Commit Graph

4962 Commits

Author SHA1 Message Date
Brendan Dahl
6ea56f35ab Add aria-labels to XFA form elements. (bug 1723422) 2021-08-03 15:58:33 -07:00
Tim van der Meij
85be62c684
Merge pull request #13854 from Snuffleupagus/issue-13851
Prevent breaking errors when an optional content group is undefined (issue 13851)
2021-08-03 23:34:34 +02:00
Tim van der Meij
ad90fe90ed
Merge pull request #13848 from Snuffleupagus/rm-lgtm
Remove the LGTM configuration and inline disable comments (issue 13829)
2021-08-03 23:13:05 +02:00
Jonas Jenwald
766299016f Remove the isEOF helper function and slightly re-factor EOF
Given how trivial the `isEOF` function is, we can simply inline the check at the various call-sites and remove the function (which ought to be ever so slightly more efficient as well).
Furthermore, this patch also changes the `EOF` primitive itself to a `Symbol` instead of an Object since that has the nice benefit of making it unclonable (thus preventing *accidentally* trying to send `EOF` from the worker-thread).
2021-08-03 20:19:32 +02:00
Calixte Denizet
be1ee155d1 XFA - Avoid to put something in very small areas
- it aims to fix #13855.
2021-08-03 17:05:29 +02:00
Jonas Jenwald
d5e14d3dc3 Prevent breaking errors when an optional content group is undefined (issue 13851)
In the referenced PDF document *most* of the form `/Form` XObjects don't have an `/OC` entry, which thus causes the runtime failure during rendering.
2021-08-03 15:59:29 +02:00
Jonas Jenwald
8fef8630fe Remove the LGTM configuration and inline disable comments (issue 13829)
Given that the GitHub Advanced Security workflow now covers everything that LGTM does, but generally faster and with better GitHub-integration, there's no longer much point in also running LGTM separately.
As a follow-up to this patch, we should also disable/remove the LGTM-integration from the PDF.js repository.
2021-08-03 11:14:49 +02:00
Jonas Jenwald
705d1cfad3 Remove useless assignment of availableSpace in the src/core/xfa/template.js file (issue 13829, 13835) 2021-08-03 10:58:57 +02:00
Tim van der Meij
10a1db6980
Merge pull request #13824 from Snuffleupagus/issue-13823
When no "V" entry exists, let the fieldValue fallback to the "DV" entry (issue 13823)
2021-07-30 22:48:38 +02:00
Tim van der Meij
67f4c34f63
Merge pull request #13822 from Snuffleupagus/ReadableStreams-cancel-no-Uncaught_promise
Prevent "Uncaught promise" messages in the console when cancelling (some) `ReadableStream`s
2021-07-30 22:09:29 +02:00
Tim van der Meij
99b14a9da0
Merge pull request #13813 from Snuffleupagus/rm-closure-API
Remove a couple of closures in the `src/display/api.js` file
2021-07-30 21:55:45 +02:00
Jonas Jenwald
ff71be793d When no "V" entry exists, let the fieldValue fallback to the "DV" entry (issue 13823) 2021-07-30 16:17:42 +02:00
Calixte Denizet
7bb5331087 XFA - Avoid an error when an exdata is a string (bug 1723114) 2021-07-30 14:43:53 +02:00
Jonas Jenwald
1df9da949e Prevent "Uncaught promise" messages in the console when cancelling (some) ReadableStreams
While fixing issue 13794, I noticed that cancelling the `ReadableStream` returned by the `PDFPageProxy.streamTextContent`-method could lead to "Uncaught promise" messages in the console.[1]
Generally speaking, we don't really care about errors when *cancelling* a `ReadableStream` and it thus seems reasonable to simply suppress any output in those cases.

---
[1] Although, after that issue was fixed you'd now need to set the API-option `stopAtErrors = true` to actually trigger this.
2021-07-30 14:27:38 +02:00
Jonas Jenwald
5fac0a4350 Simplify some code related to fallbackWorkerSrc and getMainThreadWorkerMessageHandler 2021-07-30 11:34:47 +02:00
Jonas Jenwald
4c679d80ac Remove the closure used with the InternalRenderTask class
This patch utilizes the same approach as used in lots of other parts of the code-base, which thus *slightly* reduces the size of this code.
2021-07-30 11:34:47 +02:00
Jonas Jenwald
b18620ac0f Remove the closure used with the PDFDocumentLoadingTask class
This patch utilizes the same approach as used in lots of other parts of the code-base, which thus *slightly* reduces the size of this code.

By removing some of the (current) indirection, we can also simplify the JSDocs a little bit. Looking at the `gulp jsdoc` output, this actually seem to *improve* the documentation for this class.
2021-07-30 11:34:47 +02:00
Brendan Dahl
4ad5c5d52a
Merge pull request #13808 from brendandahl/pattern-cache-v2
Improve caching of shading patterns. (bug 1721949)
2021-07-28 11:17:16 -07:00
Brendan Dahl
c836e1f0fb Improve caching of shading patterns. (bug 1721949)
The PDF in bug 1721949 uses many unique pattern objects
that references the same shading many times. This caused
a new canvas pattern to be created and cached many times
driving up memory use.

