bz2.js 12 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343
  1. /*! bz2 (C) 2019-present SheetJS LLC */
  2. 'use strict';
  3. (function bz2() {
  4. // https://www.ncbi.nlm.nih.gov/IEB/ToolBox/CPP_DOC/lxr/source/src/util/compress/bzip2/crctable.c
  5. const crc32Table = [
  6. 0x00000000, 0x04c11db7, 0x09823b6e, 0x0d4326d9, 0x130476dc, 0x17c56b6b, 0x1a864db2, 0x1e475005,
  7. 0x2608edb8, 0x22c9f00f, 0x2f8ad6d6, 0x2b4bcb61, 0x350c9b64, 0x31cd86d3, 0x3c8ea00a, 0x384fbdbd,
  8. 0x4c11db70, 0x48d0c6c7, 0x4593e01e, 0x4152fda9, 0x5f15adac, 0x5bd4b01b, 0x569796c2, 0x52568b75,
  9. 0x6a1936c8, 0x6ed82b7f, 0x639b0da6, 0x675a1011, 0x791d4014, 0x7ddc5da3, 0x709f7b7a, 0x745e66cd,
  10. 0x9823b6e0, 0x9ce2ab57, 0x91a18d8e, 0x95609039, 0x8b27c03c, 0x8fe6dd8b, 0x82a5fb52, 0x8664e6e5,
  11. 0xbe2b5b58, 0xbaea46ef, 0xb7a96036, 0xb3687d81, 0xad2f2d84, 0xa9ee3033, 0xa4ad16ea, 0xa06c0b5d,
  12. 0xd4326d90, 0xd0f37027, 0xddb056fe, 0xd9714b49, 0xc7361b4c, 0xc3f706fb, 0xceb42022, 0xca753d95,
  13. 0xf23a8028, 0xf6fb9d9f, 0xfbb8bb46, 0xff79a6f1, 0xe13ef6f4, 0xe5ffeb43, 0xe8bccd9a, 0xec7dd02d,
  14. 0x34867077, 0x30476dc0, 0x3d044b19, 0x39c556ae, 0x278206ab, 0x23431b1c, 0x2e003dc5, 0x2ac12072,
  15. 0x128e9dcf, 0x164f8078, 0x1b0ca6a1, 0x1fcdbb16, 0x018aeb13, 0x054bf6a4, 0x0808d07d, 0x0cc9cdca,
  16. 0x7897ab07, 0x7c56b6b0, 0x71159069, 0x75d48dde, 0x6b93dddb, 0x6f52c06c, 0x6211e6b5, 0x66d0fb02,
  17. 0x5e9f46bf, 0x5a5e5b08, 0x571d7dd1, 0x53dc6066, 0x4d9b3063, 0x495a2dd4, 0x44190b0d, 0x40d816ba,
  18. 0xaca5c697, 0xa864db20, 0xa527fdf9, 0xa1e6e04e, 0xbfa1b04b, 0xbb60adfc, 0xb6238b25, 0xb2e29692,
  19. 0x8aad2b2f, 0x8e6c3698, 0x832f1041, 0x87ee0df6, 0x99a95df3, 0x9d684044, 0x902b669d, 0x94ea7b2a,
  20. 0xe0b41de7, 0xe4750050, 0xe9362689, 0xedf73b3e, 0xf3b06b3b, 0xf771768c, 0xfa325055, 0xfef34de2,
  21. 0xc6bcf05f, 0xc27dede8, 0xcf3ecb31, 0xcbffd686, 0xd5b88683, 0xd1799b34, 0xdc3abded, 0xd8fba05a,
  22. 0x690ce0ee, 0x6dcdfd59, 0x608edb80, 0x644fc637, 0x7a089632, 0x7ec98b85, 0x738aad5c, 0x774bb0eb,
  23. 0x4f040d56, 0x4bc510e1, 0x46863638, 0x42472b8f, 0x5c007b8a, 0x58c1663d, 0x558240e4, 0x51435d53,
  24. 0x251d3b9e, 0x21dc2629, 0x2c9f00f0, 0x285e1d47, 0x36194d42, 0x32d850f5, 0x3f9b762c, 0x3b5a6b9b,
  25. 0x0315d626, 0x07d4cb91, 0x0a97ed48, 0x0e56f0ff, 0x1011a0fa, 0x14d0bd4d, 0x19939b94, 0x1d528623,
  26. 0xf12f560e, 0xf5ee4bb9, 0xf8ad6d60, 0xfc6c70d7, 0xe22b20d2, 0xe6ea3d65, 0xeba91bbc, 0xef68060b,
  27. 0xd727bbb6, 0xd3e6a601, 0xdea580d8, 0xda649d6f, 0xc423cd6a, 0xc0e2d0dd, 0xcda1f604, 0xc960ebb3,
  28. 0xbd3e8d7e, 0xb9ff90c9, 0xb4bcb610, 0xb07daba7, 0xae3afba2, 0xaafbe615, 0xa7b8c0cc, 0xa379dd7b,
  29. 0x9b3660c6, 0x9ff77d71, 0x92b45ba8, 0x9675461f, 0x8832161a, 0x8cf30bad, 0x81b02d74, 0x857130c3,
  30. 0x5d8a9099, 0x594b8d2e, 0x5408abf7, 0x50c9b640, 0x4e8ee645, 0x4a4ffbf2, 0x470cdd2b, 0x43cdc09c,
  31. 0x7b827d21, 0x7f436096, 0x7200464f, 0x76c15bf8, 0x68860bfd, 0x6c47164a, 0x61043093, 0x65c52d24,
  32. 0x119b4be9, 0x155a565e, 0x18197087, 0x1cd86d30, 0x029f3d35, 0x065e2082, 0x0b1d065b, 0x0fdc1bec,
