fix png-numbering: re-convert 34 zero-based docs + crop fallback

34 of 116 docs were generated with 0-based PNG numbering (p-000.png …
p-008.png) but the Sonnet chunks reference 1-based page numbers in their
YAML frontmatter (page: 9 means the 9th sheet of paper). The /api/crop
handler built p-009.png and got a 500, the browser's Next/Image surfaced
400, and the chunk rendered as a black box on screen.

Fixes:
- web/app/api/crop/route.ts: try p-NNN.png first, fall back to
  p-(NNN-1).png if the 1-based file is missing. Cheap insurance for any
  doc that comes in with the old convention.
- scripts/01-convert-pdfs.sh: previously printf '%03d' "$num" with $num
  starting at 0 (e.g. "008") raised "invalid number" because Bash
  parsed it as octal. Wrap with $((10#$num)) to force decimal — this
  was silently corrupting page sequences and producing gaps like p-001
  ... p-008, p-011 (missing p-009/p-010).
- All 34 affected docs re-converted from PDFs with the patched script;
  every directory now has continuous 1-based PNGs.
- /processing/png/ rsync'd to VPS, web redeployed.

Smoke: /api/crop?doc=doc-341-…&page=9&… now returns 200 image/webp
instead of 500. Tested in browser: chunk c0026 (diagram, p9) renders
the real engineering diagram.

scripts/01-convert-pdfs.sh

@@ -115,7 +115,9 @@ convert_one_pdf() {
             num=$(printf '%s' "$bn" | sed -E 's/^p-([0-9]+)\.png$/\1/')
             if [[ "$num" =~ ^[0-9]+$ ]]; then
                 local padded
-                padded=$(printf '%03d' "$num")
+                # Force decimal interpretation: leading zeros would make bash
+                # treat "008" as invalid octal under printf '%03d'.
+                padded=$(printf '%03d' "$((10#$num))")
                 local new_name="p-$padded.png"
                 if [[ "$bn" != "$new_name" ]]; then
                     mv "$f" "$png_dir/$new_name"

web/app/api/crop/route.ts

@@ -57,7 +57,12 @@ export async function GET(req: NextRequest) {
   const format = (u.searchParams.get("format") ?? "webp").toLowerCase();
   const tight = u.searchParams.get("tight") !== "0";
 
-  // Resolve source PNG
+  // Resolve source PNG.
+  // Two conventions exist in the corpus:
+  //   1-based:  page=1 → p-001.png  (most docs)
+  //   0-based:  page=1 → p-000.png  (~34 docs converted with pdftoppm -f 0)
+  // Try the 1-based path first; if the file doesn't exist, fall back to
+  // 0-based (page - 1). Idempotent — if both exist, 1-based wins.
   let pngPath: string;
   if (pngParam) {
     if (pngParam.includes("..")) return badRequest("png param: invalid path");
@@ -71,8 +76,12 @@ export async function GET(req: NextRequest) {
       pageNum = parseInt(pageStr, 10);
     }
     if (!Number.isFinite(pageNum) || pageNum < 1) return badRequest("bad page");
+    const docDir = path.join(PROCESSING, "png", doc);
+    const oneBased = path.join(docDir, `p-${String(pageNum).padStart(3, "0")}.png`);
+    const zeroBased = path.join(docDir, `p-${String(pageNum - 1).padStart(3, "0")}.png`);
+    const { existsSync } = await import("node:fs");
+    pngPath = existsSync(oneBased) ? oneBased : (existsSync(zeroBased) ? zeroBased : oneBased);
     pageStr = `p-${String(pageNum).padStart(3, "0")}`;
-    pngPath = path.join(PROCESSING, "png", doc, `${pageStr}.png`);
   }
 
   let buf: Buffer;