@loaders.gl/parquet 5.0.0-alpha.2 → 5.0.0-alpha.3

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 (326) hide show
  1. package/LICENSE +1 -1
  2. package/dist/bundled.d.ts +1 -1
  3. package/dist/bundled.d.ts.map +1 -1
  4. package/dist/bundled.js.map +1 -1
  5. package/dist/delta-format.d.ts +9 -0
  6. package/dist/delta-format.d.ts.map +1 -0
  7. package/dist/delta-format.js +12 -0
  8. package/dist/delta-format.js.map +1 -0
  9. package/dist/delta-source-loader-types.d.ts +27 -0
  10. package/dist/delta-source-loader-types.d.ts.map +1 -0
  11. package/dist/delta-source-loader-types.js +28 -0
  12. package/dist/delta-source-loader-types.js.map +1 -0
  13. package/dist/delta-source.d.ts +48 -0
  14. package/dist/delta-source.d.ts.map +1 -0
  15. package/dist/delta-source.js +222 -0
  16. package/dist/delta-source.js.map +1 -0
  17. package/dist/delta-types.d.ts +38 -0
  18. package/dist/delta-types.d.ts.map +1 -0
  19. package/dist/delta-types.js +5 -0
  20. package/dist/delta-types.js.map +1 -0
  21. package/dist/geoparquet-loader.js +1 -1
  22. package/dist/iceberg-rest-catalog.d.ts +49 -0
  23. package/dist/iceberg-rest-catalog.d.ts.map +1 -0
  24. package/dist/iceberg-rest-catalog.js +78 -0
  25. package/dist/iceberg-rest-catalog.js.map +1 -0
  26. package/dist/iceberg-table-source.d.ts +63 -0
  27. package/dist/iceberg-table-source.d.ts.map +1 -0
  28. package/dist/iceberg-table-source.js +797 -0
  29. package/dist/iceberg-table-source.js.map +1 -0
  30. package/dist/iceberg-types.d.ts +150 -0
  31. package/dist/iceberg-types.d.ts.map +1 -0
  32. package/dist/iceberg-types.js +5 -0
  33. package/dist/iceberg-types.js.map +1 -0
  34. package/dist/index.cjs +15005 -3143
  35. package/dist/index.cjs.map +3 -3
  36. package/dist/index.d.ts +7 -3
  37. package/dist/index.d.ts.map +1 -1
  38. package/dist/index.js +5 -1
  39. package/dist/index.js.map +1 -1
  40. package/dist/lib/arrow/convert-schema-from-parquet.d.ts +3 -1
  41. package/dist/lib/arrow/convert-schema-from-parquet.d.ts.map +1 -1
  42. package/dist/lib/arrow/convert-schema-from-parquet.js +94 -19
  43. package/dist/lib/arrow/convert-schema-from-parquet.js.map +1 -1
  44. package/dist/lib/arrow/create-nested-arrow-vector.js +89 -0
  45. package/dist/lib/arrow/create-nested-arrow-vector.js.map +1 -1
  46. package/dist/lib/constants.js +1 -1
  47. package/dist/lib/encoders/encode-table-to-parquet-js.d.ts.map +1 -1
  48. package/dist/lib/encoders/encode-table-to-parquet-js.js +202 -7
  49. package/dist/lib/encoders/encode-table-to-parquet-js.js.map +1 -1
  50. package/dist/lib/geo/geoparquet-covering.d.ts +19 -0
  51. package/dist/lib/geo/geoparquet-covering.d.ts.map +1 -0
  52. package/dist/lib/geo/geoparquet-covering.js +137 -0
  53. package/dist/lib/geo/geoparquet-covering.js.map +1 -0
  54. package/dist/lib/geo/geospatial-metadata.d.ts.map +1 -1
  55. package/dist/lib/geo/geospatial-metadata.js +39 -13
  56. package/dist/lib/geo/geospatial-metadata.js.map +1 -1
  57. package/dist/lib/parquet-bloom-filter-planner.d.ts +16 -0
  58. package/dist/lib/parquet-bloom-filter-planner.d.ts.map +1 -0
