Adds functionality to accept Queueing Strategy in
sendWithStream method. Using Queueing Strategy we
can control the data that is enqueued into the sink,
and hence regulated the flow of chunks from worker
to main thread.
Adds capability in pull and cancel methods.
Adds ready and desiredSize property in streamSink.
Adds unit test for ReadableStream and sendWithStream.
Fix inconsistent spacing and trailing commas in objects in `src/core/` files, so we can enable the `comma-dangle` and `object-curly-spacing` ESLint rules later on
http://eslint.org/docs/rules/comma-danglehttp://eslint.org/docs/rules/object-curly-spacing
Given that we currently have quite inconsistent object formatting, fixing this in *one* big patch probably wouldn't be feasible (since I cannot imagine anyone wanting to review that); hence I've opted to try and do this piecewise instead.
Please note: This patch was created automatically, using the ESLint `--fix` command line option. In a couple of places this caused lines to become too long, and I've fixed those manually; please refer to the interdiff below for the only hand-edits in this patch.
```diff
diff --git a/src/display/canvas.js b/src/display/canvas.js
index 5739f6f2..4216b2d2 100644
--- a/src/display/canvas.js
+++ b/src/display/canvas.js
@@ -2071,7 +2071,7 @@ var CanvasGraphics = (function CanvasGraphicsClosure() {
var map = [];
for (var i = 0, ii = positions.length; i < ii; i += 2) {
map.push({ transform: [scaleX, 0, 0, scaleY, positions[i],
- positions[i + 1]], x: 0, y: 0, w: width, h: height, });
+ positions[i + 1]], x: 0, y: 0, w: width, h: height, });
}
this.paintInlineImageXObjectGroup(imgData, map);
},
diff --git a/src/display/svg.js b/src/display/svg.js
index 9eb05dfa..2aa21482 100644
--- a/src/display/svg.js
+++ b/src/display/svg.js
@@ -458,7 +458,11 @@ SVGGraphics = (function SVGGraphicsClosure() {
for (var x = 0; x < fnArrayLen; x++) {
var fnId = fnArray[x];
- opList.push({ 'fnId': fnId, 'fn': REVOPS[fnId], 'args': argsArray[x], });
+ opList.push({
+ 'fnId': fnId,
+ 'fn': REVOPS[fnId],
+ 'args': argsArray[x],
+ });
}
return opListToTree(opList);
},
```
Please note that the `glyphlist.js` and `unicode.js` files are converted to CommonJS modules instead, since Babel cannot handle files that large and they are thus excluded from transpilation.
PR 7341 added special handling for `nameddest`s that look like pageNumbers, to prevent issues since we previously *incorrectly* supported specifying a pageNumber directly in the hash; i.e. `#10` versus the correct `#page=10` format.
Since this behaviour wasn't correct, PR 7757 fixed and deprecated the old format, which means that we no longer need to maintain the `nameddest` hack in multiple files.
Added test for ReadableStream.
Adds ref-implementation license-header in streams-lib
and change gulp task to copy external/streams/ in build/
external/streams/ and build/dist/external/streams folder.
Adds README.md and LICENSE.md
For some reason, we're putting all kind of images *except* JPEG into the `imageCache` in `evaluator.js`.[1]
This means that in the PDF file in issue 8380, we'll keep sending the *same* two small images[2] to the main-thread and decoding them over and over. This is obviously hugely inefficient!
As can be seen from the discussion in the issue, the performance becomes *extremely* bad if the user has the addon "Adblock Plus" installed. However, even in a clean Firefox profile, the performance isn't that great.
This patch not only addresses the performance implications of the "Adblock Plus" addon together with that particular PDF file, but it *also* improves the rendering times considerably for *all* users.
Locally, with a clean profile, the rendering times are reduced from `~2000 ms` to `~500 ms` for my setup!
Obviously, the general structure of the PDF file and its operator sequence is still hugely inefficient, however I'd say that the performance with this patch is good enough to consider the issue (as it stands) resolved.[3]
Fixes 8380.
---
[1] Not technically true, since inline images are cached from `parser.js`, but whatever :-)
[2] The two JPEG images have dimensions 1x2, respectively 4x2.
[3] To make this even more efficient, a new state would have to be added to the `QueueOptimizer`. Given that PDF files this stupid fortunately aren't too common, I'm not convinced that it's worth doing.
Currently these methods accept a large number of parameters, which creates quite unwieldy call-sites. When invoking them, you have to remember not only what arguments to supply, but also the correct order, to avoid runtime errors.
Furthermore, since some of the parameters are optional, you also have to remember to pass e.g. `null` or `undefined` for those ones.
Also, adding new parameters to these methods (which happens occasionally), often becomes unnecessarily tedious (based on personal experience).
Please note that I do *not* think that we need/should convert *every* single method in `evaluator.js` (or elsewhere in `/core` files) to take parameter objects. However, in my opinion, once a method starts relying on approximately five parameter (or even more), passing them in individually becomes quite cumbersome.
With these changes, I obviously needed to update the `evaluator_spec.js` unit-tests. The main change there, except the new method signatures[1], is that it's now re-using *one* `PartialEvalutor` instance, since I couldn't see any compelling reason for creating a new one in every single test.
*Note:* If this patch is accepted, my intention is to (time permitting) see if it makes sense to convert additional methods in `evaluator.js` (and other `/core` files) in a similar fashion, but I figured that it'd be a good idea to limit the initial scope somewhat.
---
[1] A fun fact here, note how the `PartialEvaluator` signature used in `evaluator_spec.js` wasn't even correct in the current `master`.
Note that by using `let` instead of `var` in `PartialEvaluator.setGState` and `TranslatedFont.loadType3Data`, we can get rid of further `bind` usages since `let` is block-scoped.
Also, the fact that `bind` wasn't used in the `Font` case inside of `setGState` is actually a bug which has been present ever since PR 5205, where a closure was replaced by a standard loop.[1]
---
[1] I'm not aware of any bugs caused by this, but that is probably more a happy accident than anything else, since e.g. just removing the `bind` from the `SMask` case without using block-scoped variables causes test failures.
This is a regression from commit 3888a993b1.
It turns out the even though we have a `URL` polyfill, it's still dependent on the existence of native `URL.{createObjectURL, revokeObjectURL}` functions.
Since no such thing exists in Node.js, our `createObjectURL` utility function breaks there.
Please see http://eslint.org/docs/rules/object-shorthand.
