ai-spector 0.9.21 → 0.9.24

package/dist/graph/doc-extract.js

@@ -1,821 +0,0 @@
-import { basicDesignListChapterDocumentId, basicDesignListChapterFileToPatch, detailSectionsToPatch, parseDetailSections, snippetAfterHeading, } from "./detail-sections.js";
-import { BASIC_DESIGN_LIST_DOCUMENT_IDS, DEFAULT_BD_LIST_DOC, DEFAULT_LISTED_IN, PER_DOMAIN_TEMPLATE_DOC_BD, } from "./defaults.js";
-import { scanBasicDesignListDocuments } from "../registry/build.js";
-import { mergeStructurePatches, registryDocumentToStructurePatch, } from "../registry/structure-patch.js";
-import { extractBlockquoteAfterBoldField, extractBoldFieldDeep, extractBoldFieldInParagraph, findHeadingText, parseMarkdown, textContent, } from "../markdown/parse.js";
-import { visit } from "unist-util-visit";
-/**
- * Builtin fallback for per-domain template doc IDs.
- * Prefer PackManifest.perDomainTemplates when a pack config is available.
- * @deprecated Use packConfig?.perDomainTemplates instead.
- */
-const BUILTIN_PER_DOMAIN_TEMPLATE_DOC = {
-    useCase: "doc.srs.uc-detail",
-    feature: "doc.srs.feature-detail",
-};
-const UC_ID_RE = /\b(UC-\d+)\b/gi;
-const F_ID_RE = /\b(F-\d+)\b/gi;
-const TABLE_ROW_UC = /^\|\s*(UC-\d+)\s*\|/im;
-const TABLE_ROW_F = /^\|\s*(F-\d+)\s*\|/im;
-const OVERVIEW_HEADING = /use case overview/i;
-const FEATURE_DESC_HEADING = /^1\.\s*description$/i;
-/** Normalize UC-1 → UC-01, F-2 → F-02; leaves placeholders (UC-XX) unchanged */
-export function normalizeDomainId(raw) {
-    const t = raw.trim();
-    if (/XX$/i.test(t)) {
-        return t;
-    }
-    const uc = t.match(/^UC-(\d+)$/i);
-    if (uc) {
-        return `UC-${uc[1].padStart(2, "0")}`;
-    }
-    const f = t.match(/^F-(\d+)$/i);
-    if (f) {
-        return `F-${f[1].padStart(2, "0")}`;
-    }
-    return t;
-}
-function isRealDomainId(id) {
-    return /^(UC|F)-\d+$/i.test(id) && !/XX$/i.test(id);
-}
-function normalizeRelativePath(relativePath) {
-    return relativePath.replace(/\\/g, "/").replace(/^\.\//, "");
-}
-/** Classify markdown under docs/srs as list chapter vs per-domain detail file. */
-export function classifySrsDetailFile(relativePath) {
-    const p = normalizeRelativePath(relativePath).toLowerCase();
-    if (p.endsWith("/3-use-cases.md") || p === "docs/srs/3-use-cases.md") {
-        return null;
-    }
-    if (p.endsWith("/4-system-features.md") || p === "docs/srs/4-system-features.md") {
-        return null;
-    }
-    if (/\/03-use-cases\//.test(p) || /\/uc-\d+/.test(p)) {
-        return "useCaseDetail";
-    }
-    if (/\/04-system-features\/f-\d+/.test(p)) {
-        return "featureDetail";
-    }
-    return null;
-}
-/** Classify markdown under docs/basic-design as list chapter vs per-endpoint/per-screen detail. */
-export function classifyBasicDesignDetailFile(relativePath) {
-    const p = normalizeRelativePath(relativePath).toLowerCase();
-    if (p === "docs/basic-design/api-list.md" ||
-        p.endsWith("/api-list.md") ||
-        p === "docs/basic-design/db-design.md" ||
-        p.endsWith("/db-design.md") ||
-        p === "docs/basic-design/list-screens.md" ||
-        p.endsWith("/list-screens.md") ||
-        p.includes("/screens/list-screens.md")) {
-        return null;
-    }
-    if (/\/api\/[^/]+\.md$/.test(p)) {
-        return "apiDetail";
-    }
-    if (/\/screens\/[^/]+\.md$/.test(p)) {
-        return "screenDetail";
-    }
-    return null;
-}
-export function documentIdForDomainDetail(kind, domainId, nodePrefix) {
-    const prefix = nodePrefix ?? "doc.srs";
-    const norm = normalizeDomainId(domainId);
-    return kind === "useCaseDetail"
-        ? `${prefix}.uc-${norm}`
-        : `${prefix}.f-${norm}`;
-}
-function slugFromFilename(relativePath) {
-    const base = normalizeRelativePath(relativePath)
-        .split("/")
-        .pop()
