kindle web DRM has been cracked

[milahu]

good news!

someone (pixelmelt) has cracked the text layer obfuscation
so we can avoid OCR (openAI API calls)
and we get the exact character positions

https://www.reddit.com/r/programming/comments/1o8ob82/bypassing_amazons_kindle_web_drm_because_their/

https://blog.pixelmelt.dev/kindle-web-drm/

in my fork i capture all http traffic
so it should be possible to decode the book text from these files

  const context = await chromium.launchPersistentContext(userDataDir, {
    // ...
    // https://playwright.dev/docs/api/class-browsertype#browser-type-launch-persistent-context-option-record-har
    recordHar: {
      path: path.join(outDir, 'requests.har'),
      mode: 'full',
      // no. this throws "RangeError: Invalid string length" in context.close
      // https://github.com/microsoft/playwright/issues/36707
      // content: 'embed',
      content: 'attach',
      // urlFilter: '**amazon.com/**'
    },
  })

alternative to #12