For the most part, these changes are of the search-and-replace kind, and the previously enabled `no-undef` rule should complement the tests in helping ensure that no stupid errors crept into to the patch.
Please refer to the JBIG2 standard, see https://www.itu.int/rec/dologin_pub.asp?lang=e&id=T-REC-T.88-200002-I!!PDF-E&type=items.
In particular, section "6.3.5.3 Fixed templates and adaptive templates" mentions that the offsets should be *subtracted*; where the offsets are defined according to "Table 6" under section "6.3.2 Input parameters".
Fixes 7145.
Fixes 7308.
Fixes 7401.
Fixes 7850.
Fixes 8270.
Please see http://eslint.org/docs/rules/object-shorthand.
For the most part, these changes are of the search-and-replace kind, and the previously enabled `no-undef` rule should complement the tests in helping ensure that no stupid errors crept into to the patch.
Please see http://eslint.org/docs/rules/object-shorthand.
For the most part, these changes are of the search-and-replace kind, and the previously enabled `no-undef` rule should complement the tests in helping ensure that no stupid errors crept into to the patch.
[api-minor] Always allow e.g. rendering to continue even if there are errors, and add a `stopAtErrors` parameter to `getDocument` to opt-out of this behaviour (issue 6342, issue 3795, bug 1130815)
This patch implements support for line annotations. Other viewers only
show the popup annotation when hovering over the line, which may have
any orientation. To make this possible, we render an invisible line (SVG
element) over the line on the canvas that acts as the trigger for the
popup annotation. This invisible line has the same starting coordinates,
ending coordinates and width of the line on the canvas.
Other PDF readers, e.g. Adobe Reader and PDFium (in Chrome), will attempt to render as much of a page as possible even if there are errors present.
Currently we just bail as soon the first error is hit, which means that we'll usually not render anything in these cases and just display a blank page instead.
NOTE: This patch changes the default behaviour of the PDF.js API to always attempt to recover as much data as possible, even when encountering errors during e.g. `getOperatorList`/`getTextContent`, which thus improve our handling of corrupt PDF files and allow the default viewer to handle errors slightly more gracefully.
In the event that an API consumer wishes to use the old behaviour, where we stop parsing as soon as an error is encountered, the `stopAtErrors` parameter can be set at `getDocument`.
Fixes, inasmuch it's possible since the PDF files are corrupt, e.g. issue 6342, issue 3795, and [bug 1130815](https://bugzilla.mozilla.org/show_bug.cgi?id=1130815) (and probably others too).
- When the minified version is used the resolver of the worker can not find it properly and throws 404 error.
- The problem was that:
- It was getting the current name of the file.
- It was replacing **.js** by **.worker.js**
- When it was loading the unminified version it was working fine because:
- *pdf.js - .js + .worker.js* = **pdf.worker.js**
- When it was loading the minified version it didtn't work because:
- *pdf.min.js - .js + .worker.js* = **pdf.min.worker.js**
- **pdf.min.worker.js** doesn't exist the real file name is **pdf.worker.min.js**
The display layer is responsible for creating the HTML elements for the
annotations from the core layer. If we need to ignore border styling for
the containers of certain elements, the display layer should do so and
not the core layer. I noticed this during the implementation of line
annotations, for which we actually need the original border width in the
display layer, even though we ignore it for the container. If we set the
border style to zero in the core layer, this becomes impossible.
To prevent this, this patch moves the container border removal code from
the core layer to the display layer. This makes the core layer output
the unchanged annotation data and lets the display layer remove any
border styling if necessary.
The existing implementation of fakeRequestAnimationFrame
did not return a timer ID, so the frame could not be cancelled
if you wanted to cancel it. But if you do want to cancel it,
it needs to be cancelled through clearTimeout instead of
cancelAnimationFrame, because the timer IDs are different.
Signed-off-by: Jonathan Barnes <jbarnes@pivotal.io>
I happened to notice that the error handling wasn't that great, which I missed previously since there were no unit-tests for failure to load built-in CMap files.
Hence this patch, which improves the error handling *and* adds tests.
I found that PR 8105 unfortunately causes a *very serious* performance regression in long PDF documents where the `Pages` tree only has one level; my apologies for this!
Obviously we cannot revert that PR, since that would cause more issues than it solves. Hence it seems to me that the only viable solution here, is to add a simple `RefSetCache` to reduce the amount of redundant lookups.
Previously in PR 8105 caching was thought to be unnecessary, but as it turns out I don't think that we really have a choice in the matter any more.
For reasons I don't pretend to understand, we're passing around `xref` arguments to a bunch of methods despite `this.xref` being available in `PartialEvaluator`.
This patch is a small first small step towards cleaning up the, often unwieldy, signatures of methods in `PartialEvaluator`.
I really cannot understand why this change is necessary, since modern browsers such as Firefox and Chrome work just fine with the old code.
Hence this is patch is yet another "hack" that's needed just because IE apparently cannot just work like you'd expect.
For consistency, the Node factory used in the CMap unit-tests is changed as well.
Fixes 8193.
*My apologies for inadvertently breaking this in PR 8064; apparently we don't have any tests that cover this use-case :(*
Without this patch `getTextContent` will fail if called before `getOperatorList`, since loading of fonts during text-extraction may require fetching of built-in CMap files.
*Please note:* The `text` test added here, which uses an already existing PDF file, fails without this patch.
In core/document.js: `PDFDocument.prototype.parse` accesses a dictionary
property, which could throw if the underlying data is not yet available.
In core/obj.js: `get Catalog.prototype.metadata` calls
`stream.getBytes`, which can throw MissingDataException too when the
stream is a ChunkedStream.
Similar to other `try-catch` statements in `/core` code, we must re-throw `MissingDataException` to prevent issues with missing data during document loading.
Note that I'm not sure if/how we can test this, which is why the patch doesn't include any test(s).
Fixes 8180.
*After browsing through (a version of) the JPEG specification, see https://www.w3.org/Graphics/JPEG/itu-t81.pdf, I hope that this patch makes sense.*
Note that while issue 7828 became a problem after PR 7661, it isn't really a regression from than PR. The explanation is rather that we're now relying on `core/jpg.js` instead of the Native Image decoder in more situations than before, which thus exposed an *existing* issue in our JPEG decoder.
Another factor also seems to be that in many JPEG images, the DRI (Define Restart Interval) marker isn't present, in which case this bug won't manifest either.
