@bentopdf/pymupdf-wasm 0.11.12 → 0.11.15

package/README.md

@@ -2,6 +2,42 @@
 
 PyMuPDF compiled to WebAssembly for full PDF manipulation in the browser.
 
+## Notice
+
+This package is a modified version of PyMuPDF, originally developed by Artifex Software, Inc.
+
+It has been adapted for WebAssembly (WASM) and dynamic loading.
+
+## Attribution
+
+PyMuPDF is copyright © Artifex Software, Inc.
+This package is distributed under the GNU Affero General Public License v3.0.
+
+This project is not affiliated with or endorsed by Artifex Software, Inc.
+
+## Source Code Availability
+
+This program is licensed under the GNU Affero General Public License v3.0.
+If you interact with this program over a network, you are entitled to
+receive the complete corresponding source code.
+
+The source code for this package is available at:
+https://github.com/alam00000/bentopdf-pymupdf-wasm
+
+This package includes the complete **Corresponding Source** (build scripts and configuration) in the `build_scripts/` directory.
+
+### Build Instructions
+
+To rebuild the WASM binary from source:
+1. Download the source code from the repository or this package.
+2. Navigate to the `build_scripts/` directory.
+3. Follow the instructions in `build_scripts/README.md` (uses Docker).
+
+## License
+
+This project is licensed under the [AGPL-3.0-only](LICENSE) license.
+See the License section below for details on included components.
+
 ## Features
 
