@gmod/cram - npm Package Compare versions

Comparing version 4.0.4 to 4.0.5

dist/cramFile/codecs/_base.d.ts

@@ -26,3 +26,3 @@ import { CramFileBlock } from '../file';
     constructor(parameters: TParameters, dataType: TResult);
-    abstract decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): DataTypeMapping[TResult];
+    abstract decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): DataTypeMapping[TResult] | undefined;
 }

dist/cramFile/codecs/byteArrayLength.js

@@ -15,7 +15,8 @@ "use strict";
         const lengthCodec = this._getLengthCodec();
-        const arrayLength = lengthCodec.decode(slice, coreDataBlock, blocksByContentId, cursors);
+        const arrayLength = lengthCodec.decode(slice, coreDataBlock, blocksByContentId, cursors) || 0;
         const dataCodec = this._getDataCodec();
         const data = new Uint8Array(arrayLength);
         for (let i = 0; i < arrayLength; i += 1) {
-            data[i] = dataCodec.decode(slice, coreDataBlock, blocksByContentId, cursors);
+            data[i] =
+                dataCodec.decode(slice, coreDataBlock, blocksByContentId, cursors) || 0;
         }
@@ -22,0 +23,0 @@ return data;

dist/cramFile/codecs/external.d.ts

@@ -8,5 +8,5 @@ import CramCodec, { Cursor, Cursors } from './_base';
     constructor(parameters: ExternalCramEncoding['parameters'], dataType: 'int' | 'byte');
-    decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): number;
+    decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): number | undefined;
     _decodeInt(contentBlock: CramFileBlock, cursor: Cursor): number;
     _decodeByte(contentBlock: CramFileBlock, cursor: Cursor): number;
 }

dist/cramFile/codecs/external.js

@@ -26,7 +26,4 @@ "use strict";
         const contentBlock = blocksByContentId[blockContentId];
-        if (!contentBlock) {
-            throw new errors_1.CramMalformedError(`no block found with content ID ${blockContentId}}`);
-        }
         const cursor = cursors.externalBlocks.getCursor(blockContentId);
-        return this._decodeData(contentBlock, cursor);
+        return contentBlock ? this._decodeData(contentBlock, cursor) : undefined;
     }
@@ -33,0 +30,0 @@ _decodeInt(contentBlock, cursor) {

dist/cramFile/container/index.d.ts

@@ -8,3 +8,8 @@ import CramSlice from '../slice';
     constructor(file: CramFile, filePosition: number);
-    getHeader(): Promise<({
+    getHeader(): Promise<{
+        _size: number;
+        _endPosition: number;
+        numLandmarks: number;
+        landmarks: number[];
+        crc32?: number | undefined;
         length: number;
@@ -15,20 +20,6 @@ refSeqId: number;
         numBlocks: number;
-        numLandmarks: number;
         numBases: number | undefined;
         recordCounter: number;
         numRecords: number;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        numLandmarks: number;
-        landmarks: number[];
-        crc32?: number | undefined;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        _size: number;
-        _endPosition: number;
-    }) | undefined>;
+    }>;
     getCompressionHeaderBlock(): Promise<{
@@ -53,7 +44,12 @@ parsedContent: {
         crc32?: number;
-    } | null | undefined>;
-    getFirstBlock(): Promise<import("../file").CramFileBlock | undefined>;
+    } | null>;
+    getFirstBlock(): Promise<import("../file").CramFileBlock>;
     getCompressionScheme(): Promise<CramContainerCompressionScheme | undefined>;
     getSlice(slicePosition: number, sliceSize: number): CramSlice;
-    _readContainerHeader(position: number): Promise<({
+    _readContainerHeader(position: number): Promise<{
+        _size: number;
+        _endPosition: number;
+        numLandmarks: number;
+        landmarks: number[];
+        crc32?: number | undefined;
         length: number;
@@ -64,20 +60,6 @@ refSeqId: number;
         numBlocks: number;
-        numLandmarks: number;
         numBases: number | undefined;
         recordCounter: number;
         numRecords: number;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        numLandmarks: number;
-        landmarks: number[];
-        crc32?: number | undefined;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        _size: number;
-        _endPosition: number;
-    }) | undefined>;
+    }>;
 }

dist/cramFile/container/index.js

@@ -24,3 +24,3 @@ "use strict";
         // header
-        if (!containerHeader?.numRecords) {
+        if (!containerHeader.numRecords) {
             return null;
@@ -31,5 +31,2 @@ }
         const block = await this.getFirstBlock();
-        if (block === undefined) {
-            return undefined;
-        }
         if (block.contentType !== 'COMPRESSION_HEADER') {
@@ -46,5 +43,2 @@ throw new errors_1.CramMalformedError(`invalid content type ${block.contentType} in compression header block`);
         const containerHeader = await this.getHeader();
-        if (!containerHeader) {
-            return undefined;
-        }
         return this.file.readBlock(containerHeader._endPosition);
@@ -70,7 +64,2 @@ }
         const { cramContainerHeader1, cramContainerHeader2 } = sectionParsers;
-        const { size: fileSize } = await this.file.stat();
-        if (position >= fileSize) {
-            console.warn(`pos:${position}>=fileSize:${fileSize} in cram container`);
-            return undefined;
-        }
         // parse the container header. do it in 2 pieces because you cannot tell
@@ -81,7 +70,2 @@ // how much to buffer until you read numLandmarks
         const numLandmarksSize = (0, util_1.itf8Size)(header1.numLandmarks);
-        if (position + header1.length >= fileSize) {
-            // header indicates container goes beyond fileSize
-            console.warn(`container at ${position} is beyond fileSize:${fileSize}, skipping`);
-            return undefined;
-        }
         const bytes2 = await this.file.read(cramContainerHeader2.maxLength(header1.numLandmarks), position + header1._size - numLandmarksSize);
@@ -92,7 +76,8 @@ const header2 = (0, util_1.parseItem)(bytes2, cramContainerHeader2.parser);
         }
-        const completeHeader = Object.assign(header1, header2, {
+        return {
+            ...header1,
+            ...header2,
             _size: header1._size + header2._size - numLandmarksSize,
             _endPosition: header1._size + header2._size - numLandmarksSize + position,
-        });
-        return completeHeader;
+        };
     }
@@ -99,0 +84,0 @@ }

dist/cramFile/file.d.ts

@@ -35,3 +35,2 @@ import QuickLRU from 'quick-lru';
     constructor(args: CramFileArgs);
-    stat(): Promise<import("generic-filehandle2").Stats>;
     read(length: number, position: number): Promise<Uint8Array<ArrayBuffer>>;
@@ -56,6 +55,9 @@ getDefinition(): Promise<{
      * @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
      */
     containerCount(): Promise<number | undefined>;
     getContainerAtPosition(position: number): CramContainer;
-    readBlockHeader(position: number): Promise<({
+    readBlockHeader(position: number): Promise<{
         uncompressedSize: number;
@@ -69,3 +71,3 @@ compressedSize: number;
         _size: number;
-    }) | undefined>;
+    }>;
     _parseSection<T>(section: {
@@ -77,8 +79,8 @@ maxLength: number;
         };
-    }, position: number, size?: number, preReadBuffer?: Uint8Array): Promise<(T & {
+    }, position: number, size?: number, preReadBuffer?: Uint8Array): Promise<T & {
         _endPosition: number;
         _size: number;
-    }) | undefined>;
+    }>;
     _uncompress(compressionMethod: CompressionMethod, inputBuffer: Uint8Array, uncompressedSize: number): Promise<any>;
-    readBlock(position: number): Promise<CramFileBlock | undefined>;
+    readBlock(position: number): Promise<CramFileBlock>;
 }

dist/cramFile/file.js

@@ -94,7 +94,2 @@ "use strict";
     }
-    // can just stat this object like a filehandle
-    stat() {
-        return this.file.stat();
-    }
-    // can just stat this object like a filehandle
     read(length, position) {
@@ -120,15 +115,10 @@ return this.file.read(length, position);
         const firstBlock = await firstContainer.getFirstBlock();
-        if (firstBlock === undefined) {
-            return (0, sam_1.parseHeaderText)('');
-        }
-        else {
-            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_1.parseHeaderText)(text);
-        }
+        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_1.parseHeaderText)(text);
     }
@@ -143,4 +133,2 @@ async getHeaderText() {
         let position = sectionParsers.cramFileDefinition.maxLength;
-        const { size: fileSize } = await this.file.stat();
-        const { cramContainerHeader1 } = sectionParsers;
         // skip with a series of reads to the proper container
@@ -151,10 +139,7 @@ let currentContainer;
             // and have not found that container, it does not exist
-            if (position + cramContainerHeader1.maxLength + 8 >= fileSize) {
-                return undefined;
-            }
+            // if (position + cramContainerHeader1.maxLength + 8 >= fileSize) {
+            //   return undefined
+            // }
             currentContainer = this.getContainerAtPosition(position);
             const currentHeader = await currentContainer.getHeader();
-            if (!currentHeader) {
-                throw new errors_1.CramMalformedError(`container ${containerNumber} not found in file`);
-            }
             // if this is the first container, read all the blocks in the container
@@ -167,5 +152,2 @@ // to determine its length, because we cannot trust the container
                     const block = await this.readBlock(position);
-                    if (block === undefined) {
-                        return undefined;
-                    }
                     position = block._endPosition;
@@ -194,2 +176,5 @@ }
      * @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
      */
@@ -199,31 +184,30 @@ async containerCount() {
         const sectionParsers = (0, sectionParsers_1.getSectionParsers)(majorVersion);
-        const { size: fileSize } = await this.file.stat();
-        const { cramContainerHeader1 } = sectionParsers;
         let containerCount = 0;
         let position = sectionParsers.cramFileDefinition.maxLength;
-        while (position + cramContainerHeader1.maxLength + 8 < fileSize) {
-            const currentHeader = await this.getContainerAtPosition(position).getHeader();
-            if (!currentHeader) {
-                break;
-            }
-            // 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);
-                    if (block === undefined) {
-                        return undefined;
+        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;
                     }
-                    position = block._endPosition;
                 }
+                else {
+                    // otherwise, just traverse to the next container using the container's
+                    // length
+                    position += currentHeader._size + currentHeader.length;
+                }
+                containerCount += 1;
             }
-            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;
@@ -238,6 +222,2 @@ }
         const { cramBlockHeader } = sectionParsers;
-        const { size: fileSize } = await this.file.stat();
-        if (position + cramBlockHeader.maxLength >= fileSize) {
-            return undefined;
-        }
         const buffer = await this.file.read(cramBlockHeader.maxLength, position);
@@ -247,13 +227,3 @@ return (0, util_2.parseItem)(buffer, cramBlockHeader.parser, 0, position);
     async _parseSection(section, position, size = section.maxLength, preReadBuffer) {
-        let buffer;
-        if (preReadBuffer) {
-            buffer = preReadBuffer;
-        }
-        else {
-            const { size: fileSize } = await this.file.stat();
-            if (position + size >= fileSize) {
-                return undefined;
-            }
-            buffer = await this.file.read(size, position);
-        }
+        const buffer = preReadBuffer ?? (await this.file.read(size, position));
         const data = (0, util_2.parseItem)(buffer, section.parser, 0, position);
@@ -314,5 +284,2 @@ if (data._size !== size) {
         const blockHeader = await this.readBlockHeader(position);
-        if (blockHeader === undefined) {
-            return undefined;
-        }
         const blockContentPosition = blockHeader._endPosition;
@@ -332,5 +299,2 @@ const d = await this.file.read(blockHeader.compressedSize, blockContentPosition);
             const crc = await this._parseSection(sectionParsers.cramBlockCrc32, blockContentPosition + blockHeader.compressedSize);
-            if (crc === undefined) {
-                return undefined;
-            }
             block.crc32 = crc.crc32;
@@ -337,0 +301,0 @@ // check the block data crc32