To fix, I've changed the cache in the worker to key off the
shading object and instead send the shading and matrix
separately. While that worked well to fix the above bug,
there could be PDFs that use many shading that could
cause memory issues, so I've also added a LRU cache
on the main thread for canvas patterns. This should prevent
memory use from getting too high.
2021-07-28 10:29:20 -07:00
Calixte Denizet
4a4591bd2c XFA - Fix font scale factors (bug 1720888)
- All the scale factors in for the substitution font were wrong because of different glyph positions between Liberation and the other ones:
    - regenerate all the factors
  - Text may have polish chars for example and in this case the glyph widths were wrong:
    - treat substitution font as a composite one
    - add a map glyphIndex to unicode for Liberation in order to generate width array for cid font
2021-07-28 19:10:42 +02:00
Calixte Denizet
92f4cc52a6 XFA - Add a transparent blue background on all text fields for consistency 2021-07-28 14:47:29 +02:00
Calixte Denizet
76d882b560 XFA - Fix auto-sized fields (bug 1722030)
- In order to better compute text fields size, use line height with no gaps (and consequently guessed height for text are slightly better in general).
  - Fix default background color in fields.
2021-07-28 09:43:15 +02:00
Tim van der Meij
336a74a0e5
Merge pull request #13796 from Snuffleupagus/issue-13794
Allow `StreamsSequenceStream.readBlock` to skip sub-streams with errors (issue 13794)
2021-07-27 22:25:58 +02:00
calixteman
45f3804737
Merge pull request #13807 from calixteman/fulltext
XFA - Get the full value when binding and not only the 1st line (bug 1718725)
2021-07-27 22:22:37 +02:00
Tim van der Meij
e51cbe63bf
Merge pull request #13801 from Snuffleupagus/AnnotationLayer-check-navigator
Access `navigator` safely in the `src/display/annotation_layer.js` file
2021-07-27 22:10:27 +02:00
Calixte Denizet
bd6f55186d XFA - Get the full value when binding and not only the 1st line (bug 1718725) 2021-07-27 20:25:33 +02:00
Jonas Jenwald
4b3ab1472c Access navigator safely in the src/display/annotation_layer.js file
For code that's part of the core library, rather than e.g. the `web/`-folder, we should always be careful about *directly* accessing any DOM methods.
The `navigator` is one such structure, which shouldn't be assumed to always be available and we should thus check that it's actually present.[1]

Hence this patch re-factors the `navigator.platform` access, in the `AnnotationLayer`-code, to ensure that it's generally safe. Furthermore, to reduce unnecessary repeated string-matching to determine the current platform, we're now using a shadowed getter which is evaluated only once instead (at first access).

---
[1] Note e.g. the `isSyncFontLoadingSupported` getter, in the `src/display/font_loader.js` file.
2021-07-27 09:40:42 +02:00
Calixte Denizet
959120e6c9 XFA - Elements created outside of XML must have all their properties (bug 1722029)
- an Image element was created, attached to its parent but the $globalData property was not set and that led to an error.
  - the pdf in bug 1722029 has 27 rendered rows (checked in Acrobat) when only one was displayed: this patch some binding issues around the occur element.
2021-07-26 19:38:52 +02:00
Jonas Jenwald
885e7a8aa4 Allow StreamsSequenceStream.readBlock to skip sub-streams with errors (issue 13794)
This patch makes use of the existing `ignoreErrors` option, thus allowing a page to continue parsing/rendering even if (some of) its sub-streams are corrupt. Obviously this may cause *part* of a page to be broken/missing, however it should be better than (potentially) rendering nothing.
Also, to the best of my knowledge, this is the first bug of its kind that we've encountered.

To avoid having to pass in a bunch of, for a `BaseStream`-instance, mostly unrelated parameters when initializing a `StreamsSequenceStream`-instance, I settled on utilizing a callback function instead to allow conditional Error-suppression.
Note that the `StreamsSequenceStream`-class is a *special* stream-implementation that we only use when the `/Contents`-entry, in the `/Page`-dictionary, consists of an Array with streams.
2021-07-26 16:42:50 +02:00
Jonas Jenwald
e1fa845293 Only define *existing* methods, when converting the OPS format to method-names on the CanvasGraphics.prototype
There's no good reason, as far as I can tell, to explicitly define a bunch of methods to be `undefined`, which the current unconditional "copying" of methods will do.
Note that of the `OPS` ~23 percent don't, for various reasons, have an associated method on the `CanvasGraphics.prototype`.
2021-07-25 13:28:28 +02:00
Jonas Jenwald
fbaafdc4e8 Remove the remaining closure in the src/display/canvas.js file
For e.g. the `gulp mozcentral` command, the *built* `pdf.js` file decreases from `304 607` to `301 295` bytes with this patch. The improvement comes mostly from having less overall indentation in the code.
2021-07-25 13:14:58 +02:00
Jonas Jenwald
833f27c677 Disable a LGTM warning, again (PR 13787 follow-up)
Apparently I didn't put one of the disable comments on the *correct* line, since I didn't read the instructions carefully enough, so let's try again.

Note that, most unfortunately, disabling of warnings isn't applied until *after* a patch has been merged.
2021-07-25 10:32:40 +02:00
Tim van der Meij
41a2b5c809
Merge pull request #13787 from Snuffleupagus/lgtm-fix-warnings
Fix (most) LGTM warnings
2021-07-24 15:20:07 +02:00
Tim van der Meij
7b6767d415
Merge pull request #13784 from Snuffleupagus/issue-13783
When parsing corrupt documents, avoid inserting obviously broken data in the XRef-table (issue 13783)
2021-07-24 14:37:39 +02:00
Tim van der Meij
687cfcecd4
Merge pull request #13786 from Snuffleupagus/rm-more-src-core-closures
Remove a couple of small closures in `src/core/` code
2021-07-24 14:26:57 +02:00
Jonas Jenwald
70bac87fed Fix (most) LGTM warnings
Most of the warnings we don't really care about, and those are simply white-listed using inline comments; however two cases prompted actual code changes:

 - In `src/display/pattern_helper.js` the branch in question is indeed unreachable, and should thus be safe to remove. (This code originated in PR 4192, which is now over seven years ago.)

 - In `test/test.js`, the function in question indeed doesn't accept any arguments. (The patch also re-formats a string just above, which didn't seem worthy of a separated patch.)