  59. package/dist/lib/parquet-bloom-filter-planner.js +57 -0
  60. package/dist/lib/parquet-bloom-filter-planner.js.map +1 -0
  61. package/dist/lib/parquet-bloom-filter.d.ts +30 -0
  62. package/dist/lib/parquet-bloom-filter.d.ts.map +1 -0
  63. package/dist/lib/parquet-bloom-filter.js +289 -0
  64. package/dist/lib/parquet-bloom-filter.js.map +1 -0
  65. package/dist/lib/parquet-encryption.d.ts +80 -0
  66. package/dist/lib/parquet-encryption.d.ts.map +1 -0
  67. package/dist/lib/parquet-encryption.js +190 -0
  68. package/dist/lib/parquet-encryption.js.map +1 -0
  69. package/dist/lib/parquet-page-index.d.ts +34 -5
  70. package/dist/lib/parquet-page-index.d.ts.map +1 -1
  71. package/dist/lib/parquet-page-index.js +114 -28
  72. package/dist/lib/parquet-page-index.js.map +1 -1
  73. package/dist/lib/parquet-predicate.d.ts +5 -1
  74. package/dist/lib/parquet-predicate.d.ts.map +1 -1
  75. package/dist/lib/parquet-predicate.js +42 -14
  76. package/dist/lib/parquet-predicate.js.map +1 -1
  77. package/dist/lib/parquet-source-worker-decoder.d.ts +4 -1
  78. package/dist/lib/parquet-source-worker-decoder.d.ts.map +1 -1
  79. package/dist/lib/parquet-source-worker-decoder.js +83 -17
  80. package/dist/lib/parquet-source-worker-decoder.js.map +1 -1
  81. package/dist/lib/parquet-source-worker-types.d.ts +24 -0
  82. package/dist/lib/parquet-source-worker-types.d.ts.map +1 -1
  83. package/dist/lib/parsers/get-parquet-schema.d.ts.map +1 -1
  84. package/dist/lib/parsers/get-parquet-schema.js +3 -1
  85. package/dist/lib/parsers/get-parquet-schema.js.map +1 -1
  86. package/dist/lib/parsers/parse-parquet-to-arrow-js.d.ts.map +1 -1
  87. package/dist/lib/parsers/parse-parquet-to-arrow-js.js +327 -18
  88. package/dist/lib/parsers/parse-parquet-to-arrow-js.js.map +1 -1
  89. package/dist/lib/parsers/parse-parquet-to-arrow.d.ts.map +1 -1
  90. package/dist/lib/parsers/parse-parquet-to-arrow.js +5 -5
  91. package/dist/lib/parsers/parse-parquet-to-arrow.js.map +1 -1
  92. package/dist/lib/parsers/parse-parquet-to-columns.d.ts.map +1 -1
  93. package/dist/lib/parsers/parse-parquet-to-columns.js +3 -1
  94. package/dist/lib/parsers/parse-parquet-to-columns.js.map +1 -1
  95. package/dist/lib/parsers/parse-parquet-to-json.d.ts.map +1 -1
  96. package/dist/lib/parsers/parse-parquet-to-json.js +18 -2
  97. package/dist/lib/parsers/parse-parquet-to-json.js.map +1 -1
  98. package/dist/lib/slice-parquet-batch.d.ts +9 -0
  99. package/dist/lib/slice-parquet-batch.d.ts.map +1 -0
  100. package/dist/lib/slice-parquet-batch.js +28 -0
  101. package/dist/lib/slice-parquet-batch.js.map +1 -0
  102. package/dist/lib/sources/parquet-range-file.d.ts +1 -1
  103. package/dist/lib/sources/parquet-range-file.d.ts.map +1 -1
  104. package/dist/lib/sources/parquet-range-file.js +28 -22
  105. package/dist/lib/sources/parquet-range-file.js.map +1 -1
  106. package/dist/lib/utils/normalize-parquet-options.d.ts.map +1 -1
  107. package/dist/lib/utils/normalize-parquet-options.js +15 -1
  108. package/dist/lib/utils/normalize-parquet-options.js.map +1 -1
  109. package/dist/parquet-dataset-source.cjs +15766 -4896
