@jbrowse/plugin-bed 4.3.0 → 5.0.0-beta.10

package/esm/util.js

@@ -1,31 +1,29 @@
-import { generateBedMethylFeature, isBedMethylFeature, } from "./generateBedMethylFeature.js";
-import { generateRepeatMaskerFeature, isRepeatMaskerDescriptionField, } from "./generateRepeatMaskerFeature.js";
-import { generateUcscTranscript, isUcscTranscript, } from "./generateUcscTranscript.js";
-function defaultParser(fields, splitLine) {
-    const obj = {};
-    let hasBlockCount = false;
-    for (const [i, element] of splitLine.entries()) {
-        const field = fields[i];
-        if (field) {
-            obj[field] = element;
-            if (field === 'blockCount') {
-                hasBlockCount = true;
-            }
+import BED from '@gmod/bed';
+import { parseLineByLine } from '@jbrowse/core/util/parseLineByLine';
+import { copyExcept } from './copyExcept.js';
+import { generateBedMethylFeature, isBedMethylFeature, } from './generateBedMethylFeature.js';
+import { generateRastairMethylFeature, isRastairMethylFeature, } from './generateRastairMethylFeature.js';
+import { parseRepeatMaskerDescription } from './generateRepeatMaskerFeature.js';
+import { generateUcscTranscript, isUcscTranscript, } from './generateUcscTranscript.js';
+// Heuristically split features and header from a buffered BED file, bucketing
+// data lines by the first tab-separated column (refName).
+export function bucketBedLines(buffer, statusCallback) {
+    const features = {};
+    const headerLines = [];
+    parseLineByLine(buffer, line => {
+        if (line.startsWith('#')) {
+            headerLines.push(line);
         }
-    }
-    if (hasBlockCount) {
-        const { blockStarts, blockCount, chromStarts, thickEnd, thickStart, blockSizes, ...rest } = obj;
-        return {
-            ...rest,
-            blockStarts: arrayify(blockStarts),
-            chromStarts: arrayify(chromStarts),
-            blockSizes: arrayify(blockSizes),
-            thickStart: thickStart ? +thickStart : undefined,
-            thickEnd: thickEnd ? +thickEnd : undefined,
-            blockCount: blockCount ? +blockCount : undefined,
-        };
-    }
-    return obj;
+        else if (!line.startsWith('track') && !line.startsWith('browser')) {
+            // 'track'/'browser' are UCSC directive lines, not data (matches
+            // filterBedHeaderLines in spreadsheet-view)
+            const tab = line.indexOf('\t');
+            const refName = tab === -1 ? line : line.slice(0, tab);
+            (features[refName] ??= []).push(line);
+        }
+        return true;
+    }, statusCallback);
+    return { header: headerLines.join('\n'), features };
 }
 export function makeBlocks({ start, uniqueId, refName, chromStarts, blockCount, blockSizes, blockStarts, }) {
     const subfeatures = [];
@@ -45,24 +43,55 @@ export function makeBlocks({ start, uniqueId, refName, chromStarts, blockCount,
     }
     return subfeatures;
 }
-export function featureData({ line, colRef, colStart, colEnd, scoreColumn, parser, uniqueId, names, disableGeneHeuristic, }) {
-    const splitLine = line.split('\t');
-    const refName = splitLine[colRef];
-    const start = Number.parseInt(splitLine[colStart], 10);
-    const end = Number.parseInt(splitLine[colEnd], 10) + (colStart === colEnd ? 1 : 0);
-    return featureData2({
-        splitLine,
-        refName,
+export function parseNamesFromHeader(header) {
+    const defs = header.split(/\n|\r\n|\r/).filter(Boolean);
+    const defline = defs.at(-1);
+    // Strip the comment marker if there is one, rather than assuming it: a
+    // header kept for tabix with `-S N` instead of `#` is a bare row, and
+    // chopping its first character would rename its first column rather than
+    // fail to find one.
