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)
The code responsible for highlighting/scrolling thumbnails is quite old, and parts of it hasn't really changed much over time.
In particular, the `scrollThumbnailIntoView` method is currently using `document.querySelector` in order to find both the new/current thumbnail element. This seems totally unnessary, since we can easily keep track of the current thumbnail (similar to the "regular" viewer) and thus avoid unnecessary DOM look-ups.
Furthermore, note how the `PDFThumbnailView` constructor is currently highlighting the *first* thumbnail. This is yet another leftover from older viewer code, which we can now remove and instead take care of in `PDFThumbnailViewer.setDocument`.
Given that `PDFThumbnailView` does not, nor should it, have any concept of which thumbnail is currently selected, this change also improves the general structure of this code a tiny bit.
Please refer to the specification, found at https://www.w3.org/Graphics/JPEG/itu-t81.pdf#page=49
Given how the JPEG decoder is currently implemented, we need to know the value of the scanLines parameter (among others) *before* parsing of the SOS (Start of Scan) data begins.
Hence the best solution I could come up with here, is to re-parse the image in the *hopefully* rare case of JPEG images that include a DNL (Define Number of Lines) marker.
Fixes 8614.
This works by making `PartialEvaluator.buildPaintImageXObject` wait for the success/failure of `loadJpegStream` on the API side *before* parsing continues.
Please note that in practice, it should be quite rare for the browser to fail loading/decoding of a JPEG image. In the general case, it should thus not be completely surprising if even `src/core/jpg.js` will fail to decode the image.
Since `loadJpegStream` is only used at a *single* spot in the code-base, and given that it's very heavily tailored to the calling code (since it relies on the data structure of `PDFObjects`), this patch simply inlines the code in `src/display/api.js` instead.
This method currently requires a fair number of parameters, which creates quite unwieldy call-sites. When invoking `buildPaintImageXObject`, you have to remember not only which arguments to supply, but also the correct order, to prevent run-time errors.
After the latest update of the minimum supported Firefox version, the development addon is no longer being signed. Hence this patch attempts to address that, by pinning the `minVersion` to a specific version number instead.
*Please note:* The version numbers were taken from https://addons.mozilla.org/en-US/firefox/pages/appversions/, so it ought to have worked.
At this point in time, trying to keep the development addon compatible with prior Firefox versions is already quite difficult and will become even harder very soon.
Please keep in mind that since Firefox 57, only WebExtensions are allowed/possible to install. The only exceptions are Firefox Nightly, with the `xpinstall.signatures.required` preference[1] set to `false`, and the (as of this writing) current ESR release.[2]
With the current compatibility situation, we thus need to effectively support both Nightly *and* ESR in the addon, while trying to keep up with current/upcoming changes in `mozilla-central`. With old addons no longer being officially supported, the amount of old code being removed/refactored is now increasing quite quickly.
*Please note:* The changes proposed here was to a large extent prompted by bugs such as:
- https://bugzilla.mozilla.org/show_bug.cgi?id=1431533 (already landed)
- https://bugzilla.mozilla.org/show_bug.cgi?id=767640 (already landed)
- https://bugzilla.mozilla.org/show_bug.cgi?id=1432992 (still open)
Upstreaming all of those changes for the `MOZCENTRAL` version of PDF.js, while also keeping the Firefox addon running in older versions of the browser, would quickly become non-trivial.
Since we're using the ruleset from `eslint-plugin-mozilla` for the addon files, which is enforced in `mozilla-central`, we must ensure that the neccessary changes are upstreamed to the PDF.js repo such that the `mozilla-central` version of PDF.js can still be updated without failures.
Trying to feature detect, for the `FIREFOX` build target, some of the changes in the referenced bugs would probably become both quite messy and (not to mention) difficult. E.g. with the upcoming and automatically defined `Cc, Ci, Cu, Cr` variables, any sort of feature detection might be tricky since those need to be defined in the global scope of the files in question.
Finally, given the amount of effort that we'd now need to spend to even attempt to keep the Firefox addon compatible, I just don't think it's worth the effort any more. Especially since the number of people that have, thus far, been doing this work is *very* low and those resources would be better spend elsewhere.
Unfortunately, this probably means that the development addon will no longer be compatible with release versions of Seamonkey. However the README has already mentioned, for quite some time, that support isn't guaranteed.
*In closing:* For all of the reasons mentioned above, I thus propose that we reduce the maintenance burden of the Firefox addon by only supporting the current Firefox Nightly.
---
[1] While the preference exists, and can be toggled, its value is ignored in non-Nightly versions of Firefox.
[2] There's unbranded builds of e.g. the release/beta versions of Firefox, where old and non-WebExtensions addons can be installed. However those builds can probably be assumed to be officially unsupported, and thus not recommended for users.
This commit is the first step for extracting a base class for the
`AES128Cipher` and the `AES256Cipher` classes. The objective here is to
make code changes (not altering the logic) to make the implementations
as similar as possible as found by creating a diff of both classes.
In particular, we extract the key size and cycles of repetitions
constants since they are different for AES-128 and AES-256. Moreover, we
rename functions to be similar.
In the `AES256Cipher` class, there was an additional assignment to
`this` in the decryption function. However, this was unnecessary because
the assignment would also be done when the loop was exited.
In the JPEG images in the referenced PDF file, the DHT (Define Huffman Tables) segments contain more data than expected based on the length parameter.
Fixes 9425.
Despite this patch removing the `disableWorker` option itself, please note that we'll still fallback to loading the worker file(s) on the main-thread when running in environments without proper Web Worker support.
Furthermore it's still possible, even with this patch, to force the use of fake workers by manually loading the necessary file using a `<script>` tag on the main-thread.[1]
That way, the functionality of the now removed `SINGLE_FILE` build target and the resulting `build/pdf.combined.js` file can still be achieved simply by adding e.g. `<script src="build/pdf.worker.js"></script>` to the HTML (obviously with the path adjusted as needed).
Finally note that the `disableWorker` option is a performance footgun, and unfortunately many existing third-party examples actually use it without providing any sort of warning/justification.
---
[1] This approach is used in the default viewer, since certain kind of debugging may be easier if the code is running directly on the main-thread.
This method returns the currently used `workerSrc`, which thus allows obtaining the fallback `workerSrc` value (e.g. when the option wasn't set by the user).
Please note that this build target, and the resulting `build/pdf.combined.js` file, is equivalent to setting the `PDFJS.disableWorker` option to `true` which is a performance footgun.
https://crbug.com/362061 was fixed in Chrome 36, and the lowest
supported Chrome version in the extension is Chrome 49, so the
work-around for a filesystem:-bug in chromecom can be removed.