-        ?.replace(/\.md$/i, "") ?? "unknown";
-    return base
-        .toLowerCase()
-        .replace(/[^a-z0-9]+/g, "-")
-        .replace(/^-|-$/g, "")
-        .slice(0, 64);
-}
-export function documentIdForBasicDesignDetail(kind, relativePath) {
-    const slug = slugFromFilename(relativePath);
-    return kind === "apiDetail" ? `doc.bd.api-${slug}` : `doc.bd.screen-${slug}`;
-}
-const ENDPOINT_SPEC_HEADING = /###\s*\d+(?:\.\d+)?\s*`([A-Z]+)\s+([^`]+)`/i;
-const API_DETAIL_TITLE = /^#\s*API Detail:\s*(.+)/im;
-const SCREEN_SECTION_HEADING = /^##\s*\d+\.\s*Screen:\s*(.+)/im;
-/** Title and optional related features from per-endpoint / per-screen detail markdown. */
-export function extractBasicDesignDetailMeta(content, kind, relativePath) {
-    const featureIds = new Set();
-    for (const m of content.matchAll(/\bF-\d+\b/gi)) {
-        const id = normalizeDomainId(m[0]);
-        if (isRealDomainId(id)) {
-            featureIds.add(id);
-        }
-    }
-    const root = parseMarkdown(content);
-    const fIdVal = extractBoldFieldDeep(root, /^Feature ID$/i);
-    if (fIdVal) {
-        const m = fIdVal.match(/\b(F-\d+)\b/i);
-        if (m?.[1] && isRealDomainId(m[1]))
-            featureIds.add(normalizeDomainId(m[1]));
-    }
-    if (kind === "apiDetail") {
-        const endpoint = content.match(ENDPOINT_SPEC_HEADING);
-        if (endpoint) {
-            return {
-                title: `${endpoint[1]} ${endpoint[2]}`.trim(),
-                featureIds: [...featureIds],
-            };
-        }
-        const titleLine = content.match(API_DETAIL_TITLE);
-        if (titleLine?.[1]?.trim()) {
-            return { title: titleLine[1].trim(), featureIds: [...featureIds] };
-        }
-        return { title: slugFromFilename(relativePath), featureIds: [...featureIds] };
-    }
-    const screen = content.match(SCREEN_SECTION_HEADING);
-    if (screen?.[1]?.trim()) {
-        return { title: screen[1].trim(), featureIds: [...featureIds] };
-    }
-    return { title: slugFromFilename(relativePath), featureIds: [...featureIds] };
-}
-function buildBasicDesignDetailInstancePatch(opts) {
-    const path = normalizeRelativePath(opts.relativePath);
-    const templateDocId = opts.kind === "apiDetail"
-        ? PER_DOMAIN_TEMPLATE_DOC_BD.api
-        : PER_DOMAIN_TEMPLATE_DOC_BD.screen;
-    const listAnchorId = opts.kind === "apiDetail"
-        ? DEFAULT_BD_LIST_DOC.apiList
-        : DEFAULT_BD_LIST_DOC.screenList;
-    const nodes = [
-        {
-            id: opts.docId,
-            type: "document",
-            output: path,
-            perDomain: opts.kind,
-            title: opts.title,
-            ...(opts.description ? { description: opts.description } : {}),
-        },
-    ];
-    const edges = [
-        { type: "rendersTo", from: opts.docId, to: path },
-        { type: "rendersTo", from: templateDocId, to: path },
-        { type: "partOf", from: opts.docId, to: templateDocId },
-        { type: "contains", from: listAnchorId, to: opts.docId },
-    ];
-    for (const featureId of opts.featureIds) {
-        edges.push({ type: "tracesTo", from: featureId, to: opts.docId });
-    }
-    const sections = parseDetailSections(opts.content, opts.docId);
-    const sectionPatch = detailSectionsToPatch(opts.docId, sections, opts.content);
-    nodes.push(...sectionPatch.nodes);
-    edges.push(...sectionPatch.edges);
-    return { version: 1, nodes, edges };
-}
-function primaryDomainIdFromDetailContent(content, kind) {
-    const root = parseMarkdown(content);
-    if (kind === "useCaseDetail") {
-        const idVal = extractBoldFieldDeep(root, /^Use Case ID$/i);
-        if (idVal) {
-            const m = idVal.match(/\b(UC-\d+)\b/i);
-            if (m?.[1] && isRealDomainId(m[1])) {
-                const nameVal = extractBoldFieldDeep(root, /^Use Case Name$/i);