This now leaves only *one* warning in the LGTM report, however that one is a false positive that we'll need to report upstream.
2021-07-24 14:23:59 +02:00
Tim van der Meij
9854b85dc1
Merge pull request #13775 from Snuffleupagus/Dict-merge-refactor
Remove some duplication in the `Dict.merge` method
2021-07-24 14:21:41 +02:00
Jonas Jenwald
ebbbc973a5 Remove the closure used with the PostScriptToken class
This patch uses the same approach as used in lots of other parts of the code-base, which thus *slightly* reduces the size of this code.
2021-07-24 13:05:46 +02:00
Jonas Jenwald
81009d42cf Remove the closure used with the PostScriptStack class
This patch uses the same approach as used in lots of other parts of the code-base, which thus *slightly* reduces the size of this code.
2021-07-24 12:59:53 +02:00
Jonas Jenwald
b82c802dff When parsing corrupt documents, avoid inserting obviously broken data in the XRef-table (issue 13783)
In cases where even the very *first* attempt at reading from an object will throw, simply ignoring such objects will help improve rendering of *some* corrupt documents.
Note that this will lead to more parsing in some cases, but considering that this only applies to *corrupt* documents that shouldn't be a big deal.
2021-07-23 18:10:53 +02:00
Jonas Jenwald
51f0a81085
Merge pull request #13770 from brendandahl/cache-pattern
Improve performance of reused patterns.
2021-07-23 10:43:23 +02:00
Brendan Dahl
da1af02ac8 Improve performance of reused patterns.
Bug 1721218 has a shading pattern that was used thousands of times.
To improve performance of this PDF:
 - add a cache for patterns in the evaluator and only send the IR form once
   to the main thread (this also makes caching in canvas easier)
 - cache the created canvas radial/axial patterns
 - for shading fill radial/axial use the pattern directly instead of creating temporary
   canvas
2021-07-22 16:47:40 -07:00
Calixte Denizet
a51c4a3a0f XFA - A field without an ui must provide a default one (bug 1718245) 2021-07-22 20:31:25 +02:00
Jonas Jenwald
e1ee3835cd Remove some duplication in the Dict.merge method
Currently the `!mergeSubDicts` code-path is essentially just duplicated code, which we can easily avoid by simply moving that check. (This may lead to ever so slightly more parsing for this case, but the difference ought to be negligible in practice.)
2021-07-22 14:01:43 +02:00
Jonas Jenwald
2cf90cd9ad
Merge pull request #13766 from Snuffleupagus/issue-13751
XFA - Handle `startIndex` correctly in the `Template.$toHTML` method (issue 13751)
2021-07-21 18:58:29 +02:00
Calixte Denizet
5555114bb3 XFA - Remove namespace from nodes under xfa:data node
- in real life some xfa contains xml like <xfa:data><xfa:Foo><xfa:Bar>...</xfa:data>
    since there are no Foo or Bar in the xfa namespace the JS representation are empty
    and that leads to errors.
  - so the idea is to make all nodes under xfa:data namespace agnostic which means
    that ns are removed from nodes in the parser but only xfa:data descendants.
2021-07-21 17:11:31 +02:00
Jonas Jenwald
7d1c19f8bd XFA - Handle startIndex correctly in the Template.$toHTML method (issue 13751)
*Please note:* The PDF document in issue 13751 is *dynamically* created (in e.g. Adobe Reader), with pages added when certain buttons are clicked, hence this patch simply fixes the breaking error and nothing more.

It looks like the current code contains a little bit too much copy-and-paste from the *similar* `index` branch above, since we cannot set the `startIndex` to a negative value. Note how it's being used to initialize the loop-variable, which is then used to lookup values in an Array and accessing the `-1`th element of an Array obviously makes no sense.
2021-07-21 16:17:13 +02:00
Jonas Jenwald
6c9b6bc599
Merge pull request #13764 from Snuffleupagus/issue-13748
XFA - Add a missing method to `XFAAttribute`, to prevent breaking errors (issue 13748)
2021-07-20 18:55:23 +02:00
Jonas Jenwald
c2fe493abe XFA - Add a missing method to XFAAttribute, to prevent breaking errors (issue 13748)
*This is yet another case where I've got no idea if the patch is correct, but it does at least fix a breaking error :-)*

Note how in the [`Binder._bindValue` method](683ce66a48/src/core/xfa/bind.js (L92-L93)), we're assuming that if a `data`-value exists then it'll also be possible to actually access it. For the `XFAAttribute`-implementation however, the second method is missing and that's what causes the breaking errors in issue 13748.

Please note that another possible way of "fixing" the error wouldn't been to simply change the exists-check to return `false`, and I could see that being a preferred solution.
However, the reason for submitting the current patch is that we get *fewer* warnings about Nodes with mis-matched types this way.
2021-07-20 17:41:05 +02:00
Calixte Denizet
1d07ef597e XFA - Must use bindItems element even if there is no direct binding (bug 1720907) 2021-07-20 17:07:32 +02:00
Jonas Jenwald
cf7978d507 XFA - Prevent breaking errors in Binder, when searchNode doesn't return data (issue 13756)
As can be seen in the code (see below), the `searchNode` helper function will return `null` in some cases and all of its call-sites should protect against that before attempting to access the returned data.
While only one of these changes were necessary to fix the breaking errors in issue 13756, in order to prevent future bugs I've added similar defensive code throughout this file.

