@link-assistant/web-capture 1.12.0 → 2.0.1

package/CHANGELOG.md

@@ -1,5 +1,22 @@
 # @link-assistant/web-capture
 
+## 2.0.1
+
+### Patch Changes
+
+- 724c1a9: Fix CI/CD false positives, false negatives, and warnings (#177):
+
+  - Restore the `browser-commander` 0.10.0 pin required by open blocker #160 and repair the npm lock so `npm ci` installs again after the Dependabot update.
+  - Use Node's built-in `fetch` instead of `node-fetch`, removing the deprecated `node-domexception` package from the install. Transport diagnostics now include the underlying network error (for example `fetch failed: connect ECONNREFUSED ...`) and no longer report server-side network failures as possible CORS blocks.
+  - Load the `docx` library only when a DOCX is generated. Its bundled browser polyfills read `globalThis.localStorage` on import, so on Node 25+ every process that imported the server printed "ExperimentalWarning: localStorage is not available because --localstorage-file was not provided".
+  - Publish only the runtime files (`bin/`, `src/`, `README.md`, `CHANGELOG.md`) instead of the tests and fixtures, normalize `bin` and `repository` so `npm publish` no longer auto-corrects them, and install the pre-commit hook from the repository root instead of printing ".git can't be found" on every install.
+
+## 2.0.0
+
+### Major Changes
+
+- b055a30: Require Node.js 26.10 or newer and update Playwright to 1.64. Use current Trixie browser images, native Jest ESM and reproducibly locked release tooling. Add locked Python examples and weekly dependency/security maintenance across all ecosystems.
+
 ## 1.12.0
 
 ### Minor Changes

package/README.md

@@ -19,7 +19,8 @@ A CLI and microservice to fetch URLs and render them as:
 
 ## Installation
 
-Node.js 22.22.1 or newer is required. CI and Docker use Node.js 24. Use `npm ci`
+Node.js 26.10.0 or newer is required. CI uses 26.11.1 and the latest published
+Node Docker image uses 26.10.0 on Debian Trixie. Use `npm ci`
 for a locked repository install; Docker and CI share `package-lock.json` and its
 security overrides.
 

package/bin/web-capture.js

@@ -1110,11 +1110,9 @@ async function captureUrl(url, options) {
         await import('../src/lib.js');
       const { retry } = await import('../src/retry.js');
       const cheerio = await import('cheerio');
-      const nodeFetch = await import('node-fetch');
 
+      // retry() defaults: 3 retries from a 1000ms base delay.
       const html = await retry(() => fetchHtml(absoluteUrl), {
-        retries: 3,
-        baseDelay: 1000,
         onRetry: (err, attempt) => {
           console.error(`Retry ${attempt} fetching page: ${err.message}`);
         },
@@ -1207,14 +1205,14 @@ async function captureUrl(url, options) {
       if (imageMap.size > 0) {
         for (const [imgUrl, localPath] of imageMap) {
           try {
-            const fetchFn = nodeFetch.default || nodeFetch;
-            const resp = await retry(() => fetchFn(imgUrl), {
+            const resp = await retry(() => fetch(imgUrl), {
               retries: 2,
               baseDelay: 500,
             });
             if (resp.ok) {
-              const buffer = await resp.buffer();
-              archive.append(buffer, { name: localPath });
+              archive.append(Buffer.from(await resp.arrayBuffer()), {
+                name: localPath,
+              });
             }
           } catch {
             /* skip failed image downloads */

