wiki-graph-core 0.3.1 → 0.4.1

package/dist/retrieval/query/archive-view/source-evidence/ranges.js

@@ -0,0 +1,209 @@
+import { getTextStreamIndex } from "../text-streams.js";
+function createNodeEvidenceRanges(node) {
+  const ranges = /* @__PURE__ */ new Map();
+  for (const [chapterId, sentenceIndex] of node.sentenceIds) {
+    const key = `${chapterId}`;
+    const current = ranges.get(key);
+    if (current === void 0) {
+      ranges.set(key, [sentenceIndex, sentenceIndex]);
+    } else {
+      ranges.set(key, [
+        Math.min(current[0], sentenceIndex),
+        Math.max(current[1], sentenceIndex)
+      ]);
+    }
+  }
+  return [...ranges.entries()].map(([key, [start, end]]) => {
+    const chapterId = Number(key);
+    return {
+      chapterId,
+      endSentenceIndex: end,
+      startSentenceIndex: start
+    };
+  });
+}
+async function createMentionEvidenceRanges(document, mentions) {
+  return await Promise.all(
+    mentions.map(async (mention) => {
+      const startSentenceIndex = mention.sentenceIndex ?? await findSentenceIndexAtOffset(
+        document,
+        mention.chapterId,
+        mention.rangeStart
+      );
+      const endSentenceIndex = mention.sentenceIndex ?? await findSentenceIndexAtOffset(
+        document,
+        mention.chapterId,
+        Math.max(0, mention.rangeEnd - 1)
+      );
+      return {
+        chapterId: mention.chapterId,
+        endSentenceIndex,
+        startSentenceIndex
+      };
+    })
+  );
+}
+function createMentionLinkEvidenceRanges(_document, links) {
+  return links.flatMap(
+    (link) => link.evidenceSentenceIds.map(([chapterId, sentenceIndex]) => ({
+      chapterId,
+      endSentenceIndex: sentenceIndex,
+      startSentenceIndex: sentenceIndex
+    }))
+  );
+}
+function mergeSourceEvidenceRanges(ranges) {
+  const rangesBySource = /* @__PURE__ */ new Map();
+  for (const range of ranges) {
+    const key = `${range.chapterId}`;
+    const sourceRanges = rangesBySource.get(key) ?? [];
+    sourceRanges.push({
+      end: range.endSentenceIndex,
+      ...range.score === void 0 ? {} : { score: range.score },
+      start: range.startSentenceIndex
+    });
+    rangesBySource.set(key, sourceRanges);
+  }
+  return [...rangesBySource.entries()].flatMap(([key, ranges2]) => {
+    const chapterId = Number(key);
+    return mergeEvidenceRanges(ranges2).map(
+      ({ end, score, start }) => ({
+        chapterId,
+        endSentenceIndex: end,
+        ...score === void 0 ? {} : { score },
+        startSentenceIndex: start
+      })
+    );
+  });
+}
+function mergeSourceEvidenceRangesInInputOrder(ranges) {
+  const merged = [];
+  for (const range of ranges) {
+    const overlappingIndexes = merged.map(
+      (existing, index) => areMergeableSourceEvidenceRanges(existing, range) ? index : -1
+    ).filter((index) => index >= 0);
+    if (overlappingIndexes.length === 0) {
+      merged.push(range);
+      continue;
+    }
+    const firstIndex = overlappingIndexes[0] ?? 0;
+    const overlapping = overlappingIndexes.flatMap((index) => {
+      const existing = merged[index];
+      return existing === void 0 ? [] : [existing];
+    });
+    const mergedRange = mergeSourceEvidenceRangeGroup([...overlapping, range]);
+    merged[firstIndex] = mergedRange;
+    for (const index of overlappingIndexes.slice(1).reverse()) {
+      merged.splice(index, 1);
+    }
+  }
+  return merged;
+}
+function areMergeableSourceEvidenceRanges(left, right) {
+  return left.chapterId === right.chapterId && right.startSentenceIndex <= left.endSentenceIndex + 1 && left.startSentenceIndex <= right.endSentenceIndex + 1;
+}
+function mergeSourceEvidenceRangeGroup(ranges) {
+  const [first] = ranges;
+  if (first === void 0) {
+    throw new Error("Internal error: cannot merge empty evidence range group.");
+  }
+  const scores = ranges.map((range) => range.score).filter((score) => score !== void 0);
+  return {
+    chapterId: first.chapterId,
+    endSentenceIndex: Math.max(
+      ...ranges.map((range) => range.endSentenceIndex)
+    ),
+    ...scores.length === 0 ? {} : { score: Math.max(...scores) },
+    startSentenceIndex: Math.min(
+      ...ranges.map((range) => range.startSentenceIndex)
+    )
+  };
+}
+async function createExpandedSourceEvidenceRanges(document, ranges, sourceContext, context) {
+  const expanded = await Promise.all(
+    ranges.map(async (range) => {
+      if (sourceContext <= 0) {
+        return range;
+      }
+      const sourceIndex = await getTextStreamIndex(
+        document,
+        range.chapterId,
+        "source",
+        context
+      );
+      const lastSentenceIndex = Math.max(0, sourceIndex.sentences.length - 1);
+      return {
+        ...range,
+        endSentenceIndex: clampInteger(
+          range.endSentenceIndex + sourceContext,
+          range.startSentenceIndex,
+          lastSentenceIndex
+        ),
+        startSentenceIndex: clampInteger(
+          range.startSentenceIndex - sourceContext,
+          0,
+          lastSentenceIndex
+        )
+      };
+    })
+  );
+  return mergeSourceEvidenceRangesInInputOrder(expanded);
+}
+async function findSentenceIndexAtOffset(document, chapterId, offset) {
+  const serial = document.getSerialFragments(chapterId);
+  const sentences = serial.listSentences === void 0 ? [] : await serial.listSentences();
+  if (sentences.length === 0) {
+    return 0;
+  }
+  let cursor = 0;
+  for (let index = 0; index < sentences.length; index += 1) {
+    const sentence = sentences[index];
+    if (sentence === void 0) {
+      continue;
+    }
+    const nextCursor = cursor + sentence.text.length;
+    if (offset <= nextCursor) {
+      return index;
+    }
+    cursor = nextCursor + 1;
+  }
+  return Math.max(0, sentences.length - 1);
+}
+function mergeEvidenceRanges(ranges) {
+  const sortedRanges = [...ranges].map((range) => ({
+    ...range.score === void 0 ? {} : { score: range.score },
+    end: Math.max(range.start, range.end),
+    start: Math.min(range.start, range.end)
+  })).sort(
+    (left, right) => left.start === right.start ? left.end - right.end : left.start - right.start
+  );
+  const mergedRanges = [];
+  for (const { end, score, start } of sortedRanges) {
+    const last = mergedRanges.at(-1);
+    if (last === void 0 || start > last.end + 1) {
+      mergedRanges.push({
+        end,
+        ...score === void 0 ? {} : { score },
+        start
+      });
+    } else {
+      last.end = Math.max(last.end, end);
+      if (score !== void 0) {
+        last.score = Math.max(last.score ?? score, score);
+      }
+    }
+  }
+  return mergedRanges;
+}
+function clampInteger(value, min, max) {
+  return Math.max(min, Math.min(max, Math.floor(value)));
+}
+export {
+  createExpandedSourceEvidenceRanges,
+  createMentionEvidenceRanges,
+  createMentionLinkEvidenceRanges,
+  createNodeEvidenceRanges,
+  mergeSourceEvidenceRanges,
+  mergeSourceEvidenceRangesInInputOrder
+};
+//# sourceMappingURL=ranges.js.map

