This might make debugging intermittent failures a bit easier in the
future because it allows us to spot unexpected differences in the number
of tests being run and allows us to run the tests locally in the same
order in case of intermittent failures.
*The [api-minor] label probably ought to have been added to the original PR, given the changes to the `createAnnotationLayerBuilder` signature (if nothing else).*
This patch fixes the following things:
- Let the `AnnotationLayer.render` method create an `AnnotationStorage`-instance if none was provided, thus making the parameter *properly* optional. This not only fixes the reference tests, it also prevents issues when the viewer components are used.
- Stop exporting `AnnotationStorage` in the official API, i.e. the `src/pdf.js` file, since it's no longer necessary given the change above. Generally speaking, unless absolutely necessary we probably shouldn't export unused things in the API.
- Fix a number of JSDocs `typedef`s, in `src/display/` and `web/` code, to actually account for the new `annotationStorage` parameter.
- Update `web/interfaces.js` to account for the changes in `createAnnotationLayerBuilder`.
- Initialize the storage, in `AnnotationStorage`, using `Object.create(null)` rather than `{}` (which is the PDF.js default).
The f1040-annotations reftest started hanging after #12107. We traced
this to `TypeError: can't access property "getOrCreateValue", storage is
undefined`.
We essentially need to add `annotationStorage` to the parameters in
test/driver.js.
This patch should *hopefully* remove the `Unhandled promise rejection: ...` errors, by returning the "final" promise. Also, by pausing/delaying of rendering slightly the likelihood of the test failing in the first place should thus be reduced.
This uses a similar format to the reference-test logging, and will help determine in *exactly* which browser the failure occurred (since the tests run concurrently).
When the old `Dict.getAll()` method was removed, it was replaced with a `Dict.getKeys()` call and `Dict.get(...)` calls (in a loop).
While this pattern obviously makes a lot of sense in many cases, there's some instances where we actually want the *raw* `Dict` values (i.e. `Ref`s where applicable). In those cases, `Dict.getRaw(...)` calls are instead used within the loop. However, by introducing a new `Dict.getRawValues()` method we can reduce the number of (strictly unnecessary) function calls by simply getting the *raw* `Dict` values directly.
This patch will help pathological cases the most, with issue 2813 being a particularily problematic example. While there's only *four* `/ExtGState` resources, there's a total `29062` of `setGState` operators. Even though parsing of a single `/ExtGState` resource is quite fast, having to re-parse them thousands of times does add up quite significantly.
For simplicity we'll only cache "simple" `/ExtGState` resource, since e.g. the general `SMask` case cannot be easily cached (without re-factoring other code, which may have undesirable effects on general parsing).
By caching "simple" `/ExtGState` resource, we thus improve performance by:
- Not having to fetch/validate/parse the same `/ExtGState` data over and over.
- Handling of repeated `setGState` operators becomes *synchronous* during the `OperatorList` building, instead of having to defer to the event-loop/microtask-queue since the `/ExtGState` parsing is done asynchronously.
---
Obviously I had intended to include (standard) benchmark results with this patch, but for reasons I don't understand the test run-time (even with `master`) of the document in issue 2813 is *a lot* slower than in the development viewer (making normal benchmarking infeasible).
However, testing this manually in the development viewer (using `pdfBug=Stats`) shows a *reduction* of `~10 %` in the rendering time of the PDF document in issue 2813.
Originally there weren't any (generally) good ways to handle errors gracefully, on the worker-side, however that's no longer the case and we can simply fallback to the existing `ignoreErrors` functionality instead.
Also, please note that the "no `/XObject` found"-scenario should be *extremely* unlikely in practice and would only occur in corrupt/broken documents.
Note that the `PartialEvaluator.getOperatorList` case is especially bad currently, since we'll simply (attempt to) send the data as-is to the main-thread. This is quite bad, since in a corrupt/broken document the data *could* contain anything and e.g. be unclonable (which would cause breaking errors).
Also, we're (obviously) not attempting to do anything with this "raw" `OPS.paintXObject` data on the main-thread and simply ensuring that we never send it definately seems like the correct approach.
This special-case was added in PR 1992, however it became unnecessary with the changes in PR 4824 since all of the ColorSpace parsing is now done on the worker-thread (with only RGB-data being sent to the main-thread).
Note how the `getFontID`-method in `src/core/fonts.js` is *completely* global, rather than properly tied to the current document. This means that if you repeatedly open and parse/render, and then close, even the *same* PDF document the `fontID`s will still be incremented continuously.
For comparison the `createObjId` method, on `idFactory`, will always create a *consistent* id, assuming of course that the document and its pages are parsed/rendered in the same order.
In order to address this inconsistency, it thus seems reasonable to add a new `createFontId` method on the `idFactory` and use that when obtaining `fontID`s. (When the current `getFontID` method was added the `idFactory` didn't actually exist yet, which explains why the code looks the way it does.)
*Please note:* Since the document id is (still) part of the `loadedName`, it's thus not possible for different documents to have identical font names.
This moves, and slightly simplifies, code that's currently residing in the unit-test utils into the actual library, such that it's bundled with `GENERIC`-builds and used in e.g. the API-code.