 - 07955fa1d3/src/core/xfa/som.js (L169)
 - 07955fa1d3/src/core/xfa/som.js (L239)
 - 07955fa1d3/src/core/xfa/som.js (L254)
2021-07-19 18:07:07 +02:00
Tim van der Meij
07955fa1d3
Merge pull request #13735 from Snuffleupagus/bug-1720411
Ensure that the field value, for checkboxes, refers to an existing appearance state (bug 1720411)
2021-07-18 13:48:34 +02:00
Tim van der Meij
668c58d68d
Merge pull request #13746 from Snuffleupagus/getOperatorList-intent
[api-minor] Add `intent` support to the `PDFPageProxy.getOperatorList` method (issue 13704)
2021-07-18 13:28:08 +02:00
Jonas Jenwald
481af097b4 Convert PDFFunction to a standard class with static methods
For e.g. `gulp mozcentral`, the *built* `pdf.worker.js` file decreases from `1 837 608` to `1 834 907` bytes with this patch-series.
The improvement comes first of all from less overall indentation in `PDFFunction`, and secondly from the removal of (now) unnecessary indirection in the code.
2021-07-17 16:46:57 +02:00
Jonas Jenwald
d35fe3e796 Remove the IR (internal representation) part of the PDFFunction parsing
*This follows the exact same princial as PR 12083, but for the `PDFFunction` parsing instead.*

Given that the IR format is completely unused now, all that the current code does is add a bunch of unnecessary indirection/overhead to the handling of PDF-functions.
2021-07-17 16:44:58 +02:00
Jonas Jenwald
03cf28bf17 [api-minor] Add intent support to the PDFPageProxy.getOperatorList method (issue 13704)
With this patch, the `PDFPageProxy.getOperatorList` method will now return `PDFOperatorList`-instances that also include Annotation-operatorLists (when those exist). Hence this closes a small, but potentially confusing, gap between the `render` and `getOperatorList` methods.

Previously we've been somewhat reluctant to do this, as explained below, but given that there's actual use-cases where it's required probably means that we'll *have* to implement it now.
Since we still need the ability to separate "normal" rendering operations from direct `getOperatorList` calls in the worker-thread, this API-change unfortunately causes the *internal* renderingIntent to become a bit "messy" which is indeed unfortunate (note the `"oplist-"` strings in various spots). As-is I suppose that it's not all that bad, but we may want to consider changing the *internal* renderingIntent to e.g. a bitfield in the future.

Besides fixing issue 13704, this patch would also be necessary if someone ever tries to implement e.g. issue 10165 (since currently `PDFPageProxy.getOperatorList` doesn't include Annotation-operatorLists).

*Please note:* This patch is *also* tagged "api-minor" for a second reason, which is that we're now including the Annotation-id in the `beginAnnotation` argument. The reason for this is to allow correlating the Annotation-data returned by `PDFPageProxy.getAnnotations`, with its corresponding operatorList-data (for those Annotations that have it).
2021-07-16 17:16:30 +02:00
Jonas Jenwald
da808aeab3 Ensure that the field value, for checkboxes, refers to an existing appearance state (bug 1720411)
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1720411
2021-07-16 13:11:48 +02:00
Calixte Denizet
3fb30ddde5 XFA - Checkboxes must be printed (bug 1720182)
- to avoid future regressions, annotationStorage is passed to the xfa render in reftests.
2021-07-16 11:32:03 +02:00
calixteman
4b2e0d0d01
Merge pull request #13732 from calixteman/rect
XFA - A rectangle must have the width of its parent but without inner margins
2021-07-15 22:30:25 +02:00
Jonas Jenwald
3838c4e27c Re-factor the handling of *empty* Name-instances (PR 13612 follow-up)
When working on PR 13612, I mostly prioritized a simple solution that didn't require touching a lot of code. However, while working on PR 13735 I started to realize that the static `Name.empty` construction really wasn't a good idea.

In particular, having a special `Name`-instance where the `name`-property isn't actually a String is confusing (to put it mildly) and can easily lead to issues elsewhere. The only reason for not simply allowing the `name`-property to be an *empty* string, in PR 13612, was to avoid having to touch a lot of existing code. However, it turns out that this is only limited to a few methods in the `PartialEvaluator` and a few of the `BaseLocalCache`-implementations, all of which can be easily re-factored to handle *empty* `Name`-instances.

All-in-all, I think that this patch is even an *overall* improvement since we're now validating (what should always be) `Name`-data better in the `PartialEvaluator`.
This is what I ought to have done from the start, sorry about the code churn here!
2021-07-15 12:00:42 +02:00
calixteman
64f86de5cb
Merge pull request #13734 from calixteman/print_issue
XFA - Cannot print fields with no names
2021-07-14 19:45:53 +02:00
Calixte Denizet
019699acfb XFA - Cannot print fields with no names
- it was not possible to print pdf file in issue #13500.
2021-07-14 17:38:35 +02:00
Calixte Denizet
5081167e7f XFA - A rectangle must have the width of its parent but without inner margins
- it aims to fix #13584;
  - to avoid bad rendering because of clipping just set overflow to visible on SVG element.
2021-07-14 16:46:13 +02:00
Jonas Jenwald
b6c6a0cb7c Avoid all rendering breaking completely when CanvasPattern.setTransform() is unsupported
*Please note:* This patch doesn't fix rendering of (various) patterns in browsers/environments without full `CanvasPattern.setTransform()` support, but it at least prevents outright failures and thus allows the rest of the page to render.

This patch provides a temporary work-around for Firefox 78 ESR[1], and for Node.js environments (see issue 13724), where rendering is currently completely broken.

