@gmod/cram 3.0.7 → 4.0.0

package/src/cramFile/sectionParsers.ts

@@ -1,23 +1,22 @@
 import { TupleOf } from '../typescript'
-import { parseItf8, parseLtf8 } from './util'
 import { DataSeriesEncodingMap } from './codecs/dataSeriesTypes'
 import { CramEncoding } from './encoding'
+import { parseItf8, parseLtf8 } from './util'
 
 export function cramFileDefinition() {
   return {
-    parser: (buffer: Buffer, _startOffset = 0) => {
-      const b = buffer
+    parser: (b: Uint8Array, _startOffset = 0) => {
       const dataView = new DataView(b.buffer, b.byteOffset, b.length)
+      const decoder = new TextDecoder('utf8')
       let offset = 0
-      const magic = buffer.subarray(offset, offset + 4).toString()
+      const magic = decoder.decode(b.subarray(offset, offset + 4))
       offset += 4
       const majorVersion = dataView.getUint8(offset)
       offset += 1
       const minorVersion = dataView.getUint8(offset)
       offset += 1
-      const fileId = b
-        .subarray(offset, offset + 20)
-        .toString()
+      const fileId = decoder
+        .decode(b.subarray(offset, offset + 20))
         .replaceAll('\0', '')
       offset += 20
       return {
@@ -34,7 +33,7 @@ export function cramFileDefinition() {
   }
 }
 export function cramBlockHeader() {
-  const parser = (buffer: Buffer, _startOffset = 0) => {
+  const parser = (buffer: Uint8Array, _startOffset = 0) => {
     const b = buffer
     const dataView = new DataView(b.buffer, b.byteOffset, b.length)
     let offset = 0
@@ -97,7 +96,7 @@ export function cramBlockHeader() {
 
 export function cramBlockCrc32() {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const b = buffer
       const dataView = new DataView(b.buffer, b.byteOffset, b.length)
       const crc32 = dataView.getUint32(offset, true)
@@ -115,8 +114,13 @@ export function cramBlockCrc32() {
 
 export type CramTagDictionary = string[][]
 
-function makeTagSet(buffer: Buffer, stringStart: number, stringEnd: number) {
-  const str = buffer.toString('utf8', stringStart, stringEnd)
+function makeTagSet(
+  buffer: Uint8Array,
+  stringStart: number,
+  stringEnd: number,
+) {
+  const decoder = new TextDecoder('utf8')
+  const str = decoder.decode(buffer.subarray(stringStart, stringEnd))
   const tags = []
   for (let i = 0; i < str.length; i += 3) {
     tags.push(str.slice(i, i + 3))
@@ -126,7 +130,7 @@ function makeTagSet(buffer: Buffer, stringStart: number, stringEnd: number) {
 
 export function cramTagDictionary() {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const [size, newOffset1] = parseItf8(buffer, offset)
       offset += newOffset1
       const subbuf = buffer.subarray(offset, offset + size)
@@ -169,7 +173,7 @@ export interface CramPreservationMap {
 
 export function cramPreservationMap() {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const b = buffer
       const dataView = new DataView(b.buffer, b.byteOffset, b.length)
       const [mapSize, newOffset1] = parseItf8(buffer, offset)
@@ -284,7 +288,7 @@ function cramUnmappedSliceHeader(majorVersion: number) {
   maxLength += 5 * 2
   maxLength += 16
 
-  const parser = (buffer: Buffer, offset: number) => {
+  const parser = (buffer: Uint8Array, offset: number) => {
     const [numRecords, newOffset1] = parseItf8(buffer, offset)
     offset += newOffset1
     let recordCounter = 0
@@ -348,7 +352,7 @@ function cramMappedSliceHeader(majorVersion: number) {
   maxLength += 16 // MD5
 