  110. package/dist/parquet-dataset-source.cjs.map +3 -3
  111. package/dist/parquet-dataset-source.d.ts +31 -4
  112. package/dist/parquet-dataset-source.d.ts.map +1 -1
  113. package/dist/parquet-dataset-source.js +296 -74
  114. package/dist/parquet-dataset-source.js.map +1 -1
  115. package/dist/parquet-js-loader-types.d.ts +2 -0
  116. package/dist/parquet-js-loader-types.d.ts.map +1 -1
  117. package/dist/parquet-js-loader-types.js +2 -1
  118. package/dist/parquet-js-loader-types.js.map +1 -1
  119. package/dist/parquet-js-loader.d.ts +1 -0
  120. package/dist/parquet-js-loader.d.ts.map +1 -1
  121. package/dist/parquet-js-loader.js +2 -1
  122. package/dist/parquet-js-loader.js.map +1 -1
  123. package/dist/parquet-js-writer.d.ts +23 -2
  124. package/dist/parquet-js-writer.d.ts.map +1 -1
  125. package/dist/parquet-js-writer.js +1 -1
  126. package/dist/parquet-js-writer.js.map +1 -1
  127. package/dist/parquet-loader-base.js +1 -1
  128. package/dist/parquet-loader-options.d.ts +11 -0
  129. package/dist/parquet-loader-options.d.ts.map +1 -1
  130. package/dist/parquet-loader-options.js.map +1 -1
  131. package/dist/parquet-source-capabilities.d.ts +8 -1
  132. package/dist/parquet-source-capabilities.d.ts.map +1 -1
  133. package/dist/parquet-source-capabilities.js +10 -0
  134. package/dist/parquet-source-capabilities.js.map +1 -1
  135. package/dist/parquet-source-loader-types.d.ts +10 -0
  136. package/dist/parquet-source-loader-types.d.ts.map +1 -1
  137. package/dist/parquet-source-loader-types.js +11 -1
  138. package/dist/parquet-source-loader-types.js.map +1 -1
  139. package/dist/parquet-source-loader.cjs +14911 -4316
  140. package/dist/parquet-source-loader.cjs.map +3 -3
  141. package/dist/parquet-source-loader.d.ts +42 -4
  142. package/dist/parquet-source-loader.d.ts.map +1 -1
  143. package/dist/parquet-source-loader.js +585 -38
  144. package/dist/parquet-source-loader.js.map +1 -1
  145. package/dist/parquet-source-types.d.ts +180 -30
  146. package/dist/parquet-source-types.d.ts.map +1 -1
  147. package/dist/parquet-source-worker.js +7055 -6989
  148. package/dist/parquet-source-worker.js.map +4 -4
  149. package/dist/parquet-worker-node.cjs +8505 -1041
  150. package/dist/parquet-worker-node.cjs.map +4 -4
  151. package/dist/parquet-worker-url.js +5 -1
  152. package/dist/parquet-worker-url.js.map +1 -1
  153. package/dist/parquet-worker.js +289 -218
  154. package/dist/parquet-worker.js.map +4 -4
  155. package/dist/parquetjs/codecs/declare.d.ts +15 -0
  156. package/dist/parquetjs/codecs/declare.d.ts.map +1 -1
  157. package/dist/parquetjs/codecs/declare.js +10 -0
  158. package/dist/parquetjs/codecs/declare.js.map +1 -1
  159. package/dist/parquetjs/codecs/delta.d.ts.map +1 -1
  160. package/dist/parquetjs/codecs/delta.js +356 -7
  161. package/dist/parquetjs/codecs/delta.js.map +1 -1
  162. package/dist/parquetjs/codecs/index.d.ts.map +1 -1
  163. package/dist/parquetjs/codecs/index.js +4 -0
  164. package/dist/parquetjs/codecs/index.js.map +1 -1
  165. package/dist/parquetjs/codecs/plain.d.ts +1 -1