-                return { id: normalizeDomainId(m[1]), title: nameVal?.trim() };
-            }
-        }
-        const hText = findHeadingText(root, /\bUC-\d+\b/);
-        if (hText) {
-            const hm = hText.match(/\b(UC-\d+)\b\s*[:\-–—]\s*(.+)$/i);
-            if (hm?.[1] && isRealDomainId(hm[1])) {
-                return { id: normalizeDomainId(hm[1]), title: hm[2]?.trim() };
-            }
-        }
-        return null;
-    }
-    const idVal = extractBoldFieldDeep(root, /^Feature ID$/i);
-    if (idVal) {
-        const m = idVal.match(/\b(F-\d+)\b/i);
-        if (m?.[1] && isRealDomainId(m[1])) {
-            const hText = findHeadingText(root, /\bF-\d+\b/);
-            let title;
-            if (hText) {
-                const hm = hText.match(/\b(F-\d+)\b\s*[:\-–—]\s*(.+)$/i);
-                title = hm?.[2]?.trim();
-            }
-            return { id: normalizeDomainId(m[1]), title };
-        }
-    }
-    const hText = findHeadingText(root, /\bF-\d+\b/);
-    if (hText) {
-        const hm = hText.match(/\b(F-\d+)\b\s*[:\-–—]\s*(.+)$/i);
-        if (hm?.[1] && isRealDomainId(hm[1])) {
-            return { id: normalizeDomainId(hm[1]), title: hm[2]?.trim() };
-        }
-    }
-    return null;
-}
-/** Rich title/description from generated detail markdown (not list-table stubs). */
-export function extractDetailDomainMeta(content, kind) {
-    const primary = primaryDomainIdFromDetailContent(content, kind);
-    if (!primary)
-        return null;
-    const root = parseMarkdown(content);
-    const meta = {};
-    if (primary.title)
-        meta.title = primary.title;
-    const nameVal = extractBoldFieldDeep(root, /^Use Case Name$/i);
-    if (nameVal?.trim())
-        meta.title = nameVal.trim();
-    if (kind === "useCaseDetail") {
-        const brief = extractBlockquoteAfterBoldField(root, /^Brief Description$/i);
-        if (brief)
-            meta.description = brief;
-        const priorityVal = extractBoldFieldDeep(root, /^Priority$/i);
-        if (priorityVal && /^(High|Medium|Low)$/i.test(priorityVal.trim())) {
-            meta.priority = priorityVal.trim();
-        }
-        if (!meta.description) {
-            const sections = parseDetailSections(content, "doc.temp");
-            const overview = sections.find((s) => OVERVIEW_HEADING.test(s.heading));
-            if (overview)
-                meta.description = snippetAfterHeading(content, overview.heading);
-        }
-    }
-    else {
-        const purpose = extractBlockquoteAfterBoldField(root, /^(?:Feature Purpose|Purpose)$/i);
-        if (purpose)
-            meta.description = purpose;
-        if (!meta.description) {
-            const sections = parseDetailSections(content, "doc.temp");
-            const descSec = sections.find((s) => FEATURE_DESC_HEADING.test(s.heading));
-            if (descSec)
-                meta.description = snippetAfterHeading(content, descSec.heading);
-        }
-    }
-    return meta;
-}
-function listedInForPath(relativePath, kind) {
-    const p = relativePath.replace(/\\/g, "/").toLowerCase();
-    if (kind === "useCase") {
-        if (p.includes("use-case") || p.includes("use-cases") || /\/3-/.test(p)) {
-            return DEFAULT_LISTED_IN.useCase;
-        }
-    }
-    if (kind === "feature") {
-        if (p.includes("basic-design") ||
-            p.includes("system-features") ||
-            p.includes("04-system-features") ||
-            /\/4-/.test(p)) {
-            return DEFAULT_LISTED_IN.feature;
-        }
-    }
-    if (kind === "actor") {
-        if (p.includes("overall-description") || /\/2-/.test(p)) {
-            return DEFAULT_LISTED_IN.actor;
-        }
-    }
-    return DEFAULT_LISTED_IN[kind];
-}
-function emptyParsed() {
-    return {
-        useCases: new Map(),