package/package.json

@@ -1,14 +1,20 @@
 {
   "name": "@link-assistant/web-capture",
-  "version": "1.12.0",
+  "version": "2.0.1",
   "description": "CLI and microservice to render web pages as HTML, Markdown, or PNG",
   "main": "src/index.js",
   "bin": {
-    "web-capture": "./bin/web-capture.js"
+    "web-capture": "bin/web-capture.js"
   },
+  "files": [
+    "bin/",
+    "src/",
+    "CHANGELOG.md",
+    "README.md"
+  ],
   "scripts": {
     "start": "docker compose up -d",
-    "dev": "nodemon bin/web-capture.js -- --serve",
+    "dev": "node --watch bin/web-capture.js --serve",
     "test": "node --experimental-vm-modules ./node_modules/.bin/jest",
     "test:watch": "node --experimental-vm-modules ./node_modules/.bin/jest --watch",
     "test:e2e:process": "node --experimental-vm-modules ./node_modules/.bin/jest tests/e2e/process.test.js",
@@ -16,16 +22,16 @@
     "test:e2e": "npm run test:e2e:process && npm run test:e2e:docker",
     "test:all": "npm run build && npm test && npm run test:e2e",
     "build": "docker compose down && docker compose build && npm start",
-    "examples:python": "python3 examples/python/html_download.py && python3 examples/python/markdown_download.py && python3 examples/python/image_download.py",
+    "examples:python": "uv run --project .. --locked python examples/python/html_download.py && uv run --project .. --locked python examples/python/markdown_download.py && uv run --project .. --locked python examples/python/image_download.py",
     "examples:javascript": "node examples/js/html_download.js && node examples/js/markdown_download.js && node examples/js/image_download.js",
     "examples": "npm run build && npm run examples:javascript && npm run examples:python",
-    "lint": "eslint .",
+    "lint": "node ../scripts/lint.mjs",
     "lint:fix": "eslint . --fix",
     "format": "prettier --write .",
     "format:check": "prettier --check .",
     "check:duplication": "jscpd . --baseline-from-ref origin/main --fail-on-new-clones --fail-on-empty",
     "check": "npm run lint && npm run format:check && npm run check:duplication",
-    "prepare": "husky || true",
+    "prepare": "node ../scripts/install-git-hooks.mjs",
     "changeset": "changeset",
     "changeset:version": "node ../scripts/changeset-version.mjs",
     "changeset:publish": "changeset publish",
@@ -42,19 +48,16 @@
     "iconv-lite": "^0.7.3",
     "lino-arguments": "^0.3.0",
     "log-lazy": "^1.0.4",
-    "node-fetch": "^3.3.2",
-    "playwright": "^1.63.0",
-    "puppeteer": "^25.12.0",
-    "puppeteer-core": "^25.12.0",
+    "playwright": "^1.64.0",
+    "puppeteer": "^25.13.0",
+    "puppeteer-core": "^25.13.0",
     "turndown": "^7.2.4",
     "turndown-plugin-gfm": "^1.0.2"
   },
   "devDependencies": {
-    "@babel/core": "^8.0.7",
-    "@babel/preset-env": "^8.0.7",
     "@changesets/cli": "^3.0.3",
+    "@eslint/js": "^10.0.1",
     "@types/archiver": "^8.0.0",
-    "babel-jest": "^30.5.2",
     "eslint": "^10.12.0",
     "eslint-config-prettier": "^10.1.8",
     "eslint-plugin-prettier": "^5.5.6",
@@ -67,8 +70,7 @@
     "nock": "^15.0.1",
     "prettier": "^3.9.9",
     "supertest": "^7.3.1",
-    "unzipper": "^0.12.5",
-    "@eslint/js": "^10.0.1"
+    "unzipper": "^0.12.5"
   },
   "type": "module",
   "keywords": [
@@ -80,14 +82,14 @@
   "license": "Unlicense",
   "repository": {
     "type": "git",
-    "url": "https://github.com/link-assistant/web-capture"
+    "url": "git+https://github.com/link-assistant/web-capture.git"
   },
   "engines": {
-    "node": ">=22.22.1"
+    "node": ">=26.10.0"
   },
   "lint-staged": {
     "*.{js,mjs,cjs}": [
-      "eslint --fix --max-warnings 0",
+      "eslint --fix --no-warn-ignored",
       "prettier --write",
       "prettier --check"
     ],
@@ -105,6 +107,18 @@
   "overrides": {
     "@istanbuljs/load-nyc-config": {
       "js-yaml": "^4.1.1"
+    },
+    "cheerio": {
+      "encoding-sniffer": "^1.0.2"
+    },
+    "babel-plugin-istanbul": {
+      "test-exclude": "^8.0.0"
     }
+  },
+  "packageManager": "npm@12.2.0",
+  "allowScripts": {
+    "@parcel/watcher": true,
+    "puppeteer": true,
+    "unrs-resolver": true
   }
 }

package/src/archive.js

@@ -16,7 +16,6 @@
  */
 
 import { ZipArchive } from 'archiver';
-import fetch from 'node-fetch';
 import * as cheerio from 'cheerio';
 import { URL } from 'url';
 import {
@@ -200,7 +199,7 @@ export async function archiveHandler(req, res) {
             baseDelay: 500,
           });
           if (imgResp.ok) {
-            const buffer = await imgResp.buffer();
+            const buffer = Buffer.from(await imgResp.arrayBuffer());
             archive.append(buffer, { name: localPath });
           }
         } catch {

package/src/binary-download.js

@@ -0,0 +1,13 @@
+import { retry } from './retry.js';
+
+// Shared image-download policy; keep HTTP errors distinct from network retries.
+export async function downloadBinary(url) {
+  const response = await retry(() => fetch(url), {
+    retries: 3,
+    baseDelay: 1000,
+  });
+  if (!response.ok) {
+    throw new Error(`HTTP ${response.status}`);
+  }
+  return Buffer.from(await response.arrayBuffer());
+}

package/src/docx.js

@@ -9,18 +9,8 @@
  *   engine - 'puppeteer' or 'playwright' (only used for browser rendering fallback)
  */
 