 - **Open** PDF, XPS, EPUB, and images
@@ -25,246 +61,19 @@ npm install @bentopdf/pymupdf-wasm
 ```javascript
 import { PyMuPDF } from '@bentopdf/pymupdf-wasm';
 
-// Initialize with path to assets
-const pymupdf = new PyMuPDF('/assets/pymupdf/');
-
-// Preload (optional, speeds up first operation)
-await pymupdf.load();
-
-// Open a PDF
-const doc = await pymupdf.open(pdfFile);
-
-// Extract text
-const text = doc.getPage(0).getText();
-
-// Save with modifications
-const blob = doc.saveAsBlob();
-doc.close();
-```
-
-## API Reference
-
-### PyMuPDF Class
-
-```javascript
-const pymupdf = new PyMuPDF(assetPath);
-
-// Document operations
-const doc = await pymupdf.open(file);
-const doc = await pymupdf.openUrl('https://example.com/doc.pdf');
-const doc = await pymupdf.create(); // Empty PDF
-
-// Utilities
-const merged = await pymupdf.merge([pdf1, pdf2, pdf3]);
-const [part1, part2] = await pymupdf.split(pdf, [
-  { start: 0, end: 4 },
-  { start: 5, end: 9 }
-]);
-const text = await pymupdf.extractText(pdf);
-const image = await pymupdf.renderPage(pdf, 0, 150); // 150 DPI
-
-// PDF to DOCX
-const docx = await pymupdf.pdfToDocx(pdf);
-
-// File to PDF conversion
-// Supports: XPS, EPUB, MOBI, FB2, CBZ, SVG, images (JPEG, PNG, BMP, GIF, TIFF, WEBP)
-const pdfFromXps = await pymupdf.xpsToPdf(xpsFile);
-const pdfFromEpub = await pymupdf.epubToPdf(epubFile);
-const pdfFromImage = await pymupdf.imageToPdf(imageFile);
-const pdfFromSvg = await pymupdf.svgToPdf(svgFile);
-const pdfFromImages = await pymupdf.imagesToPdf([img1, img2, img3]);
-const pdfFromAny = await pymupdf.convertToPdf(file, { filetype: 'svg' });
-
-// PDF to other formats
-const images = await pymupdf.pdfToImages(pdf, { format: 'png', dpi: 300 });
-const svgs = await pymupdf.pdfToSvg(pdf);
-const text = await pymupdf.pdfToText(pdf);
-const html = await pymupdf.pdfToHtml(pdf);
-const json = await pymupdf.pdfToJson(pdf);
-const xml = await pymupdf.pdfToXml(pdf);
-```
-
-### Document Operations
-
-```javascript
-// Properties
-doc.pageCount;
-doc.metadata;    // { title, author, ... }
-doc.isEncrypted;
-
-// Page access
-const page = doc.getPage(0);
-for (const page of doc.pages()) { ... }
-
-// Modify
-doc.deletePage(5);
-doc.insertBlankPage(0);
-doc.movePage(3, 0);
-
-// Merge another PDF
-const other = await pymupdf.open(otherPdf);
-doc.insertPdf(other);
-
-// Save
-const pdf = doc.save();
-const blob = doc.saveAsBlob();
-doc.close();
-```
-
-### Page Operations
-
-```javascript
-const page = doc.getPage(0);
-
-// Properties
-page.width;
-page.height;
-page.rotation;
-page.setRotation(90);
-
-// Text
-const text = page.getText();
-const rects = page.searchFor("keyword");
-page.insertText({ x: 100, y: 100 }, "Hello", { fontsize: 14 });
-
-// Images
-const images = page.getImages();
-const img = page.extractImage(images[0].xref);
-page.insertImage(rect, imageData);
-
-// Annotations
-page.addHighlight(rect, { r: 1, g: 1, b: 0 });
-page.addTextAnnotation({ x: 100, y: 100 }, "Note");
-const annots = page.getAnnotations();
-
-// Render
-const png = await page.toImage({ dpi: 300 });
-const svg = page.toSvg();
-
-// Redaction
-page.addRedaction(rect);
-page.applyRedactions();
-```
-
-### Security
-
-```javascript
-// Encrypt
-const pdf = doc.save({
-  encryption: {
-    ownerPassword: 'secret',
-    userPassword: 'user123',
-    permissions: {
-      print: true,
-      copy: false
-    }
-  }
+const pymupdf = new PyMuPDF({
+  assetPath: '/assets/pymupdf/',
+  ghostscriptUrl: 'https://cdn.jsdelivr.net/npm/@bentopdf/gs-wasm@0.1.0/' // Optional: for RGB conversion
 });
 
-// Decrypt
-if (doc.needsPass) {
-  doc.authenticate('password');
-}
-```
-
-### Forms
-
-```javascript
-if (doc.isFormPdf) {
-  const fields = doc.getFormFields();
-  doc.setFormField('name', 'John Doe');
-  doc.setFormField('agree', true);
-}
-```
-
-## Asset Files
-
-Copy the following files to your assets directory:
-
-```
-assets/pymupdf/
-├── pyodide.js
-├── pyodide.asm.js
-├── pyodide.asm.wasm
-├── pyodide-lock.json
-├── python_stdlib.zip
-├── pymupdf-*.whl
-├── fonttools-*.whl
-├── lxml-*.whl
-├── numpy-*.whl
-├── opencv_python-*.whl
-├── pdf2docx-*.whl
-├── python_docx-*.whl
-└── typing_extensions-*.whl
-```
-
-## About
-
-This package was ported to work with [BentoPDF](https://bentopdf.com), an open-source PDF toolkit. Maintenance and updates will be focused on features required by BentoPDF.
-
-- Website: [bentopdf.com](https://bentopdf.com)
-- GitHub: [https://github.com/alam00000/bentopdf](https://github.com/alam00000/bentopdf)
-
-## License
-
-This project is licensed under the **GNU Affero General Public License v3.0 (AGPL-3.0)**.
-
-### Copyright Notices
-
-#### BentoPDF
-
-```
-Copyright (C) 2025 BentoPDF Contributors
-
-This program is free software: you can redistribute it and/or modify
-it under the terms of the GNU Affero General Public License as published
-by the Free Software Foundation, either version 3 of the License, or
-(at your option) any later version.
-
-This program is distributed in the hope that it will be useful,
-but WITHOUT ANY WARRANTY; without even the implied warranty of
-MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-GNU Affero General Public License for more details.
-
-You should have received a copy of the GNU Affero General Public License
-along with this program. If not, see <https://www.gnu.org/licenses/>.
-```
-
-#### PyMuPDF
-
-```
-Copyright (C) 2004-2025 Artifex Software, Inc.
-
-PyMuPDF is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).
-PyMuPDF is a Python binding for MuPDF
-
-Source code: https://github.com/pymupdf/PyMuPDF
-License: https://github.com/pymupdf/PyMuPDF/blob/master/COPYING
-```
-
-#### Ghostscript
-
-```
-Copyright (C) 2001-2025 Artifex Software, Inc.
-
-Ghostscript is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).
-Ghostscript is an interpreter for the PostScript language and PDF files.
-
-Source code: https://github.com/ArtifexSoftware/ghostpdl
-License: https://www.ghostscript.com/licensing/
+await pymupdf.load();
 ```
 