---
[1] Please note that the `createMatrix` helper function doesn't actually work as intended. The reason is that it's not `DOMMatrix` itself which is unsupported in older Firefox versions, but rather calling `CanvasPattern.setTransform(...)` with a `DOMMatrix`-argument.
Furthermore, the `createSVGMatrix` fallback won't actually help either since that method doesn't accept any parameters and would thus require *manually* specifying the matrix-state; see e.g. https://developer.mozilla.org/en-US/docs/Web/API/CanvasPattern/setTransform#examples
Finally, given that it's less than a month to the [Firefox 91 ESR release](https://wiki.mozilla.org/RapidRelease/Calendar) and that as-is all patterns are completely broken e.g. when using the latest viewer in Firefox 78 ESR, I'm just not convinced that it's worth the "hassle" of providing a more proper work-around.
2021-07-13 19:36:06 +02:00
Calixte Denizet
dd55e76f5d XFA - Avoid to have containers not pushed in the html
- it aims to fix issue #13668.
2021-07-12 21:34:58 +02:00
calixteman
140c2bc563
Revert "XFA - Avoid to have containers not pushed in the html" 2021-07-12 09:46:38 +02:00
calixteman
b6445ddc08
Merge pull request #13716 from calixteman/layout7
XFA - Avoid to have containers not pushed in the html
2021-07-12 09:31:27 +02:00
Calixte Denizet
9bbc194846 XFA - Support assist element 2021-07-11 21:01:18 +02:00
Calixte Denizet
fccc6c2242 XFA - Avoid to have containers not pushed in the html
- it aims to fix issue #13668.
2021-07-11 19:14:44 +02:00
Calixte Denizet
690b5d1941 XFA - Use fake MyriadPro as a fallback for missing fonts
- aims to fix #13597.
2021-07-11 13:52:13 +02:00
calixteman
d416b23898
Merge pull request #13705 from calixteman/lineheight3
XFA - Fix text positions (bug 1718741)
2021-07-10 14:19:03 +02:00
Jonas Jenwald
700b79a305 XFA - Always compute the transformed BBox values in checkDimensions (PR 13691 follow-up)
This way we ensure that these BBox values are *always* defined as expected for every `case`-block, and we also don't need to duplicate the lookup in multiple places. (Also, the patch removes a couple of unnecessary line-breaks in existing comments.)

Fixes https://github.com/mozilla/pdf.js/pull/13691#pullrequestreview-702356627, which was flagged by LGTM.
2021-07-10 11:24:05 +02:00
calixteman
a4f60fc417
Merge pull request #13708 from calixteman/xfa_tab
XFA - Add support for traversal and traverse element
2021-07-09 21:59:50 +02:00
Calixte Denizet
ccac125623 XFA - Add support for traversal and traverse element
- For now, just implement the "next" target in using tabindex attribute of html elements.
2021-07-09 20:50:25 +02:00
calixteman
caaf77375f
Merge pull request #13703 from eltociear/patch-4
XFA - Fix typo in factory.js
2021-07-09 20:12:55 +02:00
Calixte Denizet
58e1f51688 XFA - Fix text positions (bug 1718741)
- font line height is taken into account by acrobat when it isn't with masterpdfeditor: I extracted a font from a pdf, modified some ascent/descent properties thanks to ttx and the reinjected the font in the pdf: only Acrobat is taken it into account. So in this patch, line heights for some substituted fonts are added.
  - it seems that Acrobat is using a line height of 1.2 when the line height in the font is not enough (it's the only way I found to fix correctly bug 1718741).
   - don't use flex in wrapper container (which was causing an horizontal overflow in the above bug).
   - consequently, the above fixes introduced a lot of small regressions, so in order to see real improvements on reftests, I fixed the regressions in this patch:
     - replace margin by padding in some case where padding is a part of a container dimensions;
     - remove some flex display: some containers are wrongly sized when rendered;
     - set letter-spacing to 0.01px: it helps to be sure that text is not broken because of not enough width in Firefox.
2021-07-09 18:11:12 +02:00
Ikko Ashimine
dba30eac7b
XFA - Fix typo in factory.js
occured -> occurred
2021-07-09 21:51:37 +09:00
Calixte Denizet
ad195e0f05 XFA - Scale correctly images 2021-07-08 20:28:49 +02:00
calixteman
c33bf0b5e8
Merge pull request #13692 from calixteman/bind_global
XFA - Correctly bind global data (bug 1718725)
2021-07-08 12:42:22 +02:00
Jonas Jenwald
df6107714a XFA - remove unnecessary check in the handleBreak function (PR 13687 follow-up)
With the changes in PR 13687 we're now checking if `target` is defined *twice* in a row, which shouldn't be necessary :-)

(I noticed this when glancing at the unofficial LGTM results; maybe we should re-evalute the decision to not integrate that into the CI.)
2021-07-07 23:22:29 +02:00
calixteman
fbd2f28618
Merge pull request #13693 from calixteman/textfields
XFA - Enable disabled fields (bug 1719464)
2021-07-07 21:43:21 +02:00
Calixte Denizet
d9a776caf8 XFA - Enable disabled fields (bug 1719464)
- it's a workaround in waiting for JS implementation to let the use fill manually some fields.
2021-07-07 19:11:36 +02:00
Calixte Denizet
8a06df9253 XFA - Handle correctly nested containers with lr-tb layout (bug 1718670)
- and avoid to push a field with no dimensions when we have some available space in width in a parent.
2021-07-07 18:54:32 +02:00
Calixte Denizet
778800a53a XFA - Correctly bind global data (bug 1718725) 2021-07-07 17:36:56 +02:00
Jonas Jenwald
37d2808977 Merge the supplemental font data files used with XFA documents
When XFA support was added, the size of the *built* `pdf.worker.js` file increased quite a bit. Hence I think that it makes sense to, where easily possible, do what we can to (slightly) reduce the size of the PDF.js library.

