@reportwright/pdf 0.1.0-beta.8 → 0.1.0-beta.9

package/README.md

@@ -1133,6 +1133,13 @@ bidi algorithm, hyphenation, kerning of the standard fonts, WOFF fonts; popup an
 
 ## Changes
 
+0.1.0-beta.9: stamping every page (`page.draw` then `save()`) is about 35% faster on a 1,000-page file: a page's open
+graphics states are counted with a byte scan instead of a full parse, and FlateDecode inflates whole, valid zlib
+streams with Node's zlib (damaged, truncated, raw or over-budget streams still go through fflate as before); the output
+is byte-identical. `bulk` starts an item's `itemTimeoutMs` clock when its worker is ready, so a respawned worker's
+start-up no longer fails good files; a stopped export (`timeoutMs` or `signal`) aborts its WritableStream even when
+the stop lands between two calls.
+
 0.1.0-beta.2: a left-to-right shaped cluster in an ActualText span has its text only in the ActualText, so MuPDF
 no longer reads it twice (pdf.js, which ignores ActualText, now misses those clusters; poppler, PDFium and
 `extractText` read them); `extractText` drops WJ (U+2060) and reads a code a ToUnicode skips as nothing (U+FFFD

package/dist/index.js

@@ -12059,9 +12059,9 @@ function encryptor(opt, doc) {
     const v = iv();
     return sealSync(k, v, b) ?? aesCbc(k, v, b).then((c) => concat([v, c]));
   };
-  let ready = false;
+  let ready2 = false;
   setup.then(() => {
-    ready = true;
+    ready2 = true;
   }, () => {
   });
   const LIT = (x) => {
@@ -12105,12 +12105,12 @@ function encryptor(opt, doc) {
     dict(n, d) {
       if (n === self.ref) return d;
       if (!meta && /^<< \/Type \/Metadata\b/.test(d) && !d.includes("/Filter")) return d.replace(/>>\s*$/, "/Filter /Crypt /DecodeParms << /Name /Identity >> >>");
-      return ready ? dictNow(n, d) : setup.then(() => dictNow(n, d));
+      return ready2 ? dictNow(n, d) : setup.then(() => dictNow(n, d));
     },
     /** @param {number} n @param {Uint8Array} b @param {string} d */
     stream(n, b, d) {
       if (!meta && /\/Type\s*\/Metadata\b/.test(d)) return b;
-      return ready ? seal(keyOf2(n), b) : setup.then(() => seal(keyOf2(n), b));
+      return ready2 ? seal(keyOf2(n), b) : setup.then(() => seal(keyOf2(n), b));
     }
   };
   return self;
@@ -12920,9 +12920,9 @@ function sinkOf(budget) {
   };
 }
 function flate(data, budget) {
-  const run2 = (Z) => {
+  const run2 = (Z2) => {
     const s = sinkOf(budget);
-    const z = new Z((c) => s.push(c));
+    const z = new Z2((c) => s.push(c));
     let ok = true;
     try {
       for (let at = 0; at < data.length; at += CHUNK2) z.push(data.subarray(at, Math.min(data.length, at + CHUNK2)), at + CHUNK2 >= data.length);
@@ -12936,6 +12936,18 @@ function flate(data, budget) {
     }
     return { out: s.out(), ok, n: s.n };
   };
+  const Z = builtin("node:zlib");
+  if (Z && data.length) {
+    let out = null;
+    try {
+      out = Z.inflateSync(data, { maxOutputLength: Math.max(1, Math.min(2 ** 30, budget.maxStream, budget.maxDecoded - budget.decoded)) });
+    } catch {
+    }
+    if (out) {
+      budget.take(out.length, 0);
+      return new Uint8Array(out.buffer, out.byteOffset, out.length);
+    }
+  }
   const F2 = fflateNow();
   if (!F2) throw new Error("internal: FlateDecode before fflate was loaded");
   const r = run2(F2.Unzlib);
@@ -15843,6 +15855,7 @@ var init_fill = __esm({
 var save_exports = {};
 __export(save_exports, {
   MAX_OPEN: () => MAX_OPEN,
+  countSaves: () => countSaves,
   openSaves: () => openSaves,