dist/cramFile/sectionParsers.js

@@ -566,4 +566,5 @@ "use strict";
             offset += 4;
-            // reference sequence identifier, -1 for unmapped reads, -2 for multiple
-            // reference sequences
+            // reference sequence identifier:
+            // -1 for unmapped reads,
+            // -2 for multiple reference sequences
             const [refSeqId, newOffset1] = (0, util_1.parseItf8)(buffer, offset);
@@ -570,0 +571,0 @@ offset += newOffset1;

dist/cramFile/slice/decodeRecord.d.ts

@@ -7,6 +7,6 @@ import { Cursors, DataTypeMapping } from '../codecs/_base';
 import { CramFileBlock } from '../file';
-export type DataSeriesDecoder = <T extends DataSeriesEncodingKey>(dataSeriesName: T) => DataTypeMapping[DataSeriesTypes[T]];
+export type DataSeriesDecoder = <T extends DataSeriesEncodingKey>(dataSeriesName: T) => DataTypeMapping[DataSeriesTypes[T]] | undefined;
 export default function decodeRecord(slice: CramSlice, decodeDataSeries: DataSeriesDecoder, compressionScheme: CramContainerCompressionScheme, sliceHeader: SliceHeader, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors, majorVersion: number, recordNumber: number): {
     readLength: number;
-    sequenceId: number;
+    sequenceId: number | undefined;
     cramFlags: number;
@@ -13,0 +13,0 @@ flags: number;

dist/cramFile/slice/decodeRecord.js

@@ -183,5 +183,4 @@ "use strict";
     let flags = decodeDataSeries('BF');
-    // note: the C data type of compressionFlags is byte in cram v1
-    // and int32 in cram v2+, but that does not matter for us here
-    // in javascript land.
+    // note: the C data type of compressionFlags is byte in cram v1 and int32 in
+    // cram v2+, but that does not matter for us here in javascript land.
     const cramFlags = decodeDataSeries('CF');
@@ -262,3 +261,7 @@ if (!(0, sectionParsers_1.isMappedSliceHeader)(sliceHeader.parsedContent)) {
         tags[tagName] =
-            typeof tagData === 'number' ? tagData : parseTagData(tagType, tagData);
+            tagData === undefined
+                ? undefined
+                : typeof tagData === 'number'
+                    ? tagData
+                    : parseTagData(tagType, tagData);
     }
@@ -265,0 +268,0 @@ let readFeatures;

dist/cramFile/slice/index.js

@@ -142,9 +142,3 @@ "use strict";
         const containerHeader = await this.container.getHeader();
-        if (!containerHeader) {
-            throw new Error('no container header detected');
-        }
         const header = await this.file.readBlock(containerHeader._endPosition + this.containerPosition);
-        if (header === undefined) {
-            throw new Error('block header undefined');
-        }
         if (header.contentType === 'MAPPED_SLICE_HEADER') {
@@ -170,5 +164,2 @@ const content = (0, util_1.parseItem)(header.content, sectionParsers.cramMappedSliceHeader.parser, 0, containerHeader._endPosition);
             const block = await this.file.readBlock(blockPosition);
-            if (block === undefined) {
-                throw new Error('block undefined');
-            }
             blocks[i] = block;
@@ -305,5 +296,3 @@ blockPosition = blocks[i]._endPosition;
             }
-            // console.log(dataSeriesName, Object.getPrototypeOf(codec))
-            const decoded = codec.decode(this, coreDataBlock, blocksByContentId, cursors);
-            return decoded;
+            return codec.decode(this, coreDataBlock, blocksByContentId, cursors);
         };
@@ -333,7 +322,16 @@ const records = new Array(sliceHeader.parsedContent.numRecords);
         // interpret `recordsToNextFragment` attributes to make standard `mate`
-        // objects Resolve mate pair cross-references between records in this slice
+        // objects
+        //
+        // Resolve mate pair cross-references between records in this slice
         for (let i = 0; i < records.length; i += 1) {
-            const { mateRecordNumber } = records[i];
-            if (mateRecordNumber !== undefined && mateRecordNumber >= 0) {
-                associateIntraSliceMate(records, i, records[i], records[mateRecordNumber]);
+            const r = records[i];
+            // check for !!r added after removal  of "stat" file size check: found
+            // some undefined entries
+            if (r) {
+                const { mateRecordNumber } = r;
+                if (mateRecordNumber !== undefined &&
+                    mateRecordNumber >= 0 &&
+                    records[mateRecordNumber]) {
+                    associateIntraSliceMate(records, i, r, records[mateRecordNumber]);
+                }
             }
@@ -340,0 +338,0 @@ }

dist/cramFile/util.d.ts

@@ -0,1 +1,6 @@
+export declare const TWO_PWR_16_DBL: number;
+export declare const TWO_PWR_32_DBL: number;
+export declare const TWO_PWR_64_DBL: number;
+export declare const TWO_PWR_24_DBL: number;
+export declare const TWO_PWR_56_DBL: number;
 export declare function itf8Size(v: number): 1 | 2 | 3 | 4 | 5;
@@ -2,0 +7,0 @@ export declare function parseItf8(buffer: Uint8Array, initialOffset: number): readonly [number, number];