  166. package/dist/parquetjs/codecs/plain.d.ts.map +1 -1
  167. package/dist/parquetjs/codecs/plain.js +153 -12
  168. package/dist/parquetjs/codecs/plain.js.map +1 -1
  169. package/dist/parquetjs/codecs/rle.d.ts +4 -0
  170. package/dist/parquetjs/codecs/rle.d.ts.map +1 -1
  171. package/dist/parquetjs/codecs/rle.js +406 -3
  172. package/dist/parquetjs/codecs/rle.js.map +1 -1
  173. package/dist/parquetjs/compression.d.ts +9 -0
  174. package/dist/parquetjs/compression.d.ts.map +1 -1
  175. package/dist/parquetjs/compression.js +86 -67
  176. package/dist/parquetjs/compression.js.map +1 -1
  177. package/dist/parquetjs/encoder/dictionary-planner.d.ts.map +1 -1
  178. package/dist/parquetjs/encoder/dictionary-planner.js +5 -2
  179. package/dist/parquetjs/encoder/dictionary-planner.js.map +1 -1
  180. package/dist/parquetjs/encoder/parquet-encoder.d.ts +43 -1
  181. package/dist/parquetjs/encoder/parquet-encoder.d.ts.map +1 -1
  182. package/dist/parquetjs/encoder/parquet-encoder.js +691 -34
  183. package/dist/parquetjs/encoder/parquet-encoder.js.map +1 -1
  184. package/dist/parquetjs/parquet-thrift/BoundingBox.d.ts +29 -0
  185. package/dist/parquetjs/parquet-thrift/BoundingBox.d.ts.map +1 -0
  186. package/dist/parquetjs/parquet-thrift/BoundingBox.js +85 -0
  187. package/dist/parquetjs/parquet-thrift/BoundingBox.js.map +1 -0
  188. package/dist/parquetjs/parquet-thrift/ColumnChunk.d.ts +5 -0
  189. package/dist/parquetjs/parquet-thrift/ColumnChunk.d.ts.map +1 -1
  190. package/dist/parquetjs/parquet-thrift/ColumnChunk.js +35 -0
  191. package/dist/parquetjs/parquet-thrift/ColumnChunk.js.map +1 -1
  192. package/dist/parquetjs/parquet-thrift/ColumnCryptoMetaData.d.ts +35 -0
  193. package/dist/parquetjs/parquet-thrift/ColumnCryptoMetaData.d.ts.map +1 -0
  194. package/dist/parquetjs/parquet-thrift/ColumnCryptoMetaData.js +135 -0
  195. package/dist/parquetjs/parquet-thrift/ColumnCryptoMetaData.js.map +1 -0
  196. package/dist/parquetjs/parquet-thrift/ColumnMetaData.d.ts +10 -0
  197. package/dist/parquetjs/parquet-thrift/ColumnMetaData.d.ts.map +1 -1
  198. package/dist/parquetjs/parquet-thrift/ColumnMetaData.js +73 -0
  199. package/dist/parquetjs/parquet-thrift/ColumnMetaData.js.map +1 -1
  200. package/dist/parquetjs/parquet-thrift/EncryptionAlgorithm.d.ts +22 -0
  201. package/dist/parquetjs/parquet-thrift/EncryptionAlgorithm.d.ts.map +1 -0
  202. package/dist/parquetjs/parquet-thrift/EncryptionAlgorithm.js +111 -0
  203. package/dist/parquetjs/parquet-thrift/EncryptionAlgorithm.js.map +1 -0
  204. package/dist/parquetjs/parquet-thrift/FileCryptoMetaData.d.ts +16 -0
  205. package/dist/parquetjs/parquet-thrift/FileCryptoMetaData.d.ts.map +1 -0
  206. package/dist/parquetjs/parquet-thrift/FileCryptoMetaData.js +57 -0
  207. package/dist/parquetjs/parquet-thrift/FileCryptoMetaData.js.map +1 -0
  208. package/dist/parquetjs/parquet-thrift/FileMetaData.d.ts +5 -0
  209. package/dist/parquetjs/parquet-thrift/FileMetaData.d.ts.map +1 -1
  210. package/dist/parquetjs/parquet-thrift/FileMetaData.js +35 -0
  211. package/dist/parquetjs/parquet-thrift/FileMetaData.js.map +1 -1