  33. 0x3793a651, 0x3352bbe6, 0x3e119d3f, 0x3ad08088, 0x2497d08d, 0x2056cd3a, 0x2d15ebe3, 0x29d4f654,
  34. 0xc5a92679, 0xc1683bce, 0xcc2b1d17, 0xc8ea00a0, 0xd6ad50a5, 0xd26c4d12, 0xdf2f6bcb, 0xdbee767c,
  35. 0xe3a1cbc1, 0xe760d676, 0xea23f0af, 0xeee2ed18, 0xf0a5bd1d, 0xf464a0aa, 0xf9278673, 0xfde69bc4,
  36. 0x89b8fd09, 0x8d79e0be, 0x803ac667, 0x84fbdbd0, 0x9abc8bd5, 0x9e7d9662, 0x933eb0bb, 0x97ffad0c,
  37. 0xafb010b1, 0xab710d06, 0xa6322bdf, 0xa2f33668, 0xbcb4666d, 0xb8757bda, 0xb5365d03, 0xb1f740b4,
  38. ];
  39. // generated from 1 << i, except for 32
  40. const masks = [
  41. 0x00000000, 0x00000001, 0x00000003, 0x00000007,
  42. 0x0000000f, 0x0000001f, 0x0000003f, 0x0000007f,
  43. 0x000000ff, 0x000001ff, 0x000003ff, 0x000007ff,
  44. 0x00000fff, 0x00001fff, 0x00003fff, 0x00007fff,
  45. 0x0000ffff, 0x0001ffff, 0x0003ffff, 0x0007ffff,
  46. 0x000fffff, 0x001fffff, 0x003fffff, 0x007fffff,
  47. 0x00ffffff, 0x01ffffff, 0x03ffffff, 0x07ffffff,
  48. 0x0fffffff, 0x1fffffff, 0x3fffffff, -0x80000000,
  49. ];
  50. function createOrderedHuffmanTable(lengths) {
  51. const z = [];
  52. for (let i = 0; i < lengths.length; i += 1) {
  53. z.push([i, lengths[i]]);
  54. }
  55. z.push([lengths.length, -1]);
  56. const table = [];
  57. let start = z[0][0];
  58. let bits = z[0][1];
  59. for (let i = 0; i < z.length; i += 1) {
  60. const finish = z[i][0];
  61. const endbits = z[i][1];
  62. if (bits) {
  63. for (let code = start; code < finish; code += 1) {
  64. table.push({ code, bits, symbol: undefined });
  65. }
  66. }
  67. start = finish;
  68. bits = endbits;
  69. if (endbits === -1) {
  70. break;
  71. }
  72. }
  73. table.sort((a, b) => ((a.bits - b.bits) || (a.code - b.code)));
  74. let tempBits = 0;
  75. let symbol = -1;
  76. const fastAccess = [];
  77. let current;
  78. for (let i = 0; i < table.length; i += 1) {
  79. const t = table[i];
  80. symbol += 1;
  81. if (t.bits !== tempBits) {
  82. symbol <<= t.bits - tempBits;
  83. tempBits = t.bits;
  84. current = fastAccess[tempBits] = {};
  85. }
  86. t.symbol = symbol;
  87. current[symbol] = t;
  88. }
  89. return {
  90. table,
  91. fastAccess,
  92. };
  93. }
  94. function bwtReverse(src, primary) {
  95. if (primary < 0 || primary >= src.length) {
  96. throw RangeError('Out of bound');
  97. }
  98. const unsorted = src.slice();
  99. src.sort((a, b) => a - b);
  100. const start = {};
  101. for (let i = src.length - 1; i >= 0; i -= 1) {
  102. start[src[i]] = i;
  103. }
  104. const links = [];
  105. for (let i = 0; i < src.length; i += 1) {
  106. links.push(start[unsorted[i]]++); // eslint-disable-line no-plusplus
  107. }
  108. let i;
  109. const first = src[i = primary];
  110. const ret = [];
  111. for (let j = 1; j < src.length; j += 1) {
  112. const x = src[i = links[i]];
  113. if (x === undefined) {
  114. ret.push(255);
  115. } else {
  116. ret.push(x);
  117. }
  118. }