According to https://www.w3.org/Graphics/JPEG/itu-t81.pdf#page=89 (at the bottom of the page):
"NOTE – The final restart interval may be smaller than the size specified by the DRI marker segment, as it includes only the number of MCUs remaining in the scan."
Furthermore, according to https://www.w3.org/Graphics/JPEG/itu-t81.pdf#page=39 (in the middle of the page):
"[...] If restart is enabled and the restart interval is defined to be Ri, each entropy-coded segment except the last one shall contain Ri MCUs. The last one shall contain whatever number of MCUs completes the scan."
Based on the above, it thus seem to me that we should simply ensure that we're not attempting to continue to parse Scan data once we've found all MCUs (Minimum Coded Unit) of the image.
Fixes 7828.
I happened to notice that some inequalities had the wrong order, and was surprised since I thought that the `yoda` rule should have caught that.
However, reading http://eslint.org/docs/rules/yoda#options a bit more closely than previously, it's quite obvious that the `onlyEquality` option does *exactly* what its name suggests. Hence I think that it makes sense to adjust the options such that only ranges are allowed instead.
This patch gets rid of the only case in the code-base where we're throwing a plain `string`, rather than an `Error`, which besides better/more consistent error handling also allows us to enable the [`no-throw-literal`](http://eslint.org/docs/rules/no-throw-literal) ESLint rule.
This patch is another step towards enabling Babel. Since we're moving
towards ES6 modules, we will not be using UMD headers anymore, so we can
remove the validation.
*This is something that I noticed while working on PR 8126, which is (more) fallout from PR 6065.*
In general, it's actually *not* correct to return `Dict.empty` as the default value for non-existent properties. Please note that a prior PR, see https://github.com/mozilla/pdf.js/pull/5957#issuecomment-103112698, asked for that behaviour but I don't think that's right.
Obviously for properties that are (or should) be `Dict`s it makes sense, however certain properties can be e.g. Strings or Arrays instead. In the latter case, returning `Dict.empty` is just plain wrong, and it's quite fascinating that this hasn't caused any errors in practice. (The existing validation in the various getters has actually saved us here.)
Also, when looking at this code again, it seemed unnecessary to duplicate the `MAX_LOOP_COUNT` check since we could just return immediately instead.
When using content security headers to restrict connections to the same origin,
you may not make connections to `example.com`. This feature detection also
works with a request to the current location.
It appears that I accidentally broke this in PR 6065, sorry about that!
The issue in this particular PDF file is that there's `/Rotate` entries on different levels of the `/Pages` tree. We're supposed to use the `/Rotate` entry in the `/Page` dict (which is `0`), but because of an incorrect condition we instead ended up with the one from the `/Pages` dict (which is `180`).
Fixes 8125.
Ideally, the `Annotation` class should not have anything to do with the
page's operator list. How annotations are added to the page's operator
list is logic that belongs in `src/core/document.js` instead where the
operator list is constructed.
Moreover, some comments have been added to clarify the intent of the
code.
Even though the PDF specification does not state that `Opt` fields are
inheritable, in practice there are PDF generators that let annotations
inherit the options from a parent.
This fixes something that I noticed while working with the code in `Catalog.getPageDict` when debugging issue 8088.
Note that while I don't have an example where this patch really matters, given that e.g. `PartialEvaluator.hasBlendModes` depends on the `objId` to avoid cyclic references this patch could potentially help for some PDF files.
As discussed on IRC, we need to check all nodes at the *bottom* of the tree to ensure that we find the correct `Page` dict.
Furthermore, this patch also gets rid of the caching present in a previous version, since it's not clear if that really helps.
Note that this patch purposely adds an `eq` test, using a reduced test-case, so that we can be sure that the algorithm actually finds the correct `Page` dict for each `pageIndex`.
Fixes 8088.
[api-minor] Refactor fetching of built-in CMaps to utilize a factory on the `display` side instead, to allow users of the API to provide a custom CMap loading factory (e.g. for use with Node.js)
Currently the built-in CMap files are loaded in `src/core/cmap.js` using `XMLHttpRequest` directly. For some environments that might be a problem, hence this patch refactors that to instead use a factory to load built-in CMaps on the main thread and message the data to the worker thread.
This is inspired by other recent work, e.g. the addition of the `CanvasFactory`, and to a large extent on the IRC discussion starting at http://logs.glob.uno/?c=mozilla%23pdfjs&s=12+Oct+2016&e=12+Oct+2016#c53010.
Previously, we had a function called `getDefaultAppearance`. This name,
however, is misleading as the method gets the normal appearance (in the
`N` entry) and not the default appearance (in the `DA` entry). Moreover,
it was not entirely clear how it works just from reading the code. It
primarily lacks comments and explicit error case handling.
This patch improves the situation by fixing the issues mentioned above
and making this function a proper method of the `Annotation` class, just
like e.g., `setColor` and `setBorderStyle`.
This patch basically reverts one aspect of TrueType (3, 1) cmap parsing to the state prior to PR 4259. After that PR, a number of regressions occurred in this particular code-path, which necessitated a number of follow-ups such as PRs 5703, 5743, and 6425.
The empirical data suggests, at least to me, that we should always prefer a (3, 1) cmap for TrueType fonts when they have an encoding, regardless of the Symbolic font flag.
Obviously this patch passes all unit/font/reference tests locally, and I made sure that all the PRs mentioned above landed with test-cases included.
However, in my opinion, there's still a very real possibility that this patch could potentially cause new regressions.
Given that the PDF file in bug 1337429 has been broken for almost *three* years before anyone noticed, and considering that the code-path in question has been the source of numerous regressions, I do *not* intend to request uplift of this patch to previous Firefox versions (assuming that it's even accepted).
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1337429.
*Please note:* The rendering of the PDF file in issue 8061 first regressed in PR 7039, and then PR 7493 exacerbated the problem even further by causing an infinite loop.
In this particular case, when errors were encountered inside of the `Lexer.getObject` method *itself*, we didn't advance the stream position. This thus caused an inifinite loop in `parseCMap`, since the exact same character was then parsed over and over again.
Fixes 8061.
Note that I initially tried to add this as a parameter to the `PDFPageProxy.render` method, such that it could be passed to `PartialEvaluator.getOperatorList`.
However, given all the different code-paths that call `getOperatorList` (there's a bunch only in `annotation.js`), this seemed to very quickly become unwieldy and thus difficult to maintain compared to simply using the existing `evaluatorOptions`.