package/dist/retrieval/query/archive-view/source-evidence/ranges.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../../../../../src/retrieval/query/archive-view/source-evidence/ranges.ts"],"sourcesContent":["import type {\n  MentionLinkRecord,\n  MentionRecord,\n  ReadonlyDocument,\n} from \"../../../../document/index.js\";\nimport type { GraphNode } from \"../../../../graph/reading.js\";\n\nimport { getTextStreamIndex } from \"../text-streams.js\";\nimport type { EvidenceReadContext, SourceEvidenceRange } from \"../types.js\";\n\nexport function createNodeEvidenceRanges(\n  node: Pick<GraphNode, \"sentenceIds\">,\n): Array<{\n  readonly chapterId: number;\n  readonly endSentenceIndex: number;\n  readonly startSentenceIndex: number;\n}> {\n  const ranges = new Map<string, [number, number]>();\n\n  for (const [chapterId, sentenceIndex] of node.sentenceIds) {\n    const key = `${chapterId}`;\n    const current = ranges.get(key);\n\n    if (current === undefined) {\n      ranges.set(key, [sentenceIndex, sentenceIndex]);\n    } else {\n      ranges.set(key, [\n        Math.min(current[0], sentenceIndex),\n        Math.max(current[1], sentenceIndex),\n      ]);\n    }\n  }\n\n  return [...ranges.entries()].map(([key, [start, end]]) => {\n    const chapterId = Number(key);\n    return {\n      chapterId,\n      endSentenceIndex: end,\n      startSentenceIndex: start,\n    };\n  });\n}\n\nexport async function createMentionEvidenceRanges(\n  document: ReadonlyDocument,\n  mentions: readonly MentionRecord[],\n): Promise<\n  Array<{\n    readonly chapterId: number;\n    readonly endSentenceIndex: number;\n    readonly startSentenceIndex: number;\n  }>\n> {\n  return await Promise.all(\n    mentions.map(async (mention) => {\n      const startSentenceIndex =\n        mention.sentenceIndex ??\n        (await findSentenceIndexAtOffset(\n          document,\n          mention.chapterId,\n          mention.rangeStart,\n        ));\n      const endSentenceIndex =\n        mention.sentenceIndex ??\n        (await findSentenceIndexAtOffset(\n          document,\n          mention.chapterId,\n          Math.max(0, mention.rangeEnd - 1),\n        ));\n\n      return {\n        chapterId: mention.chapterId,\n        endSentenceIndex,\n        startSentenceIndex,\n      };\n    }),\n  );\n}\n\nexport function createMentionLinkEvidenceRanges(\n  _document: ReadonlyDocument,\n  links: readonly MentionLinkRecord[],\n): Array<{\n  readonly chapterId: number;\n  readonly endSentenceIndex: number;\n  readonly startSentenceIndex: number;\n}> {\n  return links.flatMap((link) =>\n    link.evidenceSentenceIds.map(([chapterId, sentenceIndex]) => ({\n      chapterId,\n      endSentenceIndex: sentenceIndex,\n      startSentenceIndex: sentenceIndex,\n    })),\n  );\n}\n\nexport function mergeSourceEvidenceRanges(\n  ranges: readonly SourceEvidenceRange[],\n): SourceEvidenceRange[] {\n  const rangesBySource = new Map<\n    string,\n    Array<{\n      readonly end: number;\n      readonly score?: number;\n      readonly start: number;\n    }>\n  >();\n\n  for (const range of ranges) {\n    const key = `${range.chapterId}`;\n    const sourceRanges = rangesBySource.get(key) ?? [];\n\n    sourceRanges.push({\n      end: range.endSentenceIndex,\n      ...(range.score === undefined ? {} : { score: range.score }),\n      start: range.startSentenceIndex,\n    });\n    rangesBySource.set(key, sourceRanges);\n  }\n\n  return [...rangesBySource.entries()].flatMap(([key, ranges]) => {\n    const chapterId = Number(key);\n\n    return mergeEvidenceRanges(ranges).map(\n      ({ end, score, start }) =>\n        ({\n          chapterId,\n          endSentenceIndex: end,\n          ...(score === undefined ? {} : { score }),\n          startSentenceIndex: start,\n        }) as const,\n    );\n  });\n}\n\nexport function mergeSourceEvidenceRangesInInputOrder(\n  ranges: readonly SourceEvidenceRange[],\n): SourceEvidenceRange[] {\n  const merged: SourceEvidenceRange[] = [];\n\n  for (const range of ranges) {\n    const overlappingIndexes = merged\n      .map((existing, index) =>\n        areMergeableSourceEvidenceRanges(existing, range) ? index : -1,\n      )\n      .filter((index) => index >= 0);\n\n    if (overlappingIndexes.length === 0) {\n      merged.push(range);\n      continue;\n    }\n\n    const firstIndex = overlappingIndexes[0] ?? 0;\n    const overlapping = overlappingIndexes.flatMap((index) => {\n      const existing = merged[index];\n\n      return existing === undefined ? [] : [existing];\n    });\n    const mergedRange = mergeSourceEvidenceRangeGroup([...overlapping, range]);\n\n    merged[firstIndex] = mergedRange;\n    for (const index of overlappingIndexes.slice(1).reverse()) {\n      merged.splice(index, 1);\n    }\n  }\n\n  return merged;\n}\n\nfunction areMergeableSourceEvidenceRanges(\n  left: SourceEvidenceRange,\n  right: SourceEvidenceRange,\n): boolean {\n  return (\n    left.chapterId === right.chapterId &&\n    right.startSentenceIndex <= left.endSentenceIndex + 1 &&\n    left.startSentenceIndex <= right.endSentenceIndex + 1\n  );\n}\n\nfunction mergeSourceEvidenceRangeGroup(\n  ranges: readonly SourceEvidenceRange[],\n): SourceEvidenceRange {\n  const [first] = ranges;\n\n  if (first === undefined) {\n    throw new Error(\"Internal error: cannot merge empty evidence range group.\");\n  }\n\n  const scores = ranges\n    .map((range) => range.score)\n    .filter((score): score is number => score !== undefined);\n\n  return {\n    chapterId: first.chapterId,\n    endSentenceIndex: Math.max(\n      ...ranges.map((range) => range.endSentenceIndex),\n    ),\n    ...(scores.length === 0 ? {} : { score: Math.max(...scores) }),\n    startSentenceIndex: Math.min(\n      ...ranges.map((range) => range.startSentenceIndex),\n    ),\n  };\n}\n\nexport async function createExpandedSourceEvidenceRanges(\n  document: ReadonlyDocument,\n  ranges: readonly SourceEvidenceRange[],\n  sourceContext: number,\n  context: EvidenceReadContext,\n): Promise<SourceEvidenceRange[]> {\n  const expanded = await Promise.all(\n    ranges.map(async (range) => {\n      if (sourceContext <= 0) {\n        return range;\n      }\n\n      const sourceIndex = await getTextStreamIndex(\n        document,\n        range.chapterId,\n        \"source\",\n        context,\n      );\n      const lastSentenceIndex = Math.max(0, sourceIndex.sentences.length - 1);\n\n      return {\n        ...range,\n        endSentenceIndex: clampInteger(\n          range.endSentenceIndex + sourceContext,\n          range.startSentenceIndex,\n          lastSentenceIndex,\n        ),\n        startSentenceIndex: clampInteger(\n          range.startSentenceIndex - sourceContext,\n          0,\n          lastSentenceIndex,\n        ),\n      };\n    }),\n  );\n\n  return mergeSourceEvidenceRangesInInputOrder(expanded);\n}\n\nasync