-        features: new Map(),
-        actors: new Map(),
-        satisfies: [],
-    };
-}
-function addUseCase(parsed, id, fields) {
-    const norm = normalizeDomainId(id);
-    const prev = parsed.useCases.get(norm) ?? {};
-    const next = { ...prev };
-    if (fields?.title?.trim() && !prev.title) {
-        next.title = fields.title.trim();
-    }
-    if (fields?.description?.trim() && !prev.description) {
-        next.description = fields.description.trim();
-    }
-    if (fields?.priority?.trim() && !prev.priority) {
-        next.priority = fields.priority.trim();
-    }
-    parsed.useCases.set(norm, next);
-}
-function addFeature(parsed, id, fields) {
-    const norm = normalizeDomainId(id);
-    const prev = parsed.features.get(norm) ?? {};
-    const next = { ...prev };
-    if (fields?.title?.trim() && !prev.title) {
-        next.title = fields.title.trim();
-    }
-    if (fields?.description?.trim() && !prev.description) {
-        next.description = fields.description.trim();
-    }
-    parsed.features.set(norm, next);
-}
-function parseTableSatisfies(line, parsed) {
-    const cells = line
-        .split("|")
-        .map((c) => c.trim())
-        .filter(Boolean);
-    if (cells.length < 2) {
-        return;
-    }
-    const fMatch = cells[0]?.match(/^F-(\d+)$/i);
-    if (!fMatch) {
-        return;
-    }
-    const from = normalizeDomainId(cells[0]);
-    for (const cell of cells.slice(1)) {
-        for (const m of cell.matchAll(UC_ID_RE)) {
-            parsed.satisfies.push({ from, to: normalizeDomainId(m[1]) });
-        }
-    }
-}
-export function extractDomainFromMarkdown(content, relativePath) {
-    const parsed = emptyParsed();
-    const root = parseMarkdown(content);
-    // --- 1. Headings: ## UC-01: Title, ## F-01: Title ---
-    for (const node of root.children) {
-        if (node.type !== "heading")
-            continue;
-        const text = textContent(node);
-        const ucm = text.match(/\b(UC-\d+)\b\s*[:\-–—]\s*(.+)$/i);
-        if (ucm?.[1] && isRealDomainId(ucm[1])) {
-            addUseCase(parsed, ucm[1], { title: ucm[2]?.trim() });
-        }
-        const fm = text.match(/\b(F-\d+)\b\s*[:\-–—]\s*(.+)$/i);
-        if (fm?.[1] && isRealDomainId(fm[1])) {
-            addFeature(parsed, fm[1], { title: fm[2]?.trim() });
-        }
-    }
-    // --- 2. Bold field patterns in paragraphs (AST-based) ---
-    let pendingUcId;
-    for (const node of root.children) {
-        if (node.type === "heading" && pendingUcId) {
-            pendingUcId = undefined;
-        }
-        if (node.type !== "paragraph")
-            continue;
-        const para = node;
-        const ucIdVal = extractBoldFieldInParagraph(para, /^Use Case ID$/i);
-        if (ucIdVal) {
-            const m = ucIdVal.match(/\b(UC-\d+)\b/i);
-            if (m?.[1] && isRealDomainId(m[1])) {
-                pendingUcId = normalizeDomainId(m[1]);
-                addUseCase(parsed, pendingUcId, {});
-            }
-        }
-        if (pendingUcId) {
-            const nameVal = extractBoldFieldInParagraph(para, /^Use Case Name$/i);
-            if (nameVal) {
-                addUseCase(parsed, pendingUcId, { title: nameVal });
-                pendingUcId = undefined;
-            }
-        }
-        const fIdVal = extractBoldFieldInParagraph(para, /^Feature ID$/i);
-        if (fIdVal) {
-            const m = fIdVal.match(/\b(F-\d+)\b/i);
-            if (m?.[1] && isRealDomainId(m[1]))
-                addFeature(parsed, m[1], {});
-        }
-    }
-    // --- 3. Actor extraction from paragraphs and list items (AST-based) ---
-    visit(root, "paragraph", (para) => {
-        const actorVal = extractBoldFieldInParagraph(para, /^Primary Actor$/i);
-        if (actorVal && !/^tbd$/i.test(actorVal) && actorVal.length > 1) {