Fixes extra canvas create calls.
Fixes unnecessary call of `new DOMCanvasFactory`.
Fixes undefined error of DOMCanvasFactory.
Fixes failures in some of the tests.
Fixes expected behaviour.
Remove unused vars.
The `Driver._cleanup` method is removing all stylesheets between test runs, which causes "TypeError: styleElement.parentNode is null" console errors in `FontLoader.clear`.
As can also be seen during various tests, some of the changes I made in PR 7972 unfortunately causes console errors.
It seems that I didn't test this properly, since it *should* have been obvious to me that while tests are triggered using Node.js, the files in question are run within the *browser*.
My apologies for not testing this thoroughly, and for causing unnecessary churn in the code!
See http://eslint.org/docs/rules/brace-style.
Having the opening/closing braces on the same line can often make the code slightly more difficult to read, in particular for `if`/`else if` statements, compared to using new lines.
This patch also, for consistency with `mozilla-central`, enables the [`no-iterator`](http://eslint.org/docs/rules/no-iterator) rule. Note that this rule didn't require a single code change.
This property was added all the way back in PR 542, but hasn't actually been relied upon ever since PR 692.
Note that there's a `isStream()` utility function which replaced the property years ago, hence the `isStream` property is now dead code.
Other PDF viewers, e.g. Adobe Reader, seem to append `FileAttachment`s to their attachments views.
One obvious difference in PDF.js is that we cannot append all the annotations on document load, since that would require parsing *every* page. Despite that, it still seems like a good idea to add `FileAttachment`s, since it's thus possible to access all the various types of attachments from a single place.
*Note:* With the previous patch we display a notification when a `FileAttachment` is added to the sidebar, which thus makes appending the contents of these annotations to the sidebar slightly more visible/useful.
Please see http://eslint.org/docs/rules/no-unused-vars; note that this patch purposely uses the same rule options as in `mozilla-central`, such that it fixes part of issue 7957.
It wasn't, in my opinion, entirely straightforward to enable this rule compared to the already existing rules. In many cases a `var descriptiveName = ...` format was used (more or less) to document the code, and I choose to place the old variable name in a trailing comment to not lose that information.
I welcome feedback on these changes, since it wasn't always entirely easy to know what changes made the most sense in every situation.
Given the nature of `EOF` and `isEOF`, it seems to me that they really ought to be placed in `core/primitives.js` instead.
In general, it doesn't seem great to have to depend on the entire `core/parser.js` file for such simple primitives/helper functions.
In particular, while `core/ps_parser.js` is completely separate from `core/parser.js` with regards to its function, it still depends on the latter for just *one* primitive.
Note that compared to e.g. PR 7389, this will not reduce the number of dependencies for `core/ps_parser`, however the new dependency IMHO makes more sense.
PR 7322 added the `PdfJsNetwork.jsm` file, instead of the general `src/core/network.js` file for the Firefox addon. However, `make.js` wasn't updated to actually stop including the now obsolete network file.
Please see http://eslint.org/docs/rules/spaced-comment.
Note that the exceptions added for `line` comments are intended to still allow use of the old preprocessor without linting errors.
Also, I took the opportunity to improve the grammar slightly (w.r.t. capitalization and punctuation) for comments touched in the patch.
Further adjust the heuristics used to detect OpenType font files with CFF data, to ensure that all Type0 fonts are handled the same way regardless of font Subtype (issue 7901)
We're currently making use of `uniquePrefix`/`idCounters` in multiple files, to create unique object id's, and adding a new occurrence of them requires some care to ensure that an object id isn't accidentally reused.
Furthermore, having to pass around multiple parameters as we currently do seem like something you want to avoid.
Instead, this patch adds a factory which means that there's only *one* thing that needs to be passed around. And since it's now only necessary to call a method in order to obtain a unique object id, the details are thus abstracted away at the call-sites which avoids accidental reuse of object id's.
To test that this works as expected a very simple `Page` unit-test is added, and the existing `Annotation layer` tests are also adjusted slightly.
Using `else` after `return` is not necessary, and can often lead to unnecessarily cluttered code. By using the `no-else-return` rule in ESLint we can avoid this pattern, see http://eslint.org/docs/rules/no-else-return.
We're already passing in a, currently unused, `PdfManager` instance when initializing the `XRef`. To avoid having to pass a single `password` parameter around, we could thus simply get the `password` through the `PdfManager` instance instead.
This patch also removes the `UpdatePassword` message, in favour of using the `sendWithPromise` method of `MessageHandler`.
Furthermore, the patch also refactors the `BasePdfManager_updatePassword`/`BasePdfManager_passwordChanged` methods (in pdf_manager.js), and the `pdfManagerReady` function (in worker.js).
I recently happened to look at the code I wrote for PR 5964, which fixed [bug 1157493](https://bugzilla.mozilla.org/show_bug.cgi?id=1157493), and I quickly realized that the solution is way too simplistic.
The fact that only using the `length` of a `Differences` array worked seems more like a happy accident for a particular set of font data, but could just as easily be incorrect for other PDF files.
Note that in practice, the case where the `Encoding` entry is a regular `Dict` (and not a `Ref` or `Name`) is very rare, hence I don't think that we really need to worry about having to reparse this data.
Also, the performance of this code-block is quite a bit better by updating the `hash` with the data from the *entire* `Differences` array, instead of at every loop iteration.
Changing this particular code makes me somewhat nervous about regressions, since PR 5770 necessitated the follow-up PR 6270.
However, the patch passes all tests added in those PRs (and obviously all other tests). Furthermore, I've manually checked all the issues/bugs referenced in PRs 5770 and 6270 without finding any issues.
**Please note:** This patch fixes *only* the font bug, not the SVG conversion, present on pages two and three of the PDF file in issue 7901.
Modern browsers support styling radio buttons and checkboxes with CSS.
This makes the implementation much easier, and the fallback for older
browsers is still decent.
I haven't got an example where the current code breaks, but given all the previous cases we've seen where PDF generators use indirect objects in Arrays it makes sense to fix this pro-actively.
I've modified the relevant unit-tests slightly, and they would *not* pass without the code changes in this patch.
*Note:* `Dict_getArray` only dereferences Array elements on the "top-level", to avoid recursion issues. Furthermore if you have to loop through the Array at the call-site anyway, then using `Dict_get` in combination with `XRef_fetchIfRef` is a tiny bit more efficient.
