@genome-spy/core 0.65.0 → 0.66.1

package/dist/bundle/index-DbJ0oeYM.js

@@ -0,0 +1,631 @@
+import { A as J } from "./AbortablePromiseCache-CcuMrnn7.js";
+import { u as E, Q as U, a as K } from "./unzip-NywezaRR.js";
+import { LocalFile as F, RemoteFile as M } from "./browser-BRemItdO.js";
+class L {
+  constructor(e, t, n, i) {
+    this.minv = e, this.maxv = t, this.bin = n, this._fetchedSize = i;
+  }
+  toUniqueString() {
+    return `${this.minv}..${this.maxv} (bin ${this.bin}, fetchedSize ${this.fetchedSize()})`;
+  }
+  toString() {
+    return this.toUniqueString();
+  }
+  compareTo(e) {
+    return this.minv.compareTo(e.minv) || this.maxv.compareTo(e.maxv) || this.bin - e.bin;
+  }
+  fetchedSize() {
+    return this._fetchedSize !== void 0 ? this._fetchedSize : this.maxv.blockPosition + 65536 - this.minv.blockPosition;
+  }
+}
+class j {
+  constructor({ filehandle: e, renameRefSeqs: t = (n) => n }) {
+    this.filehandle = e, this.renameRefSeq = t;
+  }
+  async getMetadata(e = {}) {
+    const { indices: t, ...n } = await this.parse(e);
+    return n;
+  }
+  _findFirstData(e, t) {
+    return e ? e.compareTo(t) > 0 ? t : e : t;
+  }
+  async parse(e = {}) {
+    return this.parseP || (this.parseP = this._parse(e).catch((t) => {
+      throw this.parseP = void 0, t;
+    })), this.parseP;
+  }
+  async hasRefSeq(e, t = {}) {
+    var i;
+    return !!((i = (await this.parse(t)).indices[e]) != null && i.binIndex);
+  }
+  _parseNameBytes(e) {
+    let t = 0, n = 0;
+    const i = [], o = {}, r = new TextDecoder("utf8");
+    for (let s = 0; s < e.length; s += 1)
+      if (!e[s]) {
+        if (n < s) {
+          const c = this.renameRefSeq(r.decode(e.subarray(n, s)));
+          i[t] = c, o[c] = t;
+        }
+        n = s + 1, t += 1;
+      }
+    return {
+      refNameToId: o,
+      refIdToName: i
+    };
+  }
+}
+const H = 65536, Z = H * H;
+function Q(a, e = 0) {
+  const t = a[e] | a[e + 1] << 8 | a[e + 2] << 16 | a[e + 3] << 24;
+  return ((a[e + 4] | a[e + 5] << 8 | a[e + 6] << 16 | a[e + 7] << 24) >>> 0) * Z + (t >>> 0);
+}
+function ee(a, e) {
+  return e.minv.blockPosition - a.maxv.blockPosition < 65e3 && e.maxv.blockPosition - a.minv.blockPosition < 5e6;
+}
+function X(a, e) {
+  const t = [];
+  let n;
+  if (a.length === 0)
+    return a;
+  a.sort(function(i, o) {
+    const r = i.minv.blockPosition - o.minv.blockPosition;
+    return r === 0 ? i.minv.dataPosition - o.minv.dataPosition : r;
+  });
+  for (const i of a)
+    (!e || i.maxv.compareTo(e) > 0) && (n === void 0 ? (t.push(i), n = i) : ee(n, i) ? i.maxv.compareTo(n.maxv) > 0 && (n.maxv = i.maxv) : (t.push(i), n = i));
+  return t;
+}
+class O {
+  constructor(e, t) {
+    this.blockPosition = e, this.dataPosition = t;
+  }
+  toString() {
+    return `${this.blockPosition}:${this.dataPosition}`;
+  }
+  compareTo(e) {
+    return this.blockPosition - e.blockPosition || this.dataPosition - e.dataPosition;
+  }
+}
+function A(a, e = 0) {
+  return new O(a[e + 7] * 1099511627776 + a[e + 6] * 4294967296 + a[e + 5] * 16777216 + a[e + 4] * 65536 + a[e + 3] * 256 + a[e + 2], a[e + 1] << 8 | a[e]);
+}
+const te = 21582659, ne = 38359875, ie = {
+  0: "generic",
+  1: "SAM",
+  2: "VCF"
+};
+function se(a, e) {
+  return a * 2 ** e;
+}
+function W(a, e) {
+  return Math.floor(a / 2 ** e);
+}
+class q extends j {
+  constructor(e) {
+    super(e), this.maxBinNumber = 0, this.depth = 0, this.minShift = 0;
+  }
+  async lineCount(e, t = {}) {
+    const n = await this.parse(t), i = n.refNameToId[e];
+    if (i === void 0 || !n.indices[i])