-            const id = `actor.${slugActorId(actorVal)}`;
-            parsed.actors.set(id, { title: actorVal });
-        }
-    });
-    // --- 4. Table rows (line-based: handles GFM and non-GFM pipe tables) ---
-    for (const line of content.split(/\r?\n/)) {
-        const ucRow = line.match(TABLE_ROW_UC);
-        if (ucRow?.[1]) {
-            const cells = line.split("|").map((c) => c.trim());
-            const title = cells[2] || cells[1];
-            addUseCase(parsed, ucRow[1], {
-                title: title && !/^UC-/i.test(title) ? title : undefined,
-            });
-        }
-        const fRow = line.match(TABLE_ROW_F);
-        if (fRow?.[1]) {
-            const cells = line.split("|").map((c) => c.trim());
-            const title = cells[2] || cells[1];
-            addFeature(parsed, fRow[1], {
-                title: title && !/^F-/i.test(title) ? title : undefined,
-            });
-            parseTableSatisfies(line, parsed);
-        }
-        else if (/^\|.*F-\d+/i.test(line) && /UC-/i.test(line)) {
-            parseTableSatisfies(line, parsed);
-        }
-    }
-    // --- 5. Global catch-all scan (picks up any remaining UC/F references) ---
-    for (const m of content.matchAll(UC_ID_RE)) {
-        if (m[1] && isRealDomainId(m[1]))
-            addUseCase(parsed, m[1], {});
-    }
-    for (const m of content.matchAll(F_ID_RE)) {
-        if (m[1] && isRealDomainId(m[1]))
-            addFeature(parsed, m[1], {});
-    }
-    // --- 6. Rich metadata from SRS detail files ---
-    const srsDetailKind = classifySrsDetailFile(relativePath);
-    if (srsDetailKind) {
-        const rich = extractDetailDomainMeta(content, srsDetailKind);
-        if (rich) {
-            const primary = primaryDomainIdFromDetailContent(content, srsDetailKind);
-            if (primary) {
-                if (srsDetailKind === "useCaseDetail") {
-                    addUseCase(parsed, primary.id, rich);
-                }
-                else {
-                    addFeature(parsed, primary.id, rich);
-                }
-            }
-        }
-    }
-    // --- 7. Basic-design detail metadata ---
-    const bdKind = classifyBasicDesignDetailFile(relativePath);
-    if (bdKind) {
-        const meta = extractBasicDesignDetailMeta(content, bdKind, relativePath);
-        for (const featureId of meta.featureIds) {
-            addFeature(parsed, featureId, {
-                description: extractBasicDesignDetailDescription(content, bdKind),
-            });
-        }
-    }
-    return parsed;
-}
-function extractBasicDesignDetailDescription(content, kind) {
-    if (kind === "apiDetail") {
-        const meta = extractDetailDomainMeta(content, "featureDetail");
-        return meta?.description;
-    }
-    const root = parseMarkdown(content);
-    const purpose = extractBlockquoteAfterBoldField(root, /^Purpose$/i);
-    if (purpose)
-        return purpose;
-    const sections = parseDetailSections(content, "doc.temp");
-    const screenHeading = sections.find((s) => /^##\s*1\.\s*screen/i.test(s.heading));
-    if (screenHeading) {
-        return snippetAfterHeading(content, screenHeading.heading);
-    }
-    return undefined;
-}
-function slugActorId(name) {
-    return name
-        .toLowerCase()
-        .replace(/[^a-z0-9]+/g, "-")
-        .replace(/^-|-$/g, "")
-        .slice(0, 48) || "unknown";
-}
-export function parsedDomainToPatch(parsed, relativePath) {
-    const nodes = [];
-    const edges = [];
-    for (const [id, meta] of parsed.useCases) {
-        const node = { id, type: "useCase", title: meta.title ?? id };
-        if (meta.description) {
-            node.description = meta.description;
-        }
-        if (meta.priority) {
-            node.priority = meta.priority;
-        }
-        nodes.push(node);