dist/cramFile/util.js

@@ -6,2 +6,3 @@ "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
+exports.TWO_PWR_56_DBL = exports.TWO_PWR_24_DBL = exports.TWO_PWR_64_DBL = exports.TWO_PWR_32_DBL = exports.TWO_PWR_16_DBL = void 0;
 exports.itf8Size = itf8Size;
@@ -13,5 +14,8 @@ exports.parseItf8 = parseItf8;
 exports.sequenceMD5 = sequenceMD5;
-const longfn_1 = require("longfn");
 const md5_1 = __importDefault(require("md5"));
-const getBits_1 = require("./codecs/getBits");
+exports.TWO_PWR_16_DBL = 1 << 16;
+exports.TWO_PWR_32_DBL = exports.TWO_PWR_16_DBL * exports.TWO_PWR_16_DBL;
+exports.TWO_PWR_64_DBL = exports.TWO_PWR_32_DBL * exports.TWO_PWR_32_DBL;
+exports.TWO_PWR_24_DBL = 1 << 24;
+exports.TWO_PWR_56_DBL = exports.TWO_PWR_24_DBL * exports.TWO_PWR_32_DBL;
 function itf8Size(v) {
@@ -36,25 +40,30 @@ if (!(v & ~0x7f)) {
     let result;
+    // Single byte value (0xxxxxxx)
     if (countFlags < 0x80) {
         result = countFlags;
-        offset = offset + 1;
+        offset += 1;
     }
+    // Two byte value (10xxxxxx)
     else if (countFlags < 0xc0) {
-        result = ((countFlags << 8) | buffer[offset + 1]) & 0x3fff;
-        offset = offset + 2;
+        result = ((countFlags & 0x3f) << 8) | buffer[offset + 1];
+        offset += 2;
     }
+    // Three byte value (110xxxxx)
     else if (countFlags < 0xe0) {
         result =
-            ((countFlags << 16) | (buffer[offset + 1] << 8) | buffer[offset + 2]) &
-                0x1fffff;
-        offset = offset + 3;
+            ((countFlags & 0x1f) << 16) |
+                (buffer[offset + 1] << 8) |
+                buffer[offset + 2];
+        offset += 3;
     }
+    // Four byte value (1110xxxx)
     else if (countFlags < 0xf0) {
         result =
-            ((countFlags << 24) |
+            ((countFlags & 0x0f) << 24) |
                 (buffer[offset + 1] << 16) |
                 (buffer[offset + 2] << 8) |
-                buffer[offset + 3]) &
-                0x0fffffff;
-        offset = offset + 4;
+                buffer[offset + 3];
+        offset += 4;
     }
+    // Five byte value (11110xxx)
     else {
@@ -67,56 +76,53 @@ result =
                 (buffer[offset + 4] & 0x0f);
-        // x=((0xff & 0x0f)<<28) | (0xff<<20) | (0xff<<12) | (0xff<<4) | (0x0f & 0x0f);
-        // TODO *val_p = uv < 0x80000000UL ? uv : -((int32_t) (0xffffffffUL - uv)) - 1;
-        offset = offset + 5;
+        offset += 5;
     }
-    if (offset > buffer.length) {
-        throw new getBits_1.CramBufferOverrunError('Attempted to read beyond end of buffer; this file seems truncated.');
-    }
     return [result, offset - initialOffset];
 }
 function parseLtf8(buffer, initialOffset) {
-    const dataView = new DataView(buffer.buffer);
     let offset = initialOffset;
     const countFlags = buffer[offset];
-    let n;
+    let value;
+    // Single byte value < 0x80
     if (countFlags < 0x80) {
-        n = countFlags;
+        value = countFlags;
         offset += 1;
     }
+    // Two byte value < 0xC0
     else if (countFlags < 0xc0) {
-        n = ((buffer[offset] << 8) | buffer[offset + 1]) & 0x3fff;
+        value = ((countFlags << 8) | buffer[offset + 1]) & 0x3fff;
         offset += 2;
     }
+    // Three byte value < 0xE0
     else if (countFlags < 0xe0) {
-        n =
-            ((buffer[offset] << 16) |
+        value =
+            ((countFlags & 0x3f) << 16) |
                 (buffer[offset + 1] << 8) |
-                buffer[offset + 2]) &
-                0x1fffff;
-        n = ((countFlags & 63) << 16) | dataView.getUint16(offset + 1, true);
+                buffer[offset + 2];
         offset += 3;
     }
+    // Four byte value < 0xF0
     else if (countFlags < 0xf0) {
-        n =
-            ((buffer[offset] << 24) |
+        value =
+            ((countFlags & 0x1f) << 24) |
                 (buffer[offset + 1] << 16) |
                 (buffer[offset + 2] << 8) |
-                buffer[offset + 3]) &
-                0x0fffffff;
+                buffer[offset + 3];
         offset += 4;
     }
+    // Five byte value < 0xF8
     else if (countFlags < 0xf8) {
-        n =
-            ((buffer[offset] & 15) * 2 ** 32 + (buffer[offset + 1] << 24)) |
-                ((buffer[offset + 2] << 16) |
+        value =
+            (buffer[offset] & 0x0f) * exports.TWO_PWR_32_DBL +
+                ((buffer[offset + 1] << 24) |
+                    (buffer[offset + 2] << 16) |
                     (buffer[offset + 3] << 8) |
                     buffer[offset + 4]);
-        // TODO *val_p = uv < 0x80000000UL ? uv : -((int32_t) (0xffffffffUL - uv)) - 1;
         offset += 5;
     }
+    // Six byte value < 0xFC
     else if (countFlags < 0xfc) {
-        n =
-            ((((buffer[offset] & 7) << 8) | buffer[offset + 1]) * 2 ** 32 +
-                (buffer[offset + 2] << 24)) |
-                ((buffer[offset + 3] << 16) |
+        value =
+            (((buffer[offset] & 0x07) << 8) | buffer[offset + 1]) * exports.TWO_PWR_32_DBL +
+                ((buffer[offset + 2] << 24) |
+                    (buffer[offset + 3] << 16) |
                     (buffer[offset + 4] << 8) |
@@ -126,10 +132,11 @@ buffer[offset + 5]);
     }
+    // Seven byte value < 0xFE
     else if (countFlags < 0xfe) {
-        n =
-            ((((buffer[offset] & 3) << 16) |
+        value =
+            (((buffer[offset] & 0x03) << 16) |
                 (buffer[offset + 1] << 8) |
                 buffer[offset + 2]) *
-                2 ** 32 +
-                (buffer[offset + 3] << 24)) |
-                ((buffer[offset + 4] << 16) |
+                exports.TWO_PWR_32_DBL +
+                ((buffer[offset + 3] << 24) |
+                    (buffer[offset + 4] << 16) |
                     (buffer[offset + 5] << 8) |
@@ -139,11 +146,32 @@ buffer[offset + 6]);
     }
+    // Eight byte value < 0xFF
     else if (countFlags < 0xff) {
-        n = (0, longfn_1.toNumber)((0, longfn_1.fromBytesBE)(buffer.slice(offset + 1, offset + 8), false));
+        value =
+            ((buffer[offset + 1] << 24) |
+                (buffer[offset + 2] << 16) |
+                (buffer[offset + 3] << 8) |
+                buffer[offset + 4]) *
+                exports.TWO_PWR_32_DBL +
+                ((buffer[offset + 5] << 24) |
+                    (buffer[offset + 6] << 16) |
+                    (buffer[offset + 7] << 8) |
+                    buffer[offset + 8]);
         offset += 8;
     }
+    // Nine byte value
     else {
-        n = (0, longfn_1.toNumber)((0, longfn_1.fromBytesBE)(buffer.subarray(offset + 1, offset + 9), false));
+        value =
+            buffer[offset + 1] * exports.TWO_PWR_56_DBL +
+                ((buffer[offset + 2] << 24) |
+                    (buffer[offset + 3] << 16) |
+                    (buffer[offset + 4] << 8) |
+                    buffer[offset + 5]) *
+                    exports.TWO_PWR_32_DBL +
+                ((buffer[offset + 6] << 24) |
+                    (buffer[offset + 7] << 16) |
+                    (buffer[offset + 8] << 8) |
+                    buffer[offset + 9]);
         offset += 9;
     }
-    return [n, offset - initialOffset];
+    return [value, offset - initialOffset];
 }
@@ -158,6 +186,2 @@ function parseItem(buffer, parser, startBufferPosition = 0, startFilePosition = 0) {
 }
-// this would be nice as a decorator, but i'm a little worried about babel
-// support for it going away or changing. memoizes a method in the stupidest
-// possible way, with no regard for the arguments.  actually, this only works
-// on methods that take no arguments
 function tinyMemoize(_class, methodName) {
@@ -164,0 +188,0 @@ const method = _class.prototype[methodName];

esm/cramFile/codecs/_base.d.ts

@@ -26,3 +26,3 @@ import { CramFileBlock } from '../file';
     constructor(parameters: TParameters, dataType: TResult);
-    abstract decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): DataTypeMapping[TResult];
+    abstract decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): DataTypeMapping[TResult] | undefined;
 }

esm/cramFile/codecs/byteArrayLength.js

@@ -10,7 +10,8 @@ import CramCodec from './_base';
         const lengthCodec = this._getLengthCodec();
-        const arrayLength = lengthCodec.decode(slice, coreDataBlock, blocksByContentId, cursors);
+        const arrayLength = lengthCodec.decode(slice, coreDataBlock, blocksByContentId, cursors) || 0;
         const dataCodec = this._getDataCodec();
         const data = new Uint8Array(arrayLength);
         for (let i = 0; i < arrayLength; i += 1) {
-            data[i] = dataCodec.decode(slice, coreDataBlock, blocksByContentId, cursors);
+            data[i] =
+                dataCodec.decode(slice, coreDataBlock, blocksByContentId, cursors) || 0;
         }
@@ -17,0 +18,0 @@ return data;

esm/cramFile/codecs/external.d.ts

@@ -8,5 +8,5 @@ import CramCodec, { Cursor, Cursors } from './_base';
     constructor(parameters: ExternalCramEncoding['parameters'], dataType: 'int' | 'byte');
-    decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): number;
+    decode(slice: CramSlice, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors): number | undefined;
     _decodeInt(contentBlock: CramFileBlock, cursor: Cursor): number;
     _decodeByte(contentBlock: CramFileBlock, cursor: Cursor): number;
 }

esm/cramFile/codecs/external.js

 import CramCodec from './_base';
-import { CramMalformedError, CramUnimplementedError } from '../../errors';
+import { CramUnimplementedError } from '../../errors';
 import { parseItf8 } from '../util';
@@ -21,7 +21,4 @@ import { CramBufferOverrunError } from './getBits';
         const contentBlock = blocksByContentId[blockContentId];
-        if (!contentBlock) {
-            throw new CramMalformedError(`no block found with content ID ${blockContentId}}`);
-        }
         const cursor = cursors.externalBlocks.getCursor(blockContentId);
-        return this._decodeData(contentBlock, cursor);
+        return contentBlock ? this._decodeData(contentBlock, cursor) : undefined;
     }
@@ -28,0 +25,0 @@ _decodeInt(contentBlock, cursor) {

esm/cramFile/container/index.d.ts

@@ -8,3 +8,8 @@ import CramSlice from '../slice';
     constructor(file: CramFile, filePosition: number);
-    getHeader(): Promise<({
+    getHeader(): Promise<{
+        _size: number;
+        _endPosition: number;
+        numLandmarks: number;
+        landmarks: number[];
+        crc32?: number | undefined;
         length: number;
@@ -15,20 +20,6 @@ refSeqId: number;
         numBlocks: number;
-        numLandmarks: number;
         numBases: number | undefined;
         recordCounter: number;
         numRecords: number;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        numLandmarks: number;
-        landmarks: number[];
-        crc32?: number | undefined;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        _size: number;
-        _endPosition: number;
-    }) | undefined>;
+    }>;
     getCompressionHeaderBlock(): Promise<{
@@ -53,7 +44,12 @@ parsedContent: {
         crc32?: number;
-    } | null | undefined>;
-    getFirstBlock(): Promise<import("../file").CramFileBlock | undefined>;
+    } | null>;
+    getFirstBlock(): Promise<import("../file").CramFileBlock>;
     getCompressionScheme(): Promise<CramContainerCompressionScheme | undefined>;
     getSlice(slicePosition: number, sliceSize: number): CramSlice;
-    _readContainerHeader(position: number): Promise<({
+    _readContainerHeader(position: number): Promise<{
+        _size: number;
+        _endPosition: number;
+        numLandmarks: number;
+        landmarks: number[];
+        crc32?: number | undefined;
         length: number;
@@ -64,20 +60,6 @@ refSeqId: number;
         numBlocks: number;
-        numLandmarks: number;
         numBases: number | undefined;
         recordCounter: number;
         numRecords: number;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        numLandmarks: number;
-        landmarks: number[];
-        crc32?: number | undefined;
-    } & {
-        _endPosition: number;
-        _size: number;
-    } & {
-        _size: number;
-        _endPosition: number;
-    }) | undefined>;
+    }>;
 }

esm/cramFile/container/index.js

@@ -19,3 +19,3 @@ import { CramMalformedError } from '../../errors';
         // header
-        if (!containerHeader?.numRecords) {
+        if (!containerHeader.numRecords) {
             return null;
@@ -26,5 +26,2 @@ }
         const block = await this.getFirstBlock();
-        if (block === undefined) {
-            return undefined;
-        }
         if (block.contentType !== 'COMPRESSION_HEADER') {
@@ -41,5 +38,2 @@ throw new CramMalformedError(`invalid content type ${block.contentType} in compression header block`);
         const containerHeader = await this.getHeader();
-        if (!containerHeader) {
-            return undefined;
-        }
         return this.file.readBlock(containerHeader._endPosition);
@@ -65,7 +59,2 @@ }
         const { cramContainerHeader1, cramContainerHeader2 } = sectionParsers;
-        const { size: fileSize } = await this.file.stat();
-        if (position >= fileSize) {
-            console.warn(`pos:${position}>=fileSize:${fileSize} in cram container`);
-            return undefined;
-        }
         // parse the container header. do it in 2 pieces because you cannot tell
@@ -76,7 +65,2 @@ // how much to buffer until you read numLandmarks
         const numLandmarksSize = itf8Size(header1.numLandmarks);
-        if (position + header1.length >= fileSize) {
-            // header indicates container goes beyond fileSize
-            console.warn(`container at ${position} is beyond fileSize:${fileSize}, skipping`);
-            return undefined;
-        }
         const bytes2 = await this.file.read(cramContainerHeader2.maxLength(header1.numLandmarks), position + header1._size - numLandmarksSize);
@@ -87,7 +71,8 @@ const header2 = parseItem(bytes2, cramContainerHeader2.parser);
         }
-        const completeHeader = Object.assign(header1, header2, {
+        return {
+            ...header1,
+            ...header2,
             _size: header1._size + header2._size - numLandmarksSize,
             _endPosition: header1._size + header2._size - numLandmarksSize + position,
-        });
-        return completeHeader;
+        };
     }
@@ -94,0 +79,0 @@ }