-### Combined Work Notice
-
-This package combines code from multiple AGPL-3.0 licensed projects. As a combined work,
-it is distributed under the terms of the GNU Affero General Public License v3.0. The source
-code for all components is available in accordance with Section 13 of the AGPL-3.0.
-
-For the complete license text, see the [LICENSE](./LICENSE) file.
+## Credits & Copyrights
 
-## Credits
+- **PyMuPDF**: © Artifex Software, Inc. (AGPL-3.0)
+- **Ghostscript**: © Artifex Software, Inc. (AGPL-3.0)
+- **Pyodide**: © Mozilla Foundation / Michael Droettboom (MPL-2.0)
+- **pdf2docx**: © Artifex Software, Inc. (AGPL-3.0)
 
-- [PyMuPDF](https://github.com/pymupdf/PyMuPDF) - Python bindings for MuPDF (© Artifex Software, Inc.)
-- [Ghostscript](https://www.ghostscript.com/) - PostScript and PDF interpreter (© Artifex Software, Inc.)
-- [Pyodide](https://pyodide.org/) - Python in the browser
-- [pdf2docx](https://github.com/dothinking/pdf2docx) - PDF to DOCX conversion (© Artifex Software, Inc.)
+This package combines these components and is distributed under AGPL-3.0.

package/build_scripts/Dockerfile

@@ -0,0 +1,28 @@
+FROM node:20.11-bookworm-slim AS node-image
+FROM python:3.13.2-slim-bookworm
+
+RUN apt-get update \
+  && apt-get install -y --no-install-recommends \
+        bzip2 ccache f2c g++ gfortran git make \
+        patch pkg-config swig unzip wget xz-utils \
+        autoconf autotools-dev automake texinfo dejagnu \
+        build-essential libltdl-dev \
+        gnupg2 libdbus-glib-1-2 sudo sqlite3 \
+        ninja-build jq cmake bison \
+  && rm -rf /var/lib/apt/lists/*
+
+COPY --from=node-image /usr/local/bin/node /usr/local/bin/
+COPY --from=node-image /usr/local/lib/node_modules /usr/local/lib/node_modules
+RUN ln -s ../lib/node_modules/npm/bin/npm-cli.js /usr/local/bin/npm \
+    && ln -s ../lib/node_modules/npm/bin/npx-cli.js /usr/local/bin/npx
+
+RUN wget https://github.com/emscripten-core/emsdk/archive/refs/tags/4.0.9.tar.gz \
+    && tar -xzf 4.0.9.tar.gz \
+    && cd emsdk-4.0.9 \
+    && ./emsdk install 4.0.9 \
+    && ./emsdk activate 4.0.9 \
+    && cd .. \
+    && mv emsdk-4.0.9 /opt/emsdk \
+    && rm 4.0.9.tar.gz
+
+RUN python -m pip install --upgrade pip wheel pyodide-build==0.30.5

package/build_scripts/README.md

@@ -0,0 +1,31 @@
+# Build Instructions for PyMuPDF WASM
+
+These files constitute the "Corresponding Source" build scripts required by the AGPL license.
+
+## Prerequisites
+
+- Docker
+- Git
+
+## Build Steps
+
+1.  **Build with Docker**:
+    The build process is containerized. Use the provided `Dockerfile` to build the WASM artifacts.
+
+    ```bash
+    docker build -t pymupdf-wasm-build .
+    ```
+
+    This will compile PyMuPDF to WebAssembly using Emscripten within the container.
+
+2.  **Extract Artifacts**:
+    After building, you can extract the `pymupdf.wasm` and generated JS files from the container.
+
+    ```bash
+    docker run --rm -v $(pwd)/dist:/output pymupdf-wasm-build cp -r /app/dist/* /output/
+    ```
+
+## File Descriptions
+
+-   `Dockerfile`: Defines the build environment and steps for compiling PyMuPDF to WASM.
+-   `scripts/`: Helper scripts used during the build process.