-        edges.push({
-            type: "listedIn",
-            from: id,
-            to: listedInForPath(relativePath, "useCase"),
-        });
-    }
-    for (const [id, meta] of parsed.features) {
-        const node = { id, type: "feature", title: meta.title ?? id };
-        if (meta.description) {
-            node.description = meta.description;
-        }
-        nodes.push(node);
-        edges.push({
-            type: "listedIn",
-            from: id,
-            to: listedInForPath(relativePath, "feature"),
-        });
-    }
-    for (const [id, meta] of parsed.actors) {
-        nodes.push({
-            id,
-            type: "actor",
-            title: meta.title ?? id,
-            name: meta.title,
-        });
-        edges.push({
-            type: "describedIn",
-            from: id,
-            to: listedInForPath(relativePath, "actor"),
-        });
-    }
-    const seenSatisfies = new Set();
-    for (const { from, to } of parsed.satisfies) {
-        const key = `${from}->${to}`;
-        if (seenSatisfies.has(key)) {
-            continue;
-        }
-        seenSatisfies.add(key);
-        if (parsed.features.has(from) && parsed.useCases.has(to)) {
-            edges.push({ type: "satisfies", from, to });
-        }
-    }
-    return { version: 1, nodes, edges };
-}
-function buildDetailInstancePatch(opts) {
-    const path = normalizeRelativePath(opts.relativePath);
-    const nodes = [
-        {
-            id: opts.docId,
-            type: "document",
-            output: path,
-            perDomain: opts.perDomain,
-            title: opts.title ?? opts.domainId,
-            ...(opts.description ? { description: opts.description } : {}),
-        },
-        {
-            id: opts.domainId,
-            type: opts.domainType,
-            title: opts.title ?? opts.domainId,
-            ...(opts.description ? { description: opts.description } : {}),
-            ...(opts.priority ? { priority: opts.priority } : {}),
-        },
-    ];
-    const edges = [
-        { type: "rendersTo", from: opts.domainId, to: path },
-        { type: "rendersTo", from: opts.templateDocId, to: path },
-        { type: "partOf", from: opts.docId, to: opts.templateDocId },
-        { type: "describedIn", from: opts.domainId, to: opts.docId },
-        { type: "contains", from: opts.listAnchorId, to: opts.docId },
-        { type: "tracesTo", from: opts.domainId, to: opts.docId },
-    ];
-    const sections = parseDetailSections(opts.content, opts.docId);
-    const sectionPatch = detailSectionsToPatch(opts.docId, sections, opts.content);
-    nodes.push(...sectionPatch.nodes);
-    edges.push(...sectionPatch.edges);
-    // Anchor to the primary (first) section only — one definedIn is enough for validation
-    // and avoids flooding the graph with edges to every section.
-    edges.push({
-        type: "definedIn",
-        from: opts.domainId,
-        to: sections[0]?.id ?? opts.docId,
-    });
-    return { version: 1, nodes, edges };
-}
-function srsDetailFileToPatch(relativePath, content, packConfig) {
-    const kind = classifySrsDetailFile(relativePath);
-    if (!kind) {
-        return { version: 1, nodes: [], edges: [] };
-    }
-    const primary = primaryDomainIdFromDetailContent(content, kind);
-    if (!primary) {
-        return { version: 1, nodes: [], edges: [] };
-    }
-    const perDomain = kind === "useCaseDetail" ? "useCase" : "feature";
-    const rich = extractDetailDomainMeta(content, kind);
-    const nodePrefix = packConfig?.nodePrefix;
-    const templateDocId = packConfig?.perDomainTemplates?.[perDomain] ??
-        BUILTIN_PER_DOMAIN_TEMPLATE_DOC[perDomain] ??
-        `doc.srs.${perDomain === "useCase" ? "uc-detail" : "feature-detail"}`;
-    const listAnchorId = packConfig?.defaultListedIn?.[perDomain] ??