esm/cramFile/file.d.ts

@@ -35,3 +35,2 @@ import QuickLRU from 'quick-lru';
     constructor(args: CramFileArgs);
-    stat(): Promise<import("generic-filehandle2").Stats>;
     read(length: number, position: number): Promise<Uint8Array<ArrayBuffer>>;
@@ -56,6 +55,9 @@ getDefinition(): Promise<{
      * @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
      */
     containerCount(): Promise<number | undefined>;
     getContainerAtPosition(position: number): CramContainer;
-    readBlockHeader(position: number): Promise<({
+    readBlockHeader(position: number): Promise<{
         uncompressedSize: number;
@@ -69,3 +71,3 @@ compressedSize: number;
         _size: number;
-    }) | undefined>;
+    }>;
     _parseSection<T>(section: {
@@ -77,8 +79,8 @@ maxLength: number;
         };
-    }, position: number, size?: number, preReadBuffer?: Uint8Array): Promise<(T & {
+    }, position: number, size?: number, preReadBuffer?: Uint8Array): Promise<T & {
         _endPosition: number;
         _size: number;
-    }) | undefined>;
+    }>;
     _uncompress(compressionMethod: CompressionMethod, inputBuffer: Uint8Array, uncompressedSize: number): Promise<any>;
-    readBlock(position: number): Promise<CramFileBlock | undefined>;
+    readBlock(position: number): Promise<CramFileBlock>;
 }

esm/cramFile/file.js

@@ -56,7 +56,2 @@ import bzip2 from 'bzip2';
     }
-    // can just stat this object like a filehandle
-    stat() {
-        return this.file.stat();
-    }
-    // can just stat this object like a filehandle
     read(length, position) {
@@ -82,15 +77,10 @@ return this.file.read(length, position);
         const firstBlock = await firstContainer.getFirstBlock();
-        if (firstBlock === undefined) {
-            return parseHeaderText('');
-        }
-        else {
-            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 parseHeaderText(text);
-        }
+        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 parseHeaderText(text);
     }
@@ -105,4 +95,2 @@ async getHeaderText() {
         let position = sectionParsers.cramFileDefinition.maxLength;
-        const { size: fileSize } = await this.file.stat();
-        const { cramContainerHeader1 } = sectionParsers;
         // skip with a series of reads to the proper container
@@ -113,10 +101,7 @@ let currentContainer;
             // and have not found that container, it does not exist
-            if (position + cramContainerHeader1.maxLength + 8 >= fileSize) {
-                return undefined;
-            }
+            // if (position + cramContainerHeader1.maxLength + 8 >= fileSize) {
+            //   return undefined
+            // }
             currentContainer = this.getContainerAtPosition(position);
             const currentHeader = await currentContainer.getHeader();
-            if (!currentHeader) {
-                throw new CramMalformedError(`container ${containerNumber} not found in file`);
-            }
             // if this is the first container, read all the blocks in the container
@@ -129,5 +114,2 @@ // to determine its length, because we cannot trust the container
                     const block = await this.readBlock(position);
-                    if (block === undefined) {
-                        return undefined;
-                    }
                     position = block._endPosition;
@@ -156,2 +138,5 @@ }
      * @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
      */
@@ -161,31 +146,30 @@ async containerCount() {
         const sectionParsers = getSectionParsers(majorVersion);
-        const { size: fileSize } = await this.file.stat();
-        const { cramContainerHeader1 } = sectionParsers;
         let containerCount = 0;
         let position = sectionParsers.cramFileDefinition.maxLength;
-        while (position + cramContainerHeader1.maxLength + 8 < fileSize) {
-            const currentHeader = await this.getContainerAtPosition(position).getHeader();
-            if (!currentHeader) {
-                break;
-            }
-            // 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);
-                    if (block === undefined) {
-                        return undefined;
+        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;
                     }
-                    position = block._endPosition;
                 }
+                else {
+                    // otherwise, just traverse to the next container using the container's
+                    // length
+                    position += currentHeader._size + currentHeader.length;
+                }
+                containerCount += 1;
             }
-            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;
@@ -200,6 +184,2 @@ }
         const { cramBlockHeader } = sectionParsers;
-        const { size: fileSize } = await this.file.stat();
-        if (position + cramBlockHeader.maxLength >= fileSize) {
-            return undefined;
-        }
         const buffer = await this.file.read(cramBlockHeader.maxLength, position);
@@ -209,13 +189,3 @@ return parseItem(buffer, cramBlockHeader.parser, 0, position);
     async _parseSection(section, position, size = section.maxLength, preReadBuffer) {
-        let buffer;
-        if (preReadBuffer) {
-            buffer = preReadBuffer;
-        }
-        else {
-            const { size: fileSize } = await this.file.stat();
-            if (position + size >= fileSize) {
-                return undefined;
-            }
-            buffer = await this.file.read(size, position);
-        }
+        const buffer = preReadBuffer ?? (await this.file.read(size, position));
         const data = parseItem(buffer, section.parser, 0, position);
@@ -276,5 +246,2 @@ if (data._size !== size) {
         const blockHeader = await this.readBlockHeader(position);
-        if (blockHeader === undefined) {
-            return undefined;
-        }
         const blockContentPosition = blockHeader._endPosition;
@@ -294,5 +261,2 @@ const d = await this.file.read(blockHeader.compressedSize, blockContentPosition);
             const crc = await this._parseSection(sectionParsers.cramBlockCrc32, blockContentPosition + blockHeader.compressedSize);
-            if (crc === undefined) {
-                return undefined;
-            }
             block.crc32 = crc.crc32;
@@ -299,0 +263,0 @@ // check the block data crc32

esm/cramFile/sectionParsers.js

@@ -557,4 +557,5 @@ import { parseItf8, parseLtf8 } from './util';
             offset += 4;
-            // reference sequence identifier, -1 for unmapped reads, -2 for multiple
-            // reference sequences
+            // reference sequence identifier:
+            // -1 for unmapped reads,
+            // -2 for multiple reference sequences
             const [refSeqId, newOffset1] = parseItf8(buffer, offset);
@@ -561,0 +562,0 @@ offset += newOffset1;

esm/cramFile/slice/decodeRecord.d.ts

@@ -7,6 +7,6 @@ import { Cursors, DataTypeMapping } from '../codecs/_base';
 import { CramFileBlock } from '../file';
-export type DataSeriesDecoder = <T extends DataSeriesEncodingKey>(dataSeriesName: T) => DataTypeMapping[DataSeriesTypes[T]];
+export type DataSeriesDecoder = <T extends DataSeriesEncodingKey>(dataSeriesName: T) => DataTypeMapping[DataSeriesTypes[T]] | undefined;
 export default function decodeRecord(slice: CramSlice, decodeDataSeries: DataSeriesDecoder, compressionScheme: CramContainerCompressionScheme, sliceHeader: SliceHeader, coreDataBlock: CramFileBlock, blocksByContentId: Record<number, CramFileBlock>, cursors: Cursors, majorVersion: number, recordNumber: number): {
     readLength: number;
-    sequenceId: number;
+    sequenceId: number | undefined;
     cramFlags: number;
@@ -13,0 +13,0 @@ flags: number;

esm/cramFile/slice/decodeRecord.js

@@ -180,5 +180,4 @@ import { CramMalformedError } from '../../errors';
     let flags = decodeDataSeries('BF');
-    // note: the C data type of compressionFlags is byte in cram v1
-    // and int32 in cram v2+, but that does not matter for us here
-    // in javascript land.
+    // note: the C data type of compressionFlags is byte in cram v1 and int32 in
+    // cram v2+, but that does not matter for us here in javascript land.
     const cramFlags = decodeDataSeries('CF');
@@ -259,3 +258,7 @@ if (!isMappedSliceHeader(sliceHeader.parsedContent)) {
         tags[tagName] =
-            typeof tagData === 'number' ? tagData : parseTagData(tagType, tagData);
+            tagData === undefined
+                ? undefined
+                : typeof tagData === 'number'
+                    ? tagData
+                    : parseTagData(tagType, tagData);
     }
@@ -262,0 +265,0 @@ let readFeatures;

esm/cramFile/slice/index.js

@@ -137,9 +137,3 @@ import { CramArgumentError, CramMalformedError } from '../../errors';
         const containerHeader = await this.container.getHeader();
-        if (!containerHeader) {
-            throw new Error('no container header detected');
-        }
         const header = await this.file.readBlock(containerHeader._endPosition + this.containerPosition);
-        if (header === undefined) {
-            throw new Error('block header undefined');
-        }
         if (header.contentType === 'MAPPED_SLICE_HEADER') {
@@ -165,5 +159,2 @@ const content = parseItem(header.content, sectionParsers.cramMappedSliceHeader.parser, 0, containerHeader._endPosition);
             const block = await this.file.readBlock(blockPosition);
-            if (block === undefined) {
-                throw new Error('block undefined');
-            }
             blocks[i] = block;
@@ -300,5 +291,3 @@ blockPosition = blocks[i]._endPosition;
             }
-            // console.log(dataSeriesName, Object.getPrototypeOf(codec))
-            const decoded = codec.decode(this, coreDataBlock, blocksByContentId, cursors);
-            return decoded;
+            return codec.decode(this, coreDataBlock, blocksByContentId, cursors);
         };
@@ -328,7 +317,16 @@ const records = new Array(sliceHeader.parsedContent.numRecords);
         // interpret `recordsToNextFragment` attributes to make standard `mate`
-        // objects Resolve mate pair cross-references between records in this slice
+        // objects
+        //
+        // Resolve mate pair cross-references between records in this slice
         for (let i = 0; i < records.length; i += 1) {
-            const { mateRecordNumber } = records[i];
-            if (mateRecordNumber !== undefined && mateRecordNumber >= 0) {
-                associateIntraSliceMate(records, i, records[i], records[mateRecordNumber]);
+            const r = records[i];
+            // check for !!r added after removal  of "stat" file size check: found
+            // some undefined entries
+            if (r) {
+                const { mateRecordNumber } = r;
+                if (mateRecordNumber !== undefined &&
+                    mateRecordNumber >= 0 &&
+                    records[mateRecordNumber]) {
+                    associateIntraSliceMate(records, i, r, records[mateRecordNumber]);
+                }
             }
@@ -335,0 +333,0 @@ }