  212. package/dist/parquetjs/parquet-thrift/GeospatialStatistics.d.ts +18 -0
  213. package/dist/parquetjs/parquet-thrift/GeospatialStatistics.d.ts.map +1 -0
  214. package/dist/parquetjs/parquet-thrift/GeospatialStatistics.js +62 -0
  215. package/dist/parquetjs/parquet-thrift/GeospatialStatistics.js.map +1 -0
  216. package/dist/parquetjs/parquet-thrift/SizeStatistics.d.ts +26 -0
  217. package/dist/parquetjs/parquet-thrift/SizeStatistics.d.ts.map +1 -0
  218. package/dist/parquetjs/parquet-thrift/SizeStatistics.js +85 -0
  219. package/dist/parquetjs/parquet-thrift/SizeStatistics.js.map +1 -0
  220. package/dist/parquetjs/parquet-thrift/index.d.ts +6 -0
  221. package/dist/parquetjs/parquet-thrift/index.d.ts.map +1 -1
  222. package/dist/parquetjs/parquet-thrift/index.js +6 -0
  223. package/dist/parquetjs/parquet-thrift/index.js.map +1 -1
  224. package/dist/parquetjs/parser/decoders.d.ts +10 -1
  225. package/dist/parquetjs/parser/decoders.d.ts.map +1 -1
  226. package/dist/parquetjs/parser/decoders.js +318 -14
  227. package/dist/parquetjs/parser/decoders.js.map +1 -1
  228. package/dist/parquetjs/parser/parquet-reader.d.ts +78 -12
  229. package/dist/parquetjs/parser/parquet-reader.d.ts.map +1 -1
  230. package/dist/parquetjs/parser/parquet-reader.js +605 -66
  231. package/dist/parquetjs/parser/parquet-reader.js.map +1 -1
  232. package/dist/parquetjs/schema/declare.d.ts +22 -3
  233. package/dist/parquetjs/schema/declare.d.ts.map +1 -1
  234. package/dist/parquetjs/schema/declare.js.map +1 -1
  235. package/dist/parquetjs/schema/schema.d.ts.map +1 -1
  236. package/dist/parquetjs/schema/schema.js +10 -0
  237. package/dist/parquetjs/schema/schema.js.map +1 -1
  238. package/dist/parquetjs/schema/shred.d.ts.map +1 -1
  239. package/dist/parquetjs/schema/shred.js +219 -11
  240. package/dist/parquetjs/schema/shred.js.map +1 -1
  241. package/dist/parquetjs/schema/types.js +19 -21
  242. package/dist/parquetjs/schema/types.js.map +1 -1
  243. package/dist/parquetjs/schema/variant.d.ts +3 -0
  244. package/dist/parquetjs/schema/variant.d.ts.map +1 -0
  245. package/dist/parquetjs/schema/variant.js +248 -0
  246. package/dist/parquetjs/schema/variant.js.map +1 -0
  247. package/dist/parquetjs/utils/crc32.d.ts +3 -0
  248. package/dist/parquetjs/utils/crc32.d.ts.map +1 -0
  249. package/dist/parquetjs/utils/crc32.js +15 -0
  250. package/dist/parquetjs/utils/crc32.js.map +1 -0
  251. package/dist/parquetjs/utils/read-utils.d.ts +11 -1
  252. package/dist/parquetjs/utils/read-utils.d.ts.map +1 -1
  253. package/dist/parquetjs/utils/read-utils.js +21 -1
  254. package/dist/parquetjs/utils/read-utils.js.map +1 -1
  255. package/dist/unbundled.d.ts +2 -1
  256. package/dist/unbundled.d.ts.map +1 -1
  257. package/dist/unbundled.js +1 -0
  258. package/dist/unbundled.js.map +1 -1
  259. package/package.json +36 -11
  260. package/src/bundled.ts +3 -0
  261. package/src/delta-format.ts +12 -0
  262. package/src/delta-source-loader-types.ts +40 -0
  263. package/src/delta-source.ts +305 -0
  264. package/src/delta-types.ts +40 -0
  265. package/src/iceberg-rest-catalog.ts +122 -0