-import fetch from 'node-fetch';
 import * as cheerio from 'cheerio';
 import { URL } from 'url';
-import {
-  Document,
-  Packer,
-  Paragraph,
-  TextRun,
-  HeadingLevel,
-  ImageRun,
-  ExternalHyperlink,
-} from 'docx';
 import { fetchHtml, convertRelativeUrls } from './lib.js';
 
 export async function docxHandler(req, res) {
@@ -35,6 +25,19 @@ export async function docxHandler(req, res) {
     const absHtml = convertRelativeUrls(html, absoluteUrl);
     const $ = cheerio.load(absHtml);
 
+    // Loaded on demand: docx bundles a browser util-deprecate that reads
+    // globalThis.localStorage on import, which makes Node >=25 print an
+    // ExperimentalWarning for every process that merely imports the app.
+    const {
+      Document,
+      Packer,
+      Paragraph,
+      TextRun,
+      HeadingLevel,
+      ImageRun,
+      ExternalHyperlink,
+    } = await import('docx');
+
     // Remove unwanted elements
     $('style, script, noscript, nav, footer, header').remove();
 
@@ -75,7 +78,7 @@ export async function docxHandler(req, res) {
             const imgUrl = new URL(src, absoluteUrl).href;
             const imgResp = await fetch(imgUrl);
             if (imgResp.ok) {
-              const buffer = await imgResp.buffer();
+              const buffer = Buffer.from(await imgResp.arrayBuffer());
               const contentType = imgResp.headers.get('content-type') || '';
               children.push(
                 new Paragraph({

package/src/fetch.js

@@ -1,4 +1,3 @@
-import fetch from 'node-fetch';
 import { convertGoogleDriveUrl } from './lib.js';
 import { copyProxyResponseHeaders } from './proxy-headers.js';
 
@@ -14,7 +13,7 @@ export async function fetchHandler(req, res) {
     copyProxyResponseHeaders(response.headers, res);
 
     // Get the response body as buffer and send it
-    const buffer = await response.buffer();
+    const buffer = Buffer.from(await response.arrayBuffer());
     res.send(buffer);
   } catch (err) {
     console.error('Fetch error:', err);

package/src/figures.js

@@ -10,10 +10,9 @@
  * @module figures
  */
 
-import fetch from 'node-fetch';
 import * as cheerio from 'cheerio';
 import { URL } from 'url';
-import { retry } from './retry.js';
+import { downloadBinary } from './binary-download.js';
 import { createBrowser, getBrowserEngine } from './browser.js';
 import { dismissPopups, scrollToLoadContent } from './popups.js';
 
@@ -86,16 +85,7 @@ export async function downloadFigures(figures, options = {}) {
     const filename = `figure-${figure.figureNum}.${ext}`;
 
     try {
-      const resp = await retry(() => fetch(figure.src), {
-        retries: 3,
-        baseDelay: 1000,
-      });
-
-      if (!resp.ok) {
-        throw new Error(`HTTP ${resp.status}`);
-      }
-
-      const buffer = await resp.buffer();
+      const buffer = await downloadBinary(figure.src);
 
       if (options.onProgress) {
         options.onProgress(figure.figureNum, 'downloaded');

package/src/gdocs-images.js

@@ -5,7 +5,6 @@
 // the archive is self-contained. This module downloads the images and
 // rewrites the captured output accordingly.
 
-import fetch from 'node-fetch';
 import { URL } from 'url';
 
 const DEFAULT_USER_AGENT =

package/src/gdocs.js

@@ -7,7 +7,6 @@
 //   https://docs.googleapis.com/v1/documents/{DOCUMENT_ID}
 // Browser-model capture loads /edit and extracts DOCS_modelChunk data.
 