package/build_scripts/scripts/build.js

@@ -0,0 +1,48 @@
+import * as path from "path";
+import * as fs from "fs/promises";
+import esbuild from "esbuild";
+import { run } from "runish";
+
+const OUT_DIR = path.resolve("./out");
+const RELEASE_DIR = path.join(OUT_DIR, "release");
+const TSC = path.resolve("node_modules/typescript/bin/tsc");
+const { RELEASE } = process.env;
+const TARGET_DIR = RELEASE ? RELEASE_DIR : OUT_DIR;
+
+async function main() {
+  await fs.mkdir(OUT_DIR, { recursive: true });
+  if (RELEASE) {
+    await fs.rm(RELEASE_DIR, { force: true, recursive: true });
+    await fs.mkdir(RELEASE_DIR, { recursive: true });
+  }
+
+  await run(TSC, [
+    "--declaration",
+    "--emitDeclarationOnly",
+    "--outDir",
+    path.join(OUT_DIR, "types"),
+  ]);
+
+  await esbuild.build({
+    entryPoints: ["src/index.ts"],
+    outdir: TARGET_DIR,
+    bundle: true,
+    write: true,
+    format: "esm",
+    target: "es2020",
+    minify: !!RELEASE,
+  });
+
+  await esbuild.build({
+    entryPoints: ["tests/index.ts"],
+    outdir: path.join(OUT_DIR, "tests"),
+    bundle: true,
+    write: true,
+    format: "esm",
+  });
+}
+
+main().catch((err) => {
+  console.error(err);
+  process.exit(1);
+});

package/build_scripts/scripts/buildDeps.js

@@ -0,0 +1,31 @@
+import * as path from "path";
+import * as fs from "fs/promises";
+import { run } from "runish";
+
+const OUT_DIR = path.resolve("./out");
+const LIB_DIR = path.resolve("./lib");
+
+async function main() {
+  await fs.mkdir(OUT_DIR, { recursive: true });
+
+  await run("pyodide", ["build", "--exports", "whole_archive"], {
+    cwd: path.join(LIB_DIR, "PyMuPDF"),
+    env: {
+      SKIP_EMSCRIPTEN_VERSION_CHECK: "1",
+      HAVE_LIBCRYPTO: "no",
+      OS: "pyodide",
+      PYMUPDF_SETUP_FLAVOUR: "pb",
+      PYMUPDF_SETUP_MUPDF_BUILD_TESSERACT: "0",
+      PYMUPDF_SETUP_MUPDF_TESSERACT: "0",
+      ...process.env,
+    },
+  });
+
+  const whl = "pymupdf-1.26.1-cp313-none-pyodide_2025_0_wasm32.whl";
+  await fs.cp(path.join(LIB_DIR, "PyMuPDF/dist", whl), path.join(OUT_DIR, whl));
+}
+
+main().catch((err) => {
+  console.error(err);
+  process.exit(1);
+});