As an added bonus, this also brings out-of-the-box support for CMaps in e.g. the Node.js examples.
The default viewer, and thus Firefox, depends on the `RenderTask.onContinue` functionality to pause/continue rendering (such that the most visible page always renders first).
Despite this functionality thus being very important, it has however never actually been tested *at all* as far as I can tell. Hence this patch which adds a new boolean `renderTaskOnContinue` parameter (`false` by default), that can be used to force a reference-test to use the `RenderTask.onContinue` code-path in the `InternalRenderTask` class.
Note that I purposely made this new reference-test behaviour *optional*, since I didn't want to negatively affect the general runtime of the tests (given that there's a slight delay added to the rendering). Also, for e.g. benchmarking you'd most likely want to stay away from the `RenderTask.onContinue` functionality for similar reasons.
This should reduce the possibility of accidentally truncating some inline images, while *not* causing the "EI" detection to become significantly slower.[1]
There's obviously a possibility that these added checks are not sufficient to catch *every* single case of "EI" sequences within the actual inline image data, but without specific test-cases I decided against over-engineering the solution here.
*Please note:* The interpolation issues are somewhat orthogonal to the main issue here, which is the truncated image, and it's already tracked elsewhere.
---
[1] I've looked at the issue a few times, and this is the first approach that I was able to come up with that didn't cause *unacceptable* performance regressions in e.g. issue 2618.
This patch contains the following *notable* improvements:
- Changes the `ColorSpace.parse` call-sites to, where possible, pass in a reference rather than actual ColorSpace data (necessary for the next point).
- Adds (local) caching of `ColorSpace`s by `Ref`, when applicable, in addition the caching by name. This (generally) improves `ColorSpace` caching for e.g. the SMask code-paths.
- Extends the (local) `ColorSpace` caching to also apply when handling Images and Patterns, thus further reducing unneeded re-parsing.
- Adds a new `ColorSpace.parseAsync` method, almost identical to the existing `ColorSpace.parse` one, but returning a Promise instead (this simplifies some code in the `PartialEvaluator`).
*First of all, I should mention that my understanding of the finer details of the `QueueOptimizer` (and its related `CanvasGraphics` methods) is somewhat limited.*
Hence I'm not sure if there's actually a very good reason for *only* considering ImageMasks where the "skew" transformation matrix elements are zero as *repeated*, however simply looking at the code I just don't see why these elements cannot be non-zero as long as they are *all identical* for the ImageMasks.
Furthermore, looking at the *group* case (which is what we're currently falling back to), there's no particular limitation placed upon the transformation matrix elements.
While this patch obviously isn't enough to *completely* fix the issue, since there should be a visible Pattern rendered as well[1], it seem (at least to me) like enough of an improvement that submitting this is justified.
With these changes the referenced PDF document will no longer hang the *entire* browser, and rendering also finishes in a *reasonable* time (< 10 seconds for me) which seem fine given the *huge* number of identical inline images present.[2]
---
[1] Temporarily changing the Pattern to a solid color *does* render the correct/expected area, which suggests that the remaining problem is a pre-existing issue related to the Pattern-handling itself rather than the `QueueOptimizer` functionality.
[2] The document isn't exactly rendered immediately in e.g. Adobe Reader either.
Because of a really stupid `Promise`-related mistake on my part, when re-factoring `PDFImage.buildImage` during the `NativeImageDecoder` removal, we're no longer re-throwing errors occuring during image parsing/decoding as intended.
The result is that some (fairly) corrupt documents will never finish loading, and unfortunately there were apparently no sufficiently corrupt images in the test-suite to catch this.
Since this helper function is no longer used anywhere in the main code-base, but only in a couple of unit-tests, it's thus being moved to a more appropriate spot.
Finally, the implementation of `isEmptyObj` is also tweaked slightly by removing the manual loop.
On ISO/IEC 10918-6:2013 (E), section 6.1: (http://www.itu.int/rec/T-REC-T.872-201206-I/en)
"Images encoded with three components are assumed to be RGB data encoded as YCbCr unless the image contains an APP14 marker segment as specified in 6.5.3, in which case the colour encoding is considered either RGB or YCbCr according to the application data of the APP14 marker segment"
But common jpeg libraries consider RGB too if components index are ASCII R (0x52), G (0x47) and B (0x42): https://stackoverflow.com/questions/50798014/determining-color-space-for-jpeg/50861048
Issue #11931
Currently some JPEG images are decoded by the built-in PDF.js decoder in `src/core/jpg.js`, while others attempt to use the browser JPEG decoder. This inconsistency seem unfortunate for a number of reasons:
- It adds, compared to the other image formats supported in the PDF specification, a fair amount of code/complexity to the image handling in the PDF.js library.
- The PDF specification support JPEG images with features, e.g. certain ColorSpaces, that browsers are unable to decode natively. Hence, determining if a JPEG image is possible to decode natively in the browser require a non-trivial amount of parsing. In particular, we're parsing (part of) the raw JPEG data to extract certain marker data and we also need to parse the ColorSpace for the JPEG image.