esm/cramFile/util.d.ts

@@ -0,1 +1,6 @@
+export declare const TWO_PWR_16_DBL: number;
+export declare const TWO_PWR_32_DBL: number;
+export declare const TWO_PWR_64_DBL: number;
+export declare const TWO_PWR_24_DBL: number;
+export declare const TWO_PWR_56_DBL: number;
 export declare function itf8Size(v: number): 1 | 2 | 3 | 4 | 5;
@@ -2,0 +7,0 @@ export declare function parseItf8(buffer: Uint8Array, initialOffset: number): readonly [number, number];

esm/cramFile/util.js

@@ -1,4 +0,7 @@
-import { fromBytesBE, toNumber } from 'longfn';
 import md5 from 'md5';
-import { CramBufferOverrunError } from './codecs/getBits';
+export const TWO_PWR_16_DBL = 1 << 16;
+export const TWO_PWR_32_DBL = TWO_PWR_16_DBL * TWO_PWR_16_DBL;
+export const TWO_PWR_64_DBL = TWO_PWR_32_DBL * TWO_PWR_32_DBL;
+export const TWO_PWR_24_DBL = 1 << 24;
+export const TWO_PWR_56_DBL = TWO_PWR_24_DBL * TWO_PWR_32_DBL;
 export function itf8Size(v) {
@@ -23,25 +26,30 @@ if (!(v & ~0x7f)) {
     let result;
+    // Single byte value (0xxxxxxx)
     if (countFlags < 0x80) {
         result = countFlags;
-        offset = offset + 1;
+        offset += 1;
     }
+    // Two byte value (10xxxxxx)
     else if (countFlags < 0xc0) {
-        result = ((countFlags << 8) | buffer[offset + 1]) & 0x3fff;
-        offset = offset + 2;
+        result = ((countFlags & 0x3f) << 8) | buffer[offset + 1];
+        offset += 2;
     }
+    // Three byte value (110xxxxx)
     else if (countFlags < 0xe0) {
         result =
-            ((countFlags << 16) | (buffer[offset + 1] << 8) | buffer[offset + 2]) &
-                0x1fffff;
-        offset = offset + 3;
+            ((countFlags & 0x1f) << 16) |
+                (buffer[offset + 1] << 8) |
+                buffer[offset + 2];
+        offset += 3;
     }
+    // Four byte value (1110xxxx)
     else if (countFlags < 0xf0) {
         result =
-            ((countFlags << 24) |
+            ((countFlags & 0x0f) << 24) |
                 (buffer[offset + 1] << 16) |
                 (buffer[offset + 2] << 8) |
-                buffer[offset + 3]) &
-                0x0fffffff;
-        offset = offset + 4;
+                buffer[offset + 3];
+        offset += 4;
     }
+    // Five byte value (11110xxx)
     else {
@@ -54,56 +62,53 @@ result =
                 (buffer[offset + 4] & 0x0f);
-        // x=((0xff & 0x0f)<<28) | (0xff<<20) | (0xff<<12) | (0xff<<4) | (0x0f & 0x0f);
-        // TODO *val_p = uv < 0x80000000UL ? uv : -((int32_t) (0xffffffffUL - uv)) - 1;
-        offset = offset + 5;
+        offset += 5;
     }
-    if (offset > buffer.length) {
-        throw new CramBufferOverrunError('Attempted to read beyond end of buffer; this file seems truncated.');
-    }
     return [result, offset - initialOffset];
 }
 export function parseLtf8(buffer, initialOffset) {
-    const dataView = new DataView(buffer.buffer);
     let offset = initialOffset;
     const countFlags = buffer[offset];
-    let n;
+    let value;
+    // Single byte value < 0x80
     if (countFlags < 0x80) {
-        n = countFlags;
+        value = countFlags;
         offset += 1;
     }
+    // Two byte value < 0xC0
     else if (countFlags < 0xc0) {
-        n = ((buffer[offset] << 8) | buffer[offset + 1]) & 0x3fff;
+        value = ((countFlags << 8) | buffer[offset + 1]) & 0x3fff;
         offset += 2;
     }
+    // Three byte value < 0xE0
     else if (countFlags < 0xe0) {
-        n =
-            ((buffer[offset] << 16) |
+        value =
+            ((countFlags & 0x3f) << 16) |
                 (buffer[offset + 1] << 8) |
-                buffer[offset + 2]) &
-                0x1fffff;
-        n = ((countFlags & 63) << 16) | dataView.getUint16(offset + 1, true);
+                buffer[offset + 2];
         offset += 3;
     }
+    // Four byte value < 0xF0
     else if (countFlags < 0xf0) {
-        n =
-            ((buffer[offset] << 24) |
+        value =
+            ((countFlags & 0x1f) << 24) |
                 (buffer[offset + 1] << 16) |
                 (buffer[offset + 2] << 8) |
-                buffer[offset + 3]) &
-                0x0fffffff;
+                buffer[offset + 3];
         offset += 4;
     }
+    // Five byte value < 0xF8
     else if (countFlags < 0xf8) {
-        n =
-            ((buffer[offset] & 15) * 2 ** 32 + (buffer[offset + 1] << 24)) |
-                ((buffer[offset + 2] << 16) |
+        value =
+            (buffer[offset] & 0x0f) * TWO_PWR_32_DBL +
+                ((buffer[offset + 1] << 24) |
+                    (buffer[offset + 2] << 16) |
                     (buffer[offset + 3] << 8) |
                     buffer[offset + 4]);
-        // TODO *val_p = uv < 0x80000000UL ? uv : -((int32_t) (0xffffffffUL - uv)) - 1;
         offset += 5;
     }
+    // Six byte value < 0xFC
     else if (countFlags < 0xfc) {
-        n =
-            ((((buffer[offset] & 7) << 8) | buffer[offset + 1]) * 2 ** 32 +
-                (buffer[offset + 2] << 24)) |
-                ((buffer[offset + 3] << 16) |
+        value =
+            (((buffer[offset] & 0x07) << 8) | buffer[offset + 1]) * TWO_PWR_32_DBL +
+                ((buffer[offset + 2] << 24) |
+                    (buffer[offset + 3] << 16) |
                     (buffer[offset + 4] << 8) |
@@ -113,10 +118,11 @@ buffer[offset + 5]);
     }
+    // Seven byte value < 0xFE
     else if (countFlags < 0xfe) {
-        n =
-            ((((buffer[offset] & 3) << 16) |
+        value =
+            (((buffer[offset] & 0x03) << 16) |
                 (buffer[offset + 1] << 8) |
                 buffer[offset + 2]) *
-                2 ** 32 +
-                (buffer[offset + 3] << 24)) |
-                ((buffer[offset + 4] << 16) |
+                TWO_PWR_32_DBL +
+                ((buffer[offset + 3] << 24) |
+                    (buffer[offset + 4] << 16) |
                     (buffer[offset + 5] << 8) |
@@ -126,11 +132,32 @@ buffer[offset + 6]);
     }
+    // Eight byte value < 0xFF
     else if (countFlags < 0xff) {
-        n = toNumber(fromBytesBE(buffer.slice(offset + 1, offset + 8), false));
+        value =
+            ((buffer[offset + 1] << 24) |
+                (buffer[offset + 2] << 16) |
+                (buffer[offset + 3] << 8) |
+                buffer[offset + 4]) *
+                TWO_PWR_32_DBL +
+                ((buffer[offset + 5] << 24) |
+                    (buffer[offset + 6] << 16) |
+                    (buffer[offset + 7] << 8) |
+                    buffer[offset + 8]);
         offset += 8;
     }
+    // Nine byte value
     else {
-        n = toNumber(fromBytesBE(buffer.subarray(offset + 1, offset + 9), false));
+        value =
+            buffer[offset + 1] * TWO_PWR_56_DBL +
+                ((buffer[offset + 2] << 24) |
+                    (buffer[offset + 3] << 16) |
+                    (buffer[offset + 4] << 8) |
+                    buffer[offset + 5]) *
+                    TWO_PWR_32_DBL +
+                ((buffer[offset + 6] << 24) |
+                    (buffer[offset + 7] << 16) |
+                    (buffer[offset + 8] << 8) |
+                    buffer[offset + 9]);
         offset += 9;
     }
-    return [n, offset - initialOffset];
+    return [value, offset - initialOffset];
 }
@@ -145,6 +172,2 @@ export function parseItem(buffer, parser, startBufferPosition = 0, startFilePosition = 0) {
 }
-// this would be nice as a decorator, but i'm a little worried about babel
-// support for it going away or changing. memoizes a method in the stupidest
-// possible way, with no regard for the arguments.  actually, this only works
-// on methods that take no arguments
 export function tinyMemoize(_class, methodName) {
@@ -151,0 +174,0 @@ const method = _class.prototype[methodName];

package.json

 {
   "name": "@gmod/cram",
-  "version": "4.0.4",
+  "version": "4.0.5",
   "description": "read CRAM files with pure Javascript",
@@ -18,3 +18,2 @@ "license": "MIT",
   "files": [
-    "errors.js",
     "src",
@@ -28,3 +27,3 @@ "dist",
     "format": "prettier --write .",
-    "docs": "documentation readme --shallow src/indexedCramFile.ts --section=IndexedCramFile; documentation readme --shallow src/cramFile/file.ts --section=CramFile; documentation readme --shallow src/craiIndex.ts --section=CraiIndex;  documentation readme --shallow errors.ts '--section=Exception Classes'; documentation readme --shallow src/cramFile/file.ts --section=CramFile; documentation readme --shallow src/cramFile/record.ts --section=CramRecord",
+    "docs": "documentation readme --shallow src/indexedCramFile.ts --section=IndexedCramFile; documentation readme --shallow src/cramFile/file.ts --section=CramFile; documentation readme --shallow src/craiIndex.ts --section=CraiIndex;  documentation readme --shallow src/cramFile/file.ts --section=CramFile; documentation readme --shallow src/cramFile/record.ts --section=CramRecord",
     "prebuild": "yarn clean",
@@ -36,3 +35,3 @@ "clean": "rimraf dist esm",
     "postbuild": "webpack",
-    "preversion": "yarn test --run && yarn build && cp dist/errors.js errors.js",
+    "preversion": "yarn test --run && yarn build",
     "postpublish": "rm errors.js",
@@ -51,3 +50,2 @@ "postversion": "git push --follow-tags"
     "generic-filehandle2": "^1.0.0",
-    "longfn": "^1.3.1",
     "md5": "^2.2.1",
@@ -54,0 +52,0 @@ "pako": "^1.0.4",

