buildanything 2.0.0 → 2.1.2

package/src/graph/parser/page-spec.ts

@@ -0,0 +1,496 @@
+// Deterministic, schema-aware extractor for page-spec markdown files.
+// Source of truth: docs/graph/07-slice3-schema.md.
+// No LLM, no I/O. Caller passes the markdown content and path.
+
+import { ids, kebab, sha256Hex } from "../ids.js";
+import type {
+  ExtractError,
+  ExtractResult,
+  GraphEdge,
+  GraphFragment,
+  GraphNode,
+  KeyCopyNode,
+  PageSpecNode,
+  Relation,
+  ScreenComponentUseNode,
+  ScreenStateSlotNode,
+  WireframeSectionNode,
+} from "../types.js";
+
+const PRODUCED_BY = "design-ux-architect";
+const PRODUCED_AT_STEP = "3.3";
+
+interface Line {
+  n: number;
+  text: string;
+}
+
+interface Section {
+  heading: string;
+  level: number;
+  startLine: number;
+  bodyLines: Line[];
+}
+
+interface Ctx {
+  mdPath: string;
+  errors: ExtractError[];
+  nodes: GraphNode[];
+  edges: GraphEdge[];
+}
+
+function loc(line: number): string {
+  return `L${line}`;
+}
+
+function pushError(ctx: Ctx, line: number, message: string): void {
+  ctx.errors.push({ line, message });
+}
+
+function splitLines(content: string): Line[] {
+  return content.split(/\r?\n/).map((text, i) => ({ n: i + 1, text }));
+}
+
+function isHeadingAtLevel(text: string, level: number): boolean {
+  const prefix = "#".repeat(level) + " ";
+  if (!text.startsWith(prefix)) return false;
+  return text[level] !== "#";
+}
+
+function isHeadingAtOrAbove(text: string, level: number): boolean {
+  for (let l = 1; l <= level; l++) {
+    if (isHeadingAtLevel(text, l)) return true;
+  }
+  return false;
+}
+
+function partitionSections(lines: Line[], level: number, start: number, end: number): Section[] {
+  const sections: Section[] = [];
+  const headingPrefix = "#".repeat(level) + " ";
+  let i = start;
+  while (i < end) {
+    const line = lines[i];
+    if (isHeadingAtLevel(line.text, level)) {
+      const heading = line.text.slice(headingPrefix.length).trim();
+      const bodyStart = i + 1;
+      let j = bodyStart;
+      while (j < end && !isHeadingAtOrAbove(lines[j].text, level)) j++;
+      sections.push({ heading, level, startLine: line.n, bodyLines: lines.slice(bodyStart, j) });
+      i = j;
+    } else {
+      i++;
+    }
+  }
+  return sections;
+}
+
+interface TableRow {
+  cells: Record<string, string>;
+  line: number;
+}
+
+function parseTable(body: Line[]): { headers: string[]; rows: TableRow[] } | null {
+  const sepRe = /^\s*\|?\s*[-:| ]+\s*\|?\s*$/;
+  const significant = body.filter((l) => l.text.trim().length > 0);
+  let headerIdx = -1;
+  for (let i = 0; i < significant.length - 1; i++) {
+    if (
+      significant[i].text.includes("|") &&
+      significant[i + 1].text.includes("|") &&
+      sepRe.test(significant[i + 1].text)
+    ) {
+      headerIdx = i;
+      break;
+    }
+  }
+  if (headerIdx < 0) return null;
+  const headers = splitRow(significant[headerIdx].text).map((h) => h.toLowerCase());
+  const rows: TableRow[] = [];
+  for (let i = headerIdx + 2; i < significant.length; i++) {
+    const ln = significant[i];
+    if (!ln.text.includes("|")) break;
+    const cells = splitRow(ln.text);
+    if (cells.length === 0) continue;
+    const row: Record<string, string> = {};
+    for (let c = 0; c < headers.length; c++) {
+      row[headers[c]] = (cells[c] ?? "").trim();
+    }
+    rows.push({ cells: row, line: ln.n });
+  }
+  return { headers, rows };
+}
+
+function splitRow(text: string): string[] {
+  let s = text.trim();
+  if (s.startsWith("|")) s = s.slice(1);
+  if (s.endsWith("|")) s = s.slice(0, -1);
+  return s.split("|").map((c) => c.trim());
+}
+
+function makeEdge(
+  ctx: Ctx,
+  source: string,
+  target: string,
+  relation: Relation,
+  line: number,
+): GraphEdge {
+  return {
+    source,
+    target,
+    relation,
+    confidence: "EXTRACTED",
+    source_file: ctx.mdPath,
+    source_location: loc(line),
+    produced_by_agent: PRODUCED_BY,
