = rb_str_cat_cstr(t, sptr);
18
+ /// rb_gc_guard!(s); // ensure `s` (and thus `sptr`) do not get GC-ed
19
+ /// }
20
+ /// ```
21
+ ///
22
+ /// In this example, without the `rb_gc_guard!`, the last use of `s` is before the last use
23
+ /// of `sptr`. Compilers could think `s` and `t` are allowed to overlap. That would
24
+ /// eliminate `s` from the stack, while `sptr` is still in use. If our GC runs at that
25
+ /// very moment, `s` gets swept out, which also destroys `sptr`.
26
+ ///
27
+ /// In order to prevent this scenario, `rb_gc_guard!` must be placed after the last use
28
+ /// of `sptr`. Placing `rb_gc_guard!` before dereferencing `sptr` would be of no use.
29
+ ///
30
+ /// Using the `rb_gc_guard!` macro has the following advantages:
31
+ ///
32
+ /// - the intent of the macro use is clear.
33
+ ///
34
+ /// - `rb_gc_guard!` only affects its call site, without negatively affecting other systems.
35
+ ///
36
+ /// # Example
37
+ /// ```no_run
38
+ /// use rb_sys::{rb_utf8_str_new_cstr, rb_gc_guard};
39
+ ///
40
+ /// let my_string = unsafe { rb_utf8_str_new_cstr("hello world\0".as_ptr() as _) };
41
+ /// let _ = rb_gc_guard!(my_string);
42
+ /// ```
43
+ #[macro_export]
44
+ macro_rules! rb_gc_guard {
45
+ ($v:expr) => {{
46
+ unsafe {
47
+ let val: $crate::VALUE = $v;
48
+ let rb_gc_guarded_ptr = std::ptr::read_volatile(&&val);
49
+ std::arch::asm!("/* {0} */", in(reg) rb_gc_guarded_ptr);
50
+ *rb_gc_guarded_ptr
51
+ }
52
+ }};
53
+ }
@@ -1,38 +1,38 @@
1
- //! Helper to define the `ruby_abi_version` function needed for extensions.
2
- //!
3
- //! Since Ruby 3.2, gems are required to define a `ruby_abi_version` function.
4
- //! For C extensions, this is done transparently by including `ruby.h`, but for
5
- //! Rust we have to define it ourselves. This is enabled automatically by when
6
- //! compiling a gem.
7
-
8
- #[doc(hidden)]
9
- #[cfg(not(has_ruby_abi_version))]
10
- pub const __RB_SYS_RUBY_ABI_VERSION: std::os::raw::c_ulonglong = 0;
11
-
12
- #[doc(hidden)]
13
- #[cfg(has_ruby_abi_version)]
14
- pub const __RB_SYS_RUBY_ABI_VERSION: std::os::raw::c_ulonglong = crate::RUBY_ABI_VERSION as _;
15
-
16
- #[doc(hidden)]
17
- #[no_mangle]
18
- #[allow(unused)]
19
- pub extern "C" fn ruby_abi_version() -> std::os::raw::c_ulonglong {
20
- __RB_SYS_RUBY_ABI_VERSION
21
- }
22
-
23
- #[doc(hidden)]
24
- #[no_mangle]
25
- #[allow(unused)]
26
- #[cfg(ruby_engine = "truffleruby")]
27
- pub extern "C" fn rb_tr_abi_version() -> *const std::os::raw::c_char {
28
- crate::TRUFFLERUBY_ABI_VERSION.as_ptr() as *const _
29
- }
30
-
31
- #[deprecated(
32
- since = "0.9.102",
33
- note = "You no longer need to invoke this macro, the `ruby_abi_version` function is defined automatically."
34
- )]
35
- #[macro_export]
36
- macro_rules! ruby_abi_version {
37
- () => {};
38
- }
1
+ //! Helper to define the `ruby_abi_version` function needed for extensions.
2
+ //!
3
+ //! Since Ruby 3.2, gems are required to define a `ruby_abi_version` function.
4
+ //! For C extensions, this is done transparently by including `ruby.h`, but for
5
+ //! Rust we have to define it ourselves. This is enabled automatically by when
6
+ //! compiling a gem.
