UNPKG

@gmod/cram

Version:

read CRAM files with pure Javascript

306 lines 13.7 kB
"use strict"; var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) { if (k2 === undefined) k2 = k; var desc = Object.getOwnPropertyDescriptor(m, k); if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) { desc = { enumerable: true, get: function() { return m[k]; } }; } Object.defineProperty(o, k2, desc); }) : (function(o, m, k, k2) { if (k2 === undefined) k2 = k; o[k2] = m[k]; })); var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) { Object.defineProperty(o, "default", { enumerable: true, value: v }); }) : function(o, v) { o["default"] = v; }); var __importStar = (this && this.__importStar) || (function () { var ownKeys = function(o) { ownKeys = Object.getOwnPropertyNames || function (o) { var ar = []; for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k; return ar; }; return ownKeys(o); }; return function (mod) { if (mod && mod.__esModule) return mod; var result = {}; if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]); __setModuleDefault(result, mod); return result; }; })(); var __importDefault = (this && this.__importDefault) || function (mod) { return (mod && mod.__esModule) ? mod : { "default": mod }; }; Object.defineProperty(exports, "__esModule", { value: true }); const crc32_1 = __importDefault(require("crc/calculators/crc32")); const quick_lru_1 = __importDefault(require("quick-lru")); const xz_decompress_1 = require("xz-decompress"); const errors_ts_1 = require("../errors.js"); const htscodecs = __importStar(require("../htscodecs/index.js")); const io_ts_1 = require("../io.js"); const index_ts_1 = __importDefault(require("../rans/index.js")); const sam_ts_1 = require("../sam.js"); const index_ts_2 = require("../seek-bzip/index.js"); const unzip_ts_1 = require("../unzip.js"); const index_ts_3 = __importDefault(require("./container/index.js")); const sectionParsers_ts_1 = require("./sectionParsers.js"); const util_ts_1 = require("./util.js"); function bufferToStream(buf) { return new ReadableStream({ start(controller) { controller.enqueue(buf); controller.close(); }, }); } // source: https://abdulapopoola.com/2019/01/20/check-endianness-with-javascript/ function getEndianness() { const uInt32 = new Uint32Array([0x11223344]); const uInt8 = new Uint8Array(uInt32.buffer); if (uInt8[0] === 0x44) { return 0; // little-endian } else if (uInt8[0] === 0x11) { return 1; // big-endian } else { return 2; // mixed-endian? } } class CramFile { constructor(args) { this.file = (0, io_ts_1.open)(args.url, args.path, args.filehandle); this.validateChecksums = true; this.fetchReferenceSequenceCallback = args.seqFetch; this.options = { checkSequenceMD5: args.checkSequenceMD5, cacheSize: args.cacheSize ?? 20000, }; // cache of features in a slice, keyed by the slice offset. caches all of // the features in a slice, or none. the cache is actually used by the // slice object, it's just kept here at the level of the file this.featureCache = new quick_lru_1.default({ maxSize: this.options.cacheSize, }); if (getEndianness() > 0) { throw new Error('Detected big-endian machine, may be unable to run'); } } read(length, position) { return this.file.read(length, position); } // memoized async getDefinition() { const { maxLength, parser } = (0, sectionParsers_ts_1.cramFileDefinition)(); const headbytes = await this.file.read(maxLength, 0); const definition = parser(headbytes).value; if (definition.majorVersion !== 2 && definition.majorVersion !== 3) { throw new errors_ts_1.CramUnimplementedError(`CRAM version ${definition.majorVersion} not supported`); } return definition; } // memoize async getSamHeader() { const firstContainer = await this.getContainerById(0); if (!firstContainer) { throw new errors_ts_1.CramMalformedError('file contains no containers'); } const firstBlock = await firstContainer.getFirstBlock(); const content = firstBlock.content; const dataView = new DataView(content.buffer); const headerLength = dataView.getInt32(0, true); const textStart = 4; const decoder = new TextDecoder('utf8'); const text = decoder.decode(content.subarray(textStart, textStart + headerLength)); this.header = text; return (0, sam_ts_1.parseHeaderText)(text); } async getHeaderText() { await this.getSamHeader(); return this.header; } async getContainerById(containerNumber) { const { majorVersion } = await this.getDefinition(); const sectionParsers = (0, sectionParsers_ts_1.getSectionParsers)(majorVersion); let position = sectionParsers.cramFileDefinition.maxLength; // skip with a series of reads to the proper container let currentContainer; for (let i = 0; i <= containerNumber; i++) { // if we are about to go off the end of the file // and have not found that container, it does not exist // if (position + cramContainerHeader1.maxLength + 8 >= fileSize) { // return undefined // } currentContainer = this.getContainerAtPosition(position); const currentHeader = await currentContainer.getHeader(); // if this is the first container, read all the blocks in the container // to determine its length, because we cannot trust the container // header's given length due to a bug somewhere in htslib if (i === 0) { position = currentHeader._endPosition; for (let j = 0; j < currentHeader.numBlocks; j++) { const block = await this.readBlock(position); position = block._endPosition; } } else { // otherwise, just traverse to the next container using the container's // length position += currentHeader._size + currentHeader.length; } } return currentContainer; } async checkCrc32(position, length, recordedCrc32, description) { const b = await this.file.read(length, position); // this shift >>> 0 is equivalent to crc32(b).unsigned but uses the // internal calculator of crc32 to avoid accidentally importing buffer // https://github.com/alexgorbatchev/crc/blob/31fc3853e417b5fb5ec83335428805842575f699/src/define_crc.ts#L5 const calculatedCrc32 = (0, crc32_1.default)(b) >>> 0; if (calculatedCrc32 !