function findSentenceIndexAtOffset(\n  document: ReadonlyDocument,\n  chapterId: number,\n  offset: number,\n): Promise<number> {\n  const serial = document.getSerialFragments(chapterId);\n  const sentences =\n    serial.listSentences === undefined ? [] : await serial.listSentences();\n\n  if (sentences.length === 0) {\n    return 0;\n  }\n  let cursor = 0;\n\n  for (let index = 0; index < sentences.length; index += 1) {\n    const sentence = sentences[index];\n\n    if (sentence === undefined) {\n      continue;\n    }\n\n    const nextCursor = cursor + sentence.text.length;\n\n    if (offset <= nextCursor) {\n      return index;\n    }\n\n    cursor = nextCursor + 1;\n  }\n\n  return Math.max(0, sentences.length - 1);\n}\n\nfunction mergeEvidenceRanges(\n  ranges: readonly {\n    readonly end: number;\n    readonly score?: number;\n    readonly start: number;\n  }[],\n): readonly {\n  readonly end: number;\n  readonly score?: number;\n  readonly start: number;\n}[] {\n  const sortedRanges = [...ranges]\n    .map((range) => ({\n      ...(range.score === undefined ? {} : { score: range.score }),\n      end: Math.max(range.start, range.end),\n      start: Math.min(range.start, range.end),\n    }))\n    .sort((left, right) =>\n      left.start === right.start\n        ? left.end - right.end\n        : left.start - right.start,\n    );\n  const mergedRanges: Array<{\n    end: number;\n    score?: number;\n    start: number;\n  }> = [];\n\n  for (const { end, score, start } of sortedRanges) {\n    const last = mergedRanges.at(-1);\n\n    if (last === undefined || start > last.end + 1) {\n      mergedRanges.push({\n        end,\n        ...(score === undefined ? {} : { score }),\n        start,\n      });\n    } else {\n      last.end = Math.max(last.end, end);\n      if (score !== undefined) {\n        last.score = Math.max(last.score ?? score, score);\n      }\n    }\n  }\n\n  return mergedRanges;\n}\n\nfunction clampInteger(value: number, min: number, max: number): number {\n  return Math.max(min, Math.min(max, Math.floor(value)));\n}\n"],"mappings":"AAOA,SAAS,0BAA0B;AAG5B,SAAS,yBACd,MAKC;AACD,QAAM,SAAS,oBAAI,IAA8B;AAEjD,aAAW,CAAC,WAAW,aAAa,KAAK,KAAK,aAAa;AACzD,UAAM,MAAM,GAAG,SAAS;AACxB,UAAM,UAAU,OAAO,IAAI,GAAG;AAE9B,QAAI,YAAY,QAAW;AACzB,aAAO,IAAI,KAAK,CAAC,eAAe,aAAa,CAAC;AAAA,IAChD,OAAO;AACL,aAAO,IAAI,KAAK;AAAA,QACd,KAAK,IAAI,QAAQ,CAAC,GAAG,aAAa;AAAA,QAClC,KAAK,IAAI,QAAQ,CAAC,GAAG,aAAa;AAAA,MACpC,CAAC;AAAA,IACH;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,OAAO,QAAQ,CAAC,EAAE,IAAI,CAAC,CAAC,KAAK,CAAC,OAAO,GAAG,CAAC,MAAM;AACxD,UAAM,YAAY,OAAO,GAAG;AAC5B,WAAO;AAAA,MACL;AAAA,MACA,kBAAkB;AAAA,MAClB,oBAAoB;AAAA,IACtB;AAAA,EACF,CAAC;AACH;AAEA,eAAsB,4BACpB,UACA,UAOA;AACA,SAAO,MAAM,QAAQ;AAAA,IACnB,SAAS,IAAI,OAAO,YAAY;AAC9B,YAAM,qBACJ,QAAQ,iBACP,MAAM;AAAA,QACL;AAAA,QACA,QAAQ;AAAA,QACR,QAAQ;AAAA,MACV;AACF,YAAM,mBACJ,QAAQ,iBACP,MAAM;AAAA,QACL;AAAA,QACA,QAAQ;AAAA,QACR,KAAK,IAAI,GAAG,QAAQ,WAAW,CAAC;AAAA,MAClC;AAEF,aAAO;AAAA,QACL,WAAW,QAAQ;AAAA,QACnB;AAAA,QACA;AAAA,MACF;AAAA,IACF,CAAC;AAAA,EACH;AACF;AAEO,SAAS,gCACd,WACA,OAKC;AACD,SAAO,MAAM;AAAA,IAAQ,CAAC,SACpB,KAAK,oBAAoB,IAAI,CAAC,CAAC,WAAW,aAAa,OAAO;AAAA,MAC5D;AAAA,MACA,kBAAkB;AAAA,MAClB,oBAAoB;AAAA,IACtB,EAAE;AAAA,EACJ;AACF;AAEO,SAAS,0BACd,QACuB;AACvB,QAAM,iBAAiB,oBAAI,IAOzB;AAEF,aAAW,SAAS,QAAQ;AAC1B,UAAM,MAAM,GAAG,MAAM,SAAS;AAC9B,UAAM,eAAe,eAAe,IAAI,GAAG,KAAK,CAAC;AAEjD,iBAAa,KAAK;AAAA,MAChB,KAAK,MAAM;AAAA,MACX,GAAI,MAAM,UAAU,SAAY,CAAC,IAAI,EAAE,OAAO,MAAM,MAAM;AAAA,MAC1D,OAAO,MAAM;AAAA,IACf,CAAC;AACD,mBAAe,IAAI,KAAK,YAAY;AAAA,EACtC;AAEA,SAAO,CAAC,GAAG,eAAe,QAAQ,CAAC,EAAE,QAAQ,CAAC,CAAC,KAAKA,OAAM,MAAM;AAC9D,UAAM,YAAY,OAAO,GAAG;AAE5B,WAAO,oBAAoBA,OAAM,EAAE;AAAA,MACjC,CAAC,EAAE,KAAK,OAAO,MAAM,OAClB;AAAA,QACC;AAAA,QACA,kBAAkB;AAAA,QAClB,GAAI,UAAU,SAAY,CAAC,IAAI,EAAE,MAAM;AAAA,QACvC,oBAAoB;AAAA,MACtB;AAAA,IACJ;AAAA,EACF,CAAC;AACH;AAEO,SAAS,sCACd,QACuB;AACvB,QAAM,SAAgC,CAAC;AAEvC,aAAW,SAAS,QAAQ;AAC1B,UAAM,qBAAqB,OACxB;AAAA,MAAI,CAAC,UAAU,UACd,iCAAiC,UAAU,KAAK,IAAI,QAAQ;AAAA,IAC9D,EACC,OAAO,CAAC,UAAU,SAAS,CAAC;AAE/B,QAAI,mBAAmB,WAAW,GAAG;AACnC,aAAO,KAAK,KAAK;AACjB;AAAA,IACF;AAEA,UAAM,aAAa,mBAAmB,CAAC,KAAK;AAC5C,UAAM,cAAc,mBAAmB,QAAQ,CAAC,UAAU;AACxD,YAAM,WAAW,OAAO,KAAK;AAE7B,aAAO,aAAa,SAAY,CAAC,IAAI,CAAC,QAAQ;AAAA,IAChD,CAAC;AACD,UAAM,cAAc,8BAA8B,CAAC,GAAG,aAAa,KAAK,CAAC;AAEzE,WAAO,UAAU,IAAI;AACrB,eAAW,SAAS,mBAAmB,MAAM,CAAC,EAAE,QAAQ,GAAG;AACzD,aAAO,OAAO,OAAO,CAAC;AAAA,IACxB;AAAA,EACF;AAEA,SAAO;AACT;AAEA,SAAS,iCACP,MACA,OACS;AACT,SACE,KAAK,cAAc,MAAM,aACzB,MAAM,sBAAsB,KAAK,mBAAmB,KACpD,KAAK,sBAAsB,MAAM,mBAAmB;AAExD;AAEA,SAAS,8BACP,QACqB;AACrB,QAAM,CAAC,KAAK,IAAI;AAEhB,MAAI,UAAU,QAAW;AACvB,UAAM,IAAI,MAAM,0DAA0D;AAAA,EAC5E;AAEA,QAAM,SAAS,OACZ,IAAI,CAAC,UAAU,MAAM,KAAK,EAC1B,OAAO,CAAC,UAA2B,UAAU,MAAS;AAEzD,SAAO;AAAA,IACL,WAAW,MAAM;AAAA,IACjB,kBAAkB,KAAK;AAAA,MACrB,GAAG,OAAO,IAAI,CAAC,UAAU,MAAM,gBAAgB;AAAA,IACjD;AAAA,IACA,GAAI,OAAO,WAAW,IAAI,CAAC,IAAI,EAAE,OAAO,KAAK,IAAI,GAAG,MAAM,EAAE;AAAA,IAC5D,oBAAoB,KAAK;AAAA,MACvB,GAAG,OAAO,IAAI,CAAC,UAAU,MAAM,kBAAkB;AAAA,IACnD;AAAA,EACF;AACF;AAEA,eAAsB,mCACpB,UACA,QACA,eACA,SACgC;AAChC,QAAM,WAAW,MAAM,QAAQ;AAAA,IAC7B,OAAO,IAAI,OAAO,UAAU;AAC1B,UAAI,iBAAiB,GAAG;AACtB,eAAO;AAAA,MACT;AAEA,YAAM,cAAc,MAAM;AAAA,QACxB;AAAA,QACA,MAAM;AAAA,QACN;AAAA,QACA;AAAA,MACF;AACA,YAAM,oBAAoB,KAAK,IAAI,GAAG,YAAY,UAAU,SAAS,CAAC;AAEtE,aAAO;AAAA,QACL,GAAG;AAAA,QACH,kBAAkB;AAAA,UAChB,MAAM,mBAAmB;AAAA,UACzB,MAAM;AAAA,UACN;AAAA,QACF;AAAA,QACA,oBAAoB;AAAA,UAClB,MAAM,qBAAqB;AAAA,UAC3B;AAAA,UACA;AAAA,QACF;AAAA,MACF;AAAA,IACF,CAAC;AAAA,EACH;AAEA,SAAO,sCAAsC,QAAQ;AACvD;AAEA,eAAe,0BACb,UACA,WACA,QACiB;AACjB,QAAM,SAAS,SAAS,mBAAmB,SAAS;AACpD,QAAM,YACJ,OAAO,kBAAkB,SAAY,CAAC,IAAI,MAAM,OAAO,cAAc;AAEvE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO;AAAA,EACT;AACA,MAAI,SAAS;AAEb,WAAS,QAAQ,GAAG,QAAQ,UAAU,QAAQ,SAAS,GAAG;AACxD,UAAM,WAAW,UAAU,KAAK;AAEhC,QAAI,aAAa,QAAW;AAC1B;AAAA,IACF;AAEA,UAAM,aAAa,SAAS,SAAS,KAAK;AAE1C,QAAI,UAAU,YAAY;AACxB,aAAO;AAAA,IACT;AAEA,aAAS,aAAa;AAAA,EACxB;AAEA,SAAO,KAAK,IAAI,GAAG,UAAU,SAAS,CAAC;AACzC;AAEA,SAAS,oBACP,QASE;AACF,QAAM,eAAe,CAAC,GAAG,MAAM,EAC5B,IAAI,CAAC,WAAW;AAAA,IACf,GAAI,MAAM,UAAU,SAAY,CAAC,IAAI,EAAE,OAAO,MAAM,MAAM;AAAA,IAC1D,KAAK,KAAK,IAAI,MAAM,OAAO,MAAM,GAAG;AAAA,IACpC,OAAO,KAAK,IAAI,MAAM,OAAO,MAAM,GAAG;AAAA,EACxC,EAAE,EACD;AAAA,IAAK,CAAC,MAAM,UACX,KAAK,UAAU,MAAM,QACjB,KAAK,MAAM,MAAM,MACjB,KAAK,QAAQ,MAAM;AAAA,EACzB;AACF,QAAM,eAID,CAAC;AAEN,aAAW,EAAE,KAAK,OAAO,MAAM,KAAK,cAAc;AAChD,UAAM,OAAO,aAAa,GAAG,EAAE;AAE/B,QAAI,SAAS,UAAa,QAAQ,KAAK,MAAM,GAAG;AAC9C,mBAAa,KAAK;AAAA,QAChB;AAAA,QACA,GAAI,UAAU,SAAY,CAAC,IAAI,EAAE,MAAM;AAAA,QACvC;AAAA,MACF,CAAC;AAAA,IACH,OAAO;AACL,WAAK,MAAM,KAAK,IAAI,KAAK,KAAK,GAAG;AACjC,UAAI,UAAU,QAAW;AACvB,aAAK,QAAQ,KAAK,IAAI,KAAK,SAAS,OAAO,KAAK;AAAA,MAClD;AAAA,IACF;AAAA,EACF;AAEA,SAAO;AACT;AAEA,SAAS,aAAa,OAAe,KAAa,KAAqB;AACrE,SAAO,KAAK,IAAI,KAAK,KAAK,IAAI,KAAK,KAAK,MAAM,KAAK,CAAC,CAAC;AACvD;","names":["ranges"]}