   save: () => save2,
   saveTo: () => saveTo
@@ -15852,7 +15865,10 @@ function openSaves(data, budget) {
   let depth2 = 0;
   for (; ; ) {
     const v = parseValue(lx);
-    if (v === void 0) break;
+    if (v === void 0) {
+      if (lx.pos < data.length) continue;
+      break;
+    }
     if (!(v instanceof Op)) continue;
     if (v.v === "q") depth2++;
     else if (v.v === "Q") depth2 = Math.max(0, depth2 - 1);
@@ -15867,6 +15883,80 @@ function openSaves(data, budget) {
   }
   return depth2;
 }
+function countSaves(b, budget) {
+  const e = b.length;
+  if (e >= 4e6) return openSaves(b, budget);
+  const st = [];
+  let depth2 = 0, i = 0;
+  for (; ; ) {
+    for (; ; ) {
+      while (i < e && CLASS[b[i]] === 1) i++;
+      if (i < e && b[i] === 37) {
+        while (i < e && b[i] !== 10 && b[i] !== 13) i++;
+        continue;
+      }
+      break;
+    }
+    if (i >= e) return depth2;
+    budget.tick();
+    const c = b[i];
+    if (c === 91 || c === 60 && b[i + 1] === 60) {
+      if (st.length >= MAX_DEPTH4) return openSaves(b, budget);
+      st.push(c === 91 ? 1 : 2);
+      i += c === 91 ? 1 : 2;
+      continue;
+    }
+    if (c === 93 || c === 62) {
+      if (!st.length) {
+        i += c === 62 && b[i + 1] === 62 ? 2 : 1;
+        continue;
+      }
+      if (st[st.length - 1] === (c === 93 ? 1 : 2)) st.pop();
+      i += c === 62 && b[i + 1] === 62 ? 2 : 1;
+      continue;
+    }
+    if (c === 60) {
+      const s2 = i;
+      i++;
+      while (i < e && b[i] !== 62) i++;
+      if (i < e) i++;
+      budget.tick(1 + (i - s2 >> 11));
+      continue;
+    }
+    if (c === 40) {
+      const s2 = i;
+      i++;
+      for (let d = 1; i < e; i++) {
+        const x = b[i];
+        if (x === 92) i++;
+        else if (x === 40) d++;
+        else if (x === 41 && --d === 0) break;
+      }
+      i = Math.min(i + 1, e);
+      budget.tick(1 + (i - s2 >> 10));
+      continue;
+    }
+    if (c === 47) {
+      i++;
+      while (i < e && CLASS[b[i]] === 0) i++;
+      continue;
+    }
+    if (CLASS[c] === 2) {
+      i++;
+      continue;
+    }
+    const s = i;
+    while (i < e && CLASS[b[i]] === 0) i++;
+    if (c >= 48 && c <= 57 || c === 43 || c === 45 || c === 46) continue;
+    const w = i - s > 5 ? "" : latin12(b.subarray(s, i));
+    if (w === "true" || w === "false" || w === "null") continue;
+    if (w === "R" && st.length) return openSaves(b, budget);
+    if (w === "BI") return openSaves(b, budget);
+    st.length = 0;
+    if (w === "q") depth2++;
+    else if (w === "Q") depth2 = Math.max(0, depth2 - 1);
+  }
+}
 function options(self, o = {}) {
   if (o == null || typeof o !== "object") throw new TypeError(`save: options are { ${KEYS2.join(", ")} }`);
   for (const k of Object.keys(o)) if (!KEYS2.includes(k)) throw new TypeError(`save: unknown option ${JSON.stringify(k.slice(0, 40))} (${KEYS2.join(", ")})`);
@@ -15999,7 +16089,7 @@ async function run(self, sink, o) {
   });
   const qRef = { n: 0 };
   const made = /* @__PURE__ */ new Map();
-  const ready = async (v, kind) => {
+  const ready2 = async (v, kind) => {
     let h = made.get(v);
     if (!h) {
       h = kind === "image" ? await pdf.embedImage(v) : typeof v === "string" ? pdf.standardFont(v) : await pdf.embedFont(v);
@@ -16038,8 +16128,8 @@ async function run(self, sink, o) {
     const xobjs = [];
     for (const dr of p.draws) {
       const fonts = {}, images = {};
-      for (const [n2, v] of Object.entries(dr.fonts)) fonts[n2] = await ready(v, "font");