-import fetch from 'node-fetch';
 import * as he from 'he';
 import * as cheerio from 'cheerio';
 import { convertHtmlToMarkdown } from './lib.js';

package/src/github.js

@@ -1,4 +1,3 @@
-import fetch from 'node-fetch';
 import { URL } from 'node:url';
 
 const GITHUB_API_BASE = 'https://api.github.com';

package/src/lib.js

@@ -1,6 +1,5 @@
 /* eslint-disable no-useless-escape */
 // Common logic for the web-capture microservice
-import fetch from 'node-fetch';
 import * as cheerio from 'cheerio';
 import TurndownService from 'turndown';
 import iconv from 'iconv-lite';
@@ -151,7 +150,7 @@ export async function fetchGoogleDriveImage(url) {
   }
 
   return {
-    buffer: await response.buffer(),
+    buffer: Buffer.from(await response.arrayBuffer()),
     contentType,
     downloadUrl,
     filename: `google-drive-${fileId}.${imageExtensionForContentType(contentType)}`,

package/src/localize-images.js

@@ -13,9 +13,8 @@
  * @module localize-images
  */
 
-import fetch from 'node-fetch';
 import { URL } from 'url';
-import { retry } from './retry.js';
+import { downloadBinary } from './binary-download.js';
 
 /**
  * Extract image references from markdown text.
@@ -165,16 +164,7 @@ export async function localizeImages(markdownText, options = {}) {
     }
 
     try {
-      const resp = await retry(() => fetch(image.url), {
-        retries: 3,
-        baseDelay: 1000,
-      });
-
-      if (!resp.ok) {
-        throw new Error(`HTTP ${resp.status}`);
-      }
-
-      const buffer = await resp.buffer();
+      const buffer = await downloadBinary(image.url);
       downloadedCount++;
 
       replacements.push({

package/src/markdown.js

@@ -18,7 +18,6 @@ import { applyImageMode } from './extract-images.js';
 import { createBrowser, getBrowserEngine } from './browser.js';
 import { retry } from './retry.js';
 import { ZipArchive } from 'archiver';
-import fetch from 'node-fetch';
 import * as cheerio from 'cheerio';
 
 const INLINE_MARKDOWN_LINE_LIMIT = 1500;

package/src/shared-dialog.js

@@ -8,7 +8,6 @@
  * unsupported diagnostic instead of guessed content.
  */
 
-import fetch from 'node-fetch';
 import * as he from 'he';
 import { URL } from 'node:url';
 

package/src/stream.js

@@ -1,4 +1,3 @@
-import fetch from 'node-fetch';
 import { convertGoogleDriveUrl } from './lib.js';
 import { copyProxyResponseHeaders } from './proxy-headers.js';
 

package/src/transport.js

@@ -1,5 +1,3 @@
-import fetch from 'node-fetch';
-
 export const RECEIPT_HEADERS = [
   'cache-control',
   'content-encoding',
@@ -22,22 +20,42 @@ function selectedHeaders(headers) {
   return selected;
 }
 
+const TIMEOUT_CAUSE_CODES = new Set([
+  'ETIMEDOUT',
+  'UND_ERR_CONNECT_TIMEOUT',
+  'UND_ERR_HEADERS_TIMEOUT',
+  'UND_ERR_BODY_TIMEOUT',
+]);
+
 function classifyTransportError(error, signal) {
   if (signal?.aborted || error?.name === 'AbortError') {
     return 'cancelled';
   }
-  if (error?.name === 'TimeoutError' || error?.type === 'request-timeout') {
+  if (
+    error?.name === 'TimeoutError' ||
+    TIMEOUT_CAUSE_CODES.has(error?.cause?.code)
+  ) {
     return 'timeout';
   }
-  if (error instanceof TypeError) {
+  // Node's built-in fetch rejects every network failure with
+  // TypeError('fetch failed') and the real reason in `cause`. Browsers give a
+  // cause-less TypeError for CORS and network errors alike.
+  if (error?.name === 'TypeError' && !error.cause) {
     return 'cors_or_transport';
   }
   return 'transport';
 }
 
+/** Message of `error` including its `cause`, e.g. "fetch failed: getaddrinfo ENOTFOUND host". */
+export function describeTransportError(error) {
+  const message = error?.message || String(error);
+  const cause = error?.cause?.message;
+  return cause && !message.includes(cause) ? `${message}: ${cause}` : message;
+}
+
 export class CaptureTransportError extends Error {
   constructor(error, url, signal) {
-    super(error?.message || String(error));
+    super(describeTransportError(error));
     this.name = 'CaptureTransportError';
     this.cause = error;
     this.diagnostics = {

package/src/txt.js

@@ -1,4 +1,3 @@
-import fetch from 'node-fetch';
 import { URL } from 'node:url';
 import {
   fetchHtml,

package/.changeset/README.md

@@ -1,8 +0,0 @@
-# Changesets
-
-Hello and welcome! This folder has been automatically generated by `@changesets/cli`, a build tool that works
-with multi-package repos, or single-package repos to help you version and publish your code. You can
-find the full documentation for it [in our repository](https://github.com/changesets/changesets)
-
-We have a quick list of common questions to get you started engaging with this project in
-[our documentation](https://github.com/changesets/changesets/blob/main/docs/common-questions.md)