*Please note that most of the necessary code adjustments were made in PR 7890.*
ESLint has a number of advantageous properties, compared to JSHint. Among those are:
- The ability to find subtle bugs, thanks to more rules (e.g. PR 7881).
- Much more customizable in general, and many rules allow fine-tuned behaviour rather than the just the on/off rules in JSHint.
- Many more rules that can help developers avoid bugs, and a lot of rules that can be used to enforce a consistent coding style. The latter should be particularily useful for new contributors (and reduce the amount of stylistic review comments necessary).
- The ability to easily specify exactly what rules to use/not to use, as opposed to JSHint which has a default set. *Note:* in future JSHint version some of the rules we depend on will be removed, according to warnings in http://jshint.com/docs/options/, so we wouldn't be able to update without losing lint coverage.
- More easily disable one, or more, rules temporarily. In JSHint this requires using a numeric code, which isn't very user friendly, whereas in ESLint the rule name is simply used instead.
By default there's no rules enabled in ESLint, but there are some default rule sets available. However, to prevent linting failures if we update ESLint in the future, it seemed easier to just explicitly specify what rules we want.
Obviously this makes the ESLint config file somewhat bigger than the old JSHint config file, but given how rarely that one has been updated over the years I don't think that matters too much.
I've tried, to the best of my ability, to ensure that we enable the same rules for ESLint that we had for JSHint. Furthermore, I've also enabled a number of rules that seemed to make sense, both to catch possible errors *and* various style guide violations.
Despite the ESLint README claiming that it's slower that JSHint, https://github.com/eslint/eslint#how-does-eslint-performance-compare-to-jshint, locally this patch actually reduces the runtime for `gulp` lint (by approximately 20-25%).
A couple of stylistic rules that would have been nice to enable, but where our code currently differs to much to make it feasible:
- `comma-dangle`, controls trailing commas in Objects and Arrays (among others).
- `object-curly-spacing`, controls spacing inside of Objects.
- `spaced-comment`, used to enforce spaces after `//` and `/*. (This is made difficult by the fact that there's still some usage of the old preprocessor left.)
Rules that I indend to look into possibly enabling in follow-ups, if it seems to make sense: `no-else-return`, `no-lonely-if`, `brace-style` with the `allowSingleLine` parameter removed.
Useful links:
- http://eslint.org/docs/user-guide/configuring
- http://eslint.org/docs/rules/
This patch refactors the `CropBox` code to combine fetching and
validation code in a getter, like we already did for the `MediaBox`
property. Combined with variable name changes, this improves readability
of the code and makes the `view` getter simpler as well.
I've not actually, thus far, come across a PDF file that this patch fixes. However, given the string of recent patches that has fixed issues with indirect objects in arrays, I think that it makes sense to proactively avoid any issues in this code.
This patch avoids the creation of extra arrays when initializing an
array with default (zero) values. Doing this additionally makes the code
more readable by allocating enough space for the number of color
components.
For Arabic characters, the Unicode character codes are mapped to Unicode
character types using the character codes for indexing. However, the
character code 0x061D is undefined (and therefore invalid) in the
Unicode standard. The imported list does not contain this entry, but not
having it in the list breaks indexing for items after it. Therefore, put
an empty string on its position to make indexing work properly and issue
a warning in the unlikely event that we encounter this character.
*This patch fixes something that I noticed while debugging https://bugzilla.mozilla.org/show_bug.cgi?id=1308536.*
The PDF file contains a font called "NuptialScript", which unfortunately is not embedded. Since that is a non-standard font we will not be able to render it entirely correct. However, by adding "NuptialScript" to the `getNonStdFontMap`, we can at least improve the rendering slightly by using an italic (serif) fallback font.
This patch adds support for non-embedded Arial Black fonts, that use a `Arial-Black...` format for the font names.
Also, this patch changes `canvas.js` such that we always render Arial Black fonts with the maximum weight, which actually improves a number of existing test-cases. This should thus explain the test "failures", which are clear improvements compared with e.g. Adobe Reader.
Fixes 7835.
For `PartialEvaluator_getTextContent`, the same `args` Array should be re-used for every `EvaluatorPreprocessor_read` call. Hence we want to ensure that it's not accidentally replaced with `null` in `EvaluatorPreprocessor_read`, since otherwise corrupt PDF files (with too few arguments for certain commands) will cause errors in `PartialEvaluator_getTextContent`.
Perhaps a micro-optimization, but this patch also changes two `!args` comparisons to `args === null`, since that should be a tiny bit more efficient.
By only allowing very specific type of `JavaScript` actions, and also utilizing the existing `URL` validation, this patch shouldn't pose too much risk.
Fixes one of the points in issue 3897 (with the PDF file taken from issue 3438).
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=843699 (probably, since that bug doesn't contain a test-case).
It seems that certain bad PDF generators can create badly encoded "Prefix" entries for Page Labels, one example being http://ukjewishfilm.org/wp-content/uploads/2015/09/Jewish-Film-Festival-Programme-ONLINE.pdf.
Unfortunately I didn't come across such a PDF file while adding the API support for Page Labels, but with them now being used in the viewer I just found this issue. With this patch, we now display the Page Labels in the same way as Adobe Reader.
The original code is difficult to read and, more importantly, performs
actions that are not described in the specification. It replaces empty
names with a backtick and an index, but this behavior is not described
in the specification. While the specification is not entirely clear
about what should happen in this case, it does specify that the `T`
field is optional and that multiple field dictionaries may have the same
fully qualified name, so to achieve this it makes the most sense to
ignore missing `T` fields during construction of the field name. This is
the most specification-compliant solution and, judging by opened issue #6623, also the required and expected behavior.
In general we neither want, nor can, support arbitrary `Launch` actions. But in practice, all the cases we've seen so far just contains relative URLs to other PDF files. Building on PR 7689, we can thus at least support basic `Launch` actions.
Note that in `FIREFOX/MOZCENTRAL/CHROME` builds of the standard viewer the `docBaseUrl` parameter will be set by default, since in that case it makes sense to use the current URL as a base.
For the `GENERIC` viewer, or the API itself, it doesn't make sense to try and set the `docBaseUrl` by default. However, custom deployments/implementations may still find the parameter useful.
Note that this will automatically reject any relative URL.
To make the API more useful to consumers, URLs that are rejected will be available via the `unsafeUrl` property in the data object returned by `PDFPageProxy_getAnnotations`.