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       // L0
       const [refSeqId, newOffset1] = parseItf8(buffer, offset)
       offset += newOffset1
@@ -419,12 +423,13 @@ function cramMappedSliceHeader(majorVersion: number) {
 
 function cramEncoding() {
   return {
-    parser: (buffer: Buffer, offset: number) => cramEncodingSub(buffer, offset),
+    parser: (buffer: Uint8Array, offset: number) =>
+      cramEncodingSub(buffer, offset),
   }
 }
 
 function cramEncodingSub(
-  buffer: Buffer,
+  buffer: Uint8Array,
   offset: number,
 ): { value: Value; offset: number } {
   const b = buffer
@@ -542,7 +547,7 @@ function cramEncodingSub(
 
 function cramDataSeriesEncodingMap() {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const [mapSize, newOffset1] = parseItf8(buffer, offset)
       offset += newOffset1
       const [mapCount, newOffset2] = parseItf8(buffer, offset)
@@ -572,7 +577,7 @@ function cramDataSeriesEncodingMap() {
 
 function cramTagEncodingMap() {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const [mapSize, newOffset1] = parseItf8(buffer, offset)
       offset += newOffset1
       const [mapCount, newOffset2] = parseItf8(buffer, offset)
@@ -604,7 +609,7 @@ function cramTagEncodingMap() {
 
 function cramCompressionHeader() {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       // TODO: if we want to support CRAM v1, we will need to refactor
       // compression header into 2 parts to parse the landmarks, like the
       // container header
@@ -644,7 +649,7 @@ function cramContainerHeader1(majorVersion: number) {
   maxLength += 5 + 5
   return {
     maxLength,
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const b = buffer
       const dataView = new DataView(b.buffer, b.byteOffset, b.length)
       // byte size of the container data (blocks)
@@ -704,7 +709,7 @@ function cramContainerHeader1(majorVersion: number) {
 
 function cramContainerHeader2(majorVersion: number) {
   return {
-    parser: (buffer: Buffer, offset: number) => {
+    parser: (buffer: Uint8Array, offset: number) => {
       const b = buffer
       const dataView = new DataView(b.buffer, b.byteOffset, b.length)
       const [numLandmarks, newOffset1] = parseItf8(buffer, offset)

package/src/cramFile/slice/decodeRecord.ts

@@ -1,4 +1,9 @@
 import { CramMalformedError } from '../../errors'
+import { Cursors, DataTypeMapping } from '../codecs/_base'
+import { DataSeriesEncodingKey } from '../codecs/dataSeriesTypes'
+import CramContainerCompressionScheme, {
+  DataSeriesTypes,
+} from '../container/compressionScheme'
 import {
   BamFlagsDecoder,
   CramFlagsDecoder,
@@ -6,13 +11,8 @@ import {
   ReadFeature,
 } from '../record'
 import CramSlice, { SliceHeader } from './index'
-import { isMappedSliceHeader } from '../sectionParsers'
-import CramContainerCompressionScheme, {
-  DataSeriesTypes,
-} from '../container/compressionScheme'
 import { CramFileBlock } from '../file'
-import { Cursors, DataTypeMapping } from '../codecs/_base'
-import { DataSeriesEncodingKey } from '../codecs/dataSeriesTypes'
+import { isMappedSliceHeader } from '../sectionParsers'
 
 /**
  * given a Buffer, read a string up to the first null character
@@ -42,37 +42,37 @@ function parseTagValueArray(buffer: Uint8Array) {
 
   if (arrayType === 'c') {
     const arr = new Int8Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else if (arrayType === 'C') {
     const arr = new Uint8Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else if (arrayType === 's') {
     const arr = new Int16Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else if (arrayType === 'S') {
     const arr = new Uint16Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else if (arrayType === 'i') {
     const arr = new Int32Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else if (arrayType === 'I') {
     const arr = new Uint32Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else if (arrayType === 'f') {
     const arr = new Float32Array(buffer.buffer)
-    for (let i = 0; i < length; i += 1) {
+    for (let i = 0; i < length; i++) {
       array[i] = arr[i]!
     }
   } else {
@@ -158,7 +158,7 @@ function decodeReadFeatures(
     return data
   }
 
-  for (let i = 0; i < readFeatureCount; i += 1) {
+  for (let i = 0; i < readFeatureCount; i++) {
     const code = String.fromCharCode(decodeDataSeries('FC'))
 
     const readPosDelta = decodeDataSeries('FP')
@@ -315,7 +315,7 @@ export default function decodeRecord(
   // TN = tag names
   const TN = compressionScheme.getTagNames(TLindex)!
   const ntags = TN.length
-  for (let i = 0; i < ntags; i += 1) {
+  for (let i = 0; i < ntags; i++) {
     const tagId = TN[i]!
     const tagName = tagId.slice(0, 2)
     const tagType = tagId.slice(2, 3)
@@ -381,14 +381,14 @@ export default function decodeRecord(
     qualityScores = null
   } else {
     const bases = new Array(readLength) as number[]
-    for (let i = 0; i < bases.length; i += 1) {
+    for (let i = 0; i < bases.length; i++) {
       bases[i] = decodeDataSeries('BA')
     }
     readBases = String.fromCharCode(...bases)
 