The supplemental font data files (added for XFA rendering), containing rescale-factors respectively widths, seem like an excellent candidate here since they're not particularly large in either line-count or file sizes.
In this patch these files are instead merged into a *single* file per font, rather than four different ones, and even with these changes the resulting source files don't become all that large.[1]

For e.g. the `gulp mozcentral` build, this reduces the size of the *built* `pdf.worker.js` file by more than `3 kB`. Given the overall simplicity of the patch, that kind of size decrease definitely seem worthwhile to me.

---
[1] Especially when compared to truly large files such as e.g. `glyphlist.js`, `metrics.js`, and `unicode.js`.
2021-07-07 11:56:34 +02:00
calixteman
1eb9a3e9eb
Merge pull request #13687 from calixteman/failing_som
XFA - Don't fail xfa loading because of a JS subexpression in SOM expressions
2021-07-07 11:50:26 +02:00
Calixte Denizet
0486d24e36 XFA - Don't fail xfa loading because of a JS subexpression in SOM expressions
- Fix for one pdf in bug 1717668 (PDFIUM-292-0.pdf).
2021-07-07 10:47:53 +02:00
Jonas Jenwald
05ebb6329b
Merge pull request #13683 from brendandahl/mask-fixes
Fix transformations when painting image masks and tiling patterns.
2021-07-07 10:24:01 +02:00
Brendan Dahl
a52c0c6988 Fix transformations when painting image masks and tiling patterns.
Previously, when we filled image masks we didn't copy over the current transformation,
this caused patterns to be misaligned when painted. Now we create a temporary
canvas with the mask and have the transform copied over and offset it relative to
where the mask would be painted. We also weren't properly offsetting tiling patterns.
This isn't usually noticeable since patters repeat, but in the case of #13561 the pattern
is only drawn once and has to be in the correct position to line up with the mask image.

These fixes broke #11473, but highlighted that we were drawing that correctly by
accident and not correctly handling negative bounding boxes on tiling patterns.

Fixes #6297,  #13561, #13441

Partially fixes #1344 (still blurry but boxes are in correct position now)
2021-07-06 17:29:32 -07:00
Calixte Denizet
c47f0f0f40 XFA - Default background in rectangle is white
- Fix a typo in order to open the pdf in issue #13679
  - After fixing the fill default color there wer some regressions because of z-index
    and when fixing z-index there were some regressions because of borders
  - So fix the borders rendering.
2021-07-06 21:17:20 +02:00
Calixte Denizet
5f76b6370c XFA - Layout correctly a subform with row layout (bug 1718740)
- Fix issues with subformSet elements which are not a real container.
2021-07-06 14:11:25 +02:00
calixteman
ba2d685166
Merge pull request #13673 from calixteman/images2
XFA - An image can be a stream in the pdf (bug 1718521)
2021-07-06 09:53:29 +02:00
calixteman
b9e84ba70e
Merge pull request #13665 from calixteman/reserve
XFA - Fix indentation for justified paragraph
2021-07-05 15:45:59 +02:00
Calixte Denizet
5cdee80c8e XFA - An image can be a stream in the pdf (bug 1718521) - hrefs can be found in catalog > Names > XFAImages 2021-07-05 14:06:23 +02:00
calixteman
02c069481e
Merge pull request #13653 from calixteman/lineheight
XFA - Improve text layout
2021-07-05 13:32:42 +02:00
calixteman
783cbc1793
Revert "XFA - An image can be a stream in the pdf (bug 1718521)" 2021-07-05 12:47:14 +02:00
calixteman
b370d4714f
Merge pull request #13654 from calixteman/images
XFA - An image can be a stream in the pdf (bug 1718521)
2021-07-05 12:04:34 +02:00
Jonas Jenwald
819be0e78b Fix the remaining ESLint operator-assignment errors 2021-07-04 15:23:56 +02:00
Jonas Jenwald
901b24e8af Enable the ESLint operator-assignment rule
This patch was generated automatically, using the `gulp lint --fix` command.

Please find additional details about the ESLint rule at https://eslint.org/docs/rules/operator-assignment
2021-07-04 12:57:45 +02:00
Jonas Jenwald
de80590157
Merge pull request #13662 from calixteman/no_arial
XFA - Don't use system font when a font is not embeded but there is a substitution
2021-07-04 11:35:50 +02:00
Calixte Denizet
9b5574d3ef XFA - Fix indentation for justified paragraph
- and ceil the reserve for a caption to avoid to split it;
  - both issues are present in the pdf in issue #13633.
2021-07-03 18:07:01 +02:00
Calixte Denizet
5744dd773d XFA - Don't use system font when a font is not embeded but there is a substitution
- always use a font coming from pdf.js when there is one: this way we don't use a system font which could looks wrong.
2021-07-03 15:13:56 +02:00
Jonas Jenwald
661c60ecc9 [api-minor] Support accessing both the original and modified PDF fingerprint
The PDF.js API has only ever supported accessing the original file ID, however the second one that (should) exist in *modified* documents have thus far been completely inaccessible through the API.
That seems like a simple oversight, caused e.g. by the viewer not needing it, since it really shouldn't hurt to provide API-users with the ability to check if a PDF document has been modified since its creation.[1]

Please refer to https://www.adobe.com/content/dam/acom/en/devnet/pdf/pdfs/PDF32000_2008.pdf#G13.2261661 for additional information.

For an example of how to update existing code to use the new API, please see the changes in the `web/app.js` file included in this patch.

*Please note:* While I'm not sure if we'll ever be able to remove the old `PDFDocumentProxy.fingerprint` getter, given that it's existed since "forever", that probably isn't a big deal given that it's now limited to only `GENERIC`-builds.