+    produced_at_step: PRODUCED_AT_STEP,
+  };
+}
+
+function truncLabel(text: string): string {
+  return text.length > 80 ? text.slice(0, 77) + "..." : text;
+}
+
+function sortNodes(nodes: GraphNode[]): GraphNode[] {
+  return [...nodes].sort((a, b) => (a.id < b.id ? -1 : a.id > b.id ? 1 : 0));
+}
+
+function sortEdges(edges: GraphEdge[]): GraphEdge[] {
+  return [...edges].sort((a, b) => {
+    const k = (e: GraphEdge): string =>
+      `${e.relation} ${e.source} ${e.target} ${e.source_location ?? ""}`;
+    return k(a) < k(b) ? -1 : k(a) > k(b) ? 1 : 0;
+  });
+}
+
+// ---------------------------------------------------------------------------
+// Section parsers
+// ---------------------------------------------------------------------------
+
+function findH2(sections: Section[], name: string): Section | undefined {
+  return sections.find((s) => s.heading.trim().toLowerCase() === name.toLowerCase());
+}
+
+function parseRoute(section: Section | undefined): string | null {
+  if (!section) return null;
+  for (const line of section.bodyLines) {
+    const t = line.text.trim();
+    if (!t) continue;
+    const stripped = t.replace(/^`+|`+$/g, "").trim();
+    if (!stripped || /^n\/a$/i.test(stripped) || /^modal$/i.test(stripped)) return null;
+    return stripped;
+  }
+  return null;
+}
+
+function parseWireframe(
+  ctx: Ctx,
+  section: Section,
+  screenName: string,
+  pageSpecId: string,
+): string | null {
+  // Find fenced code block in bodyLines (including sub-headings like ### Desktop)
+  let inFence = false;
+  let fenceContent: string[] = [];
+  let foundFence = false;
+  for (const line of section.bodyLines) {
+    if (!inFence && /^\s*```/.test(line.text)) {
+      inFence = true;
+      foundFence = true;
+      continue;
+    }
+    if (inFence && /^\s*```/.test(line.text)) {
+      inFence = false;
+      break;
+    }
+    if (inFence) {
+      fenceContent.push(line.text);
+    }
+  }
+  if (!foundFence) {
+    pushError(ctx, section.startLine, `## ASCII Wireframe section has no fenced code block`);
+    return null;
+  }
+  const wireframeText = fenceContent.join("\n");
+
+  // Scan for [SectionName] markers
+  const markerRe = /\[([A-Z][A-Za-z0-9 :_-]*)\]/g;
+  const seen = new Set<string>();
+  const ordered: string[] = [];
+  let m: RegExpExecArray | null;
+  while ((m = markerRe.exec(wireframeText)) !== null) {
+    const name = m[1].trim();
+    if (!seen.has(name)) {
+      seen.add(name);
+      ordered.push(name);
+    }
+  }
+
+  for (let i = 0; i < ordered.length; i++) {
+    const sectionName = ordered[i];
+    const node: WireframeSectionNode = {
+      id: ids.wireframeSection(screenName, sectionName, i),
+      label: sectionName,
+      entity_type: "wireframe_section",
+      source_file: ctx.mdPath,
+      source_location: loc(section.startLine),
+      confidence: "EXTRACTED",
+      section_name: sectionName,
+      parent_page_spec_id: pageSpecId,
+      order: i,
+      prose: "",
+    };
+    ctx.nodes.push(node);
+    ctx.edges.push(makeEdge(ctx, pageSpecId, node.id, "has_section", section.startLine));
+  }
+
+  return wireframeText;
+}
+
+function parseContentHierarchy(ctx: Ctx, section: Section): string[] | null {
+  // Try table first
+  const table = parseTable(section.bodyLines);
+  if (table) {
+    const sectionCol = table.headers.find((h) => h === "section" || h === "section name");
+    if (sectionCol && table.rows.length > 0) {
+      const entries = table.rows.map((r) => r.cells[sectionCol]!.trim()).filter((s) => s);
+      if (entries.length > 0) return entries;
+    }
+  }
+  // Try bulleted/numbered list
+  const listRe = /^\s*(?:\d+\.|[-*])\s+(.+)$/;
+  const entries: string[] = [];
+  for (const line of section.bodyLines) {
+    const m = line.text.match(listRe);
+    if (m) entries.push(m[1].trim());
+  }
+  if (entries.length > 0) return entries;
+  pushError(ctx, section.startLine, `## Content Hierarchy is empty`);
+  return null;
+}
+