+      return -1;
+    const { stats: r } = n.indices[i];
+    return r ? r.lineCount : -1;
+  }
+  indexCov() {
+    throw new Error("CSI indexes do not support indexcov");
+  }
+  parseAuxData(e, t) {
+    const n = new DataView(e.buffer), i = n.getInt32(t, !0), o = i & 65536 ? "zero-based-half-open" : "1-based-closed", r = ie[i & 15];
+    if (!r)
+      throw new Error(`invalid Tabix preset format flags ${i}`);
+    const s = {
+      ref: n.getInt32(t + 4, !0),
+      start: n.getInt32(t + 8, !0),
+      end: n.getInt32(t + 12, !0)
+    }, c = n.getInt32(t + 16, !0), m = c ? String.fromCharCode(c) : void 0, f = n.getInt32(t + 20, !0), x = n.getInt32(t + 24, !0), { refIdToName: d, refNameToId: l } = this._parseNameBytes(e.subarray(t + 28, t + 28 + x));
+    return {
+      refIdToName: d,
+      refNameToId: l,
+      skipLines: f,
+      metaChar: m,
+      columnNumbers: s,
+      format: r,
+      coordinateType: o
+    };
+  }
+  async _parse(e = {}) {
+    const t = await E(await this.filehandle.readFile(e)), n = new DataView(t.buffer);
+    let i;
+    if (n.getUint32(0, !0) === te)
+      i = 1;
+    else if (n.getUint32(0, !0) === ne)
+      i = 2;
+    else
+      throw new Error("Not a CSI file");
+    this.minShift = n.getInt32(4, !0), this.depth = n.getInt32(8, !0), this.maxBinNumber = ((1 << (this.depth + 1) * 3) - 1) / 7;
+    const o = 2 ** (this.minShift + this.depth * 3), r = n.getInt32(12, !0), s = r && r >= 30 ? this.parseAuxData(t, 16) : {
+      refIdToName: [],
+      refNameToId: {},
+      metaChar: void 0,
+      columnNumbers: { ref: 0, start: 1, end: 2 },
+      coordinateType: "zero-based-half-open",
+      format: "generic"
+    }, c = n.getInt32(16 + r, !0);
+    let m, f = 16 + r + 4;
+    const x = new Array(c).fill(0).map(() => {
+      const d = n.getInt32(f, !0);
+      f += 4;
+      const l = {};
+      let h;
+      for (let b = 0; b < d; b += 1) {
+        const k = n.getUint32(f, !0);
+        if (k > this.maxBinNumber)
+          h = this.parsePseudoBin(t, f + 4), f += 48;
+        else {
+          const w = A(t, f + 4);
+          m = this._findFirstData(m, w);
+          const p = n.getInt32(f + 12, !0);
+          f += 16;
+          const I = new Array(p);
+          for (let g = 0; g < p; g += 1) {
+            const u = A(t, f), S = A(t, f + 8);
+            f += 16, I[g] = new L(u, S, k);
+          }
+          l[k] = I;
+        }
+      }
+      return { binIndex: l, stats: h };
+    });
+    return {
+      ...s,
+      csi: !0,
+      refCount: c,
+      maxBlockSize: 65536,
+      firstDataLine: m,
+      csiVersion: i,
+      indices: x,
+      depth: this.depth,
+      maxBinNumber: this.maxBinNumber,
+      maxRefLength: o
+    };
+  }
+  parsePseudoBin(e, t) {
+    return {
+      lineCount: Q(e, t + 28)
+    };
+  }
+  async blocksForRange(e, t, n, i = {}) {
+    t < 0 && (t = 0);
+    const o = await this.parse(i), r = o.refNameToId[e];
+    if (r === void 0)
+      return [];
+    const s = o.indices[r];
+    if (!s)
+      return [];
+    const c = this.reg2bins(t, n), m = [];
+    for (const [f, x] of c)
+      for (let d = f; d <= x; d++)
+        if (s.binIndex[d])
+          for (const l of s.binIndex[d])
+            m.push(new L(l.minv, l.maxv, d));
+    return X(m, new O(0, 0));
+  }
+  /**
+   * calculate the list of bins that may overlap with region [beg,end) (zero-based half-open)
+   */
+  reg2bins(e, t) {
+    e -= 1, e < 1 && (e = 1), t > 2 ** 50 && (t = 2 ** 34), t -= 1;
+    let n = 0, i = 0, o = this.minShift + this.depth * 3;