---
[1] Although this obviously depends on the PDF software following the specification, by updating the second file ID as intended.
2021-07-03 13:56:33 +02:00
Tim van der Meij
f9d506cf50
Merge pull request #13658 from Snuffleupagus/cloneValue
Don't attempt to structure clone unsupported types with workers disabled
2021-07-03 13:01:46 +02:00
Jonas Jenwald
bdf6f733bf Don't attempt to structure clone unsupported types with workers disabled
Please refer to https://developer.mozilla.org/en-US/docs/Web/API/Web_Workers_API/Structured_clone_algorithm

Based on that information, and manually testing our code, the implementation in `cloneValue` has the following shortcomings:
 - Attempting to clone `function`s is only prevented when they're part of an Object, but is currently allowed when they occur standalone.
 - Cloning of `Symbol`s is currently not prevented, which it should be since the native structured clone algorithm throws.
 - Any disallowed types should be checked first, to reduce the risk of future changes accidentally allowing something that shouldn't be supported.
2021-07-03 11:56:33 +02:00
Jonas Jenwald
909ff8e29f Replace instanceof Object with typeof checks
Using `instanceof Object` is generally problematic, since it's not guaranteed to always do the right thing for all Objects.
(I stumbled upon this while working on another patch, when I noticed that the `outlineView` was broken with workers disabled.)
2021-07-03 11:30:46 +02:00
Calixte Denizet
f16828be49 XFA - An image can be a stream in the pdf (bug 1718521)
- hrefs can be found in catalog > Names > XFAImages
2021-07-02 20:34:10 +02:00
Calixte Denizet
f7d3b22480 XFA - Improve text layout
- support paragraph margins, line height, letter spacing, ...
  - compute missing dimensions from fields based almost on the dimensions of caption contents.
2021-07-02 17:53:32 +02:00
calixteman
d80651e572
Merge pull request #13598 from calixteman/dhl
XFA - Remove empty pages
2021-06-30 20:43:07 +02:00
calixteman
a8a5c5f10b
Merge pull request #13648 from calixteman/xfa_bg
XFA - Don't fill when the fill element is not visible (bug 1718735)
2021-06-30 18:12:13 +02:00
Calixte Denizet
08e08d5852 XFA - Don't fill when the fill element is not visible (bug 1718735) 2021-06-30 17:14:08 +02:00
Calixte Denizet
ff440d13e7 XFA - Remove empty pages
- it aims to fix #13583;
  - fix the switch to breakBefore target;
  - force the layout of an unsplittable element on an empty page;
  - don't fail when there is horizontal overflow (except in lr-tb);
  - handle correctly overflow in the same content area (bug 1717805, bug 1717668);
  - fix a typo in radial gradient first argument.
2021-06-30 16:32:27 +02:00
Tim van der Meij
6307349e31
Merge pull request #13640 from Snuffleupagus/issue-6759
Add non-PRODUCTION/TESTING overflow `assert`s to various string helper-functions (issue 6759)
2021-06-29 21:22:34 +02:00
calixteman
f35e4cc9ab
Merge pull request #13645 from calixteman/bug1718241
XFA - Choice list has no selected value by default (bug 1718241)
2021-06-28 23:46:59 +02:00
calixteman
04dc902933
Merge pull request #13644 from calixteman/xfa_missing_fonts
XFA - Support non-embedded fonts without a Widths entry
2021-06-28 23:46:09 +02:00
Calixte Denizet
70bb672dcd XFA - Support non-embedded fonts without a Widths entry
- some pdf use some fonts which are not embedded or they don't have any width array or don't have any css info (e.g. for standard fonts or Arial).
  - so add widths arrays for Liberation fonts in order to compute the ones for other fonts in using scale factors array.
2021-06-28 23:05:08 +02:00
Calixte Denizet
1de133a7c9 XFA - Choice list has no selected value by default 2021-06-28 22:10:26 +02:00
Calixte Denizet
71d17b0cc4 XFA - Implement aspect property on image element
- it aims to fix issue #13634;
  - move some img-related functions in test/drivers.js in order to have images in xfa reftests.
2021-06-28 20:43:39 +02:00
Calixte Denizet
b261446981 XFA - Fix width of a container with lr-tb layout (bug 1718037) 2021-06-28 17:47:04 +02:00
calixteman
03dff1c5f5
Merge pull request #13639 from calixteman/old_break
XFA - Replace deprecated break element (bug 1718053)
2021-06-28 17:44:03 +02:00
calixteman
191db4145e
Merge pull request #13642 from calixteman/quotes
XFA - Remove quotes of font name in xhtml
2021-06-28 13:25:47 +02:00
Calixte Denizet
677332aa7b XFA - Remove quotes of font name in xhtml 2021-06-27 18:05:12 +02:00
Jonas Jenwald
273d8cb746 Add non-PRODUCTION/TESTING overflow asserts to various string helper-functions (issue 6759) 2021-06-27 16:06:30 +02:00
Calixte Denizet
257de0e8c5 XFA - Replace deprecated break element (bug 1718053)
- the break element has been deprecated in XFA 2.4 but some old documents can use it, so replace it with one (or more) of its possible substitutions:
    - breakBefore;
    - breakAfter;
    - overflow.
2021-06-27 15:03:00 +02:00
Jonas Jenwald
d02146b13b Add a OnProgressParameters typedef to reduce (some) duplication in src/display/api.js 2021-06-27 11:55:53 +02:00
Jonas Jenwald
ea4b162328 Use the RefProxy typedef in more JSDoc comments in src/display/api.js 2021-06-27 11:34:59 +02:00
Tim van der Meij
d7f8a0e9b9
Merge pull request #13628 from Snuffleupagus/issue-13626
Check that TrueType (3, 0) cmap tables, for symbolic fonts, are sorted correctly (issue 13626)
2021-06-26 14:17:11 +02:00
Calixte Denizet
429ffdcd2f XFA - Save filled data in the pdf when downloading the file (Bug 1716288)
- when binding (after parsing) we get a map between some template nodes and some data nodes;
  - so set user data in input handlers in using data node uids in the annotation storage;
  - to save the form, just put the value we have in the storage in the correct data nodes, serialize the xml as a string and then write the string at the end of the pdf using src/core/writer.js;
  - fix few bugs around data bindings:
    - the "Off" issue in Bug 1716980.
2021-06-25 18:57:01 +02:00
Jonas Jenwald
50edd5da63 Suppress OTS warnings about the caretOffset in the hhea-table
- https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6hhea.html
 - https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6head.html
2021-06-25 17:02:02 +02:00
Jonas Jenwald
185be678ec Check that TrueType (3, 0) cmap tables, for symbolic fonts, are sorted correctly (issue 13626)
According to a comment in `readCmapTable`, we're assuming that the cmap tables (when more than one exist) are sorted in ascending order. If that's not the case, keep checking the following cmap tables in order to fix the referenced issue.
2021-06-25 16:56:00 +02:00
Brendan Dahl
d7fdb72a3f
Merge pull request #13619 from calixteman/bg
XFA - Add back empty subforms (which can have a background)
2021-06-23 16:21:28 -07:00
Brendan Dahl
f4f00a9bc6
Merge pull request #13618 from calixteman/bind_root
XFA - Always bind root subform on root data
2021-06-23 13:14:12 -07:00
Tim van der Meij
f74562b19c
Merge pull request #13613 from Snuffleupagus/xfa-printing-tweaks
[api-minor] Slightly tweak/improve various code related to XFA-printing
2021-06-23 21:56:29 +02:00
Tim van der Meij
ad4b2ce021
Merge pull request #13612 from Snuffleupagus/issue-13610
Support corrupt documents with *empty* `Name`-entries (issue 13610)
2021-06-23 21:49:02 +02:00
Calixte Denizet
b836616667 XFA - Always bind root subform on root data
- it partially fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1717805 (on the data side at least but there is still a layout issue).
2021-06-23 20:46:41 +02:00
Calixte Denizet
f168998d93 XFA - Add back empty subforms (which can have a background) 2021-06-23 19:42:36 +02:00
Calixte Denizet
e82446fa5a XFA - Get line height from the font
- when the CSS line-height property is set to 'normal' then the value depends of the user agent. So use a line height based on the font itself and if for any reasons this value is not available use 1.2 as default.
  - it's a partial fix for https://bugzilla.mozilla.org/show_bug.cgi?id=1717681.
2021-06-23 14:11:10 +02:00
Jonas Jenwald
87be43c193 [api-minor] Add a new getXfaPageViewport helper function to support printing
This patch provides an overall simpler *and* more consistent way of handling the `viewport` parameter during printing of XFA forms, since it's now again guaranteed to always be an instance of `PageViewport`.
Furthermore, for anyone attempting to e.g. implement custom printing of XFA forms this probably cannot hurt either.
2021-06-23 08:17:58 +02:00
Jonas Jenwald
6467907318 Support corrupt documents with *empty* Name-entries (issue 13610)
Apparently some really bad PDF software can create documents with *empty* `Name`-entries, which we thus need to somehow deal with.
While I don't know if this patch is necessarily the best solution, it should at least ensure that the *empty* `Name`-instance cannot accidentally match a proper `Name`-instance (and it doesn't require changes to a lot of existing code).[1]