package/build_scripts/scripts/download.js

@@ -0,0 +1,72 @@
+import * as path from "path";
+import * as fs from "fs/promises";
+import { run } from "runish";
+import { existsSync } from "fs";
+
+const LIB_DIR = path.resolve("./lib");
+const OUT_DIR = path.resolve("./out");
+
+async function main() {
+  await fs.mkdir(LIB_DIR, { recursive: true });
+
+  const libs = [
+    [
+      "PyMuPDF",
+      "https://github.com/pymupdf/PyMuPDF",
+      "4a53405a51d29f2f620c0c7659b7c4d404a9f9c0",
+    ],
+  ];
+  for (const [name, repo, hash, callback] of libs) {
+    process.chdir(LIB_DIR);
+    const cloned = await gitClone(name, repo, hash);
+    if (cloned && callback) await callback();
+  }
+
+  await fs.mkdir(OUT_DIR, { recursive: true });
+
+  const assets = [
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/pyodide.js",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/python_stdlib.zip",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/pyodide.asm.wasm",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/pyodide-lock.json",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/pyodide.asm.js",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/lxml-5.4.0-cp313-cp313-pyodide_2025_0_wasm32.whl",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/typing_extensions-4.12.2-py3-none-any.whl",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/numpy-2.2.5-cp313-cp313-pyodide_2025_0_wasm32.whl",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/opencv_python-4.11.0.86-cp313-cp313-pyodide_2025_0_wasm32.whl",
+    "https://cdn.jsdelivr.net/pyodide/v0.28.0a3/full/fonttools-4.56.0-py3-none-any.whl",
+    "https://files.pythonhosted.org/packages/d0/00/1e03a4989fa5795da308cd774f05b704ace555a70f9bf9d3be057b680bcf/python_docx-1.2.0-py3-none-any.whl",
+    "https://files.pythonhosted.org/packages/b5/f9/6d567df395c0409baf2b4dd9cd30d1e977c70672fe7ec2a684af1e6aa41c/pdf2docx-0.5.8-py3-none-any.whl",
+  ];
+  for (let url of assets) {
+    const name = url.split("/").at(-1);
+    if (name === "pyodide.js") url = url.replace(/.js$/, () => ".mjs");
+    await download(name, url);
+  }
+}
+
+async function gitClone(name, repo, hash) {
+  if (existsSync(name)) return;
+
+  console.log(`git cloning ${name} - ${repo} - ${hash}`);
+  await run("git", ["init", name]);
+  process.chdir(path.join(LIB_DIR, name));
+  await run("git", ["fetch", "--depth", "1", repo, hash]);
+  await run("git", ["checkout", "FETCH_HEAD"]);
+  return true;
+}
+
+async function download(name, url) {
+  const filePath = path.join(OUT_DIR, name);
+
+  if (existsSync(filePath)) return;
+
+  console.log(`downloading ${name} - ${url}`);
+  const buf = await fetch(url).then((x) => x.arrayBuffer());
+  await fs.writeFile(filePath, Buffer.from(buf));
+}
+
+main().catch((err) => {
+  console.error(err);
+  process.exit(1);
+});

package/dist/index.js

@@ -1128,17 +1128,21 @@ elif t != "null":
 };
 
 // src/pymupdf.ts
-import loadGhostscriptWASM from "@bentopdf/gs-wasm";
-async function convertPdfToRgb(pdfData) {
+async function convertPdfToRgb(pdfData, gsBaseUrl) {
+  if (!gsBaseUrl) {
+    throw new Error("Ghostscript URL not configured. Cannot perform RGB conversion.");
+  }
   console.log("[convertPdfToRgb] Starting Ghostscript RGB conversion...");
   console.log("[convertPdfToRgb] Input size:", pdfData.length);
+  console.log("[convertPdfToRgb] GS base URL:", gsBaseUrl);
+  const normalizedGsUrl = gsBaseUrl.endsWith("/") ? gsBaseUrl : `${gsBaseUrl}/`;
+  const libraryUrl = `${normalizedGsUrl}dist/index.js`;
+  const { loadGhostscriptWASM } = await import(
+    /* @vite-ignore */
+    libraryUrl
+  );
   const gs = await loadGhostscriptWASM({
-    locateFile: (path) => {
-      if (path.endsWith(".wasm")) {
-        return "/ghostscript-wasm/gs.wasm";
-      }
-      return path;
-    },
+    baseUrl: `${normalizedGsUrl}assets/`,
     print: (text) => console.log("[GS RGB]", text),
     printErr: (text) => console.error("[GS RGB Error]", text)
   });
@@ -1257,8 +1261,10 @@ var PyMuPDF = class {
     this.docCounter = 0;
     if (typeof options === "string") {
       this.assetPath = options;
+      this.ghostscriptUrl = "";
     } else {
       this.assetPath = options?.assetPath ?? "./";
+      this.ghostscriptUrl = options?.ghostscriptUrl ?? "";
     }
     if (!this.assetPath.endsWith("/")) {
       this.assetPath += "/";
@@ -1429,7 +1435,7 @@ def deskew_image(img_array, angle):
     let pdfData = new Uint8Array(buf);
     console.log("[pdfToDocx] Converting PDF to RGB colorspace with Ghostscript...");
     try {
-      const rgbData = await convertPdfToRgb(pdfData);
+      const rgbData = await convertPdfToRgb(pdfData, this.ghostscriptUrl);
       pdfData = rgbData;
       console.log("[pdfToDocx] RGB conversion complete");
     } catch (e) {
@@ -2312,19 +2318,121 @@ doc.scrub(
     reset_responses=${scrubResetResponses ? "True" : "False"},
 )
 