-      for (const [n2, v] of Object.entries(dr.images)) images[n2] = await ready(v, "image");
+      for (const [n2, v] of Object.entries(dr.fonts)) fonts[n2] = await ready2(v, "font");
+      for (const [n2, v] of Object.entries(dr.images)) images[n2] = await ready2(v, "image");
       xobjs.push({ ref: appearance(page, W, H, (ap) => dr.fn(ap, { fonts, images, pdf })), m: `1 0 0 1 ${numText(b.mediaBox[0])} ${numText(b.mediaBox[1])}` });
     }
     const list = OR(d.get("Annots"));
@@ -16080,7 +16170,7 @@ async function run(self, sink, o) {
     }
     if (xobjs.length) {
       const res2 = OR(pg.inh.Resources), xo = res2 instanceof Map ? OR(res2.get("XObject")) : null;
-      const open = openSaves(contentBytes(src, d.get("Contents")), src.budget);
+      const open = countSaves(contentBytes(src, d.get("Contents")), src.budget);
       if (open > MAX_OPEN) throw new PdfReadError("E_MALFORMED", `Page ${k + 1} leaves ${open > 1e6 ? "over a million" : open} graphics states open (more than ${MAX_OPEN}): it cannot be drawn on or flattened safely`);
       let tail = "Q\n".repeat(1 + open), i = 0;
       for (const x of xobjs) {
@@ -17660,7 +17750,7 @@ async function bulk(inputs, task, o = {}) {
     );
     const arm = () => {
       clearTimeout(timer);
-      timer = flight.length ? setTimeout(() => dead({ name: "Error", code: "E_TIMEOUT", message: `the item took longer than itemTimeoutMs (${itemTimeoutMs} ms)` }), itemTimeoutMs) : null;
+      timer = flight.length && ready.has(worker) ? setTimeout(() => dead({ name: "Error", code: "E_TIMEOUT", message: `the item took longer than itemTimeoutMs (${itemTimeoutMs} ms)` }), itemTimeoutMs) : null;
     };
     const fill2 = () => {
       while (queue.length && (!flight.length || flight.length < DEPTH && bytes < BATCH_BYTES)) {
@@ -17681,6 +17771,11 @@ async function bulk(inputs, task, o = {}) {
       }
     };
     const onMessage = (m) => {
+      if (m?.ready === true && !("id" in m)) {
+        ready.add(worker);
+        arm();
+        return;
+      }
       flight = flight.filter((i) => i !== m.id);
       bytes -= sizeOf(list[m.id]);
       const { id, ...r } = m;
@@ -17760,8 +17855,9 @@ function __bulkWorker() {
       }
     });
   });
+  parentPort.postMessage({ ready: true });
 }
-var TASKS, BULK_TASKS, load, errorOf, DEPTH, BATCH_BYTES, MAX_WORKERS, IDLE_MS, idle, sizeOf;
+var TASKS, BULK_TASKS, load, errorOf, DEPTH, BATCH_BYTES, MAX_WORKERS, IDLE_MS, idle, ready, sizeOf;
 var init_bulk = __esm({
   "src/bulk.js"() {
     init_pdf();
@@ -17796,6 +17892,7 @@ var init_bulk = __esm({
     MAX_WORKERS = 64;
     IDLE_MS = 2e3;
     idle = /* @__PURE__ */ new Map();
+    ready = /* @__PURE__ */ new WeakSet();
     sizeOf = (input) => Array.isArray(input) ? input.reduce((n, b) => n + (b?.byteLength || 0), 0) : input?.byteLength || 0;
   }
 });
@@ -17880,6 +17977,10 @@ function createPdf(sink, options2 = {}) {
   });
   let ended = false;
   const live = () => {
+    if (ctl.signal.aborted) {
+      cleanup();
+      out.abort(ctl.signal.reason);
+    }
     ctl.signal.throwIfAborted();
     if (doc.failed) throw doc.failed;
   };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@reportwright/pdf",
-  "version": "0.1.0-beta.8",
+  "version": "0.1.0-beta.9",
   "description": "A streaming PDF writer for Node and browsers with no runtime dependencies: flat memory, PDF/A-1b to 3a, PDF/UA-1, Factur-X, embedded TrueType with Unicode text.",
   "type": "module",
   "license": "MIT",