-        (perDomain === "useCase" ? DEFAULT_LISTED_IN.useCase : DEFAULT_LISTED_IN.feature);
-    return buildDetailInstancePatch({
-        relativePath,
-        content,
-        docId: documentIdForDomainDetail(kind, primary.id, nodePrefix),
-        domainId: primary.id,
-        domainType: perDomain === "useCase" ? "useCase" : "feature",
-        perDomain,
-        templateDocId,
-        listAnchorId,
-        title: rich?.title ?? primary.title,
-        description: rich?.description,
-        priority: rich?.priority,
-    });
-}
-function basicDesignDetailFileToPatch(relativePath, content) {
-    const kind = classifyBasicDesignDetailFile(relativePath);
-    if (!kind) {
-        return { version: 1, nodes: [], edges: [] };
-    }
-    const meta = extractBasicDesignDetailMeta(content, kind, relativePath);
-    return buildBasicDesignDetailInstancePatch({
-        relativePath,
-        content,
-        docId: documentIdForBasicDesignDetail(kind, relativePath),
-        kind,
-        title: meta.title,
-        description: extractBasicDesignDetailDescription(content, kind),
-        featureIds: meta.featureIds,
-    });
-}
-/** Document nodes + sections for per-UC/F SRS or per-endpoint/per-screen basic-design files. */
-export function detailFileToPatch(relativePath, content, options) {
-    if (options?.includeListChapterMarkdown !== false) {
-        const listChapter = basicDesignListChapterFileToPatch(relativePath, content);
-        if (listChapter.nodes.length > 0 || listChapter.edges.length > 0) {
-            return listChapter;
-        }
-    }
-    else if (basicDesignListChapterDocumentId(relativePath)) {
-        return { version: 1, nodes: [], edges: [] };
-    }
-    const srs = srsDetailFileToPatch(relativePath, content, options?.packConfig);
-    if (srs.nodes.length > 0 || srs.edges.length > 0) {
-        return srs;
-    }
-    return basicDesignDetailFileToPatch(relativePath, content);
-}
-export function mergeParsedDomains(domains) {
-    const out = emptyParsed();
-    for (const d of domains) {
-        for (const [id, meta] of d.useCases) {
-            addUseCase(out, id, meta);
-        }
-        for (const [id, meta] of d.features) {
-            addFeature(out, id, meta);
-        }
-        for (const [id, meta] of d.actors) {
-            out.actors.set(id, meta);
-        }
-        out.satisfies.push(...d.satisfies);
-    }
-    return out;
-}
-/** Per-domain template shells (outputPattern — no section coverage required). */
-export function basicDesignPerDomainTemplateNodes() {
-    return [
-        {
-            id: PER_DOMAIN_TEMPLATE_DOC_BD.api,
-            type: "document",
-            outputPattern: "docs/basic-design/api/",
-            perDomain: "apiDetail",
-            title: "API Detail (template)",
-        },
-        {
-            id: PER_DOMAIN_TEMPLATE_DOC_BD.screen,
-            type: "document",
-            outputPattern: "docs/basic-design/screens/",
-            perDomain: "screenDetail",
-            title: "Screen Detail (template)",
-        },
-    ];
-}
-/** @deprecated Prefer {@link basicDesignAnchorStructurePatch} — documents only, no sections. */
-export function basicDesignAnchorDocumentNodes() {
-    return [
-        {
-            id: DEFAULT_BD_LIST_DOC.apiList,
-            type: "document",
-            output: "docs/basic-design/api-list.md",
-            title: "API List",
-        },
-        {
-            id: DEFAULT_BD_LIST_DOC.screenList,
-            type: "document",
-            output: "docs/basic-design/list-screens.md",
-            title: "Screen Map",
-        },
-        {
-            id: DEFAULT_BD_LIST_DOC.dbDesign,
-            type: "document",
-            output: "docs/basic-design/db-design.md",
-            title: "Database Design",
-        },
-    ];
-}
-/** True when any basic-design list chapter document has no child sections in the graph. */
-export function basicDesignListChaptersNeedSections(graph) {
-    for (const docId of BASIC_DESIGN_LIST_DOCUMENT_IDS) {
-        const doc = graph.nodesById.get(docId);
-        if (!doc || doc.type !== "document") {
-            continue;
-        }
-        const hasSection = (graph.outEdges.get(docId) ?? []).some((e) => e.type === "contains" &&
-            graph.nodesById.get(e.to)?.type === "section");
-        if (!hasSection) {
-            return true;
-        }
-    }
-    return false;
-}
-/** List chapters + section trees from templates; used when bootstrap omitted basic-design registry docs. */