-# 2. Image compression
+# 2. Image compression (safe per-xref approach to avoid MuPDF buffer overflow
+#    with shared image xrefs across many pages \u2014 bypasses doc.rewrite_images())
 if ${compressImages ? "True" : "False"}:
-    doc.rewrite_images(
-        dpi_threshold=${dpiThreshold},
-        dpi_target=${dpiTarget},
-        quality=${imageQuality},
-        lossy=${processLossy ? "True" : "False"},
-        lossless=${processLossless ? "True" : "False"},
-        bitonal=${processBitonal ? "True" : "False"},
-        color=${processColor ? "True" : "False"},
-        gray=${processGray ? "True" : "False"},
-        set_to_gray=${convertToGray ? "True" : "False"},
-    )
+    import math as _math
+    import sys as _sys
+
+    _dpi_target = ${dpiTarget}
+    _dpi_threshold = ${dpiThreshold}
+    _quality = ${imageQuality}
+    _set_to_gray = ${convertToGray ? "True" : "False"}
+    _process_lossy = ${processLossy ? "True" : "False"}
+    _process_lossless = ${processLossless ? "True" : "False"}
+    _process_bitonal = ${processBitonal ? "True" : "False"}
+    _process_color = ${processColor ? "True" : "False"}
+    _process_gray = ${processGray ? "True" : "False"}
+
+    # Phase 1: Collect unique image xrefs and smask info
+    _xref_info = {}
+    for _page in doc:
+        for _img in _page.get_images(full=True):
+            _xref, _smask = _img[0], _img[1]
+            if _xref > 0:
+                _xref_info.setdefault(_xref, {"smask": _smask, "min_dpi": float("inf")})
+
+    # Phase 2: Calculate effective DPI for each xref across all page usages
+    for _page in doc:
+        for _info in _page.get_image_info(hashes=False, xrefs=True):
+            _xref = _info.get("xref", 0)
+            if _xref not in _xref_info:
+                continue
+            _bbox = _info.get("bbox")
+            _w = _info.get("width", 0)
+            _h = _info.get("height", 0)
+            if _bbox and _w > 0 and _h > 0:
+                _disp_w = abs(_bbox[2] - _bbox[0])
+                _disp_h = abs(_bbox[3] - _bbox[1])
+                if _disp_w > 0 and _disp_h > 0:
+                    _dpi = min(_w / _disp_w * 72, _h / _disp_h * 72)
+                    if _dpi < _xref_info[_xref]["min_dpi"]:
+                        _xref_info[_xref]["min_dpi"] = _dpi
+
+    _effective_threshold = max(_dpi_threshold or 0, (_dpi_target or 0) + 10) if _dpi_target else None
+
+    # Phase 3: Rewrite each image xref individually
+    for _xref, _meta in _xref_info.items():
+        _min_dpi = _meta["min_dpi"]
+        _smask_xref = _meta["smask"]
+
+        _needs_downscale = bool(
+            _dpi_target and _effective_threshold
+            and _min_dpi != float("inf")
+            and _min_dpi > _effective_threshold
+        )
+        if not _needs_downscale and _quality is None and not _set_to_gray:
+            continue
+
+        try:
+            # Check image type filters (match rewrite_images behavior)
+            _xref_obj = doc.xref_object(_xref)
+            _is_lossy = "/DCTDecode" in _xref_obj or "/JPXDecode" in _xref_obj
+            _is_lossless = not _is_lossy
+            if _is_lossy and not _process_lossy:
+                continue
+            if _is_lossless and not _process_lossless:
+                continue
+
+            _pix = pymupdf.Pixmap(doc, _xref)
+
+            # Check colorspace filters
+            _n = _pix.colorspace.n if _pix.colorspace else 0
+            _is_bitonal = (_pix.colorspace and _n == 1 and doc.xref_get_key(_xref, "BitsPerComponent")[1] == "1")
+            _is_gray = (_n == 1 and not _is_bitonal)
+            _is_color = (_n >= 3)
+            if _is_bitonal and not _process_bitonal:
+                _pix = None
+                continue
+            if _is_gray and not _process_gray:
+                _pix = None
+                continue
+            if _is_color and not _process_color:
+                _pix = None
+                continue
+
+            if _set_to_gray and _pix.colorspace and _pix.colorspace.n > 1:
+                _pix = pymupdf.Pixmap(pymupdf.csGRAY, _pix)
+            elif _pix.alpha:
+                _pix = pymupdf.Pixmap(_pix.colorspace or pymupdf.csRGB, _pix)
+
+            if _needs_downscale:
+                _ratio = _min_dpi / _dpi_target
+                _shrink_n = max(0, min(7, int(_math.log2(_ratio))))
+                if _shrink_n > 0:
+                    _pix.shrink(_shrink_n)
+
+            _q = _quality if _quality is not None else 85
+            _jpeg_bytes = _pix.tobytes("jpeg", jpg_quality=_q)
+
+            _cs_name = (
+                "/DeviceGray"
+                if _pix.colorspace and _pix.colorspace.n == 1
+                else "/DeviceRGB"
+            )
+            _smask_entry = f"/SMask {_smask_xref} 0 R " if _smask_xref else ""
+            _new_obj = (
+                f"<</Type /XObject /Subtype /Image /BitsPerComponent 8"
+                f" /ColorSpace {_cs_name} /Filter /DCTDecode"
+                f" /Height {_pix.height} /Width {_pix.width}"
+                f" {_smask_entry}>>"
+            )
+            doc.update_object(_xref, _new_obj)
+            doc.update_stream(_xref, _jpeg_bytes, compress=0)
+            _pix = None
+
+        except Exception as _e:
+            _sys.stderr.write(f"[pymupdf-wasm] safe_rewrite_images xref {_xref}: {_e}\\n")
 