- While some JPEG images may, for all intents and purposes, appear to be natively supported there's still cases where the browser may fail to decode some JPEG images. In order to support those cases, we've had to implement a fallback to the PDF.js JPEG decoder if there's any issues during the native decoding. This also means that it's no longer possible to simply send the JPEG image to the main-thread and continue parsing, but you now need to actually wait for the main-thread to indicate success/failure first.
In practice this means that there's a code-path where the worker-thread is forced to wait for the main-thread, while the reverse should *always* be the case.
- The native decoding, for anything except the *simplest* of JPEG images, result in increased peak memory usage because there's a handful of short-lived copies of the JPEG data (see PR 11707).
Furthermore this also leads to data being *parsed* on the main-thread, rather than the worker-thread, which you usually want to avoid for e.g. performance and UI-reponsiveness reasons.
- Not all environments, e.g. Node.js, fully support native JPEG decoding. This has, historically, lead to some issues and support requests.
- Different browsers may use different JPEG decoders, possibly leading to images being rendered slightly differently depending on the platform/browser where the PDF.js library is used.
Originally the implementation in `src/core/jpg.js` were unable to handle all of the JPEG images in the test-suite, but over the last couple of years I've fixed (hopefully) all of those issues.
At this point in time, there's two kinds of failure with this patch:
- Changes which are basically imperceivable to the naked eye, where some pixels in the images are essentially off-by-one (in all components), which could probably be attributed to things such as different rounding behaviour in the browser/PDF.js JPEG decoder.
This type of "failure" accounts for the *vast* majority of the total number of changes in the reference tests.
- Changes where the JPEG images now looks *ever so slightly* blurrier than with the native browser decoder. For quite some time I've just assumed that this pointed to a general deficiency in the `src/core/jpg.js` implementation, however I've discovered when comparing two viewers side-by-side that the differences vanish at higher zoom levels (usually around 200% is enough).
Basically if you disable [this downscaling in canvas.js](8fb82e939c/src/display/canvas.js (L2356-L2395)), which is what happens when zooming in, the differences simply vanish!
Hence I'm pretty satisfied that there's no significant problems with the `src/core/jpg.js` implementation, and the problems are rather tied to the general quality of the downscaling algorithm used. It could even be seen as a positive that *all* images now share the same downscaling behaviour, since this actually fixes one old bug; see issue 7041.
Currently image resources, as opposed to e.g. font resources, are handled exclusively on a page-specific basis. Generally speaking this makes sense, since pages are separate from each other, however there's PDF documents where many (or even all) pages actually references exactly the same image resources (through the XRef table). Hence, in some cases, we're decoding the *same* images over and over for every page which is obviously slow and wasting both CPU and memory resources better used elsewhere.[1]
Obviously we cannot simply treat all image resources as-if they're used throughout the entire PDF document, since that would end up increasing memory usage too much.[2]
However, by introducing a `GlobalImageCache` in the worker we can track image resources that appear on more than one page. Hence we can switch image resources from being page-specific to being document-specific, once the image resource has been seen on more than a certain number of pages.
In many cases, such as e.g. the referenced issue, this patch will thus lead to reduced memory usage for image resources. Scrolling through all pages of the document, there's now only a few main-thread copies of the same image data, as opposed to one for each rendered page (i.e. there could theoretically be *twenty* copies of the image data).
While this obviously benefit both CPU and memory usage in this case, for *very* large image data this patch *may* possibly increase persistent main-thread memory usage a tiny bit. Thus to avoid negatively affecting memory usage too much in general, particularly on the main-thread, the `GlobalImageCache` will *only* cache a certain number of image resources at the document level and simply fallback to the default behaviour.
Unfortunately the asynchronous nature of the code, with ranged/streamed loading of data, actually makes all of this much more complicated than if all data could be assumed to be immediately available.[3]
*Please note:* The patch will lead to *small* movement in some existing test-cases, since we're now using the built-in PDF.js JPEG decoder more. This was done in order to simplify the overall implementation, especially on the main-thread, by limiting it to only the `OPS.paintImageXObject` operator.
---
[1] There's e.g. PDF documents that use the same image as background on all pages.
[2] Given that data stored in the `commonObjs`, on the main-thread, are only cleared manually through `PDFDocumentProxy.cleanup`. This as opposed to data stored in the `objs` of each page, which is automatically removed when the page is cleaned-up e.g. by being evicted from the cache in the default viewer.
[3] If the latter case were true, we could simply check for repeat images *before* parsing started and thus avoid handling *any* duplicate image resources.
Having `assert` calls without a message string isn't very helpful when debugging, and it turns out that it's easy enough to make use of ESLint to enforce better `assert` call-sites.
In a couple of cases the `assert` calls were changed to "regular" throwing of errors instead, since that seemed more appropriate.
Please find additional details about the ESLint rule at https://eslint.org/docs/rules/no-restricted-syntax
This should ensure that a page will always render successfully, even if there's errors during the Annotation fetching/parsing.
Additionally the `OperatorList.addOpList` method is also adjusted to ignore invalid data, to make it slightly more robust.