README.md

@@ -145,2 +145,3 @@ # @gmod/cram
 - [CramRecord](#cramrecord)
+  - [Parameters](#parameters)
   - [isPaired](#ispaired)
@@ -165,3 +166,3 @@ - [isProperlyPaired](#isproperlypaired)
   - [addReferenceSequence](#addreferencesequence)
-    - [Parameters](#parameters)
+    - [Parameters](#parameters-1)
 
@@ -172,2 +173,24 @@ #### CramRecord
 
+##### Parameters
+
+- `$0` **any** 
+
+  - `$0.flags`  
+  - `$0.cramFlags`  
+  - `$0.readLength`  
+  - `$0.mappingQuality`  
+  - `$0.lengthOnRef`  
+  - `$0.qualityScores`  
+  - `$0.mateRecordNumber`  
+  - `$0.readBases`  
+  - `$0.readFeatures`  
+  - `$0.mateToUse`  
+  - `$0.readGroupId`  
+  - `$0.readName`  
+  - `$0.sequenceId`  
+  - `$0.uniqueId`  
+  - `$0.templateSize`  
+  - `$0.alignmentStart`  
+  - `$0.tags`  
+
 ##### isPaired
@@ -288,3 +311,3 @@
 Annotates this feature with the given reference sequence basepair information.
-This will add a `sub` and a `ref` item to base subsitution read features given
+This will add a `sub` and a `ref` item to base substitution read features given
 the actual substituted and reference base pairs, and will make the
@@ -296,12 +319,12 @@ `getReadSequence()` method work.
 - `refRegion`
-  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)**
+  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)** 
 
   - `refRegion.start`
-    **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+    **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)** 
   - `refRegion.end`
-    **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+    **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)** 
   - `refRegion.seq`
-    **[string](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/String)**
+    **[string](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/String)** 
 
-- `compressionScheme` **CramContainerCompressionScheme**
+- `compressionScheme` **CramContainerCompressionScheme** 
 
@@ -344,5 +367,5 @@ Returns
 - `args`
-  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)**
+  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)** 
 
-  - `args.cram` **CramFile**
+  - `args.cram` **CramFile** 
   - `args.index` **Index-like** object that supports
@@ -372,3 +395,9 @@ getEntriesForRange(seqId,start,end) -> Promise\[Array\[index entries]]
   end of the range of interest. 1-based closed coordinates.
-- `opts` (optional, default `{}`)
+- `opts` **{viewAsPairs:
+  [boolean](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Boolean)?,
+  pairAcrossChr:
+  [boolean](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Boolean)?,
+  maxInsertSize:
+  [number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)?}**
+  (optional, default `{}`)
 
@@ -380,3 +409,3 @@ #### hasDataForReferenceSequence
 - `seqId`
-  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)** 
 
@@ -394,38 +423,11 @@ Returns
 
-- [constructor](#constructor)
-  - [Parameters](#parameters)
 - [containerCount](#containercount)
 
-#### constructor
+#### containerCount
 
-##### Parameters
+Returns
+**[Promise](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Promise)<([number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)
+|
+[undefined](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/undefined))>**&#x20;
 
-- `args`
-  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)**
-
-  - `args.filehandle`
-    **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)?**
-    a filehandle that implements the stat() and read() methods of the Node
-    filehandle API <https://nodejs.org/api/fs.html#fs_class_filehandle>
-  - `args.path`
-    **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)?**
-    path to the cram file
-  - `args.url`
-    **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)?**
-    url for the cram file. also supports file:// urls for local files
-  - `args.seqFetch`
-    **[function](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Statements/function)?**
-    a function with signature `(seqId, startCoordinate, endCoordinate)` that
-    returns a promise for a string of sequence bases
-  - `args.cacheSize`
-    **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)?**
-    optional maximum number of CRAM records to cache. default 20,000
-  - `args.checkSequenceMD5`
-    **[boolean](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Boolean)?**
-    default true. if false, disables verifying the MD5 checksum of the reference
-    sequence underlying a slice. In some applications, this check can cause an
-    inconvenient amount (many megabases) of sequences to be fetched.
-
-#### containerCount
-
 ### CraiIndex
@@ -449,9 +451,9 @@
 - `args`
-  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)**
+  **[object](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Object)**&#x20;
 
   - `args.path`
-    **[string](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/String)?**
+    **[string](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/String)?**&#x20;
   - `args.url`
-    **[string](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/String)?**
-  - `args.filehandle` **FileHandle?**
+    **[string](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/String)?**&#x20;
+  - `args.filehandle` **FileHandle?**&#x20;
 
@@ -463,3 +465,3 @@ #### hasDataForReferenceSequence
 - `seqId`
-  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**&#x20;
 
@@ -478,7 +480,7 @@ Returns
 - `seqId`
-  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**&#x20;
 - `queryStart`
-  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**&#x20;
 - `queryEnd`
-  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**
+  **[number](https://developer.mozilla.org/docs/Web/JavaScript/Reference/Global_Objects/Number)**&#x20;
 
@@ -485,0 +487,0 @@ Returns

src/cramFile/codecs/_base.ts

@@ -44,3 +44,3 @@ import { CramFileBlock } from '../file'
     cursors: Cursors,
-  ): DataTypeMapping[TResult]
+  ): DataTypeMapping[TResult] | undefined
 }

src/cramFile/codecs/byteArrayLength.ts

@@ -35,8 +35,4 @@ import CramCodec, { Cursors } from './_base'
     const lengthCodec = this._getLengthCodec()
-    const arrayLength = lengthCodec.decode(
-      slice,
-      coreDataBlock,
-      blocksByContentId,
-      cursors,
-    )
+    const arrayLength =
+      lengthCodec.decode(slice, coreDataBlock, blocksByContentId, cursors) || 0
 
@@ -46,8 +42,4 @@ const dataCodec = this._getDataCodec()
     for (let i = 0; i < arrayLength; i += 1) {
-      data[i] = dataCodec.decode(
-        slice,
-        coreDataBlock,
-        blocksByContentId,
-        cursors,
-      )
+      data[i] =
+        dataCodec.decode(slice, coreDataBlock, blocksByContentId, cursors) || 0
     }
@@ -54,0 +46,0 @@

src/cramFile/codecs/external.ts

 import CramCodec, { Cursor, Cursors } from './_base'
-import { CramMalformedError, CramUnimplementedError } from '../../errors'
+import { CramUnimplementedError } from '../../errors'
 import { CramFileBlock } from '../file'
@@ -42,9 +42,5 @@ import CramSlice from '../slice'
     const contentBlock = blocksByContentId[blockContentId]
-    if (!contentBlock) {
-      throw new CramMalformedError(
-        `no block found with content ID ${blockContentId}}`,
-      )
-    }
+
     const cursor = cursors.externalBlocks.getCursor(blockContentId)
-    return this._decodeData(contentBlock, cursor)
+    return contentBlock ? this._decodeData(contentBlock, cursor) : undefined
   }
@@ -51,0 +47,0 @@

src/cramFile/container/index.ts

@@ -24,3 +24,3 @@ import { CramMalformedError } from '../../errors'
     // header
-    if (!containerHeader?.numRecords) {
+    if (!containerHeader.numRecords) {
       return null
@@ -32,5 +32,2 @@ }
     const block = await this.getFirstBlock()
-    if (block === undefined) {
-      return undefined
-    }
     if (block.contentType !== 'COMPRESSION_HEADER') {
@@ -56,5 +53,2 @@ throw new CramMalformedError(
     const containerHeader = await this.getHeader()
-    if (!containerHeader) {
-      return undefined
-    }
     return this.file.readBlock(containerHeader._endPosition)
@@ -84,9 +78,3 @@ }
     const { cramContainerHeader1, cramContainerHeader2 } = sectionParsers
-    const { size: fileSize } = await this.file.stat()
 
-    if (position >= fileSize) {
-      console.warn(`pos:${position}>=fileSize:${fileSize} in cram container`)
-      return undefined
-    }
-
     // parse the container header. do it in 2 pieces because you cannot tell
@@ -100,9 +88,2 @@ // how much to buffer until you read numLandmarks
     const numLandmarksSize = itf8Size(header1.numLandmarks)
-    if (position + header1.length >= fileSize) {
-      // header indicates container goes beyond fileSize
-      console.warn(
-        `container at ${position} is beyond fileSize:${fileSize}, skipping`,
-      )
-      return undefined
-    }
 
@@ -124,8 +105,8 @@ const bytes2 = await this.file.read(
 
-    const completeHeader = Object.assign(header1, header2, {
+    return {
+      ...header1,
+      ...header2,
       _size: header1._size + header2._size - numLandmarksSize,
       _endPosition: header1._size + header2._size - numLandmarksSize + position,
-    })
-
-    return completeHeader
+    }
   }
@@ -132,0 +113,0 @@ }

src/cramFile/file.ts

@@ -105,8 +105,2 @@ import bzip2 from 'bzip2'
 
-  // can just stat this object like a filehandle
-  stat() {
-    return this.file.stat()
-  }
-
-  // can just stat this object like a filehandle
   read(length: number, position: number) {
@@ -137,16 +131,13 @@ return this.file.read(length, position)
     const firstBlock = await firstContainer.getFirstBlock()
-    if (firstBlock === undefined) {
-      return parseHeaderText('')
-    } else {
-      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 parseHeaderText(text)
-    }
+
+    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 parseHeaderText(text)
   }
@@ -163,4 +154,2 @@
     let position = sectionParsers.cramFileDefinition.maxLength
-    const { size: fileSize } = await this.file.stat()
-    const { cramContainerHeader1 } = sectionParsers
 