package/.changeset/config.json

@@ -1,11 +0,0 @@
-{
-  "$schema": "https://unpkg.com/@changesets/config@3.1.1/schema.json",
-  "changelog": "@changesets/cli/changelog",
-  "commit": false,
-  "fixed": [],
-  "linked": [],
-  "access": "public",
-  "baseBranch": "main",
-  "updateInternalDependencies": "patch",
-  "ignore": []
-}

package/.dockerignore

@@ -1,6 +0,0 @@
-node_modules
-reports
-coverage
-*.log
-.captures
-capture-replay.json

package/.jscpd.json

@@ -1,19 +0,0 @@
-{
-  "minTokens": 30,
-  "minLines": 5,
-  "ignore": [
-    "**/node_modules/**",
-    "**/build/**",
-    "**/dist/**",
-    "**/*.min.js",
-    "**/coverage/**",
-    "**/.changeset/**",
-    "**/package-lock.json",
-    "**/pnpm-lock.yaml",
-    "**/yarn.lock",
-    "**/reports/**"
-  ],
-  "reporters": ["console", "html"],
-  "output": "./reports/jscpd",
-  "mode": "weak"
-}

package/.lenv

@@ -1,8 +0,0 @@
-# web-capture default configuration
-# This file uses Links Notation format (key: value)
-
-# Server port
-PORT: 3000
-
-# Browser engine (puppeteer or playwright)
-BROWSER_ENGINE: puppeteer

package/.prettierignore

@@ -1,8 +0,0 @@
-node_modules
-coverage
-dist
-*.min.js
-package-lock.json
-.eslintcache
-CLAUDE.md
-tests/fixtures/deepwiki/*.md

package/.prettierrc

@@ -1,10 +0,0 @@
-{
-  "semi": true,
-  "trailingComma": "es5",
-  "singleQuote": true,
-  "printWidth": 80,
-  "tabWidth": 2,
-  "useTabs": false,
-  "arrowParens": "always",
-  "endOfLine": "lf"
-}

package/Dockerfile

@@ -1,52 +0,0 @@
-FROM node:24-bookworm
-
-WORKDIR /app
-
-# Install dependencies for headless Chrome (used by both Puppeteer and Playwright)
-# Retry loop handles transient network errors from Debian mirrors in CI
-RUN apt-get update && \
-    for i in 1 2 3; do \
-      apt-get install -y \
-        chromium \
-        fonts-liberation \
-        libasound2 \
-        libatk-bridge2.0-0 \
-        libatk1.0-0 \
-        libatspi2.0-0 \
-        libcups2 \
-        libdbus-1-3 \
-        libdrm2 \
-        libgbm1 \
-        libgtk-3-0 \
-        libnspr4 \
-        libnss3 \
-        libxcomposite1 \
-        libxdamage1 \
-        libxfixes3 \
-        libxrandr2 \
-        libxshmfence1 \
-        xdg-utils \
-        --no-install-recommends && break || \
-      { if [ "$i" -eq 3 ]; then exit 1; fi; \
-        echo "apt-get install attempt $i failed, retrying..."; \
-        apt-get update && sleep 5; }; \
-    done && \
-    rm -rf /var/lib/apt/lists/*
-
-# Set Puppeteer environment variables
-ENV PUPPETEER_SKIP_DOWNLOAD=true
-ENV PUPPETEER_EXECUTABLE_PATH=/usr/bin/chromium
-ENV PUPPETEER_ARGS="--no-sandbox --disable-setuid-sandbox --disable-dev-shm-usage"
-
-# Set Playwright environment variables
-ENV PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1
-ENV PLAYWRIGHT_CHROMIUM_EXECUTABLE_PATH=/usr/bin/chromium
-
-COPY package.json package-lock.json ./
-RUN npm ci
-
-COPY . .
-
-EXPOSE 3000
-
-ENTRYPOINT ["node", "bin/web-capture.js", "--serve"]

package/babel.config.cjs

@@ -1,13 +0,0 @@
-module.exports = {
-  presets: [
-    [
-      '@babel/preset-env',
-      {
-        targets: {
-          node: 'current',
-        },
-        modules: false,
-      },
-    ],
-  ],
-};

package/docker-compose.yml

@@ -1,11 +0,0 @@
-services:
-  web-capture:
-    build:
-      context: .
-      dockerfile: Dockerfile
-    ports:
-      - '3000:3000'
-    extra_hosts:
-      - 'host.docker.internal:host-gateway'
-    environment:
-      - PORT=3000