+    return defline?.includes('\t')
+        ? defline
+            .replace(/^#/, '')
+            .split('\t')
+            .map(f => f.trim())
+        : undefined;
+}
+// shared by BedAdapter/BedpeAdapter's getNames(): explicit columnNames config
+// wins, else fall back to parsing the file's own header line
+export async function resolveColumnNames(columnNames, getHeader) {
+    if (columnNames.length) {
+        return columnNames;
+    }
+    return parseNamesFromHeader(await getHeader());
+}
+// A header names the columns but says nothing about their types, so @gmod/bed
+// takes the types from the standard BED columns of the same name — that is what
+// splits `blockSizes` into numbers and leaves an unrecognized column a string.
+// Names win over a configured autoSql, as they did when this file parsed the
+// named case itself.
+export function makeParser({ autoSql, columnNames, }) {
+    return columnNames?.length ? new BED({ columnNames }) : new BED({ autoSql });
+}
+// Decode a BED line's locus (0-based half-open) from its column layout, so the
+// off-by-one rules live in one place instead of being re-derived at each call
+// site. `oneBased` shifts a 1-based-closed start back a base. `hasEndColumn`
+// false (tabix `-b` with no `-e`, e.g. point/GWAS data) means width-1 features.
+// A start/end column collision in a 0-based file is also a point feature (+1).
+export function bedFeatureLocus({ splitLine, colRef, colStart, colEnd, oneBased = false, hasEndColumn = true, }) {
+    const start = +splitLine[colStart] - (oneBased ? 1 : 0);
+    return {
+        refName: splitLine[colRef],
         start,
-        end,
-        parser,
-        uniqueId,
-        scoreColumn,
-        names,
-        disableGeneHeuristic,
-    });
+        end: hasEndColumn
+            ? +splitLine[colEnd] + (colStart === colEnd && !oneBased ? 1 : 0)
+            : start + 1,
+    };
 }
-function parseStrand(strand) {
+// Takes either form: `@gmod/bed`'s parseLine has already run its own
+// parseStrand and hands over ±1/0, while the methylation and bedpe paths read
+// the raw column. `.` and a missing column are unstranded, i.e. 0.
+export function parseStrand(strand) {
     if (strand === '-' || strand === -1) {
         return -1;
     }
@@ -71,24 +100,19 @@ function parseStrand(strand) {
     }
     return 0;
 }
-export function featureData2({ splitLine, refName, start, end, parser, uniqueId, scoreColumn, names, disableGeneHeuristic, }) {
-    const data = names
-        ? defaultParser(names, splitLine)
-        : parser.parseLine(splitLine, { uniqueId });
-    const { strand: strand2, score: score2, chrom: _1, chromStart: _2, chromEnd: _3, ...rest } = data;
-    const score = scoreColumn ? +data[scoreColumn] : score2 ? +score2 : undefined;
-    const strand = parseStrand(strand2);
-    const subfeatures = rest.blockCount
-        ? makeBlocks({
-            start,
-            uniqueId,
-            refName,
-            chromStarts: rest.chromStarts,
-            blockCount: rest.blockCount,
-            blockSizes: rest.blockSizes,
-            blockStarts: rest.blockStarts,
-        })
-        : undefined;
+const BED_OWN = new Set(['strand', 'score', 'chrom', 'chromStart', 'chromEnd']);
+const REPEAT_OWN = new Set([
+    ...BED_OWN,
+    'description',
+    'chromStarts',
+    'blockSizes',
+    'blockStarts',
+    'blockCount',
+    'thickStart',
+    'thickEnd',
+]);
+export function featureData({ splitLine, refName, start, end, parser, uniqueId, scoreColumn, names, disableGeneHeuristic, }) {
+    // bedMethyl detection runs on raw splitLine before parsing
     if (isBedMethylFeature({ splitLine, start, end })) {
         return generateBedMethylFeature({
             splitLine,
@@ -98,59 +122,76 @@ export function featureData2({ splitLine, refName, start, end, parser, uniqueId,
             end,
         });
     }