 # 3. Font subsetting
 if ${subsetFonts ? "True" : "False"}:

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@bentopdf/pymupdf-wasm",
-    "version": "0.11.12",
+    "version": "0.11.15",
     "description": "PyMuPDF compiled to WebAssembly - Full PDF manipulation in the browser",
     "type": "module",
     "main": "dist/index.js",
@@ -19,16 +19,14 @@
     "files": [
         "dist",
         "types",
-        "assets"
+        "assets",
+        "build_scripts"
     ],
     "scripts": {
-        "build": "esbuild src/index.ts --bundle --format=esm --outfile=dist/index.js --external:./assets/* --external:@bentopdf/gs-wasm",
-        "dev": "esbuild src/index.ts --bundle --format=esm --outfile=dist/index.js --watch --external:@bentopdf/gs-wasm",
+        "build": "esbuild src/index.ts --bundle --format=esm --outfile=dist/index.js --external:./assets/*",
+        "dev": "esbuild src/index.ts --bundle --format=esm --outfile=dist/index.js --watch",
         "typecheck": "tsc --noEmit"
     },
-    "peerDependencies": {
-        "@bentopdf/gs-wasm": "*"
-    },
     "repository": {
         "type": "git",
         "url": "git+https://github.com/alam00000/bentopdf-pymupdf-wasm.git"
@@ -46,7 +44,10 @@
         "extract"
     ],
     "author": "BentoPDF",
-    "license": "AGPL-3.0",
+    "contributors": [
+        "Artifex Software, Inc."
+    ],
+    "license": "AGPL-3.0-only",
     "bugs": "https://github.com/alam00000/bentopdf-pymupdf-wasm/issues",
     "devDependencies": {
         "esbuild": "^0.21.2",