  119. ret.push(first);
  120. ret.reverse();
  121. return ret;
  122. }
  123. function decompress(bytes, checkCRC = false) {
  124. let index = 0;
  125. let bitfield = 0;
  126. let bits = 0;
  127. const read = (n) => {
  128. if (n >= 32) {
  129. const nd = n >> 1;
  130. return read(nd) * (1 << nd) + read(n - nd);
  131. }
  132. while (bits < n) {
  133. bitfield = (bitfield << 8) + bytes[index];
  134. index += 1;
  135. bits += 8;
  136. }
  137. const m = masks[n];
  138. const r = (bitfield >> (bits - n)) & m;
  139. bits -= n;
  140. bitfield &= ~(m << bits);
  141. return r;
  142. };
  143. const magic = read(16);
  144. if (magic !== 0x425A) { // 'BZ'
  145. throw new Error('Invalid magic');
  146. }
  147. const method = read(8);
  148. if (method !== 0x68) { // h for huffman
  149. throw new Error('Invalid method');
  150. }
  151. let blocksize = read(8);
  152. if (blocksize >= 49 && blocksize <= 57) { // 1..9
  153. blocksize -= 48;
  154. } else {
  155. throw new Error('Invalid blocksize');
  156. }
  157. let out = new Uint8Array(bytes.length * 1.5);
  158. let outIndex = 0;
  159. let newCRC = -1;
  160. while (true) {
  161. const blocktype = read(48);
  162. const crc = read(32) | 0;
  163. if (blocktype === 0x314159265359) {
  164. if (read(1)) {
  165. throw new Error('do not support randomised');
  166. }
  167. const pointer = read(24);
  168. const used = [];
  169. const usedGroups = read(16);
  170. for (let i = 1 << 15; i > 0; i >>= 1) {
  171. if (!(usedGroups & i)) {
  172. for (let j = 0; j < 16; j += 1) {
  173. used.push(false);
  174. }
  175. continue; // eslint-disable-line no-continue
  176. }
  177. const usedChars = read(16);
  178. for (let j = 1 << 15; j > 0; j >>= 1) {
  179. used.push(!!(usedChars & j));
  180. }
  181. }
  182. const groups = read(3);
  183. if (groups < 2 || groups > 6) {
  184. throw new Error('Invalid number of huffman groups');
  185. }
  186. const selectorsUsed = read(15);
  187. const selectors = [];
  188. const mtf = Array.from({ length: groups }, (_, i) => i);
  189. for (let i = 0; i < selectorsUsed; i += 1) {
  190. let c = 0;
  191. while (read(1)) {
  192. c += 1;
  193. if (c >= groups) {
  194. throw new Error('MTF table out of range');
  195. }
  196. }
  197. const v = mtf[c];
  198. for (let j = c; j > 0; mtf[j] = mtf[--j]) { // eslint-disable-line no-plusplus
  199. // nothing
  200. }
  201. selectors.push(v);
  202. mtf[0] = v;
  203. }
  204. const symbolsInUse = used.reduce((a, b) => a + b, 0) + 2;
  205. const tables = [];
  206. for (let i = 0; i < groups; i += 1) {
  207. let length = read(5);
  208. const lengths = [];
  209. for (let j = 0; j < symbolsInUse; j += 1) {
  210. if (length < 0 || length > 20) {
  211. throw new Error('Huffman group length outside range');
  212. }
  213. while (read(1)) {
  214. length -= (read(1) * 2) - 1;
  215. }
  216. lengths.push(length);
  217. }
  218. tables.push(createOrderedHuffmanTable(lengths));
  219. }