---
[1] I briefly considered using a `Symbol` rather than an Object, but quickly decided against that since the former one [is not clonable](https://developer.mozilla.org/en-US/docs/Web/API/Web_Workers_API/Structured_clone_algorithm#supported_types) and `Name`-instances may be sent to the API.
2021-06-22 16:55:44 +02:00
Calixte Denizet
aca102a35e XFA - Add margins if needed after having layout some text 2021-06-22 10:36:01 +02:00
calixteman
b886b6c995
Merge pull request #13604 from calixteman/xfa_proto_propr
XFA - A prototype can have a property which needs itself to resolve a proto
2021-06-22 09:58:29 +02:00
calixteman
e84b3bbf6e
Merge pull request #13592 from calixteman/xfa_print_only
XFA - Don't display print-only elements
2021-06-21 21:02:53 +02:00
Calixte Denizet
72c32b3498 XFA - A prototype can have a property which needs itself to resolve a proto 2021-06-21 17:26:29 +02:00
calixteman
56a75f8b26 Revert "Revert "XFA - Fix the way to select page on breaking"" - and fix the error which caused the backout: add an $extra property when creating html. - switch to next content area when breaking on page area. 2021-06-21 17:07:31 +02:00
calixteman
a9385bbb52
Revert "XFA - Fix the way to select page on breaking" 2021-06-21 15:45:04 +02:00
calixteman
da19997781
Merge pull request #13573 from calixteman/bug1716838
XFA - Fix the way to select page on breaking
2021-06-21 15:06:03 +02:00
Calixte Denizet
7aea8faa34 XFA - Fix the way to select page on breaking
- it aims to fix https://bugzilla.mozilla.org/show_bug.cgi?id=1716838.
  - some fonts in the pdf in the bug where bold when they shouldn't so write the font properties in the html to avoid to use some wrong inherited ones.
2021-06-21 12:45:23 +02:00
calixteman
34acf29403
Merge pull request #13593 from calixteman/xfa_rect_border
XFA - Don't display invisible rectangle borders
2021-06-21 12:24:23 +02:00
Calixte Denizet
d99a7c070f XFA - Don't display print-only elements
- partial fix for https://bugzilla.mozilla.org/show_bug.cgi?id=1716980.
2021-06-21 10:08:10 +02:00
Calixte Denizet
7cb92a64b1 XFA - Add support for access property
- it's a partial fix for https://bugzilla.mozilla.org/show_bug.cgi?id=1716816.
2021-06-21 09:56:28 +02:00