  266. package/src/iceberg-table-source.ts +1037 -0
  267. package/src/iceberg-types.ts +165 -0
  268. package/src/index.ts +50 -1
  269. package/src/lib/arrow/convert-schema-from-parquet.ts +113 -19
  270. package/src/lib/arrow/create-nested-arrow-vector.ts +116 -0
  271. package/src/lib/encoders/encode-table-to-parquet-js.ts +268 -7
  272. package/src/lib/geo/geoparquet-covering.ts +201 -0
  273. package/src/lib/geo/geospatial-metadata.ts +42 -15
  274. package/src/lib/parquet-bloom-filter-planner.ts +86 -0
  275. package/src/lib/parquet-bloom-filter.ts +345 -0
  276. package/src/lib/parquet-encryption.ts +344 -0
  277. package/src/lib/parquet-page-index.ts +180 -34
  278. package/src/lib/parquet-predicate.ts +52 -16
  279. package/src/lib/parquet-source-worker-decoder.ts +116 -20
  280. package/src/lib/parquet-source-worker-types.ts +25 -0
  281. package/src/lib/parsers/get-parquet-schema.ts +3 -1
  282. package/src/lib/parsers/parse-parquet-to-arrow-js.ts +397 -17
  283. package/src/lib/parsers/parse-parquet-to-arrow.ts +4 -5
  284. package/src/lib/parsers/parse-parquet-to-columns.ts +3 -1
  285. package/src/lib/parsers/parse-parquet-to-json.ts +18 -2
  286. package/src/lib/slice-parquet-batch.ts +32 -0
  287. package/src/lib/sources/parquet-range-file.ts +35 -29
  288. package/src/lib/utils/normalize-parquet-options.ts +17 -1
  289. package/src/parquet-dataset-source.ts +397 -85
  290. package/src/parquet-js-loader-types.ts +1 -0
  291. package/src/parquet-js-loader.ts +2 -1
  292. package/src/parquet-js-writer.ts +30 -0
  293. package/src/parquet-loader-options.ts +11 -0
  294. package/src/parquet-source-capabilities.ts +18 -1
  295. package/src/parquet-source-loader-types.ts +10 -0
  296. package/src/parquet-source-loader.ts +929 -52
  297. package/src/parquet-source-types.ts +197 -28
  298. package/src/parquet-worker-url.ts +6 -1
  299. package/src/parquetjs/codecs/declare.ts +27 -0
  300. package/src/parquetjs/codecs/delta.ts +466 -6
  301. package/src/parquetjs/codecs/index.ts +4 -0
  302. package/src/parquetjs/codecs/plain.ts +177 -11
  303. package/src/parquetjs/codecs/rle.ts +573 -3
  304. package/src/parquetjs/compression.ts +102 -76
  305. package/src/parquetjs/encoder/dictionary-planner.ts +7 -2
  306. package/src/parquetjs/encoder/parquet-encoder.ts +1016 -38
  307. package/src/parquetjs/parquet-thrift/BoundingBox.ts +108 -0
  308. package/src/parquetjs/parquet-thrift/ColumnChunk.ts +35 -0
  309. package/src/parquetjs/parquet-thrift/ColumnCryptoMetaData.ts +139 -0
  310. package/src/parquetjs/parquet-thrift/ColumnMetaData.ts +73 -0
  311. package/src/parquetjs/parquet-thrift/EncryptionAlgorithm.ts +127 -0
  312. package/src/parquetjs/parquet-thrift/FileCryptoMetaData.ts +60 -0
  313. package/src/parquetjs/parquet-thrift/FileMetaData.ts +35 -0
  314. package/src/parquetjs/parquet-thrift/GeospatialStatistics.ts +68 -0
  315. package/src/parquetjs/parquet-thrift/SizeStatistics.ts +98 -0
  316. package/src/parquetjs/parquet-thrift/index.ts +6 -0
  317. package/src/parquetjs/parser/decoders.ts +451 -26
  318. package/src/parquetjs/parser/parquet-reader.ts +935 -87