-export async function basicDesignAnchorStructurePatch(projectRoot) {
-    try {
-        const listDocs = await scanBasicDesignListDocuments(projectRoot);
-        const listPatches = listDocs.map((d) => registryDocumentToStructurePatch(d));
-        const templatePatch = {
-            version: 1,
-            nodes: basicDesignPerDomainTemplateNodes(),
-            edges: [],
-        };
-        return mergeStructurePatches([...listPatches, templatePatch]);
-    }
-    catch {
-        return {
-            version: 1,
-            nodes: [
-                ...basicDesignAnchorDocumentNodes(),
-                ...basicDesignPerDomainTemplateNodes(),
-            ],
-            edges: [],
-        };
-    }
-}
-export async function buildDocExtractPatch(entries, _projectRoot, options) {
-    const includeListChapterMarkdown = options?.includeListChapterMarkdown !== false;
-    const perFile = [];
-    for (const e of entries) {
-        perFile.push(extractDomainFromMarkdown(e.content, e.relativePath));
-    }
-    const merged = mergeParsedDomains(perFile);
-    const allNodesMap = new Map();
-    const allEdges = [];
-    const edgeKeys = new Set();
-    let detailDocuments = 0;
-    let detailSections = 0;
-    let srsDetailDocuments = 0;
-    let bdDetailDocuments = 0;
-    const mergePatchInto = (patch) => {
-        for (const n of patch.nodes) {
-            const existing = allNodesMap.get(n.id);
-            if (!existing) {
-                allNodesMap.set(n.id, n);
-                if (n.type === "document" && n.output && !n.outputPattern) {
-                    const isBdInstance = n.perDomain === "apiDetail" || n.perDomain === "screenDetail";
-                    const isSrsInstance = n.perDomain === "useCase" || n.perDomain === "feature";
-                    if (isBdInstance) {
-                        detailDocuments++;
-                        bdDetailDocuments++;
-                    }
-                    else if (isSrsInstance) {
-                        detailDocuments++;
-                        srsDetailDocuments++;
-                    }
-                }
-                if (n.type === "section" &&
-                    typeof n.documentId === "string" &&
-                    (n.documentId.startsWith("doc.srs.") ||
-                        n.documentId.startsWith("doc.bd.") ||
-                        allNodesMap.get(n.documentId)?.perDomain != null)) {
-                    detailSections++;
-                }
-            }
-            else {
-                const merged = { ...existing, ...n };
-                if (!n.title && existing.title) {
-                    merged.title = existing.title;
-                }
-                if (!n.description && existing.description) {
-                    merged.description = existing.description;
-                }
-                if (n.type === "document" && !n.output && existing.output) {
-                    merged.output = existing.output;
-                }
-                allNodesMap.set(n.id, merged);
-            }
-        }
-        for (const edge of patch.edges) {
-            const key = `${edge.type}:${edge.from}:${edge.to}`;
-            if (!edgeKeys.has(key)) {
-                edgeKeys.add(key);
-                allEdges.push(edge);
-            }
-        }
-    };
-    const hasBasicDesignFiles = entries.some((e) => normalizeRelativePath(e.relativePath).toLowerCase().startsWith("docs/basic-design/"));
-    if (hasBasicDesignFiles) {
-        mergePatchInto({
-            version: 1,
-            nodes: [
-                ...basicDesignAnchorDocumentNodes(),
-                ...basicDesignPerDomainTemplateNodes(),
-            ],
-            edges: [],
-        });
-    }
-    for (const e of entries) {
-        const single = extractDomainFromMarkdown(e.content, e.relativePath);
-        mergePatchInto(parsedDomainToPatch(single, e.relativePath));
-        mergePatchInto(detailFileToPatch(e.relativePath, e.content, {
-            includeListChapterMarkdown,
-        }));
-    }
-    return {
-        patch: { version: 1, nodes: [...allNodesMap.values()], edges: allEdges },
-        stats: {
-            useCases: merged.useCases.size,
-            features: merged.features.size,
-            actors: merged.actors.size,
-            detailDocuments,
-            detailSections,
-            srsDetailDocuments,
-            bdDetailDocuments,
-            filesScanned: entries.length,
-        },
-    };
-}
-//# sourceMappingURL=doc-extract.js.map