7
+
8
+ #[doc(hidden)]
9
+ #[cfg(not(has_ruby_abi_version))]
10
+ pub const __RB_SYS_RUBY_ABI_VERSION: std::os::raw::c_ulonglong = 0;
11
+
12
+ #[doc(hidden)]
13
+ #[cfg(has_ruby_abi_version)]
14
+ pub const __RB_SYS_RUBY_ABI_VERSION: std::os::raw::c_ulonglong = crate::RUBY_ABI_VERSION as _;
15
+
16
+ #[doc(hidden)]
17
+ #[no_mangle]
18
+ #[allow(unused)]
19
+ pub extern "C" fn ruby_abi_version() -> std::os::raw::c_ulonglong {
20
+ __RB_SYS_RUBY_ABI_VERSION
21
+ }
22
+
23
+ #[doc(hidden)]
24
+ #[no_mangle]
25
+ #[allow(unused)]
26
+ #[cfg(ruby_engine = "truffleruby")]
27
+ pub extern "C" fn rb_tr_abi_version() -> *const std::os::raw::c_char {
28
+ crate::TRUFFLERUBY_ABI_VERSION.as_ptr() as *const _
29
+ }
30
+
31
+ #[deprecated(
32
+ since = "0.9.102",
33
+ note = "You no longer need to invoke this macro, the `ruby_abi_version` function is defined automatically."
34
+ )]
35
+ #[macro_export]
36
+ macro_rules! ruby_abi_version {
37
+ () => {};
38
+ }
@@ -1,31 +1,31 @@
1
- #![allow(rustdoc::broken_intra_doc_links)]
2
- #![allow(non_upper_case_globals)]
3
- #![allow(non_snake_case)]
4
-
5
- //! Definitions for Ruby's special constants.
6
- //!
7
- //! Makes it easier to reference important Ruby constants, without having to dig
8
- //! around in bindgen's output.
9
-
10
- use std::ffi::c_long;
11
-
12
- use crate::{ruby_special_consts, VALUE};
13
-
14
- pub const Qfalse: ruby_special_consts = ruby_special_consts::RUBY_Qfalse;
15
- pub const Qtrue: ruby_special_consts = ruby_special_consts::RUBY_Qtrue;
16
- pub const Qnil: ruby_special_consts = ruby_special_consts::RUBY_Qnil;
17
- pub const Qundef: ruby_special_consts = ruby_special_consts::RUBY_Qundef;
18
- pub const IMMEDIATE_MASK: ruby_special_consts = ruby_special_consts::RUBY_IMMEDIATE_MASK;
19
- pub const FIXNUM_FLAG: ruby_special_consts = ruby_special_consts::RUBY_FIXNUM_FLAG;
20
- pub const FIXNUM_MIN: c_long = c_long::MIN / 2;
21
- pub const FIXNUM_MAX: c_long = c_long::MAX / 2;
22
- pub const FLONUM_MASK: ruby_special_consts = ruby_special_consts::RUBY_FLONUM_MASK;
23
- pub const FLONUM_FLAG: ruby_special_consts = ruby_special_consts::RUBY_FLONUM_FLAG;
24
- pub const SYMBOL_FLAG: ruby_special_consts = ruby_special_consts::RUBY_SYMBOL_FLAG;
25
-
26
- #[allow(clippy::from_over_into)]
27
- impl Into<VALUE> for ruby_special_consts {
28
- fn into(self) -> VALUE {
29
- self as VALUE
30
- }
31
- }
1
+ #![allow(rustdoc::broken_intra_doc_links)]
2
+ #![allow(non_upper_case_globals)]
3
+ #![allow(non_snake_case)]
4
+
5
+ //! Definitions for Ruby's special constants.
6
+ //!
7
+ //! Makes it easier to reference important Ruby constants, without having to dig
8
+ //! around in bindgen's output.