  319. package/src/parquetjs/schema/declare.ts +22 -2
  320. package/src/parquetjs/schema/schema.ts +10 -0
  321. package/src/parquetjs/schema/shred.ts +238 -11
  322. package/src/parquetjs/schema/types.ts +16 -25
  323. package/src/parquetjs/schema/variant.ts +313 -0
  324. package/src/parquetjs/utils/crc32.ts +15 -0
  325. package/src/parquetjs/utils/read-utils.ts +26 -1
  326. package/src/unbundled.ts +14 -0
@@ -3,7 +3,7 @@
3
3
  // Copyright (c) vis.gl contributors
4
4
 
5
5
  import type {ReadableFile, Stat} from '@loaders.gl/loader-utils';
6
- import {RangeRequestScheduler} from '@loaders.gl/loader-utils';
6
+ import {RangeRequestCache, RangeRequestScheduler} from '@loaders.gl/loader-utils';
7
7
 
8
8
  import type {ParquetObjectVersion, ParquetRangeRequestOptions} from '../../parquet-source-types';
9
9
 
@@ -47,6 +47,8 @@ type ContentRange = {
47
47
  total: number;
48
48
  };
49
49
 
50
+ const DEFAULT_RANGE_CACHE_BYTES = 65536;
51
+
50
52
  /** Strict HTTP range-backed file with Parquet object-version validation. */
51
53
  export class ParquetRangeFile implements ReadableFile {
52
54
  /** URL used as the file handle. */
@@ -64,8 +66,8 @@ export class ParquetRangeFile implements ReadableFile {
64
66
  private readonly scheduler: RangeRequestScheduler;
65
67
  /** Aborts all active requests when the file is closed. */
66
68
  private readonly closeController = new AbortController();
67
- /** Exact ranges already loaded during initialization. */
68
- private readonly cache = new Map<string, ArrayBuffer>();
69
+ /** Bounded cache for repeated metadata and small range reads. */
70
+ private readonly cache: RangeRequestCache;
69
71
  /** Remote object byte length discovered from `Content-Range`. */
70
72
  private fileByteLength = 0;
71
73
  /** Object validators captured by the opening range request. */
@@ -80,6 +82,14 @@ export class ParquetRangeFile implements ReadableFile {
80
82
  this.fetch = options.fetch;
81
83
  this.headers = options.headers;
82
84
  this.onTelemetry = options.onTelemetry;
85
+ this.cache = new RangeRequestCache({
86
+ maxBytes: DEFAULT_RANGE_CACHE_BYTES,
87
+ onEvent: event => {
88
+ if (event.type === 'hit') {
89
+ this.onTelemetry?.({type: 'cache-hit', cacheHits: 1});
90
+ }
91
+ }
92
+ });
83
93
  this.scheduler =
84
94
  options.rangeRequests?.scheduler ||
85
95
  new RangeRequestScheduler({
@@ -113,7 +123,7 @@ export class ParquetRangeFile implements ReadableFile {
113
123
  const result = await this.fetchExactRange(0, 4, signal, false);
114
124
  this.fileByteLength = result.contentRange.total;
115
125
  this.version = getObjectVersion(result.response.headers);
116
- this.cache.set(getRangeKey(0, 4), result.arrayBuffer);
126
+ this.cache.set(getVersionedSourceId(this.url, this.version), 0, result.arrayBuffer);
117
127
  return this;
118
128
  }
119
129
 
@@ -130,32 +140,33 @@ export class ParquetRangeFile implements ReadableFile {
130
140
  return new ArrayBuffer(0);
131
141
  }
132
142
 
133
- const cached = this.cache.get(getRangeKey(offset, length));
134
- if (cached) {
135
- this.onTelemetry?.({type: 'cache-hit', cacheHits: 1});