+    const r = [];
+    for (; n <= this.depth; o -= 3, i += se(1, n * 3), n += 1) {
+      const s = i + W(e, o), c = i + W(t, o);
+      if (c - s + r.length > this.maxBinNumber)
+        throw new Error(`query ${e}-${t} is too large for current binning scheme (shift ${this.minShift}, depth ${this.depth}), try a smaller query or a coarser index binning scheme`);
+      r.push([s, c]);
+    }
+    return r;
+  }
+}
+const re = 21578324, G = 14;
+function ae(a, e) {
+  return a += 1, e -= 1, [
+    [0, 0],
+    [1 + (a >> 26), 1 + (e >> 26)],
+    [9 + (a >> 23), 9 + (e >> 23)],
+    [73 + (a >> 20), 73 + (e >> 20)],
+    [585 + (a >> 17), 585 + (e >> 17)],
+    [4681 + (a >> 14), 4681 + (e >> 14)]
+  ];
+}
+class R extends j {
+  async lineCount(e, t = {}) {
+    var r;
+    const n = await this.parse(t), i = n.refNameToId[e];
+    return i === void 0 || !n.indices[i] ? -1 : ((r = n.indices[i].stats) == null ? void 0 : r.lineCount) ?? -1;
+  }
+  // fetch and parse the index
+  async _parse(e = {}) {
+    const t = await this.filehandle.readFile(e), n = await E(t), i = new DataView(n.buffer);
+    if (i.getUint32(0, !0) !== re)
+      throw new Error("Not a TBI file");
+    const r = i.getUint32(4, !0), s = i.getUint32(8, !0), c = s & 65536 ? "zero-based-half-open" : "1-based-closed", f = {
+      0: "generic",
+      1: "SAM",
+      2: "VCF"
+    }[s & 15];
+    if (!f)
+      throw new Error(`invalid Tabix preset format flags ${s}`);
+    const x = {
+      ref: i.getInt32(12, !0),
+      start: i.getInt32(16, !0),
+      end: i.getInt32(20, !0)
+    }, d = i.getInt32(24, !0), l = 5, h = ((1 << (l + 1) * 3) - 1) / 7, b = 2 ** (14 + l * 3), k = d ? String.fromCharCode(d) : void 0, w = i.getInt32(28, !0), p = i.getInt32(32, !0), { refNameToId: I, refIdToName: g } = this._parseNameBytes(n.slice(36, 36 + p));
+    let u = 36 + p, S;
+    return {
+      indices: new Array(r).fill(0).map(() => {
+        const T = i.getInt32(u, !0);
+        u += 4;
+        const N = {};
+        let _;
+        for (let v = 0; v < T; v += 1) {
+          const y = i.getUint32(u, !0);
+          if (u += 4, y > h + 1)
+            throw new Error("tabix index contains too many bins, please use a CSI index");
+          if (y === h + 1) {
+            const P = i.getInt32(u, !0);
+            u += 4, P === 2 && (_ = this.parsePseudoBin(n, u)), u += 16 * P;
+          } else {
+            const P = i.getInt32(u, !0);
+            u += 4;
+            const D = new Array(P);
+            for (let V = 0; V < P; V += 1) {
+              const $ = A(n, u), Y = A(n, u + 8);
+              u += 16, S = this._findFirstData(S, $), D[V] = new L($, Y, y);
+            }
+            N[y] = D;
+          }
+        }
+        const z = i.getInt32(u, !0);
+        u += 4;
+        const B = new Array(z);
+        for (let v = 0; v < z; v += 1)
+          B[v] = A(n, u), u += 8, S = this._findFirstData(S, B[v]);
+        return {
+          binIndex: N,
+          linearIndex: B,
+          stats: _
+        };
+      }),
+      metaChar: k,
+      maxBinNumber: h,
+      maxRefLength: b,
+      skipLines: w,
+      firstDataLine: S,
+      columnNumbers: x,
+      coordinateType: c,
+      format: f,
+      refIdToName: g,
+      refNameToId: I,
+      maxBlockSize: 65536
+    };
+  }
+  parsePseudoBin(e, t) {
+    return {
+      lineCount: Q(e, t + 16)
+    };
+  }
+  async blocksForRange(e, t, n, i = {}) {
+    t < 0 && (t = 0);
+    const o = await this.parse(i), r = o.refNameToId[e];
+    if (r === void 0)
+      return [];