-    else if (isRepeatMaskerDescriptionField(rest.description)) {
-        const { chromStarts: _4, blockSizes: _5, blockStarts: _6, blockCount: _7, thickStart: _8, thickEnd: _9, description, ...rest2 } = rest;
-        return generateRepeatMaskerFeature({
-            ...rest2,
-            description,
+    // rastair methylation is identified by its header column names
+    if (names !== undefined && isRastairMethylFeature(names)) {
+        return generateRastairMethylFeature({
+            splitLine,
+            names,
             uniqueId,
-            score,
+            refName,
             start,
             end,
-            strand,
-            refName,
-            subfeatures,
         });
     }
-    else if (!disableGeneHeuristic &&
-        subfeatures &&
-        isUcscTranscript({
-            strand,
-            blockCount: rest.blockCount,
-            thickStart: rest.thickStart,
-        })) {
-        return generateUcscTranscript({
-            ...rest,
+    const data = parser.parseLine(splitLine, { uniqueId });
+    const strand = parseStrand(data.strand);
+    const rawScore = scoreColumn ? data[scoreColumn] : data.score;
+    const score = rawScore === undefined ? undefined : Number(rawScore);
+    const repeat = parseRepeatMaskerDescription(data.description);
+    if (repeat) {
+        return {
+            ...copyExcept(data, REPEAT_OWN),
+            ...repeat,
+            uniqueId,
             score,
             start,
             end,
             strand,
             refName,
-            uniqueId,
-            subfeatures,
-            thickStart: rest.thickStart,
-            thickEnd: rest.thickEnd,
-        });
+        };
     }
-    else {
-        return {
-            ...rest,
+    const rest = copyExcept(data, BED_OWN);
+    const subfeatures = data.blockCount
+        ? makeBlocks({
+            start,
             uniqueId,
+            refName,
+            chromStarts: data.chromStarts,
+            blockCount: data.blockCount,
+            blockSizes: data.blockSizes,
+            blockStarts: data.blockStarts,
+        })
+        : undefined;
+    const transcriptCheck = {
+        strand,
+        blockCount: data.blockCount,
+        thickStart: data.thickStart,
+        thickEnd: data.thickEnd,
+    };
+    if (!disableGeneHeuristic &&
+        subfeatures &&
+        isUcscTranscript(transcriptCheck)) {
+        return generateUcscTranscript({
+            ...rest,
             score,
             start,
             end,
-            strand,
+            strand: transcriptCheck.strand,
             refName,
+            uniqueId,
             subfeatures,
-        };
-    }
-}
-export function arrayify(f) {
-    if (f === undefined) {
-        return undefined;
-    }
-    if (typeof f === 'string') {
-        return f.split(',').map(s => +s);
+            thickStart: transcriptCheck.thickStart,
+            thickEnd: transcriptCheck.thickEnd,
+        });
     }
-    return f;
+    return {
+        ...rest,
+        uniqueId,
+        score,
+        start,
+        end,
+        strand,
+        refName,
+        subfeatures,
+    };
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jbrowse/plugin-bed",
-  "version": "4.3.0",
+  "version": "5.0.0-beta.10",
   "type": "module",
   "description": "JBrowse 2 bed adapters, tracks, etc.",
   "keywords": [
@@ -21,14 +21,14 @@
     "esm"
   ],
   "dependencies": {
-    "@flatten-js/interval-tree": "^2.0.3",
-    "@gmod/bbi": "^9.0.9",
-    "@gmod/bed": "^2.2.1",
-    "@gmod/tabix": "^3.3.3",
-    "@jbrowse/mobx-state-tree": "^5.6.0",
-    "mobx": "^6.15.3",
+    "@gmod/bbi": "^11.3.0",
+    "@gmod/bed": "^2.3.1",
+    "@gmod/tabix": "^3.8.3",
+    "@gmod/trix": "^3.0.22",
+    "@jbrowse/mobx-state-tree": "^6.6.1",
+    "mobx": "^7.0.5",
     "rxjs": "^7.8.2",
-    "@jbrowse/core": "^4.3.0"
+    "@jbrowse/core": "^5.0.0-beta.10"
   },
   "publishConfig": {