The patch also adds a bit more validation of the data for `Named` actions.
This not only reduces code duplication, but it also allow us to easily support the same kind of URLs we currently do for Link annotations in the Outline as well.
Each well-formed SVG image has the following structure:
SVG element
- Definitions element
- Root group
- Other group 1
- ...
- Other group n
This patch factors out initialization code into a private method in such
as way that the creation of this structure is clear from the code. The
root group is the replacement for the parent group from before. We need
this group as we cannot apply the viewport transform on the SVG element
itself (this caused issues in Chrome). If other code appends groups to
the SVG image, in reality it is appending those groups to the root
group, but this detail is abstracted away by this patch.
This patch ensures that we only create transformation groups when it is
actually required and that we re-use transform groups as much as possible.
It reduces the number of transform groups for the Tracemonkey paper from
2790 to 1271, thereby making the DOM much lighter and rendering/scrolling
smoother. Moreover, it simplifies the code and prevents duplication.
Finally, we issue a warning when an unimplemented graphic state is
encountered. Before, this was ignored silently, making debugging harder.
Let `Parser_makeFilter` pass in the `DecodeParms` data to various image `Stream`s, instead of re-fetching it in various `[...]Stream.prototype.ensureBuffer` methods
In `Parser_filter` the `DecodeParms` data is fetched and passed to `Parser_makeFilter`, where we also make sure that a `Ref` is resolved to a direct object.
We can thus pass this along to the various image `Stream` constructors, to avoid the current situation where we lookup/resolve data that is already available.
Note also that we currently do *not* handle the case where `DecodeParms` is an Array entirely correct in the various image `Stream`s, and this patch fixes that for free.
While the array argument to TJ should only contain strings and numbers, other
unfortunate items are found in PDFs in the wild, e.g.:
[(Grandes) 0.0 Tc
-250.0 (Client\350les,) 0.0 Tc
-250.0 (Financements) 0.0 Tc
-250.0 (et) 0.0 Tc
-250.0 (March\351s) ] TJ
getOperatorList already properly ignores any non-string, non-numeric values in
TJ arrays; without this patch to getTextContent, returned text items can have
NaN widths due to calculations being applied to those non-numeric values.
For PDF files with multiple `/Filter`s, where the `/Length` entry is zero, we fail to render the file correctly. The reason is that `maybeLength` is `null` for the every filter except the first, and `!maybeLength` is thus truthy.
Hence it seems that we should completely ignore the `/Length` entry and also explicitly check `maybeLength === 0`.
Note that I've not (yet) come across a PDF file with this issue in the wild, but given all the stupid things PDF generators do I wouldn't be surprised if such a file actually exists. In order to prevent a possible future bug, I'm submitting this patch which includes a hand-edited PDF file that we currently cannot render correctly (but e.g. Adobe Reader can).
Directly use the hexadecimal representation, just like the
`AnnotationFlags`, to avoid calculations and to improve readability.
This allows us to simplify the unit tests for text widget annotations as
well.
When debugging issue 7643, I noticed that the `forms` tests currently doesn't look like the rendering in the viewer (with `renderInteractiveForms = true` set).
After scratching my head for a little while, I realized that PR 7633 make the implicit assumption that `Page_getOperatorList` (in `core/document.js`) is called *before* fetching the annotation with `PDFPageProxy_getAnnotations` (in `display/api.js`).
Hence this patch, that changes it so that we instead pass in the `renderInteractiveForms` parameter to `Annotation_appendToOperatorList` to ensure that it's always correctly set.
If interactive forms are enabled, then the display layer takes care of
rendering the form elements. There is no need to draw them on the canvas
as well. This also leads to issues when values are prefilled, because
the text fields are transparent, so the contents that have been rendered
onto the canvas will be visible too.
We address this issue by passing the `renderInteractiveForms` parameter
to the render task and handling it when the page is rendered (i.e., when
the canvas is rendered).
This patch improves the unit tests by testing the support for read-only
and multiline fields. Moreover, we add a reference test to ensure that
the text widgets are not only rendered, but also that their contents are
styled properly.
Finally, we perform minor improvements in `src/core/annotation.js`, for
example adding missing comments.
For embedded Type1 fonts without included `ToUnicode`/`Encoding` data, attempt to improve text selection by using the `builtInEncoding` to amend the `toUnicode` map (issue 6901, issue 7182, issue 7217, bug 917796, bug 1242142)
I intended to provide proper benchmarking results here, as outlined in https://github.com/mozilla/pdf.js/wiki/Benchmarking-your-changes, but after wasting a couple of hours over the weekend getting weird results I gave up.
It appears that there's a lot of, i.e. way too much, variance between subsequent runs of `text` tests for the results to be meaningful.
(Previously I've only benchmarked `eq` tests, so I don't know if the `text` tests has never worked well or if it's a newer problem. For reference, please see the results of back-to-back benchmark runs on the current `master` with a *very* simple manifest file: [link here].)
Instead I used `console.time/timeEnd` in `appendText` and `expandTextDivs` to be able to compare the performance with/without the patch. The entire viewer was (skip-cache) reloaded between measurements, and the result are available here: [link here].
Given the troubles I've had with benchmarking, I've not yet computed any statistics on the results (e.g. mean, variance, confidence intervals, and so on).
However, just by looking at the data I think it's safe to say that this patch first of all doesn't seem to regress the current performance. Secondly it certainly looks *very* likely that this patch actually improves the performance, especially for the one-glyph-per-text-div case (cf. issue 7224).
Re: issue 7584.
Even though this patch passes all tests (unit/font/reference) locally, including the new ones that I added in PR 7621, I'm still a bit nervous about modifying the code that choose the fallback encoding for fonts without an `/Encoding` entry.
Note that over the years this code has been changed on a number of occasions, see a possibly incomplete [list here], to deal with various cases of incorrect font data.
According to the PDF specification, see http://www.adobe.com/content/dam/Adobe/en/devnet/acrobat/pdfs/PDF32000_2008.pdf#G8.1904184, it seems that we should fallback to the `StandardEncoding` for Nonsymbolic fonts.
There's obviously a risk that fixing this particular issue *could* break other PDF files for which we don't have tests. However I've tried to change the logic as little as possible in this patch, to hopefully reduce possible breakage.
Based on debugging numerous font issue, it seems that a lot of fonts actually set the Symbolic flag, even when they are in fact *not* Symbolic. Fonts actually marked as Nonsymbolic seem to be somewhat less common, which I hope should reduce the risk of the patch somewhat.