package/dist/retrieval/query/archive-view/source-evidence/read.d.ts

@@ -0,0 +1,31 @@
+import { ReadonlyDocument } from '../../../../document/directory/types.js';
+import { EvidenceReadContext, ArchiveEvidenceItem } from '../types.js';
+import '../../../../text/source/meta.js';
+import 'zod';
+import '../../../../text/source/toc.js';
+import '../../../../text/source/types.js';
+import '../../../../document/database.js';
+import 'sqlite3';
+import '../../../../document/types.js';
+import '../../../../document/text-streams/serial.js';
+import '../../../../document/text-streams/draft.js';
+import '../../../../document/text-streams/types.js';
+import '../../../../document/stores/chunk.js';
+import '../../../../document/stores/types.js';
+import '../../../../document/stores/fragment-group.js';
+import '../../../../document/stores/graph-build-parameter.js';
+import '../../../../document/stores/mention-link.js';
+import '../../../../document/stores/mention.js';
+import '../../../../document/stores/object-metadata.js';
+import '../../../../document/stores/reading-edge.js';
+import '../../../../document/stores/serial.js';
+import '../../../../document/stores/snake/chunk.js';
+import '../../../../document/stores/snake/core.js';
+import '../../../../document/stores/snake/edge.js';
+import '../../../../graph/reading.js';
+import '../../../../document/chapter/types.js';
+
+declare function createSourceEvidenceItem(document: ReadonlyDocument, chapterId: number, startSentenceIndex: number, endSentenceIndex: number, context?: EvidenceReadContext, score?: number): Promise<ArchiveEvidenceItem>;
+declare function createEvidenceReadContext(): EvidenceReadContext;
+
+export { createEvidenceReadContext, createSourceEvidenceItem };

package/dist/retrieval/query/archive-view/source-evidence/read.js

@@ -0,0 +1,49 @@
+import { formatTextStreamRangeUri } from "../references.js";
+import { readTextStreamRange } from "../text-streams.js";
+import { requireChapter } from "../core.js";
+async function createSourceEvidenceItem(document, chapterId, startSentenceIndex, endSentenceIndex, context = createEvidenceReadContext(), score) {
+  const chapter = await getEvidenceChapter(document, chapterId, context);
+  const range = await readTextStreamRange(
+    document,
+    chapterId,
+    "source",
+    startSentenceIndex,
+    endSentenceIndex,
+    context
+  );
+  return {
+    chapterId,
+    endSentenceIndex: range.endSentenceIndex,
+    fragmentId: range.startSentenceIndex,
+    id: formatTextStreamRangeUri(
+      chapterId,
+      "source",
+      range.startSentenceIndex,
+      range.endSentenceIndex
+    ),
+    ...score === void 0 ? {} : { score },
+    source: range.text,
+    startSentenceIndex: range.startSentenceIndex,
+    title: chapter.title ?? chapter.uri,
+    type: "source"
+  };
+}
+function createEvidenceReadContext() {
+  return {
+    chapters: /* @__PURE__ */ new Map(),
+    streamIndexes: /* @__PURE__ */ new Map()
+  };
+}
+async function getEvidenceChapter(document, chapterId, context) {
+  let chapter = context.chapters.get(chapterId);
+  if (chapter === void 0) {
+    chapter = requireChapter(document, chapterId);
+    context.chapters.set(chapterId, chapter);
+  }
+  return await chapter;
+}
+export {
+  createEvidenceReadContext,
+  createSourceEvidenceItem
+};
+//# sourceMappingURL=read.js.map