+function parseStates(
+  ctx: Ctx,
+  section: Section,
+  screenName: string,
+  screenId: string,
+  pageSpecId: string,
+): void {
+  // Try table
+  const table = parseTable(section.bodyLines);
+  if (table && table.headers.includes("state") && table.headers.includes("appearance")) {
+    for (const row of table.rows) {
+      const stateName = (row.cells["state"] ?? "").trim();
+      const appearance = (row.cells["appearance"] ?? "").trim();
+      if (!stateName) continue;
+      emitStateSlot(ctx, screenName, screenId, pageSpecId, stateName, appearance, row.line);
+    }
+    return;
+  }
+  // Try bullets
+  const bulletRe = /^\s*-\s+(?:\*\*(.+?)\*\*\s*(?:—|--|-|:)?\s*(.*)$|(.+?):\s+(.*)$)/;
+  for (const line of section.bodyLines) {
+    const m = line.text.match(bulletRe);
+    if (!m) continue;
+    const stateName = (m[1] ?? m[3] ?? "").trim();
+    const appearance = (m[2] ?? m[4] ?? "").trim();
+    if (!stateName) continue;
+    emitStateSlot(ctx, screenName, screenId, pageSpecId, stateName, appearance, line.n);
+  }
+}
+
+function emitStateSlot(
+  ctx: Ctx,
+  screenName: string,
+  screenId: string,
+  pageSpecId: string,
+  stateName: string,
+  appearance: string,
+  line: number,
+): void {
+  const node: ScreenStateSlotNode = {
+    id: ids.screenStateSlot(screenName, stateName),
+    label: stateName,
+    entity_type: "screen_state_slot",
+    source_file: ctx.mdPath,
+    source_location: loc(line),
+    confidence: "EXTRACTED",
+    screen_id: screenId,
+    state_id: kebab(stateName),
+    appearance_text: appearance,
+  };
+  ctx.nodes.push(node);
+  ctx.edges.push(makeEdge(ctx, pageSpecId, node.id, "has_screen_state", line));
+}
+
+function parseKeyCopy(
+  ctx: Ctx,
+  section: Section,
+  screenName: string,
+  screenId: string,
+  pageSpecId: string,
+): boolean {
+  // Regex: capture text between first and last quote (straight or smart) on the line
+  const quoteRe = /^\s*-\s+(?:\*\*)?["\u201C\u2018](.+)["\u201D\u2019](?:\*\*)?\s*(?:\u2014|--|(?:-\s))\s*(.*)$/;
+  let count = 0;
+  for (const line of section.bodyLines) {
+    const m = line.text.match(quoteRe);
+    if (!m) continue;
+    const text = m[1].trim();
+    let placement = m[2].trim();
+    // Strip leading 'placement:' prefix
+    placement = placement.replace(/^placement:\s*/i, "").trim();
+    const node: KeyCopyNode = {
+      id: ids.keyCopy(screenName, text),
+      label: truncLabel(text),
+      entity_type: "key_copy",
+      source_file: ctx.mdPath,
+      source_location: loc(line.n),
+      confidence: "EXTRACTED",
+      screen_id: screenId,
+      text,
+      placement,
+    };
+    ctx.nodes.push(node);
+    ctx.edges.push(makeEdge(ctx, pageSpecId, node.id, "key_copy_on_screen", line.n));
+    count++;
+  }
+  if (count === 0) {
+    pushError(ctx, section.startLine, `## Key Copy yielded no parsed bullets`);
+    return false;
+  }
+  return true;
+}
+
+function parseComponentPicks(
+  ctx: Ctx,
+  section: Section,
+  screenName: string,
+  screenId: string,
+  pageSpecId: string,
+): void {
+  const table = parseTable(section.bodyLines);
+  if (!table) return;
+  const slotCol = table.headers.find((h) => h === "manifest slot" || h === "slot");
+  const sectionCol = table.headers.find((h) => h === "section" || h === "section name");
+  if (!slotCol || !sectionCol) return;
+  const propsCol = table.headers.find((h) => h === "prop overrides" || h === "props" || h === "overrides");
+  for (const row of table.rows) {
+    const sectionName = (row.cells[sectionCol] ?? "").trim();
+    let rawSlot = (row.cells[slotCol] ?? "").trim();
+    // Strip surrounding backticks
+    rawSlot = rawSlot.replace(/^`+|`+$/g, "");
+    // Strip trailing italics parenthetical: *(...)* 
+    rawSlot = rawSlot.replace(/\s*\*\([^)]*\)\*\s*$/, "").trim();
+    // Strip trailing plain parenthetical too
+    rawSlot = rawSlot.replace(/\s*\([^)]*\)\s*$/, "").trim();
+    const slot = kebab(rawSlot);