Fixes 7580.
Note that in `parseCodestream` I purposly left the `throw new Error` instances inside of the `try` block, since we don't want to throw any `Errors` while in recovery mode.
Finally somewhat unrelated to the rest of the patch, but I moved the `doNotRecover` variable declaration outside of the `try` block to avoid variable hoisting given that it's accessed inside the `catch` block.
Note that in order to prevent any possible issues, this patch does *not* try to amend the `toUnicode` data for Type1 fonts that contain either `ToUnicode` or `Encoding` entries in the font dictionary.
Fixes, or at least improves, issues/bugs such as e.g. 6658, 6901, 7182, 7217, bug 917796, bug 1242142.
Moreover, we refactor the code a bit to extract code that is shared
between the two branches and we only apply text alignment (and create
the array) when it is actually defined, since it's optional and left is
already the default.
This patch is the first step towards implementing support for
interactive forms (AcroForms). It makes it possible to render text
widget annotations exactly like Adobe Reader/Acrobat.
Everything we implement for AcroForms is disabled by default using a
preference, mainly because it is not ready to use yet, but has to
implemented in many steps to avoid complexity. The preference allows us
to work with the code while not exposing the behavior by default. Mainly
storing entered values and printing them is still absent, which would be
minimal requirements for enabling this by default.
This patch is yet another instalment in the (never ending) series of patches for PDF files that specify completely incorrect Type/Subtype for its fonts. In this case Type1/Type1C, when in fact OpenType would have been correct.
Fixes 7598.
Currently, we only support text widget annotations (field type 'Tx')
partially. However, the current code does not make this entirely clear
and does not provide a warning when an unsupported field type is
encountered, making it harder to determine why rendering fails.
Moreover, in the display layer we make no distinction between the
various types of widget annotations, causing the code for text widget
annotations to also be executed for other types of widget annotations in
a fallback situation.
This patch improves the structure of the widget annotation code. In the
core layer, we use the same structure we use for non-widget annotations
in the factory and provide a clear warning when an unsupported type is
encountered. In the display layer, we do the same and split the
`WidgetAnnotationElement` class into two classes, namely
`TextWidgetAnnotationElement` for text widget annotations and
`WidgetAnnotationElement` for other unsupported annotations as a
fallback. From this it clear that we only support text widget
annotations and nothing else.
In the case where the document was destroyed, we were rejecting the `Promise` in `JpegDecode` with a string instead of an `Error`. The patch also brings the wording more inline with other such rejections.
Use the `isInt` utility function when validating the `pageNumber` parameter in `WorkerTransport_getPage`, to make it more obvious what's actually happening. There's also a couple more unit-tests added, to ensure that we always fail in the expected way.
Finally, we can simplify the rejection handling in `WorkerTransport_getPageIndexByRef` somewhat. (Note that the only reason for using `catch` here is that since the promise is rejected on the worker side, the `reason` becomes a string instead of an `Error` which is why we "re-reject" on the display side.)
This patch avoids having to calculate the scale twice by saving it in
the properties object.
Moreover, we remove a temporary variable and place parentheses around a
calculation inside a string concatenation.
This allows us to remove the `try/catch` statements used in `src/core/stream.js` when parsing JPEG images.
As far as I can tell, the only reason for the current usage of plain `throw` is that `jpg.js` originally was external code. Given that this code now lives in our repo, this patch brings the JPEG code more in line with e.g. `src/core/jpx.js` and `src/core/jbig2.js`.
Assign the `quantizationTables` after parsing the entire JPEG image, to prevent issues when the DQT (Define Quantization Tables) marker is encountered after SOF{n} (Start of Frame) markers (issue 7406)
This patch improves performance by avoiding unnecessary type
conversions, which also help the JIT for optimizations.
Moreover, this patch fixes issues with the div expansion code where
`textScale` would be undefined in a division. Because of the `dataset`
usage, other comparisons evaluated to `true` while `false` would have
been correct. This makes the expansion mode now work correctly for cases
with, for example, each glyph in one div.
The polyfill for `WeakMap` has been provided by @yurydelendik.
We pass many parameters to `appendText` while we might as well pass the
`task` object that contains them. This saves a few lines of code and
makes the signature of `appendText` more clear. We do the same for
`expand`, which is useful for the next commit in which we replace
`div.dataset` with a `WeakMap`.
Furthermore, this patch adds a missing parameter to a comment block to
make it clear which parameters remain.
1. Expanding divs to improve text selection. (Yury)
2. Adding enhanceTextSelection as an option.
3. Moving feature functionality from text_layer_builder.js to text_layer.js.
4. Added expandTextDivs method to only load expanded divs on first click, and only show on subsequent clicks
When adding new entries to `ProblematicCharRanges`, you have to be careful to not make any mistakes since that could cause glyph mapping issues.
Currently the existing reference tests should probably help catch any errors, but based on experience I think that having a unit-test which specifically checks `ProblematicCharRanges` would be both helpful and timesaving when modifying/reviewing changes to this code.
Hence this patch which adds a function (and unit-test) that is used to validate the entries in `ProblematicCharRanges`, and also checks that we don't accidentally add more character ranges than the Private Use Area can actually contain.
The way that the validation code, and thus the unit-test, is implemented also means that we have an easy way to tell how much of the Private Use Area is potentially utilized by re-mapped characters.
Instead of having `Parser_getObj` fail unconditionally for the referenced PDF file, this patch attempts to let searching for the main trailer continue even if there are errors.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1250079.
This is similar to the existing `isCmd` and `isDict` functions, which already support similar kind of checks.
With the updated `isName` function, we'll be able to simplify many callsites from: `isName(someVariable) && someVariable.name === 'someName'` to: `isName(someVariable, 'someName')`.
This patch improves the performance of issue 5808, but I'm not sure if it's enough to call it fixed. On average, this patch reduces the number of textLayer div's by a factor of 3, and it also reduces the time spend in `getTextContent` by a factor of ~2.
The PDF file is generated by `Scribus PDF`, which for reasons I cannot understand is placing redundant `Tf` commands before *every* showText command.
Note how the PDF file also contains lots of (basically) identical fonts, but with slightly different names, which causes unnecessary font-switching. This causes some unnecessary breaking of textLayer div's, but this issue cannot be easily worked around.