9
+
10
+ use std::ffi::c_long;
11
+
12
+ use crate::{ruby_special_consts, VALUE};
13
+
14
+ pub const Qfalse: ruby_special_consts = ruby_special_consts::RUBY_Qfalse;
15
+ pub const Qtrue: ruby_special_consts = ruby_special_consts::RUBY_Qtrue;
16
+ pub const Qnil: ruby_special_consts = ruby_special_consts::RUBY_Qnil;
17
+ pub const Qundef: ruby_special_consts = ruby_special_consts::RUBY_Qundef;
18
+ pub const IMMEDIATE_MASK: ruby_special_consts = ruby_special_consts::RUBY_IMMEDIATE_MASK;
19
+ pub const FIXNUM_FLAG: ruby_special_consts = ruby_special_consts::RUBY_FIXNUM_FLAG;
20
+ pub const FIXNUM_MIN: c_long = c_long::MIN / 2;
21
+ pub const FIXNUM_MAX: c_long = c_long::MAX / 2;
22
+ pub const FLONUM_MASK: ruby_special_consts = ruby_special_consts::RUBY_FLONUM_MASK;
23
+ pub const FLONUM_FLAG: ruby_special_consts = ruby_special_consts::RUBY_FLONUM_FLAG;
24
+ pub const SYMBOL_FLAG: ruby_special_consts = ruby_special_consts::RUBY_SYMBOL_FLAG;
25
+
26
+ #[allow(clippy::from_over_into)]
27
+ impl Into<VALUE> for ruby_special_consts {
28
+ fn into(self) -> VALUE {
29
+ self as VALUE
30
+ }
31
+ }
@@ -1,179 +1,179 @@
1
- #include "ruby.h"
2
- #include "ruby/ruby.h"
3
- #include "ruby/intern.h"
4
- #include "ruby/version.h"
5
-
6
- // Check that we have Ruby API version macros
7
- #if !defined(RUBY_API_VERSION_MAJOR)
8
- #error "RUBY_API_VERSION_MAJOR is not defined, Ruby headers might not be configured correctly"
9
- #endif
10
-
11
- #if !defined(RUBY_API_VERSION_MINOR)
12
- #error "RUBY_API_VERSION_MINOR is not defined, Ruby headers might not be configured correctly"
13
- #endif
14
-
15
- // Define a macro to check for Ruby 3.3+
16
- #if RUBY_API_VERSION_MAJOR > 3 || (RUBY_API_VERSION_MAJOR == 3 && RUBY_API_VERSION_MINOR >= 3)
17
- #define RUBY_VERSION_AT_LEAST_3_3 1
18
- #else
19
- #define RUBY_VERSION_AT_LEAST_3_3 0
20
- #endif
21
-
22
- long impl_rstring_len(VALUE obj)
23
- {
24
- return RSTRING_LEN(obj);
25
- }
26
-
27
- char *
28
- impl_rstring_ptr(VALUE obj)
29
- {
30
- return RSTRING_PTR(obj);
31
- }
32
-
33
- long impl_rarray_len(VALUE obj)
34
- {
35
- return RARRAY_LEN(obj);
36
- }
37
-
38
- const VALUE *
39
- impl_rarray_const_ptr(VALUE obj)
40
- {
41
- return RARRAY_CONST_PTR(obj);
42
- }
43
-
44
- VALUE
45
- impl_rbasic_class(VALUE obj)
46
- {
47
- return RBASIC_CLASS(obj);
48
- }
49
-
50
- int impl_frozen_p(VALUE obj)
51
- {
52
- return RB_OBJ_FROZEN(obj);
53
- }
54
-
55
- int impl_special_const_p(VALUE obj)
56
- {
57
- return SPECIAL_CONST_P(obj);
58
- }
59
-
60
- int impl_bignum_positive_p(VALUE obj)
61
- {
62
- return RBIGNUM_POSITIVE_P(obj);
63
- }
64
-
65
- int impl_bignum_negative_p(VALUE obj)
66
- {
67
- return RBIGNUM_NEGATIVE_P(obj);
68
- }
69
-
70
- enum ruby_value_type
71
- impl_builtin_type(VALUE obj)
72
- {
73
- return RB_BUILTIN_TYPE(obj);
74
- }
75
-
76
- int impl_nil_p(VALUE obj)
77
- {
78
- return NIL_P(obj);
79
- }
80
-
81