+    if (!slot || !sectionName) continue;
+    const propOverrides = propsCol ? (row.cells[propsCol] ?? "").trim() : "";
+    const node: ScreenComponentUseNode = {
+      id: ids.screenComponentUse(screenName, slot, sectionName),
+      label: `${slot} @ ${sectionName}`,
+      entity_type: "screen_component_use",
+      source_file: ctx.mdPath,
+      source_location: loc(row.line),
+      confidence: "EXTRACTED",
+      screen_id: screenId,
+      slot,
+      position_in_wireframe: sectionName,
+      prop_overrides: propOverrides,
+    };
+    ctx.nodes.push(node);
+    ctx.edges.push(makeEdge(ctx, pageSpecId, node.id, "slot_used_on_screen", row.line));
+  }
+}
+
+// ---------------------------------------------------------------------------
+// Public entrypoint
+// ---------------------------------------------------------------------------
+
+export function extractPageSpec(input: { mdPath: string; mdContent: string }): ExtractResult {
+  const { mdPath, mdContent } = input;
+  const lines = splitLines(mdContent);
+  const ctx: Ctx = { mdPath, errors: [], nodes: [], edges: [] };
+
+  // Parse h1: # Page: <Screen Name>
+  const h1 = lines.find((l) => isHeadingAtLevel(l.text, 1));
+  const h1Match = h1?.text.match(/^#\s+Page:\s+(.+)$/);
+  if (!h1 || !h1Match) {
+    pushError(ctx, h1?.n ?? 1, "Missing required h1: '# Page: <Screen Name>'");
+    return { ok: false, errors: ctx.errors };
+  }
+  const screenName = h1Match[1].trim();
+  const screenId = ids.screen(screenName);
+  const pageSpecId = ids.pageSpec(screenName);
+
+  const h2Sections = partitionSections(lines, 2, 0, lines.length);
+
+  // Route (optional)
+  const route = parseRoute(findH2(h2Sections, "Route"));
+
+  // ASCII Wireframe (required)
+  const wireframeSec = findH2(h2Sections, "ASCII Wireframe");
+  if (!wireframeSec) {
+    pushError(ctx, 1, "Missing required section: '## ASCII Wireframe'");
+    return { ok: false, errors: ctx.errors };
+  }
+  const wireframeText = parseWireframe(ctx, wireframeSec, screenName, pageSpecId);
+  if (wireframeText === null) return { ok: false, errors: ctx.errors };
+
+  // Content Hierarchy (required)
+  const hierarchySec = findH2(h2Sections, "Content Hierarchy");
+  if (!hierarchySec) {
+    pushError(ctx, 1, "Missing required section: '## Content Hierarchy'");
+    return { ok: false, errors: ctx.errors };
+  }
+  const contentHierarchy = parseContentHierarchy(ctx, hierarchySec);
+  if (!contentHierarchy) return { ok: false, errors: ctx.errors };
+
+  // States (optional)
+  const statesSec = findH2(h2Sections, "States");
+  if (statesSec) parseStates(ctx, statesSec, screenName, screenId, pageSpecId);
+
+  // Key Copy (required)
+  const keyCopySec = findH2(h2Sections, "Key Copy");
+  if (!keyCopySec) {
+    pushError(ctx, 1, "Missing required section: '## Key Copy'");
+    return { ok: false, errors: ctx.errors };
+  }
+  if (!parseKeyCopy(ctx, keyCopySec, screenName, screenId, pageSpecId)) {
+    return { ok: false, errors: ctx.errors };
+  }
+
+  // Component Picks (optional)
+  const compSec = findH2(h2Sections, "Component Picks");
+  if (compSec) parseComponentPicks(ctx, compSec, screenName, screenId, pageSpecId);
+
+  // PageSpecNode
+  const pageNode: PageSpecNode = {
+    id: pageSpecId,
+    label: screenName,
+    entity_type: "page_spec",
+    source_file: mdPath,
+    source_location: loc(h1.n),
+    confidence: "EXTRACTED",
+    screen_id: screenId,
+    wireframe_text: wireframeText,
+    content_hierarchy: contentHierarchy,
+    route,
+  };
+  ctx.nodes.push(pageNode);
+
+  if (ctx.errors.length > 0) {
+    return { ok: false, errors: ctx.errors };
+  }
+
+  const fragment: GraphFragment = {
+    version: 1,
+    schema: "buildanything-slice-3",
+    source_file: mdPath,
+    source_sha: sha256Hex(mdContent),
+    produced_at: new Date().toISOString(),
+    nodes: sortNodes(ctx.nodes),
+    edges: sortEdges(ctx.edges),
+  };
+  return { ok: true, fragment, errors: [] };
+}