+    const s = o.indices[r];
+    if (!s)
+      return [];
+    (s.linearIndex.length > 0 ? s.linearIndex[t >> G >= s.linearIndex.length ? s.linearIndex.length - 1 : t >> G] : new O(0, 0)) || console.warn("querying outside of possible tabix range");
+    const m = ae(t, n), f = [];
+    for (const [b, k] of m)
+      for (let w = b; w <= k; w++)
+        if (s.binIndex[w])
+          for (const p of s.binIndex[w])
+            f.push(new L(p.minv, p.maxv, w));
+    const x = s.linearIndex.length;
+    let d;
+    const l = Math.min(t >> 14, x - 1), h = Math.min(n >> 14, x - 1);
+    for (let b = l; b <= h; ++b) {
+      const k = s.linearIndex[b];
+      k && (!d || k.compareTo(d) < 0) && (d = k);
+    }
+    return X(f, d);
+  }
+}
+class he {
+  /**
+   * @param {object} args
+   *
+   * @param {string} [args.path]
+   *
+   * @param {filehandle} [args.filehandle]
+   *
+   * @param {string} [args.tbiPath]
+   *
+   * @param {filehandle} [args.tbiFilehandle]
+   *
+   * @param {string} [args.csiPath]
+   *
+   * @param {filehandle} [args.csiFilehandle]
+   *
+   * @param {url} [args.url]
+   *
+   * @param {csiUrl} [args.csiUrl]
+   *
+   * @param {tbiUrl} [args.tbiUrl]
+   *
+   * @param {function} [args.renameRefSeqs] optional function with sig `string
+   * => string` to transform reference sequence names for the purpose of
+   * indexing and querying. note that the data that is returned is not altered,
+   * just the names of the reference sequences that are used for querying.
+   */
+  constructor({ path: e, filehandle: t, url: n, tbiPath: i, tbiUrl: o, tbiFilehandle: r, csiPath: s, csiUrl: c, csiFilehandle: m, renameRefSeqs: f, chunkCacheSize: x = 5 * 2 ** 20 }) {
+    this.cache = new U({
+      maxSize: 1e3
+    });
+    const d = f ?? ((l) => l);
+    if (t)
+      this.filehandle = t;
+    else if (e)
+      this.filehandle = new F(e);
+    else if (n)
+      this.filehandle = new M(n);
+    else
+      throw new TypeError("must provide either filehandle or path");
+    if (r)
+      this.index = new R({
+        filehandle: r,
+        renameRefSeqs: d
+      });
+    else if (m)
+      this.index = new q({
+        filehandle: m,
+        renameRefSeqs: d
+      });
+    else if (i)
+      this.index = new R({
+        filehandle: new F(i),
+        renameRefSeqs: d
+      });
+    else if (s)
+      this.index = new q({
+        filehandle: new F(s),
+        renameRefSeqs: d
+      });
+    else if (e)
+      this.index = new R({
+        filehandle: new F(`${e}.tbi`),
+        renameRefSeqs: d
+      });
+    else if (c)
+      this.index = new q({
+        filehandle: new M(c)
+      });
+    else if (o)
+      this.index = new R({
+        filehandle: new M(o)
+      });
+    else if (n)
+      this.index = new R({
+        filehandle: new M(`${n}.tbi`)
+      });
+    else
+      throw new TypeError("must provide one of tbiFilehandle, tbiPath, csiFilehandle, csiPath, tbiUrl, csiUrl");
+    this.renameRefSeq = d, this.hasCustomRenameRefSeq = f !== void 0, this.chunkCache = new J({
+      cache: new U({ maxSize: Math.floor(x / 65536) }),
+      fill: (l, h) => this.readChunk(l, { signal: h })
+    });
+  }
+  /**
+   * @param refName name of the reference sequence
+   *
+   * @param start start of the region (in 0-based half-open coordinates)
+   *
+   * @param end end of the region (in 0-based half-open coordinates)
+   *
+   * @param opts callback called for each line in the region. can also pass a
+   * object param containing obj.lineCallback, obj.signal, etc
+   *