package/dist/retrieval/query/archive-view/source-evidence/read.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../../../../../src/retrieval/query/archive-view/source-evidence/read.ts"],"sourcesContent":["import type { ReadonlyDocument } from \"../../../../document/index.js\";\nimport type { ChapterEntry } from \"../../../../document/chapter/index.js\";\n\nimport { formatTextStreamRangeUri } from \"../references.js\";\nimport { readTextStreamRange } from \"../text-streams.js\";\nimport type { ArchiveEvidenceItem, EvidenceReadContext } from \"../types.js\";\nimport { requireChapter } from \"../core.js\";\n\nexport async function createSourceEvidenceItem(\n  document: ReadonlyDocument,\n  chapterId: number,\n  startSentenceIndex: number,\n  endSentenceIndex: number,\n  context: EvidenceReadContext = createEvidenceReadContext(),\n  score?: number,\n): Promise<ArchiveEvidenceItem> {\n  const chapter = await getEvidenceChapter(document, chapterId, context);\n  const range = await readTextStreamRange(\n    document,\n    chapterId,\n    \"source\",\n    startSentenceIndex,\n    endSentenceIndex,\n    context,\n  );\n\n  return {\n    chapterId,\n    endSentenceIndex: range.endSentenceIndex,\n    fragmentId: range.startSentenceIndex,\n    id: formatTextStreamRangeUri(\n      chapterId,\n      \"source\",\n      range.startSentenceIndex,\n      range.endSentenceIndex,\n    ),\n    ...(score === undefined ? {} : { score }),\n    source: range.text,\n    startSentenceIndex: range.startSentenceIndex,\n    title: chapter.title ?? chapter.uri,\n    type: \"source\",\n  };\n}\n\nexport function createEvidenceReadContext(): EvidenceReadContext {\n  return {\n    chapters: new Map(),\n    streamIndexes: new Map(),\n  };\n}\n\nasync function getEvidenceChapter(\n  document: ReadonlyDocument,\n  chapterId: number,\n  context: EvidenceReadContext,\n): Promise<ChapterEntry> {\n  let chapter = context.chapters.get(chapterId);\n\n  if (chapter === undefined) {\n    chapter = requireChapter(document, chapterId);\n    context.chapters.set(chapterId, chapter);\n  }\n\n  return await chapter;\n}\n"],"mappings":"AAGA,SAAS,gCAAgC;AACzC,SAAS,2BAA2B;AAEpC,SAAS,sBAAsB;AAE/B,eAAsB,yBACpB,UACA,WACA,oBACA,kBACA,UAA+B,0BAA0B,GACzD,OAC8B;AAC9B,QAAM,UAAU,MAAM,mBAAmB,UAAU,WAAW,OAAO;AACrE,QAAM,QAAQ,MAAM;AAAA,IAClB;AAAA,IACA;AAAA,IACA;AAAA,IACA;AAAA,IACA;AAAA,IACA;AAAA,EACF;AAEA,SAAO;AAAA,IACL;AAAA,IACA,kBAAkB,MAAM;AAAA,IACxB,YAAY,MAAM;AAAA,IAClB,IAAI;AAAA,MACF;AAAA,MACA;AAAA,MACA,MAAM;AAAA,MACN,MAAM;AAAA,IACR;AAAA,IACA,GAAI,UAAU,SAAY,CAAC,IAAI,EAAE,MAAM;AAAA,IACvC,QAAQ,MAAM;AAAA,IACd,oBAAoB,MAAM;AAAA,IAC1B,OAAO,QAAQ,SAAS,QAAQ;AAAA,IAChC,MAAM;AAAA,EACR;AACF;AAEO,SAAS,4BAAiD;AAC/D,SAAO;AAAA,IACL,UAAU,oBAAI,IAAI;AAAA,IAClB,eAAe,oBAAI,IAAI;AAAA,EACzB;AACF;AAEA,eAAe,mBACb,UACA,WACA,SACuB;AACvB,MAAI,UAAU,QAAQ,SAAS,IAAI,SAAS;AAE5C,MAAI,YAAY,QAAW;AACzB,cAAU,eAAe,UAAU,SAAS;AAC5C,YAAQ,SAAS,IAAI,WAAW,OAAO;AAAA,EACzC;AAEA,SAAO,MAAM;AACf;","names":[]}

package/dist/retrieval/query/archive-view/source.d.ts

@@ -0,0 +1,30 @@
+export { SourceEvidenceCandidateQueryMatch, createMentionEvidencePage, createMentionEvidencePagePreview, createMentionEvidencePreview, createMentionLinkEvidencePage, createMentionLinkEvidencePagePreview, createMentionLinkEvidencePreview, createRangeEvidencePreview, createSortedMentionLinkEvidencePreview, createSortedRangeEvidencePreview, createSourceEvidencePage, createSourceEvidencePreview, filterAndSortSourceEvidenceCandidatesByFtsQuery, filterAndSortSourceEvidenceRangesByFtsQuery } from './source-evidence/core.js';
+export { SourceEvidenceCandidatePageOptions, SourceEvidencePageDisplayOptions, createSourceEvidenceCandidatePage, createSourceEvidenceCandidatePreview, createSourceEvidenceDisplayItems, createSourceEvidenceDisplayRanges } from './source-evidence/pagination.js';
+export { createExpandedSourceEvidenceRanges, createMentionEvidenceRanges, createMentionLinkEvidenceRanges, createNodeEvidenceRanges, mergeSourceEvidenceRanges, mergeSourceEvidenceRangesInInputOrder } from './source-evidence/ranges.js';
+export { createEvidenceReadContext, createSourceEvidenceItem } from './source-evidence/read.js';
+import '../../../document/types.js';
+import '../../../document/directory/types.js';
+import '../../../text/source/meta.js';
+import 'zod';
+import '../../../text/source/toc.js';
+import '../../../text/source/types.js';
+import '../../../document/database.js';
+import 'sqlite3';
+import '../../../document/text-streams/serial.js';
+import '../../../document/text-streams/draft.js';
+import '../../../document/text-streams/types.js';
+import '../../../document/stores/chunk.js';
+import '../../../document/stores/types.js';
+import '../../../document/stores/fragment-group.js';
+import '../../../document/stores/graph-build-parameter.js';
+import '../../../document/stores/mention-link.js';
+import '../../../document/stores/mention.js';
+import '../../../document/stores/object-metadata.js';
+import '../../../document/stores/reading-edge.js';
+import '../../../document/stores/serial.js';
+import '../../../document/stores/snake/chunk.js';
+import '../../../document/stores/snake/core.js';
+import '../../../document/stores/snake/edge.js';
+import './types.js';
+import '../../../graph/reading.js';
+import '../../../document/chapter/types.js';

package/dist/retrieval/query/archive-view/source.js

@@ -0,0 +1,2 @@
+export * from "./source-evidence/index.js";
+//# sourceMappingURL=source.js.map

package/dist/retrieval/query/archive-view/source.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../../../../src/retrieval/query/archive-view/source.ts"],"sourcesContent":["export * from \"./source-evidence/index.js\";\n"],"mappings":"AAAA,cAAc;","names":[]}

package/dist/retrieval/query/archive-view/text-streams.d.ts

@@ -0,0 +1,44 @@
+import { ReadonlyDocument } from '../../../document/directory/types.js';
+import { WikiGraphReference } from './references.js';
+import { ArchiveTextStreamKind, ArchiveTextStreamIndex, ArchiveSourceFragment, EvidenceReadContext } from './types.js';
+import '../../../text/source/meta.js';
+import 'zod';
+import '../../../text/source/toc.js';
+import '../../../text/source/types.js';
+import '../../../document/database.js';
+import 'sqlite3';
+import '../../../document/types.js';
+import '../../../document/text-streams/serial.js';
+import '../../../document/text-streams/draft.js';
+import '../../../document/text-streams/types.js';
+import '../../../document/stores/chunk.js';
+import '../../../document/stores/types.js';
+import '../../../document/stores/fragment-group.js';
+import '../../../document/stores/graph-build-parameter.js';
+import '../../../document/stores/mention-link.js';
+import '../../../document/stores/mention.js';
+import '../../../document/stores/object-metadata.js';
+import '../../../document/stores/reading-edge.js';
+import '../../../document/stores/serial.js';
+import '../../../document/stores/snake/chunk.js';
+import '../../../document/stores/snake/core.js';
+import '../../../document/stores/snake/edge.js';
+import '../../../graph/reading.js';
+import '../../../document/chapter/types.js';
+
+declare function readSourceFragment(document: ReadonlyDocument, serialId: number, fragmentId: number): Promise<ArchiveSourceFragment>;
+declare function createTextStreamRangeFragment(document: ReadonlyDocument, reference: Extract<WikiGraphReference, {
+    readonly type: "text-stream";
+}>): Promise<ArchiveSourceFragment>;
+declare function readTextStreamRange(document: ReadonlyDocument, chapterId: number, stream: ArchiveTextStreamKind, startSentenceIndex: number, endSentenceIndex: number, context?: EvidenceReadContext): Promise<{
+    readonly endSentenceIndex: number;
+    readonly id: string;
+    readonly startSentenceIndex: number;
+    readonly text: string;
+}>;
+declare function readTextStreamText(document: ReadonlyDocument, chapterId: number, stream: ArchiveTextStreamKind): Promise<string>;
+declare function getTextStreamIndex(document: ReadonlyDocument, chapterId: number, stream: ArchiveTextStreamKind, context?: EvidenceReadContext): Promise<ArchiveTextStreamIndex>;
+declare function createTextStreamIndex(document: ReadonlyDocument, chapterId: number, stream: ArchiveTextStreamKind): Promise<ArchiveTextStreamIndex>;
+declare function countWords(text: string): number;
+
+export { countWords, createTextStreamIndex, createTextStreamRangeFragment, getTextStreamIndex, readSourceFragment, readTextStreamRange, readTextStreamText };