@@ -172,13 +161,9 @@ // skip with a series of reads to the proper container
       // and have not found that container, it does not exist
-      if (position + cramContainerHeader1.maxLength + 8 >= fileSize) {
-        return undefined
-      }
+      // if (position + cramContainerHeader1.maxLength + 8 >= fileSize) {
+      //   return undefined
+      // }
 
       currentContainer = this.getContainerAtPosition(position)
       const currentHeader = await currentContainer.getHeader()
-      if (!currentHeader) {
-        throw new CramMalformedError(
-          `container ${containerNumber} not found in file`,
-        )
-      }
+
       // if this is the first container, read all the blocks in the container
@@ -191,5 +176,2 @@ // to determine its length, because we cannot trust the container
           const block = await this.readBlock(position)
-          if (block === undefined) {
-            return undefined
-          }
           position = block._endPosition
@@ -227,2 +209,5 @@ }
    * @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
    */
@@ -232,31 +217,30 @@ async containerCount(): Promise<number | undefined> {
     const sectionParsers = getSectionParsers(majorVersion)
-    const { size: fileSize } = await this.file.stat()
-    const { cramContainerHeader1 } = sectionParsers
 
     let containerCount = 0
     let position = sectionParsers.cramFileDefinition.maxLength
-    while (position + cramContainerHeader1.maxLength + 8 < fileSize) {
-      const currentHeader =
-        await this.getContainerAtPosition(position).getHeader()
-      if (!currentHeader) {
-        break
-      }
-      // 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)
-          if (block === undefined) {
-            return undefined
+    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
           }
-          position = block._endPosition
+        } else {
+          // otherwise, just traverse to the next container using the container's
+          // length
+          position += currentHeader._size + currentHeader.length
         }
-      } else {
-        // otherwise, just traverse to the next container using the container's
-        // length
-        position += currentHeader._size + currentHeader.length
+        containerCount += 1
       }
-      containerCount += 1
+    } catch (e) {
+      containerCount--
+      /* do nothing */
     }
@@ -275,8 +259,3 @@
     const { cramBlockHeader } = sectionParsers
-    const { size: fileSize } = await this.file.stat()
 
-    if (position + cramBlockHeader.maxLength >= fileSize) {
-      return undefined
-    }
-
     const buffer = await this.file.read(cramBlockHeader.maxLength, position)
@@ -298,12 +277,3 @@ return parseItem(buffer, cramBlockHeader.parser, 0, position)
   ) {
-    let buffer: Uint8Array
-    if (preReadBuffer) {
-      buffer = preReadBuffer
-    } else {
-      const { size: fileSize } = await this.file.stat()
-      if (position + size >= fileSize) {
-        return undefined
-      }
-      buffer = await this.file.read(size, position)
-    }
+    const buffer = preReadBuffer ?? (await this.file.read(size, position))
     const data = parseItem(buffer, section.parser, 0, position)
@@ -368,5 +338,2 @@ if (data._size !== size) {
     const blockHeader = await this.readBlockHeader(position)
-    if (blockHeader === undefined) {
-      return undefined
-    }
     const blockContentPosition = blockHeader._endPosition
@@ -399,5 +366,2 @@
       )
-      if (crc === undefined) {
-        return undefined
-      }
       block.crc32 = crc.crc32
@@ -404,0 +368,0 @@

src/cramFile/record.ts

@@ -274,3 +274,3 @@ import Constants from './constants'
     this.readName = readName
-    this.sequenceId = sequenceId
+    this.sequenceId = sequenceId!
     this.uniqueId = uniqueId
@@ -277,0 +277,0 @@ this.templateSize = templateSize

src/cramFile/sectionParsers.ts

@@ -655,7 +655,10 @@ import { TupleOf } from '../typescript'
       const dataView = new DataView(b.buffer, b.byteOffset, b.length)
+
       // byte size of the container data (blocks)
       const length = dataView.getInt32(offset, true)
       offset += 4
-      // reference sequence identifier, -1 for unmapped reads, -2 for multiple
-      // reference sequences
+
+      // reference sequence identifier:
+      // -1 for unmapped reads,
+      // -2 for multiple reference sequences
       const [refSeqId, newOffset1] = parseItf8(buffer, offset)
@@ -662,0 +665,0 @@ offset += newOffset1

src/cramFile/slice/decodeRecord.ts

@@ -214,3 +214,3 @@ import { CramMalformedError } from '../../errors'
   dataSeriesName: T,
-) => DataTypeMapping[DataSeriesTypes[T]]
+) => DataTypeMapping[DataSeriesTypes[T]] | undefined
 
@@ -228,8 +228,7 @@ export default function decodeRecord(
 ) {
-  let flags = decodeDataSeries('BF')
+  let flags = decodeDataSeries('BF')!
 
-  // note: the C data type of compressionFlags is byte in cram v1
-  // and int32 in cram v2+, but that does not matter for us here
-  // in javascript land.
-  const cramFlags = decodeDataSeries('CF')
+  // note: the C data type of compressionFlags is byte in cram v1 and int32 in
+  // cram v2+, but that does not matter for us here in javascript land.
+  const cramFlags = decodeDataSeries('CF')!
 
@@ -245,5 +244,5 @@ if (!isMappedSliceHeader(sliceHeader.parsedContent)) {
 
-  const readLength = decodeDataSeries('RL')
+  const readLength = decodeDataSeries('RL')!
   // if APDelta, will calculate the true start in a second pass
-  let alignmentStart = decodeDataSeries('AP')
+  let alignmentStart = decodeDataSeries('AP')!
   if (compressionScheme.APdelta) {
@@ -253,7 +252,7 @@ alignmentStart = alignmentStart + cursors.lastAlignmentStart
   cursors.lastAlignmentStart = alignmentStart
-  const readGroupId = decodeDataSeries('RG')
+  const readGroupId = decodeDataSeries('RG')!
 
   let readName: string | undefined
   if (compressionScheme.readNamesIncluded) {
-    readName = readNullTerminatedString(decodeDataSeries('RN'))
+    readName = readNullTerminatedString(decodeDataSeries('RN')!)
   }
@@ -275,10 +274,10 @@
     // matter for javascript
-    const mateFlags = decodeDataSeries('MF')
+    const mateFlags = decodeDataSeries('MF')!
     let mateReadName: string | undefined
     if (!compressionScheme.readNamesIncluded) {
-      mateReadName = readNullTerminatedString(decodeDataSeries('RN'))
+      mateReadName = readNullTerminatedString(decodeDataSeries('RN')!)
       readName = mateReadName
     }
-    const mateSequenceId = decodeDataSeries('NS')
-    const mateAlignmentStart = decodeDataSeries('NP')
+    const mateSequenceId = decodeDataSeries('NS')!
+    const mateAlignmentStart = decodeDataSeries('NP')!
     if (mateFlags || mateSequenceId > -1) {
@@ -293,3 +292,3 @@ mateToUse = {
 
-    templateSize = decodeDataSeries('TS')
+    templateSize = decodeDataSeries('TS')!
 
@@ -307,3 +306,3 @@ // set mate unmapped if needed
   } else if (CramFlagsDecoder.isWithMateDownstream(cramFlags)) {
-    mateRecordNumber = decodeDataSeries('NF') + recordNumber + 1
+    mateRecordNumber = decodeDataSeries('NF')! + recordNumber + 1
   }
@@ -313,3 +312,3 @@
   // cram v1
-  const TLindex = decodeDataSeries('TL')
+  const TLindex = decodeDataSeries('TL')!
   if (TLindex < 0) {
@@ -333,3 +332,7 @@ /* TODO: check nTL: TLindex >= compressionHeader.tagEncoding.size */
     tags[tagName] =
-      typeof tagData === 'number' ? tagData : parseTagData(tagType, tagData)
+      tagData === undefined
+        ? undefined
+        : typeof tagData === 'number'
+          ? tagData
+          : parseTagData(tagType, tagData)
   }
@@ -344,3 +347,3 @@
     // reading read features
-    const readFeatureCount = decodeDataSeries('FN')
+    const readFeatureCount = decodeDataSeries('FN')!
     if (readFeatureCount) {
@@ -380,7 +383,7 @@ readFeatures = decodeReadFeatures(
     // mapping quality
-    mappingQuality = decodeDataSeries('MQ')
+    mappingQuality = decodeDataSeries('MQ')!
     if (CramFlagsDecoder.isPreservingQualityScores(cramFlags)) {
       qualityScores = new Array(readLength)
       for (let i = 0; i < qualityScores.length; i++) {
-        qualityScores[i] = decodeDataSeries('QS')
+        qualityScores[i] = decodeDataSeries('QS')!
       }
@@ -394,3 +397,3 @@ }
     for (let i = 0; i < bases.length; i++) {
-      bases[i] = decodeDataSeries('BA')
+      bases[i] = decodeDataSeries('BA')!
     }
@@ -402,3 +405,3 @@ readBases = String.fromCharCode(...bases)
       for (let i = 0; i < bases.length; i++) {
-        qualityScores[i] = decodeDataSeries('QS')
+        qualityScores[i] = decodeDataSeries('QS')!
       }
@@ -405,0 +408,0 @@ }

src/cramFile/slice/index.ts

@@ -23,2 +23,9 @@ import { CramArgumentError, CramMalformedError } from '../../errors'
 
+interface RefRegion {
+  id: number
+  start: number
+  end: number
+  seq: string | null
+}
+
 /**
@@ -195,5 +202,2 @@ * @private
     const containerHeader = await this.container.getHeader()
-    if (!containerHeader) {
-      throw new Error('no container header detected')
-    }
 
@@ -203,5 +207,2 @@ const header = await this.file.readBlock(
     )
-    if (header === undefined) {
-      throw new Error('block header undefined')
-    }
     if (header.contentType === 'MAPPED_SLICE_HEADER') {
@@ -238,5 +239,2 @@ const content = parseItem(
       const block = await this.file.readBlock(blockPosition)
-      if (block === undefined) {
-        throw new Error('block undefined')
-      }
       blocks[i] = block
@@ -411,3 +409,3 @@ blockPosition = blocks[i]!._endPosition
       dataSeriesName: T,
-    ): DataTypeMapping[DataSeriesTypes[T]] => {
+    ): DataTypeMapping[DataSeriesTypes[T]] | undefined => {
       const codec = compressionScheme.getCodecForDataSeries(dataSeriesName)
@@ -419,10 +417,3 @@ if (!codec) {
       }
-      // console.log(dataSeriesName, Object.getPrototypeOf(codec))
-      const decoded = codec.decode(
-        this,
-        coreDataBlock,
-        blocksByContentId,
-        cursors,
-      )
-      return decoded
+      return codec.decode(this, coreDataBlock, blocksByContentId, cursors)
     }
@@ -466,12 +457,18 @@ const records: CramRecord[] = new Array(
     // interpret `recordsToNextFragment` attributes to make standard `mate`
-    // objects Resolve mate pair cross-references between records in this slice
+    // objects
+    //
+    // Resolve mate pair cross-references between records in this slice
     for (let i = 0; i < records.length; i += 1) {
-      const { mateRecordNumber } = records[i]!
-      if (mateRecordNumber !== undefined && mateRecordNumber >= 0) {
-        associateIntraSliceMate(
-          records,
-          i,
-          records[i]!,
-          records[mateRecordNumber]!,
-        )
+      const r = records[i]
+      // check for !!r added after removal  of "stat" file size check: found
+      // some undefined entries
+      if (r) {
+        const { mateRecordNumber } = r
+        if (
+          mateRecordNumber !== undefined &&
+          mateRecordNumber >= 0 &&
+          records[mateRecordNumber]
+        ) {
+          associateIntraSliceMate(records, i, r, records[mateRecordNumber])
+        }
       }
@@ -511,6 +508,3 @@ }
         }
-        const refRegions: Record<
-          string,
-          { id: number; start: number; end: number; seq: string | null }
-        > = {}
+        const refRegions: Record<string, RefRegion> = {}
 