+   * @returns promise that is resolved when the whole read is finished,
+   * rejected on error
+   */
+  calculateFileOffset(e, t, n, i, o) {
+    return e[n] * 256 + (i - t[n]) + o + 1;
+  }
+  async getLines(e, t, n, i) {
+    var I;
+    let o, r = {}, s;
+    typeof i == "function" ? s = i : (r = i, s = i.lineCallback, o = i.signal);
+    const c = await this.index.getMetadata(r), m = t ?? 0, f = n ?? c.maxRefLength;
+    if (!(m <= f))
+      throw new TypeError("invalid start and end coordinates. start must be less than or equal to end");
+    if (m === f)
+      return;
+    const x = await this.index.blocksForRange(e, m, f, r), d = new TextDecoder("utf8"), l = c.format === "VCF", h = {
+      ref: c.columnNumbers.ref || 0,
+      start: c.columnNumbers.start || 0,
+      end: l ? 8 : c.columnNumbers.end || 0
+    }, b = Math.max(h.ref, h.start, h.end), k = (I = c.metaChar) == null ? void 0 : I.charCodeAt(0), w = c.coordinateType === "1-based-closed" ? -1 : 0, p = !this.hasCustomRenameRefSeq;
+    for (const g of x) {
+      const { buffer: u, cpositions: S, dpositions: C } = await this.chunkCache.get(g.toString(), g, o);
+      let T = 0, N = 0;
+      const _ = d.decode(u);
+      if (u.length == _.length)
+        for (; T < _.length; ) {
+          const B = _.indexOf(`
+`, T);
+          if (B === -1)
+            break;
+          const v = _.slice(T, B);
+          if (C) {
+            const P = T + g.minv.dataPosition;
+            for (; N < C.length && P >= C[N]; )
+              N++;
+          }
+          const y = this.checkLine(e, m, f, v, h.ref, h.start, h.end, b, k, w, l, p);
+          if (y === null)
+            return;
+          y !== void 0 && s(v, this.calculateFileOffset(S, C, N, T, g.minv.dataPosition), y.start, y.end), T = B + 1;
+        }
+      else
+        for (; T < u.length; ) {
+          const B = u.indexOf(10, T);
+          if (B === -1)
+            break;
+          const v = u.slice(T, B), y = d.decode(v);
+          if (C) {
+            const D = T + g.minv.dataPosition;
+            for (; N < C.length && D >= C[N]; )
+              N++;
+          }
+          const P = this.checkLine(e, m, f, y, h.ref, h.start, h.end, b, k, w, l, p);
+          if (P === null)
+            return;
+          P !== void 0 && s(y, this.calculateFileOffset(S, C, N, T, g.minv.dataPosition), P.start, P.end), T = B + 1;
+        }
+    }
+  }
+  async getMetadata(e = {}) {
+    return this.index.getMetadata(e);
+  }
+  /**
+   * get a buffer containing the "header" region of the file, which are the
+   * bytes up to the first non-meta line
+   */
+  async getHeaderBuffer(e = {}) {
+    const { firstDataLine: t, metaChar: n, maxBlockSize: i } = await this.getMetadata(e), o = ((t == null ? void 0 : t.blockPosition) || 0) + i, r = await this.filehandle.read(o, 0, e), s = await E(r);
+    if (n) {
+      let c = -1;
+      const m = 10, f = n.charCodeAt(0);
+      for (let x = 0, d = s.length; x < d; x++) {
+        const l = s[x];
+        if (x === c + 1 && l !== f)
+          break;
+        l === m && (c = x);
+      }
+      return s.subarray(0, c + 1);
+    }
+    return s;
+  }
+  /**
+   * get a string containing the "header" region of the file, is the portion up
+   * to the first non-meta line
+   */
+  async getHeader(e = {}) {
+    const t = new TextDecoder("utf8"), n = await this.getHeaderBuffer(e);
+    return t.decode(n);
+  }
+  /**
+   * get an array of reference sequence names, in the order in which they occur
+   * in the file. reference sequence renaming is not applied to these names.