package/dist/retrieval/query/archive-view/text-streams.js

@@ -0,0 +1,160 @@
+import { createSnippet } from "./helpers.js";
+import {
+  formatChapterId,
+  formatFragmentId,
+  formatTextStreamRangeUri
+} from "./references.js";
+function createTextStreamReadContext() {
+  return {
+    chapters: /* @__PURE__ */ new Map(),
+    streamIndexes: /* @__PURE__ */ new Map()
+  };
+}
+function clampInteger(value, min, max) {
+  return Math.max(min, Math.min(max, Math.floor(value)));
+}
+async function readSourceFragment(document, serialId, fragmentId) {
+  const fragment = await document.getSerialFragments(serialId).getFragment(fragmentId);
+  const text = fragment.sentences.map((sentence) => sentence.text).join("\n");
+  return {
+    fragmentId,
+    id: formatFragmentId(serialId, fragmentId),
+    preview: createSnippet(text),
+    sentenceCount: fragment.sentences.length,
+    text,
+    wordsCount: fragment.sentences.reduce(
+      (total, sentence) => total + sentence.wordsCount,
+      0
+    )
+  };
+}
+async function createTextStreamRangeFragment(document, reference) {
+  const range = await readTextStreamRange(
+    document,
+    reference.chapterId,
+    reference.stream,
+    reference.startSentenceIndex,
+    reference.endSentenceIndex
+  );
+  return {
+    fragmentId: range.startSentenceIndex,
+    id: range.id,
+    preview: createSnippet(range.text),
+    sentenceCount: range.endSentenceIndex - range.startSentenceIndex + 1,
+    text: range.text,
+    wordsCount: countWords(range.text)
+  };
+}
+async function readTextStreamRange(document, chapterId, stream, startSentenceIndex, endSentenceIndex, context = createTextStreamReadContext()) {
+  const index = await getTextStreamIndex(document, chapterId, stream, context);
+  if (index.sentences.length === 0) {
+    throw new Error(
+      `Chapter ${formatChapterId(chapterId)} has no ${stream} text.`
+    );
+  }
+  const lastSentenceIndex = index.sentences.length - 1;
+  if (startSentenceIndex > lastSentenceIndex) {
+    throw new Error(
+      `${stream} range ${formatTextStreamRangeUri(chapterId, stream, startSentenceIndex, endSentenceIndex)} is out of bounds. Last sentence number is ${lastSentenceIndex + 1}.`
+    );
+  }
+  const start = clampInteger(startSentenceIndex, 0, lastSentenceIndex);
+  const end = clampInteger(endSentenceIndex, start, lastSentenceIndex);
+  const sentences = index.sentences.slice(start, end + 1);
+  const text = normalizeRenderedTextStreamRange(
+    await readTextStreamRawRange(document, chapterId, stream, start, end)
+  ) ?? joinTextStreamSentences(sentences);
+  return {
+    endSentenceIndex: end,
+    id: formatTextStreamRangeUri(chapterId, stream, start, end),
+    startSentenceIndex: start,
+    text
+  };
+}
+async function readTextStreamText(document, chapterId, stream) {
+  const serial = getTextStreamSerial(document, chapterId, stream);
+  const text = await serial.readText?.();
+  if (text !== void 0) {
+    return text;
+  }
+  const index = await createTextStreamIndex(document, chapterId, stream);
+  return joinTextStreamSentences(index.sentences);
+}
+async function readTextStreamRawRange(document, chapterId, stream, startSentenceIndex, endSentenceIndex) {
+  const serial = getTextStreamSerial(document, chapterId, stream);
+  return await serial.readTextInRange?.(startSentenceIndex, endSentenceIndex);
+}
+function getTextStreamSerial(document, chapterId, stream) {
+  return stream === "summary" ? document.getSummaryFragments(chapterId) : document.getSerialFragments(chapterId);
+}
+function joinTextStreamSentences(sentences) {
+  return sentences.map((sentence) => sentence.text).join("");
+}
+function normalizeRenderedTextStreamRange(text) {
+  return text?.replace(/^(?:[^\S\r\n]*(?:\r\n|\n|\r))+/u, "").replace(/(?:(?:\r\n|\n|\r)[^\S\r\n]*)+$/u, "");
+}
+async function getTextStreamIndex(document, chapterId, stream, context = createTextStreamReadContext()) {
+  const key = `${chapterId}:${stream}`;
+  let index = context.streamIndexes.get(key);
+  if (index === void 0) {
+    index = createTextStreamIndex(document, chapterId, stream);
+    context.streamIndexes.set(key, index);
+  }
+  return await index;
+}
+async function createTextStreamIndex(document, chapterId, stream) {
+  if (stream === "summary") {
+    const fragments2 = document.getSummaryFragments(chapterId);
+    const sentences2 = [];
+    for (const fragmentId of await fragments2.listFragmentIds()) {
+      const fragment = await fragments2.getFragment(fragmentId);
+      for (let index = 0; index < fragment.sentences.length; index += 1) {
+        const sentence = fragment.sentences[index];
+        if (sentence === void 0) {
+          continue;
+        }
+        sentences2.push({
+          fragmentId,
+          globalIndex: sentences2.length,
+          localIndex: index,
+          text: sentence.text,
+          wordsCount: sentence.wordsCount
+        });
+      }
+    }
+    return { sentences: sentences2 };
+  }
+  const fragments = document.getSerialFragments(chapterId);
+  const sentences = [];
+  for (const fragmentId of await fragments.listFragmentIds()) {
+    const fragment = await fragments.getFragment(fragmentId);
+    for (let index = 0; index < fragment.sentences.length; index += 1) {
+      const sentence = fragment.sentences[index];
+      if (sentence === void 0) {
+        continue;
+      }
+      sentences.push({
+        fragmentId,
+        globalIndex: sentences.length,
+        localIndex: index,
+        text: sentence.text,
+        wordsCount: sentence.wordsCount
+      });
+    }
+  }
+  return { sentences };
+}
+function countWords(text) {
+  const trimmed = text.trim();
+  return trimmed === "" ? 0 : trimmed.split(/\s+/u).length;
+}
+export {
+  countWords,
+  createTextStreamIndex,
+  createTextStreamRangeFragment,
+  getTextStreamIndex,
+  readSourceFragment,
+  readTextStreamRange,
+  readTextStreamText
+};
+//# sourceMappingURL=text-streams.js.map

