+++ This bug was initially created as a clone of Bug #1435814 +++ Changes since last update: #9447 Expose some headers to fetch requests from the PDF.js Chrome extension #9460 Use the correct stream position when reading `maxSizeOfInstructions` from the `maxp` table (issue 9458) #9459 Respect workerSrc if set #9427 Fallback to the built-in JPEG decoder when browser decoding fails, and attempt to handle JPEG images with DNL (Define Number of Lines) markers (issue 8614) #9466 Refactor `PDFThumbnailViewer.scrollThumbnailIntoView` to avoid unnecessary DOM look-ups #9453 Upstream the changes from: Bug 1432992 - Remove definitions of Ci, Cr, Cc, and Cu #9468 Enable the `mozilla/avoid-removeChild` and `mozilla/use-includes-instead-of-indexOf` ESLint rules globally #9465 Attempt to unify the `disableRange`/`contentLength` handling in the various network streams #9472 Improve the interface for `PDFFindController` #9473 Remove the non-`a.click()` case from the `DownloadManager` #9475 Remove the actual `enableHandToolOnLoad` preference (PR 9040 follow-up) #9420 Take the dictionary, and not just the image data, into account when caching inline images (issue 9398) #9479 [api-major] Refactor viewer components initialization to reduce their dependency on the global `PDFJS` object #9401 Make the test framework more resilient to errors. #9470 Ensure that `JpegImage.getData` returns the correct data length when `forceRGBoutput == true` (issue 4888) #9480 [api-major] Move the worker related options from the global `PDFJS` object Looks like bug 792808 recently made changes to PdfStreamConverter.jsm without upstreaming them :(. I've filed an issue upstream for that and will re-apply the change locally for now to un-block this update. https://github.com/mozilla/pdf.js/issues/9499

Pushed by ryanvm@gmail.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/318cb586911e Update pdf.js to version 2.0.385. r=bdahl