+   */
+  async getReferenceSequenceNames(e = {}) {
+    return (await this.getMetadata(e)).refIdToName;
+  }
+  /**
+   * @param {string} regionRefName
+   *
+   * @param {number} regionStart region start coordinate (0-based-half-open)
+   *
+   * @param {number} regionEnd region end coordinate (0-based-half-open)
+   *
+   * @param {string} line
+   *
+   * @param {number} refColumn column number for ref
+   *
+   * @param {number} startColumn column number for start
+   *
+   * @param {number} endColumn column number for end
+   *
+   * @param {number} maxColumn pre-calculated max column
+   *
+   * @param {number} metaCharCode pre-calculated metaChar code
+   *
+   * @param {number} coordinateOffset 0 or -1 for coordinate adjustment
+   *
+   * @param {boolean} isVCF whether this is VCF format
+   *
+   * @param {boolean} isIdentityRename whether renameRefSeq is the identity function
+   *
+   * @returns {{ start: number, end: number } | null | undefined} coordinates if overlapping, null if should stop processing, undefined otherwise
+   */
+  checkLine(e, t, n, i, o, r, s, c, m, f, x, d) {
+    if (m !== void 0 && i.charCodeAt(0) === m)
+      return;
+    if (i.length < 500) {
+      const I = i.split("	"), g = I[o - 1];
+      if (!(d ? g === e : this.renameRefSeq(g) === e))
+        return;
+      const S = +I[r - 1] + f;
+      if (S >= n)
+        return null;
+      let C;
+      return s === 0 || s === r ? C = S + 1 : x ? C = this._getVcfEnd(S, I[3], I[s - 1]) : C = +I[s - 1], C <= t ? void 0 : { start: S, end: C };
+    }
+    let l = -1;
+    const h = [-1];
+    for (let I = 0; I < c; I++) {
+      const g = i.indexOf("	", l + 1);
+      if (g === -1) {
+        h.push(i.length);
+        break;
+      }
+      h.push(g), l = g;
+    }
+    const b = i.slice(h[o - 1] + 1, h[o]);
+    if (!(d ? b === e : this.renameRefSeq(b) === e))
+      return;
+    const w = +i.slice(h[r - 1] + 1, h[r]) + f;
+    if (w >= n)
+      return null;
+    let p;
+    if (s === 0 || s === r ? p = w + 1 : x ? p = this._getVcfEnd(w, i.slice(h[3] + 1, h[4]), i.slice(h[s - 1] + 1, h[s])) : p = +i.slice(h[s - 1] + 1, h[s]), !(p <= t))
+      return { start: w, end: p };
+  }
+  _getVcfEnd(e, t, n) {
+    let i = e + t.length;
+    if (n.includes("SVTYPE=TRA"))
+      return e + 1;
+    if (n[0] !== ".") {
+      const r = n.indexOf("END=");
+      if (r !== -1 && (r === 0 || n[r - 1] === ";")) {
+        const s = r + 4;
+        let c = n.indexOf(";", s);
+        c === -1 && (c = n.length), i = Number.parseInt(n.slice(s, c), 10);
+      }
+    }
+    return i;
+  }
+  /**
+   * return the approximate number of data lines in the given reference
+   * sequence
+   *
+   * @param refSeq reference sequence name
+   *
+   * @returns number of data lines present on that reference sequence
+   */
+  async lineCount(e, t = {}) {
+    return this.index.lineCount(e, t);
+  }
+  /**
+   * read and uncompress the data in a chunk (composed of one or more
+   * contiguous bgzip blocks) of the file
+   */
+  async readChunk(e, t = {}) {
+    const n = await this.filehandle.read(e.fetchedSize(), e.minv.blockPosition, t);
+    return K(n, e, this.cache);
+  }
+}
+export {
+  q as CSI,
+  R as TBI,
+  he as TabixIndexedFile
+};