@stratta/mcp 0.4.0 → 0.5.0

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@stratta/mcp",
   "mcpName": "io.github.hugogebel-boop/stratta",
-  "version": "0.4.0",
+  "version": "0.5.0",
   "description": "MCP server exposing Swiss engineering norms (SIA / Eurocodes) to Claude clients via Stratta TreeRAG.",
   "license": "UNLICENSED",
   "author": "Hugo Gebel <hugo.gebel@epfl.ch>",

package/scripts/ingest-prepass.py

@@ -74,8 +74,19 @@ def detect_running_text(doc: "fitz.Document") -> set[str]:
     return {l for l, c in counter.items() if c >= 5 and len(l) > 8}
 
 
-def extract_chapters(doc: "fitz.Document") -> dict[str, dict]:
-    """Top-level chapters from PDF bookmarks (authoritative titles + pages)."""
+CHAP_UPPER_RE = re.compile(
+    r"^(\d+)\s+([A-ZÉÈÀÂÔÎÛÇ][^a-z\n]{3,120})\s*$", re.MULTILINE
+)
+ANNEX_BODY_RE = re.compile(
+    r"^(?:ANNEXE|Annexe)\s+([A-Z])(?:\s*\(([^)]+)\))?\s*(.*?)$", re.MULTILINE
+)
+
+
+def extract_chapters(
+    doc: "fitz.Document", toc_pages: set[int] | None = None
+) -> dict[str, dict]:
+    """Top-level chapters: prefer PDF bookmarks; fallback to body-text regex
+    (uppercase heading on its own line) when the PDF has no bookmarks."""
     chapters: dict[str, dict] = {}
     for _depth, title, page in doc.get_toc(simple=True):
         m = BM_ANNEX_RE.match(title)
@@ -91,6 +102,34 @@ def extract_chapters(doc: "fitz.Document") -> dict[str, dict]:
         if m:
             num, rest = m.group(1), m.group(2).strip()
             chapters[num] = {"title": rest, "pageStart": page}
+    if chapters:
+        return chapters
+
+    # Fallback: no bookmarks → scan body for uppercase chapter headings.
+    skip = toc_pages or set()
+    for pageno in range(doc.page_count):
+        if pageno in skip:
+            continue
+        text = doc[pageno].get_text("text")
+        for m in CHAP_UPPER_RE.finditer(text):
+            num = m.group(1)
+            title = clean(m.group(2))
+            if int(num) > 50:
+                continue
+            if num in chapters:
+                continue
+            chapters[num] = {"title": title, "pageStart": pageno + 1}
+        for m in ANNEX_BODY_RE.finditer(text):
+            letter = m.group(1)
+            kind = m.group(2) or ""
+            rest = clean(m.group(3) or "")
+            path = f"Annexe {letter}"
+            if path in chapters:
+                continue
+            t = path + (f" ({kind})" if kind else "")
+            if rest and rest.lower() != path.lower():
+                t += f" — {rest}"
+            chapters[path] = {"title": t, "pageStart": pageno + 1}
     return chapters
 
 
@@ -254,7 +293,7 @@ def main() -> None:
     doc = fitz.open(str(pdf_path))
     toc_pages = detect_toc_pages(doc)
     running = detect_running_text(doc)
-    chapters = extract_chapters(doc)
+    chapters = extract_chapters(doc, toc_pages)
     chap_prefixes = {k for k in chapters if k.isdigit()}
     subsections = extract_subsections(doc, chap_prefixes, toc_pages, running)
     nodes = build_tree(chapters, subsections, doc.page_count)