136
- return cached.slice(0);
137
- }
138
-
139
143
  const abortContext = createCombinedAbortSignal(signal, this.closeController.signal);
140
144
  try {
141
- return await this.scheduler.scheduleRequest({
145
+ return await this.cache.read({
142
146
  sourceId: getVersionedSourceId(this.url, this.version),
143
147
  offset,
144
148
  length,
145
149
  signal: abortContext.signal,
146
- fetchRange: async (transportOffset, transportLength, transportSignal) => {
147
- const result = await this.fetchExactRange(
148
- transportOffset,
149
- transportLength,
150
- transportSignal,
151
- true
152
- );
153
- return {
154
- arrayBuffer: result.arrayBuffer,
155
- status: result.response.status,
156
- transportBytes: result.arrayBuffer.byteLength
157
- };
158
- }
150
+ fetchRange: async (rangeOffset, rangeLength, cacheSignal) =>
151
+ await this.scheduler.scheduleRequest({
152
+ sourceId: getVersionedSourceId(this.url, this.version),
153
+ offset: rangeOffset,
154
+ length: rangeLength,
155
+ signal: cacheSignal,
156
+ fetchRange: async (transportOffset, transportLength, transportSignal) => {
157
+ const result = await this.fetchExactRange(
158
+ transportOffset,
159
+ transportLength,
160
+ transportSignal,
161
+ true
162
+ );
163
+ return {
164
+ arrayBuffer: result.arrayBuffer,
165
+ status: result.response.status,
166
+ transportBytes: result.arrayBuffer.byteLength
167
+ };
168
+ }
169
+ })
159
170
  });
160
171
  } finally {
161
172
  abortContext.dispose();
@@ -323,11 +334,6 @@ function assertObjectVersion(expected: ParquetObjectVersion, actual: ParquetObje
323
334
  }
324
335
  }
325
336
 
326
- /** Creates a cache key for one exact range. */
327
- function getRangeKey(offset: number, length: number): string {
328
- return `${offset}:${length}`;
329
- }
330
-
331
337
  /** Creates a scheduler identity that isolates object versions. */
332
338
  function getVersionedSourceId(url: string, version: ParquetObjectVersion): string {
333
339
  return `${url}#${version.etag || version.lastModified || 'unversioned'}`;
@@ -7,8 +7,24 @@ export function normalizeParquetOptions<OptionsT extends {parquet?: unknown}>(
7
7
  options: OptionsT | undefined,
8
8
  parquetDefaults: Record<string, unknown>
9
9
  ): OptionsT {
10
+ const scanOptions = (
11
+ options as {_scan?: {columns?: readonly string[]; limit?: number}} | undefined
12
+ )?._scan;
13
+ if (scanOptions) {
14
+ for (const key of Object.keys(scanOptions)) {
15
+ if (key !== 'columns' && key !== 'limit') {
16
+ throw new Error(`Parquet _scan option "${key}" is not supported`);
17
+ }
18
+ }
19
+ }
20
+ const parquetOptions = (options?.parquet as Record<string, unknown> | undefined) || {};
10
21
  return {
11
22
  ...options,
12
- parquet: {...parquetDefaults, ...(options?.parquet as Record<string, unknown> | undefined)}
23
+ parquet: {
24
+ ...parquetDefaults,
25
+ ...parquetOptions,
26
+ ...(scanOptions?.columns !== undefined ? {columns: [...scanOptions.columns]} : {}),
27
+ ...(scanOptions?.limit !== undefined ? {limit: scanOptions.limit} : {})
28
+ }
13
29
  } as unknown as OptionsT;
14
30
  }