- int impl_fixnum_p(VALUE obj)
82
- {
83
- return FIXNUM_P(obj);
84
- }
85
-
86
- int impl_static_sym_p(VALUE obj)
87
- {
88
- return STATIC_SYM_P(obj);
89
- }
90
-
91
- int impl_flonum_p(VALUE obj)
92
- {
93
- return FLONUM_P(obj);
94
- }
95
-
96
- int impl_immediate_p(VALUE obj)
97
- {
98
- return IMMEDIATE_P(obj);
99
- }
100
-
101
- int impl_rb_test(VALUE obj)
102
- {
103
- return RB_TEST(obj);
104
- }
105
-
106
- int impl_type_p(VALUE obj, enum ruby_value_type type)
107
- {
108
- return RB_TYPE_P(obj, type);
109
- }
110
-
111
- int impl_dynamic_sym_p(VALUE obj)
112
- {
113
- return RB_DYNAMIC_SYM_P(obj);
114
- }
115
-
116
- int impl_symbol_p(VALUE obj)
117
- {
118
- return RB_SYMBOL_P(obj);
119
- }
120
-
121
- int impl_float_type_p(VALUE obj)
122
- {
123
- return RB_FLOAT_TYPE_P(obj);
124
- }
125
-
126
- enum ruby_value_type
127
- impl_rb_type(VALUE obj)
128
- {
129
- return rb_type(obj);
130
- }
131
-
132
- int impl_integer_type_p(VALUE obj)
133
- {
134
- return RB_INTEGER_TYPE_P(obj);
135
- }
136
-
137
- int impl_rstring_interned_p(VALUE obj)
138
- {
139
- Check_Type(obj, T_STRING);
140
-
141
- return !(FL_TEST(obj, RSTRING_FSTR) == 0);
142
- }
143
-
144
- void impl_thread_sleep(struct timeval time)
145
- {
146
- rb_thread_wait_for(time);
147
- }
148
-
149
- // RTypedData implementations
150
- int impl_rtypeddata_p(VALUE obj)
151
- {
152
- return RTYPEDDATA_P(obj);
153
- }
154
-
155
- int impl_rtypeddata_embedded_p(VALUE obj)
156
- {
157
- #if RUBY_VERSION_AT_LEAST_3_3
158
- return RTYPEDDATA_EMBEDDED_P(obj);
159
- #else
160
- // On Ruby versions before 3.3, embedded typed data is not supported
161
- return 0;
162
- #endif
163
- }
164
-
165
- const struct rb_data_type_struct *
166
- impl_rtypeddata_type(VALUE obj)
167
- {
168
- return RTYPEDDATA_TYPE(obj);
169
- }
170
-
171
- void *
172
- impl_rtypeddata_get_data(VALUE obj)
173
- {
174
- #if RUBY_VERSION_AT_LEAST_3_3
175
- return RTYPEDDATA_GET_DATA(obj);
176
- #else
177
- return RTYPEDDATA(obj)->data;
178
- #endif
179
- }
1
+ #include "ruby.h"
2
+ #include "ruby/ruby.h"
3
+ #include "ruby/intern.h"
4
+ #include "ruby/version.h"
5
+
6
+ // Check that we have Ruby API version macros
7
+ #if !defined(RUBY_API_VERSION_MAJOR)
8
+ #error "RUBY_API_VERSION_MAJOR is not defined, Ruby headers might not be configured correctly"
9
+ #endif
10
+
11
+ #if !defined(RUBY_API_VERSION_MINOR)
12
+ #error "RUBY_API_VERSION_MINOR is not defined, Ruby headers might not be configured correctly"
13
+ #endif
14
+
15
+ // Define a macro to check for Ruby 3.3+
16
+ #if RUBY_API_VERSION_MAJOR > 3 || (RUBY_API_VERSION_MAJOR == 3 && RUBY_API_VERSION_MINOR >= 3)
17
+ #define RUBY_VERSION_AT_LEAST_3_3 1
18
+ #else
19
+ #define RUBY_VERSION_AT_LEAST_3_3 0
20
+ #endif
21
+
22
+ long impl_rstring_len(VALUE obj)
23
+ {
24
+ return RSTRING_LEN(obj);
25
+ }
26
+
27
+ char *
28
+ impl_rstring_ptr(VALUE obj)
29
+ {
30