     if (CramFlagsDecoder.isPreservingQualityScores(cramFlags)) {
       qualityScores = new Array(readLength)
-      for (let i = 0; i < bases.length; i += 1) {
+      for (let i = 0; i < bases.length; i++) {
         qualityScores[i] = decodeDataSeries('QS')
       }
     }

package/src/cramFile/slice/index.ts

@@ -1,21 +1,20 @@
 import { CramArgumentError, CramMalformedError } from '../../errors'
-import { parseItem, sequenceMD5, tinyMemoize } from '../util'
-
+import { Cursors, DataTypeMapping } from '../codecs/_base'
+import { CramBufferOverrunError } from '../codecs/getBits'
 import Constants from '../constants'
 import decodeRecord, { DataSeriesDecoder } from './decodeRecord'
-import CramRecord from '../record'
+import { DataSeriesEncodingKey } from '../codecs/dataSeriesTypes'
 import CramContainer from '../container'
+import { DataSeriesTypes } from '../container/compressionScheme'
 import CramFile, { CramFileBlock } from '../file'
+import CramRecord from '../record'
 import {
-  getSectionParsers,
-  isMappedSliceHeader,
   MappedSliceHeader,
   UnmappedSliceHeader,
+  getSectionParsers,
+  isMappedSliceHeader,
 } from '../sectionParsers'
-import { CramBufferOverrunError } from '../codecs/getBits'
-import { Cursors, DataTypeMapping } from '../codecs/_base'
-import { DataSeriesEncodingKey } from '../codecs/dataSeriesTypes'
-import { DataSeriesTypes } from '../container/compressionScheme'
+import { parseItem, sequenceMD5, tinyMemoize } from '../util'
 
 export type SliceHeader = CramFileBlock & {
   parsedContent: MappedSliceHeader | UnmappedSliceHeader

package/src/cramFile/util.ts

@@ -1,5 +1,6 @@
-import md5 from 'md5'
 import Long from 'long'
+import md5 from 'md5'
+
 import { CramBufferOverrunError } from './codecs/getBits'
 
 export function itf8Size(v: number) {
@@ -60,7 +61,8 @@ export function parseItf8(buffer: Uint8Array, initialOffset: number) {
   return [result, offset - initialOffset] as const
 }
 
-export function parseLtf8(buffer: Buffer, initialOffset: number) {
+export function parseLtf8(buffer: Uint8Array, initialOffset: number) {
+  const dataView = new DataView(buffer.buffer)
   let offset = initialOffset
   const countFlags = buffer[offset]!
   let n: number | Long
@@ -76,7 +78,7 @@ export function parseLtf8(buffer: Buffer, initialOffset: number) {
         (buffer[offset + 1]! << 8) |
         buffer[offset + 2]!) &
       0x1fffff
-    n = ((countFlags & 63) << 16) | buffer.readUInt16LE(offset + 1)
+    n = ((countFlags & 63) << 16) | dataView.getUint16(offset + 1, true)
     offset += 3
   } else if (countFlags < 0xf0) {
     n =
@@ -142,8 +144,8 @@ export function parseLtf8(buffer: Buffer, initialOffset: number) {
 }
 
 export function parseItem<T>(
-  buffer: Buffer,
-  parser: (buffer: Buffer, offset: number) => { offset: number; value: T },
+  buffer: Uint8Array,
+  parser: (buffer: Uint8Array, offset: number) => { offset: number; value: T },
   startBufferPosition = 0,
   startFilePosition = 0,
 ) {
@@ -177,3 +179,20 @@ export function tinyMemoize(_class: any, methodName: any) {
 export function sequenceMD5(seq: string) {
   return md5(seq.toUpperCase().replaceAll(/[^\u0021-\u007e]/g, ''))
 }
+
+export function sum(array: Uint8Array[]) {
+  let sum = 0
+  for (const entry of array) {
+    sum += entry.length
+  }
+  return sum
+}
+export function concatUint8Array(args: Uint8Array[]) {
+  const mergedArray = new Uint8Array(sum(args))
+  let offset = 0
+  for (const entry of args) {
+    mergedArray.set(entry, offset)
+    offset += entry.length
+  }
+  return mergedArray
+}