== recordedCrc32) { throw new errors_ts_1.CramMalformedError(`crc mismatch in ${description}: recorded CRC32 = ${recordedCrc32}, but calculated CRC32 = ${calculatedCrc32}`); } } /** * @returns {Promise[number]} the number of containers in the file * * note: this is currently used only in unit tests, and after removing file * length check, relies on a try catch to read return an error to break */ async containerCount() { const { majorVersion } = await this.getDefinition(); const sectionParsers = (0, sectionParsers_ts_1.getSectionParsers)(majorVersion); let containerCount = 0; let position = sectionParsers.cramFileDefinition.maxLength; try { // eslint-disable-next-line @typescript-eslint/no-unnecessary-condition while (true) { const currentHeader = await this.getContainerAtPosition(position).getHeader(); // if this is the first container, read all the blocks in the container, // because we cannot trust the container header's given length due to a // bug somewhere in htslib if (containerCount === 0) { position = currentHeader._endPosition; for (let j = 0; j < currentHeader.numBlocks; j++) { const block = await this.readBlock(position); position = block._endPosition; } } else { // otherwise, just traverse to the next container using the container's // length position += currentHeader._size + currentHeader.length; } containerCount += 1; } } catch (e) { containerCount--; /* do nothing */ } return containerCount; } getContainerAtPosition(position) { return new index_ts_3.default(this, position); } async readBlockHeader(position) { const { majorVersion } = await this.getDefinition(); const sectionParsers = (0, sectionParsers_ts_1.getSectionParsers)(majorVersion); const { cramBlockHeader } = sectionParsers; const buffer = await this.file.read(cramBlockHeader.maxLength, position); return (0, util_ts_1.parseItem)(buffer, cramBlockHeader.parser, 0, position); } async _parseSection(section, position, size = section.maxLength, preReadBuffer) { const buffer = preReadBuffer ?? (await this.file.read(size, position)); const data = (0, util_ts_1.parseItem)(buffer, section.parser, 0, position); if (data._size !== size) { throw new errors_ts_1.CramMalformedError(`section read error: requested size ${size} does not equal parsed size ${data._size}`); } return data; } async _uncompressPre(compressionMethod, inputBuffer, uncompressedSize) { // console.log({ compressionMethod }) if (compressionMethod === 'gzip') { return (0, unzip_ts_1.unzip)(inputBuffer); } else if (compressionMethod === 'bzip2') { return (0, index_ts_2.decode)(inputBuffer); } else if (compressionMethod === 'lzma') { const decompressedResponse = new Response(new xz_decompress_1.XzReadableStream(bufferToStream(inputBuffer))); return new Uint8Array(await decompressedResponse.arrayBuffer()); } else if (compressionMethod === 'rans') { const outputBuffer = new Uint8Array(uncompressedSize); (0, index_ts_1.default)(inputBuffer, outputBuffer); return outputBuffer; } else if (compressionMethod === 'rans4x16') { return htscodecs.r4x16_uncompress(inputBuffer); } else if (compressionMethod === 'arith') { return htscodecs.arith_uncompress(inputBuffer); } else if (compressionMethod === 'fqzcomp') { return htscodecs.fqzcomp_uncompress(inputBuffer); } else if (compressionMethod === 'tok3') { return htscodecs.tok3_uncompress(inputBuffer); } else { throw new errors_ts_1.CramUnimplementedError(`${compressionMethod} decompression not yet implemented`); } } async _uncompress(compressionMethod, inputBuffer, uncompressedSize) { const buf = await this._uncompressPre(compressionMethod, inputBuffer, uncompressedSize); if (buf.length !== uncompressedSize) { const ret = new Uint8Array(uncompressedSize); ret.set(buf, 0); return ret; } return buf; } async readBlock(position) { const { majorVersion } = await this.getDefinition(); const sectionParsers = (0, sectionParsers_ts_1.getSectionParsers)(majorVersion); const blockHeader = await this.readBlockHeader(position); const blockContentPosition = blockHeader._endPosition; const d = await this.file.read(blockHeader.compressedSize, blockContentPosition); const uncompressedData = blockHeader.compressionMethod !== 'raw' ? await this._uncompress(blockHeader.compressionMethod, d, blockHeader.uncompressedSize) : d; const block = { ...blockHeader, _endPosition: blockContentPosition, contentPosition: blockContentPosition, content: uncompressedData, }; if (majorVersion >= 3) { // parse the crc32 const crc = await this._parseSection(sectionParsers.cramBlockCrc32, blockContentPosition + blockHeader.compressedSize); block.crc32 = crc.crc32; // check the block data crc32 if (this.validateChecksums) { await this.checkCrc32(position, blockHeader._size + blockHeader.compressedSize, crc.crc32, 'block data'); } // make the endposition and size reflect the whole block block._endPosition = crc._endPosition; block._size = block.compressedSize + sectionParsers.cramBlockCrc32.maxLength; } else { block._endPosition = blockContentPosition + block.compressedSize; block._size = block.compressedSize; } return block; } } exports.default = CramFile; 'getDefinition getSectionParsers getSamHeader'.split(' ').forEach(method => { (0, util_ts_1.tinyMemoize)(CramFile, method); }); //# sourceMappingURL=file.js.map