package/dist/retrieval/query/archive-view/text-streams.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../../../../src/retrieval/query/archive-view/text-streams.ts"],"sourcesContent":["import type {\n  ReadonlyDocument,\n  ReadonlySerialTextStream,\n} from \"../../../document/index.js\";\n\nimport { createSnippet } from \"./helpers.js\";\nimport {\n  formatChapterId,\n  formatFragmentId,\n  formatTextStreamRangeUri,\n} from \"./references.js\";\nimport type { WikiGraphReference } from \"./references.js\";\nimport type {\n  ArchiveSourceFragment,\n  ArchiveTextStreamIndex,\n  ArchiveTextStreamKind,\n  ArchiveTextStreamSentence,\n  EvidenceReadContext,\n} from \"./types.js\";\n\nfunction createTextStreamReadContext(): EvidenceReadContext {\n  return {\n    chapters: new Map(),\n    streamIndexes: new Map(),\n  };\n}\n\nfunction clampInteger(value: number, min: number, max: number): number {\n  return Math.max(min, Math.min(max, Math.floor(value)));\n}\n\nexport async function readSourceFragment(\n  document: ReadonlyDocument,\n  serialId: number,\n  fragmentId: number,\n): Promise<ArchiveSourceFragment> {\n  const fragment = await document\n    .getSerialFragments(serialId)\n    .getFragment(fragmentId);\n  const text = fragment.sentences.map((sentence) => sentence.text).join(\"\\n\");\n\n  return {\n    fragmentId,\n    id: formatFragmentId(serialId, fragmentId),\n    preview: createSnippet(text),\n    sentenceCount: fragment.sentences.length,\n    text,\n    wordsCount: fragment.sentences.reduce(\n      (total, sentence) => total + sentence.wordsCount,\n      0,\n    ),\n  };\n}\n\nexport async function createTextStreamRangeFragment(\n  document: ReadonlyDocument,\n  reference: Extract<WikiGraphReference, { readonly type: \"text-stream\" }>,\n): Promise<ArchiveSourceFragment> {\n  const range = await readTextStreamRange(\n    document,\n    reference.chapterId,\n    reference.stream,\n    reference.startSentenceIndex,\n    reference.endSentenceIndex,\n  );\n\n  return {\n    fragmentId: range.startSentenceIndex,\n    id: range.id,\n    preview: createSnippet(range.text),\n    sentenceCount: range.endSentenceIndex - range.startSentenceIndex + 1,\n    text: range.text,\n    wordsCount: countWords(range.text),\n  };\n}\n\nexport async function readTextStreamRange(\n  document: ReadonlyDocument,\n  chapterId: number,\n  stream: ArchiveTextStreamKind,\n  startSentenceIndex: number,\n  endSentenceIndex: number,\n  context: EvidenceReadContext = createTextStreamReadContext(),\n): Promise<{\n  readonly endSentenceIndex: number;\n  readonly id: string;\n  readonly startSentenceIndex: number;\n  readonly text: string;\n}> {\n  const index = await getTextStreamIndex(document, chapterId, stream, context);\n  if (index.sentences.length === 0) {\n    throw new Error(\n      `Chapter ${formatChapterId(chapterId)} has no ${stream} text.`,\n    );\n  }\n\n  const lastSentenceIndex = index.sentences.length - 1;\n  if (startSentenceIndex > lastSentenceIndex) {\n    throw new Error(\n      `${stream} range ${formatTextStreamRangeUri(chapterId, stream, startSentenceIndex, endSentenceIndex)} is out of bounds. Last sentence number is ${lastSentenceIndex + 1}.`,\n    );\n  }\n\n  const start = clampInteger(startSentenceIndex, 0, lastSentenceIndex);\n  const end = clampInteger(endSentenceIndex, start, lastSentenceIndex);\n  const sentences = index.sentences.slice(start, end + 1);\n  const text =\n    normalizeRenderedTextStreamRange(\n      await readTextStreamRawRange(document, chapterId, stream, start, end),\n    ) ?? joinTextStreamSentences(sentences);\n\n  return {\n    endSentenceIndex: end,\n    id: formatTextStreamRangeUri(chapterId, stream, start, end),\n    startSentenceIndex: start,\n    text,\n  };\n}\n\nexport async function readTextStreamText(\n  document: ReadonlyDocument,\n  chapterId: number,\n  stream: ArchiveTextStreamKind,\n): Promise<string> {\n  const serial = getTextStreamSerial(document, chapterId, stream);\n  const text = await serial.readText?.();\n\n  if (text !== undefined) {\n    return text;\n  }\n\n  const index = await createTextStreamIndex(document, chapterId, stream);\n\n  return joinTextStreamSentences(index.sentences);\n}\n\nasync function readTextStreamRawRange(\n  document: ReadonlyDocument,\n  chapterId: number,\n  stream: ArchiveTextStreamKind,\n  startSentenceIndex: number,\n  endSentenceIndex: number,\n): Promise<string | undefined> {\n  const serial = getTextStreamSerial(document, chapterId, stream);\n\n  return await serial.readTextInRange?.(startSentenceIndex, endSentenceIndex);\n}\n\nfunction getTextStreamSerial(\n  document: ReadonlyDocument,\n  chapterId: number,\n  stream: ArchiveTextStreamKind,\n): ReadonlySerialTextStream {\n  return stream === \"summary\"\n    ? document.getSummaryFragments(chapterId)\n    : document.getSerialFragments(chapterId);\n}\n\nfunction joinTextStreamSentences(\n  sentences: readonly Pick<ArchiveTextStreamSentence, \"text\">[],\n): string {\n  return sentences.map((sentence) => sentence.text).join(\"\");\n}\n\nfunction normalizeRenderedTextStreamRange(\n  text: string | undefined,\n): string | undefined {\n  return text\n    ?.replace(/^(?:[^\\S\\r\\n]*(?:\\r\\n|\\n|\\r))+/u, \"\")\n    .replace(/(?:(?:\\r\\n|\\n|\\r)[^\\S\\r\\n]*)+$/u, \"\");\n}\n\nexport async function getTextStreamIndex(\n  document: ReadonlyDocument,\n  chapterId: number,\n  stream: ArchiveTextStreamKind,\n  context: EvidenceReadContext = createTextStreamReadContext(),\n): Promise<ArchiveTextStreamIndex> {\n  const key = `${chapterId}:${stream}`;\n  