     "access": "public"
@@ -38,14 +38,11 @@
     "build": "pnpm run /^build:/",
     "test": "cd ../..; jest --passWithNoTests plugins/bed",
     "prebuild": "pnpm clean",
-    "build:esm": "tsc -p tsconfig.build.esm.json",
+    "build:esm": "node ../../node_modules/typescript7/bin/tsc --build tsconfig.build.esm.json",
     "clean": "rimraf esm *.tsbuildinfo"
   },
   "types": "esm/index.d.ts",
   "exports": {
-    ".": {
-      "types": "./esm/index.d.ts",
-      "import": "./esm/index.js"
-    }
+    ".": "./esm/index.js"
   }
 }

package/esm/GuessAdapter/index.d.ts

@@ -1,2 +0,0 @@
-import type PluginManager from '@jbrowse/core/PluginManager';
-export default function GuessAdapterF(pluginManager: PluginManager): void;

package/esm/GuessAdapter/index.js

@@ -1,71 +0,0 @@
-import { testAdapter } from '@jbrowse/core/util';
-import { getFileName, makeIndex, makeIndexType, } from '@jbrowse/core/util/tracks';
-export default function GuessAdapterF(pluginManager) {
-    pluginManager.addToExtensionPoint('Core-guessAdapterForLocation', (adapterGuesser) => {
-        return (file, index, adapterHint) => {
-            const fileName = getFileName(file);
-            const indexName = index && getFileName(index);
-            if (testAdapter(fileName, /\.bedpe(\.gz)?$/i, adapterHint, 'BedpeAdapter')) {
-                return {
-                    type: 'BedpeAdapter',
-                    bedpeLocation: file,
-                };
-            }
-            else if (testAdapter(fileName, /\.bb$/i, adapterHint, 'BigBedAdapter') ||
-                testAdapter(fileName, /\.bigbed$/i, adapterHint, 'BigBedAdapter')) {
-                return {
-                    type: 'BigBedAdapter',
-                    bigBedLocation: file,
-                };
-            }
-            else if (testAdapter(fileName, /\.bed$/i, adapterHint, 'BedAdapter')) {
-                return {
-                    type: 'BedAdapter',
-                    bedLocation: file,
-                };
-            }
-            else if (testAdapter(fileName, /\.bg$/i, adapterHint, 'BedGraphAdapter')) {
-                return {
-                    type: 'BedGraphAdapter',
-                    bedGraphLocation: file,
-                };
-            }
-            else if (testAdapter(fileName, /\.bg(\.gz)?$/i, adapterHint, 'BedGraphTabixAdapter')) {
-                return {
-                    type: 'BedGraphTabixAdapter',
-                    bedGraphGzLocation: file,
-                    index: {
-                        location: index || makeIndex(file, '.tbi'),
-                        indexType: makeIndexType(indexName, 'CSI', 'TBI'),
-                    },
-                };
-            }
-            else if (testAdapter(fileName, /\.bedmethyl\.gz$/i, adapterHint, 'BedTabixAdapter') ||
-                testAdapter(fileName, /\.bed\.gz$/i, adapterHint, 'BedTabixAdapter')) {
-                return {
-                    type: 'BedTabixAdapter',
-                    bedGzLocation: file,
-                    index: {
-                        location: index || makeIndex(file, '.tbi'),
-                        indexType: makeIndexType(indexName, 'CSI', 'TBI'),
-                    },
-                };
-            }
-            else {
-                return adapterGuesser(file, index, adapterHint);
-            }
-        };
-    });
-    pluginManager.addToExtensionPoint('Core-guessTrackTypeForLocation', (trackTypeGuesser) => (adapterName, file) => {
-        if (adapterName === 'BedTabixAdapter' &&
-            file &&
-            /\.bedmethyl\.gz$/i.test(getFileName(file))) {
-            return 'MultiQuantitativeTrack';
-        }
-        return ({
-            BedpeAdapter: 'VariantTrack',
-            BedGraphAdapter: 'QuantitativeTrack',
-            BedGraphTabixAdapter: 'QuantitativeTrack',
-        }[adapterName] || trackTypeGuesser(adapterName, file));
-    });
-}