@@ -517,0 +511,0 @@ // iterate over the records to find the spans of the reference

src/cramFile/util.ts

@@ -1,5 +0,8 @@
-import { fromBytesBE, toNumber } from 'longfn'
 import md5 from 'md5'
 
-import { CramBufferOverrunError } from './codecs/getBits'
+export const TWO_PWR_16_DBL = 1 << 16
+export const TWO_PWR_32_DBL = TWO_PWR_16_DBL * TWO_PWR_16_DBL
+export const TWO_PWR_64_DBL = TWO_PWR_32_DBL * TWO_PWR_32_DBL
+export const TWO_PWR_24_DBL = 1 << 24
+export const TWO_PWR_56_DBL = TWO_PWR_24_DBL * TWO_PWR_32_DBL
 
@@ -26,22 +29,32 @@ export function itf8Size(v: number) {
   let result: number
+
+  // Single byte value (0xxxxxxx)
   if (countFlags < 0x80) {
     result = countFlags
-    offset = offset + 1
-  } else if (countFlags < 0xc0) {
-    result = ((countFlags << 8) | buffer[offset + 1]!) & 0x3fff
-    offset = offset + 2
-  } else if (countFlags < 0xe0) {
+    offset += 1
+  }
+  // Two byte value (10xxxxxx)
+  else if (countFlags < 0xc0) {
+    result = ((countFlags & 0x3f) << 8) | buffer[offset + 1]!
+    offset += 2
+  }
+  // Three byte value (110xxxxx)
+  else if (countFlags < 0xe0) {
     result =
-      ((countFlags << 16) | (buffer[offset + 1]! << 8) | buffer[offset + 2]!) &
-      0x1fffff
-    offset = offset + 3
-  } else if (countFlags < 0xf0) {
+      ((countFlags & 0x1f) << 16) |
+      (buffer[offset + 1]! << 8) |
+      buffer[offset + 2]!
+    offset += 3
+  }
+  // Four byte value (1110xxxx)
+  else if (countFlags < 0xf0) {
     result =
-      ((countFlags << 24) |
-        (buffer[offset + 1]! << 16) |
-        (buffer[offset + 2]! << 8) |
-        buffer[offset + 3]!) &
-      0x0fffffff
-    offset = offset + 4
-  } else {
+      ((countFlags & 0x0f) << 24) |
+      (buffer[offset + 1]! << 16) |
+      (buffer[offset + 2]! << 8) |
+      buffer[offset + 3]!
+    offset += 4
+  }
+  // Five byte value (11110xxx)
+  else {
     result =
@@ -53,11 +66,5 @@ ((countFlags & 0x0f) << 28) |
       (buffer[offset + 4]! & 0x0f)
-    // x=((0xff & 0x0f)<<28) | (0xff<<20) | (0xff<<12) | (0xff<<4) | (0x0f & 0x0f);
-    // TODO *val_p = uv < 0x80000000UL ? uv : -((int32_t) (0xffffffffUL - uv)) - 1;
-    offset = offset + 5
+    offset += 5
   }
-  if (offset > buffer.length) {
-    throw new CramBufferOverrunError(
-      'Attempted to read beyond end of buffer; this file seems truncated.',
-    )
-  }
+
   return [result, offset - initialOffset] as const
@@ -67,65 +74,97 @@ }
 export function parseLtf8(buffer: Uint8Array, initialOffset: number) {
-  const dataView = new DataView(buffer.buffer)
   let offset = initialOffset
   const countFlags = buffer[offset]!
-  let n: number
+  let value: number
+
+  // Single byte value < 0x80
   if (countFlags < 0x80) {
-    n = countFlags
+    value = countFlags
     offset += 1
-  } else if (countFlags < 0xc0) {
-    n = ((buffer[offset]! << 8) | buffer[offset + 1]!) & 0x3fff
+  }
+  // Two byte value < 0xC0
+  else if (countFlags < 0xc0) {
+    value = ((countFlags << 8) | buffer[offset + 1]!) & 0x3fff
     offset += 2
-  } else if (countFlags < 0xe0) {
-    n =
-      ((buffer[offset]! << 16) |
-        (buffer[offset + 1]! << 8) |
-        buffer[offset + 2]!) &
-      0x1fffff
-    n = ((countFlags & 63) << 16) | dataView.getUint16(offset + 1, true)
+  }
+  // Three byte value < 0xE0
+  else if (countFlags < 0xe0) {
+    value =
+      ((countFlags & 0x3f) << 16) |
+      (buffer[offset + 1]! << 8) |
+      buffer[offset + 2]!
     offset += 3
-  } else if (countFlags < 0xf0) {
-    n =
-      ((buffer[offset]! << 24) |
-        (buffer[offset + 1]! << 16) |
-        (buffer[offset + 2]! << 8) |
-        buffer[offset + 3]!) &
-      0x0fffffff
+  }
+  // Four byte value < 0xF0
+  else if (countFlags < 0xf0) {
+    value =
+      ((countFlags & 0x1f) << 24) |
+      (buffer[offset + 1]! << 16) |
+      (buffer[offset + 2]! << 8) |
+      buffer[offset + 3]!
     offset += 4
-  } else if (countFlags < 0xf8) {
-    n =
-      ((buffer[offset]! & 15) * 2 ** 32 + (buffer[offset + 1]! << 24)) |
-      ((buffer[offset + 2]! << 16) |
+  }
+  // Five byte value < 0xF8
+  else if (countFlags < 0xf8) {
+    value =
+      (buffer[offset]! & 0x0f) * TWO_PWR_32_DBL +
+      ((buffer[offset + 1]! << 24) |
+        (buffer[offset + 2]! << 16) |
         (buffer[offset + 3]! << 8) |
         buffer[offset + 4]!)
-    // TODO *val_p = uv < 0x80000000UL ? uv : -((int32_t) (0xffffffffUL - uv)) - 1;
     offset += 5
-  } else if (countFlags < 0xfc) {
-    n =
-      ((((buffer[offset]! & 7) << 8) | buffer[offset + 1]!) * 2 ** 32 +
-        (buffer[offset + 2]! << 24)) |
-      ((buffer[offset + 3]! << 16) |
+  }
+  // Six byte value < 0xFC
+  else if (countFlags < 0xfc) {
+    value =
+      (((buffer[offset]! & 0x07) << 8) | buffer[offset + 1]!) * TWO_PWR_32_DBL +
+      ((buffer[offset + 2]! << 24) |
+        (buffer[offset + 3]! << 16) |
         (buffer[offset + 4]! << 8) |
         buffer[offset + 5]!)
     offset += 6
-  } else if (countFlags < 0xfe) {
-    n =
-      ((((buffer[offset]! & 3) << 16) |
+  }
+  // Seven byte value < 0xFE
+  else if (countFlags < 0xfe) {
+    value =
+      (((buffer[offset]! & 0x03) << 16) |
         (buffer[offset + 1]! << 8) |
         buffer[offset + 2]!) *
-        2 ** 32 +
-        (buffer[offset + 3]! << 24)) |
-      ((buffer[offset + 4]! << 16) |
+        TWO_PWR_32_DBL +
+      ((buffer[offset + 3]! << 24) |
+        (buffer[offset + 4]! << 16) |
         (buffer[offset + 5]! << 8) |
         buffer[offset + 6]!)
     offset += 7
-  } else if (countFlags < 0xff) {
-    n = toNumber(fromBytesBE(buffer.slice(offset + 1, offset + 8), false))
-
+  }
+  // Eight byte value < 0xFF
+  else if (countFlags < 0xff) {
+    value =
+      ((buffer[offset + 1]! << 24) |
+        (buffer[offset + 2]! << 16) |
+        (buffer[offset + 3]! << 8) |
+        buffer[offset + 4]!) *
+        TWO_PWR_32_DBL +
+      ((buffer[offset + 5]! << 24) |
+        (buffer[offset + 6]! << 16) |
+        (buffer[offset + 7]! << 8) |
+        buffer[offset + 8]!)
     offset += 8
-  } else {
-    n = toNumber(fromBytesBE(buffer.subarray(offset + 1, offset + 9), false))
-
+  }
+  // Nine byte value
+  else {
+    value =
+      buffer[offset + 1]! * TWO_PWR_56_DBL +
+      ((buffer[offset + 2]! << 24) |
+        (buffer[offset + 3]! << 16) |
+        (buffer[offset + 4]! << 8) |
+        buffer[offset + 5]!) *
+        TWO_PWR_32_DBL +
+      ((buffer[offset + 6]! << 24) |
+        (buffer[offset + 7]! << 16) |
+        (buffer[offset + 8]! << 8) |
+        buffer[offset + 9]!)
     offset += 9
   }
-  return [n, offset - initialOffset] as const
+
+  return [value, offset - initialOffset] as const
 }
@@ -146,7 +185,2 @@
 }
-
-// this would be nice as a decorator, but i'm a little worried about babel
-// support for it going away or changing. memoizes a method in the stupidest
-// possible way, with no regard for the arguments.  actually, this only works
-// on methods that take no arguments
 export function tinyMemoize(_class: any, methodName: any) {
@@ -153,0 +187,0 @@ const method = _class.prototype[methodName]

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Dependency count

7

decreased by-12.5%

Number of lines in readme file

509

increased by0.39%

Worsened metrics

Total package byte prevSize

1285662

decreased by-1.87%

Number of package files

330

decreased by-0.3%

Lines of code

20554

decreased by-0.68%

Dependency changes

• - Removedlongfn@^1.3.1

• - Removedlongfn@1.3.1(transitive)