let index = context.streamIndexes.get(key);\n\n  if (index === undefined) {\n    index = createTextStreamIndex(document, chapterId, stream);\n    context.streamIndexes.set(key, index);\n  }\n\n  return await index;\n}\n\nexport async function createTextStreamIndex(\n  document: ReadonlyDocument,\n  chapterId: number,\n  stream: ArchiveTextStreamKind,\n): Promise<ArchiveTextStreamIndex> {\n  if (stream === \"summary\") {\n    const fragments = document.getSummaryFragments(chapterId);\n    const sentences: ArchiveTextStreamSentence[] = [];\n\n    for (const fragmentId of await fragments.listFragmentIds()) {\n      const fragment = await fragments.getFragment(fragmentId);\n\n      for (let index = 0; index < fragment.sentences.length; index += 1) {\n        const sentence = fragment.sentences[index];\n\n        if (sentence === undefined) {\n          continue;\n        }\n\n        sentences.push({\n          fragmentId,\n          globalIndex: sentences.length,\n          localIndex: index,\n          text: sentence.text,\n          wordsCount: sentence.wordsCount,\n        });\n      }\n    }\n\n    return { sentences };\n  }\n\n  const fragments = document.getSerialFragments(chapterId);\n  const sentences: ArchiveTextStreamSentence[] = [];\n\n  for (const fragmentId of await fragments.listFragmentIds()) {\n    const fragment = await fragments.getFragment(fragmentId);\n\n    for (let index = 0; index < fragment.sentences.length; index += 1) {\n      const sentence = fragment.sentences[index];\n\n      if (sentence === undefined) {\n        continue;\n      }\n\n      sentences.push({\n        fragmentId,\n        globalIndex: sentences.length,\n        localIndex: index,\n        text: sentence.text,\n        wordsCount: sentence.wordsCount,\n      });\n    }\n  }\n\n  return { sentences };\n}\n\nexport function countWords(text: string): number {\n  const trimmed = text.trim();\n\n  return trimmed === \"\" ? 0 : trimmed.split(/\\s+/u).length;\n}\n"],"mappings":"AAKA,SAAS,qBAAqB;AAC9B;AAAA,EACE;AAAA,EACA;AAAA,EACA;AAAA,OACK;AAUP,SAAS,8BAAmD;AAC1D,SAAO;AAAA,IACL,UAAU,oBAAI,IAAI;AAAA,IAClB,eAAe,oBAAI,IAAI;AAAA,EACzB;AACF;AAEA,SAAS,aAAa,OAAe,KAAa,KAAqB;AACrE,SAAO,KAAK,IAAI,KAAK,KAAK,IAAI,KAAK,KAAK,MAAM,KAAK,CAAC,CAAC;AACvD;AAEA,eAAsB,mBACpB,UACA,UACA,YACgC;AAChC,QAAM,WAAW,MAAM,SACpB,mBAAmB,QAAQ,EAC3B,YAAY,UAAU;AACzB,QAAM,OAAO,SAAS,UAAU,IAAI,CAAC,aAAa,SAAS,IAAI,EAAE,KAAK,IAAI;AAE1E,SAAO;AAAA,IACL;AAAA,IACA,IAAI,iBAAiB,UAAU,UAAU;AAAA,IACzC,SAAS,cAAc,IAAI;AAAA,IAC3B,eAAe,SAAS,UAAU;AAAA,IAClC;AAAA,IACA,YAAY,SAAS,UAAU;AAAA,MAC7B,CAAC,OAAO,aAAa,QAAQ,SAAS;AAAA,MACtC;AAAA,IACF;AAAA,EACF;AACF;AAEA,eAAsB,8BACpB,UACA,WACgC;AAChC,QAAM,QAAQ,MAAM;AAAA,IAClB;AAAA,IACA,UAAU;AAAA,IACV,UAAU;AAAA,IACV,UAAU;AAAA,IACV,UAAU;AAAA,EACZ;AAEA,SAAO;AAAA,IACL,YAAY,MAAM;AAAA,IAClB,IAAI,MAAM;AAAA,IACV,SAAS,cAAc,MAAM,IAAI;AAAA,IACjC,eAAe,MAAM,mBAAmB,MAAM,qBAAqB;AAAA,IACnE,MAAM,MAAM;AAAA,IACZ,YAAY,WAAW,MAAM,IAAI;AAAA,EACnC;AACF;AAEA,eAAsB,oBACpB,UACA,WACA,QACA,oBACA,kBACA,UAA+B,4BAA4B,GAM1D;AACD,QAAM,QAAQ,MAAM,mBAAmB,UAAU,WAAW,QAAQ,OAAO;AAC3E,MAAI,MAAM,UAAU,WAAW,GAAG;AAChC,UAAM,IAAI;AAAA,MACR,WAAW,gBAAgB,SAAS,CAAC,WAAW,MAAM;AAAA,IACxD;AAAA,EACF;AAEA,QAAM,oBAAoB,MAAM,UAAU,SAAS;AACnD,MAAI,qBAAqB,mBAAmB;AAC1C,UAAM,IAAI;AAAA,MACR,GAAG,MAAM,UAAU,yBAAyB,WAAW,QAAQ,oBAAoB,gBAAgB,CAAC,8CAA8C,oBAAoB,CAAC;AAAA,IACzK;AAAA,EACF;AAEA,QAAM,QAAQ,aAAa,oBAAoB,GAAG,iBAAiB;AACnE,QAAM,MAAM,aAAa,kBAAkB,OAAO,iBAAiB;AACnE,QAAM,YAAY,MAAM,UAAU,MAAM,OAAO,MAAM,CAAC;AACtD,QAAM,OACJ;AAAA,IACE,MAAM,uBAAuB,UAAU,WAAW,QAAQ,OAAO,GAAG;AAAA,EACtE,KAAK,wBAAwB,SAAS;AAExC,SAAO;AAAA,IACL,kBAAkB;AAAA,IAClB,IAAI,yBAAyB,WAAW,QAAQ,OAAO,GAAG;AAAA,IAC1D,oBAAoB;AAAA,IACpB;AAAA,EACF;AACF;AAEA,eAAsB,mBACpB,UACA,WACA,QACiB;AACjB,QAAM,SAAS,oBAAoB,UAAU,WAAW,MAAM;AAC9D,QAAM,OAAO,MAAM,OAAO,WAAW;AAErC,MAAI,SAAS,QAAW;AACtB,WAAO;AAAA,EACT;AAEA,QAAM,QAAQ,MAAM,sBAAsB,UAAU,WAAW,MAAM;AAErE,SAAO,wBAAwB,MAAM,SAAS;AAChD;AAEA,eAAe,uBACb,UACA,WACA,QACA,oBACA,kBAC6B;AAC7B,QAAM,SAAS,oBAAoB,UAAU,WAAW,MAAM;AAE9D,SAAO,MAAM,OAAO,kBAAkB,oBAAoB,gBAAgB;AAC5E;AAEA,SAAS,oBACP,UACA,WACA,QAC0B;AAC1B,SAAO,WAAW,YACd,SAAS,oBAAoB,SAAS,IACtC,SAAS,mBAAmB,SAAS;AAC3C;AAEA,SAAS,wBACP,WACQ;AACR,SAAO,UAAU,IAAI,CAAC,aAAa,SAAS,IAAI,EAAE,KAAK,EAAE;AAC3D;AAEA,SAAS,iCACP,MACoB;AACpB,SAAO,MACH,QAAQ,mCAAmC,EAAE,EAC9C,QAAQ,mCAAmC,EAAE;AAClD;AAEA,eAAsB,mBACpB,UACA,WACA,QACA,UAA+B,4BAA4B,GAC1B;AACjC,QAAM,MAAM,GAAG,SAAS,IAAI,MAAM;AAClC,MAAI,QAAQ,QAAQ,cAAc,IAAI,GAAG;AAEzC,MAAI,UAAU,QAAW;AACvB,YAAQ,sBAAsB,UAAU,WAAW,MAAM;AACzD,YAAQ,cAAc,IAAI,KAAK,KAAK;AAAA,EACtC;AAEA,SAAO,MAAM;AACf;AAEA,eAAsB,sBACpB,UACA,WACA,QACiC;AACjC,MAAI,WAAW,WAAW;AACxB,UAAMA,aAAY,SAAS,oBAAoB,SAAS;AACxD,UAAMC,aAAyC,CAAC;AAEhD,eAAW,cAAc,MAAMD,WAAU,gBAAgB,GAAG;AAC1D,YAAM,WAAW,MAAMA,WAAU,YAAY,UAAU;AAEvD,eAAS,QAAQ,GAAG,QAAQ,SAAS,UAAU,QAAQ,SAAS,GAAG;AACjE,cAAM,WAAW,SAAS,UAAU,KAAK;AAEzC,YAAI,aAAa,QAAW;AAC1B;AAAA,QACF;AAEA,QAAAC,WAAU,KAAK;AAAA,UACb;AAAA,UACA,aAAaA,WAAU;AAAA,UACvB,YAAY;AAAA,UACZ,MAAM,SAAS;AAAA,UACf,YAAY,SAAS;AAAA,QACvB,CAAC;AAAA,MACH;AAAA,IACF;AAEA,WAAO,EAAE,WAAAA,WAAU;AAAA,EACrB;AAEA,QAAM,YAAY,SAAS,mBAAmB,SAAS;AACvD,QAAM,YAAyC,CAAC;AAEhD,aAAW,cAAc,MAAM,UAAU,gBAAgB,GAAG;AAC1D,UAAM,WAAW,MAAM,UAAU,YAAY,UAAU;AAEvD,aAAS,QAAQ,GAAG,QAAQ,SAAS,UAAU,QAAQ,SAAS,GAAG;AACjE,YAAM,WAAW,SAAS,UAAU,KAAK;AAEzC,UAAI,aAAa,QAAW;AAC1B;AAAA,MACF;AAEA,gBAAU,KAAK;AAAA,QACb;AAAA,QACA,aAAa,UAAU;AAAA,QACvB,YAAY;AAAA,QACZ,MAAM,SAAS;AAAA,QACf,YAAY,SAAS;AAAA,MACvB,CAAC;AAAA,IACH;AAAA,EACF;AAEA,SAAO,EAAE,UAAU;AACrB;AAEO,SAAS,WAAW,MAAsB;AAC/C,QAAM,UAAU,KAAK,KAAK;AAE1B,SAAO,YAAY,KAAK,IAAI,QAAQ,MAAM,MAAM,EAAE;AACpD;","names":["fragments","sentences"]}