  220. const favourites = [];
  221. for (let i = 0; i < used.length - 1; i += 1) {
  222. if (used[i]) {
  223. favourites.push(i);
  224. }
  225. }
  226. let decoded = 0;
  227. let selectorPointer = 0;
  228. let t;
  229. let r;
  230. let repeat = 0;
  231. let repeatPower = 0;
  232. const buffer = [];
  233. while (true) {
  234. decoded -= 1;
  235. if (decoded <= 0) {
  236. decoded = 50;
  237. if (selectorPointer <= selectors.length) {
  238. t = tables[selectors[selectorPointer]];
  239. selectorPointer += 1;
  240. }
  241. }
  242. for (const b in t.fastAccess) {
  243. if (!Object.prototype.hasOwnProperty.call(t.fastAccess, b)) {
  244. continue; // eslint-disable-line no-continue
  245. }
  246. if (bits < b) {
  247. bitfield = (bitfield << 8) + bytes[index];
  248. index += 1;
  249. bits += 8;
  250. }
  251. r = t.fastAccess[b][bitfield >> (bits - b)];
  252. if (r) {
  253. bitfield &= masks[bits -= b];
  254. r = r.code;
  255. break;
  256. }
  257. }
  258. if (r >= 0 && r <= 1) {
  259. if (repeat === 0) {
  260. repeatPower = 1;
  261. }
  262. repeat += repeatPower << r;
  263. repeatPower <<= 1;
  264. continue; // eslint-disable-line no-continue
  265. } else {
  266. const v = favourites[0];
  267. for (; repeat > 0; repeat -= 1) {
  268. buffer.push(v);
  269. }
  270. }
  271. if (r === symbolsInUse - 1) {
  272. break;
  273. } else {
  274. const v = favourites[r - 1];
  275. // eslint-disable-next-line no-plusplus
  276. for (let j = r - 1; j > 0; favourites[j] = favourites[--j]) {
  277. // nothing
  278. }
  279. favourites[0] = v;
  280. buffer.push(v);
  281. }
  282. }
  283. const nt = bwtReverse(buffer, pointer);
  284. let i = 0;
  285. while (i < nt.length) {
  286. const c = nt[i];
  287. let count = 1;
  288. if ((i < nt.length - 4)
  289. && nt[i + 1] === c
  290. && nt[i + 2] === c
  291. && nt[i + 3] === c) {
  292. count = nt[i + 4] + 4;
  293. i += 5;
  294. } else {
  295. i += 1;
  296. }
  297. if (outIndex + count >= out.length) {
  298. const old = out;
  299. out = new Uint8Array(old.length * 2);
  300. out.set(old);
  301. }
  302. for (let j = 0; j < count; j += 1) {
  303. if (checkCRC) {
  304. newCRC = (newCRC << 8) ^ crc32Table[((newCRC >> 24) ^ c) & 0xff];
  305. }
  306. out[outIndex] = c;
  307. outIndex += 1;
  308. }
  309. }
  310. if (checkCRC) {
  311. const calculatedCRC = newCRC ^ -1;
  312. if (calculatedCRC !== crc) {
  313. throw new Error(`CRC mismatch: ${calculatedCRC} !== ${crc}`);
  314. }
  315. newCRC = -1;
  316. }
  317. } else if (blocktype === 0x177245385090) {
  318. read(bits & 0x07); // pad align
  319. break;
  320. } else {
  321. throw new Error('Invalid bz2 blocktype');
  322. }
  323. }
  324. return out.subarray(0, outIndex);
  325. }
  326. const exports = { decompress };
  327. if (typeof window !== 'undefined') {
  328. window.bz2 = exports; // eslint-disable-line no-undef
  329. } else {
  330. module.exports = exports;
  331. }
  332. }());