+ return RSTRING_PTR(obj);
31
+ }
32
+
33
+ long impl_rarray_len(VALUE obj)
34
+ {
35
+ return RARRAY_LEN(obj);
36
+ }
37
+
38
+ const VALUE *
39
+ impl_rarray_const_ptr(VALUE obj)
40
+ {
41
+ return RARRAY_CONST_PTR(obj);
42
+ }
43
+
44
+ VALUE
45
+ impl_rbasic_class(VALUE obj)
46
+ {
47
+ return RBASIC_CLASS(obj);
48
+ }
49
+
50
+ int impl_frozen_p(VALUE obj)
51
+ {
52
+ return RB_OBJ_FROZEN(obj);
53
+ }
54
+
55
+ int impl_special_const_p(VALUE obj)
56
+ {
57
+ return SPECIAL_CONST_P(obj);
58
+ }
59
+
60
+ int impl_bignum_positive_p(VALUE obj)
61
+ {
62
+ return RBIGNUM_POSITIVE_P(obj);
63
+ }
64
+
65
+ int impl_bignum_negative_p(VALUE obj)
66
+ {
67
+ return RBIGNUM_NEGATIVE_P(obj);
68
+ }
69
+
70
+ enum ruby_value_type
71
+ impl_builtin_type(VALUE obj)
72
+ {
73
+ return RB_BUILTIN_TYPE(obj);
74
+ }
75
+
76
+ int impl_nil_p(VALUE obj)
77
+ {
78
+ return NIL_P(obj);
79
+ }
80
+
81
+ int impl_fixnum_p(VALUE obj)
82
+ {
83
+ return FIXNUM_P(obj);
84
+ }
85
+
86
+ int impl_static_sym_p(VALUE obj)
87
+ {
88
+ return STATIC_SYM_P(obj);
89
+ }
90
+
91
+ int impl_flonum_p(VALUE obj)
92
+ {
93
+ return FLONUM_P(obj);
94
+ }
95
+
96
+ int impl_immediate_p(VALUE obj)
97
+ {
98
+ return IMMEDIATE_P(obj);
99
+ }
100
+
101
+ int impl_rb_test(VALUE obj)
102
+ {
103
+ return RB_TEST(obj);
104
+ }
105
+
106
+ int impl_type_p(VALUE obj, enum ruby_value_type type)
107
+ {
108
+ return RB_TYPE_P(obj, type);
109
+ }
110
+
111
+ int impl_dynamic_sym_p(VALUE obj)
112
+ {
113
+ return RB_DYNAMIC_SYM_P(obj);
114
+ }
115
+
116
+ int impl_symbol_p(VALUE obj)
117
+ {
118
+ return RB_SYMBOL_P(obj);
119
+ }
120
+
121
+ int impl_float_type_p(VALUE obj)
122
+ {
123
+ return RB_FLOAT_TYPE_P(obj);
124
+ }
125
+
126
+ enum ruby_value_type
127
+ impl_rb_type(VALUE obj)
128
+ {
129
+ return rb_type(obj);
130
+ }
131
+
132
+ int impl_integer_type_p(VALUE obj)
133
+ {
134
+ return RB_INTEGER_TYPE_P(obj);
135
+ }
136
+
137
+ int impl_rstring_interned_p(VALUE obj)
138
+ {
139
+ Check_Type(obj, T_STRING);
140
+
141
+ return !(FL_TEST(obj, RSTRING_FSTR) == 0);
142
+ }
143
+
144
+ void impl_thread_sleep(struct timeval time)
145
+ {
146
+ rb_thread_wait_for(time);
147
+ }
148
+
149
+ // RTypedData implementations
150
+ int impl_rtypeddata_p(VALUE obj)
151
+ {
152
+ return RTYPEDDATA_P(obj);
153
+ }
154
+
155
+ int impl_rtypeddata_embedded_p(VALUE obj)
156
+ {
157
+ #if RUBY_VERSION_AT_LEAST_3_3
158
+ return RTYPEDDATA_EMBEDDED_P(obj);
159
+ #else
160
+ // On Ruby versions before 3.3, embedded typed data is not supported
161
+ return 0;
162
+ #endif
163
+ }
164
+
165
+ const struct rb_data_type_struct *
166
+ impl_rtypeddata_type(VALUE obj)
167
+ {
168
+ return RTYPEDDATA_TYPE(obj);
169
+ }
170
+
171
+ void *
172
+ impl_rtypeddata_get_data(VALUE obj)
173
+ {
174
+ #if RUBY_VERSION_AT_LEAST_3_3
175
+ return RTYPEDDATA_GET_DATA(obj);
176
+ #else
177
+ return RTYPEDDATA(obj)->data;
178
+ #endif
179
+ }