[api-minor] Add a parameter to `PDFPageProxy_getTextContent` that controls whether `PartialEvaluator_getTextContent` will attempt to combine same line text items
Note that I used a separate warning message for this case, instead of utilizing the same one as in the unsupported subtype case, to more clearly indicate that the PDF file itself is to blame rather than PDF.js.
Fixes 7446.
Fonts that are not referenced by `Ref`s are very uncommon in practice, but it can unfortunately happen. In this case, we're currently not caching them in the usual way, i.e. by `Ref`, which leads to failures when a page is rendered after `cleanup` has run.
The simplest solution would have been to remove the `font.translated` workaround, but since this would have meant loading these kind of fonts over and over, the patch attempts to be a bit clever about this situation.
Note that if we instead loaded fonts per *page*, instead of per document, this issue wouldn't have existed.
Originally, I was just going to change this code to use `Ref_toString` in a couple more places. When I started reading the code, I figured that it wouldn't hurt to clean up a couple of comments. While doing this, I noticed that the logic for the (rare) `isDict(fontRef)` case could do with a few improvements.
There should be no functional changes with this patch, but given the added reference checks, we will now avoid bogus `Ref`s when resolving font aliases. In practice, as issue 7403 shows, the current code can break certain PDF files even if it's very rare.
Note that the only thing that this patch will change, is the `font.loadedName` in the case where a `fontRef` is a reference *and* the font doesn't have a descriptor. Previously for `fontRef = Ref(4, 0)` we'd get `font.loadedName = 'g_d0_f4_0'`, and with this patch `font.loadedName = g_d0_f4R`, which is actually one character shorted in most cases. (Given that `Ref_toString` contains an optimization for the `gen === 0` case, which is by far the most common `gen` value.)
In the already existing fallback case, where the `fontName` is used to when creating the `font.loadedName`, we allow any alphanumeric character. Hence I don't see how (as mentioned above) e.g. `font.loadedName = g_d0_f4R` would be an issue here.
From the discussion in issue 7445, it seems that there may be cases where an API consumer would want to get the text content as is, without combined text items.
After PR 7039, the PDF file in issue 7492 no longer renders at all, but note that text selection wasn't working correctly previously.
The problem with the PDF file in issue 7492 is that the `cMap`, in the `toUnicode` entry in the font, contains an invalid name:
```
/CMapName /-usr-share-fonts-truetype-Panton-Panton Family-Fontfabric - Panton.otf,000-UTF16 def
```
When we parse that line, things obviously break because there are spaces present in the wrong places.
To avoid that issue, the patch simply lets `parseCMap` continue when errors are encountered, to try and recover usable data. Note that by not aborting immediatly when an error is encountered, we are also able to fix the text selection.
Obviously, it could be argued that we should just immediatly reject a corrupt `cMap`. But given that they usually are correct, it seems that trying to recover as much data as possible from corrupt one can only be a good thing for both glyph mapping and text selection.
Fixes 7492.
In the PDF file in the issue, some of the glyphs end up being mapped to the Lepcha Unicode block; see https://en.wikipedia.org/wiki/Lepcha_(Unicode_block).
This didn't use to matter, but after HarfBuzz updates that improved support for Lepcha fonts, in particular https://bugzilla.mozilla.org/show_bug.cgi?id=1249861, some glyphs are now moved horizontally.
To avoid that, this patch adds the Lepcha block to the list of Unicode ranges that we skip when building the glyph mapping.
Fixes 7426.
After PR 7441, where `recoverGlyphName` is used a lot more than before, many PDF files will generate a lot of warnings the console. For normal usage, compared to debugging/development, this is probably more annoying than helpful.
Fallback to attempt to recover standard glyph names when amending the `charCodeToGlyphId` with entries from the `differences` array in `type1FontGlyphMapping` (issue 7439)
In fonts with only upper-case glyphs, that are also missing a space glyph, `get spaceWidth` won't be able to return anything useful.
By adding upper-case `I` as a fallback, we can thus improve text-selection in some PDF files.
Note that locally, the patch causes slight movement in a few existing `text` tests, but in my opinion this actually looks like slight improvements.
Fixes 7180.
Currently the `isSpace` utility function is a member of `Lexer`, which seems suboptimal, given that it's placed in `core/parser.js`. In practice, this means that in a number of `core/*.js` files we thus have an *otherwise* completely unnecessary dependency on `core/parser.js` for a one-line function.
Instead, this patch moves `isSpace` into `shared/util.js` which seems more appropriate for this kind of utility function. Not to mention that since all the affected `core/*.js` files already depends on `shared/util.js`, this doesn't incur any more file dependencies.
Currently `setGState` is completely broken, and looking through the history of that code, it seems to me that this may never have worked correctly.
This patch fixes the text-selection in `extgstate.pdf` in the test-suite, which is also added as a `text` test.
Fixes http://www.pdf-archive.com/2013/09/30/file2/file2.pdf.
Note how it's not possible to show the various Popup Annotations in the above document.
To fix that, this patch lets the Popup inherit the flags of the parent, in the special case where the parent is `viewable` *and* the Popup is not.
In general, I don't think that a Popup must have the same flags set as the parent. However, it seems very strange to have a `viewable` parent annotation, and then not being able to view the Popup.
Annoyingly the PDF specification doesn't, as far as I can find, mention anything about how this case should be handled, but this patch seem consistent with the actual behaviour in Adobe Reader.
Currently for explicit destinations, compared to named destinations, we manually try to build a hash that often times is a quite poor representation of the *actual* destination. (Currently this only, kind of, works for `\XYZ` destinations.)
For PDF files using explicit destinations, this can make it difficult/impossible to obtain a link to a specific section of the document through the URL.
Note that in practice most PDF files, especially newer ones, use named destinations and these are thus unnaffected by this patch.
This patch also fixes an existing issue in `PDFLinkService_getDestinationHash`, where a named destination consisting of only a number would not be handled correctly.
With the added, and already existing, type checks in place for destinations, I really don't think that this patch exposes any "sensitive" internal destination code not already accessible through normal hash parameters.
*Please note:* Just trying to improve the algorithm that generates the hash is unfortunately not possible in general, since there are a number of cases where it will simply never work well.
- First of all, note that `getDestinationHash` currently relies on the `_pagesRefCache`, hence it's possible that the hash returned is empty during e.g. ranged/streamed loading of a PDF file.
- Second of all, the currently computed hash is actually dependent on the document rotation. With named destinations, the fetched internal destination array is rotational invariant (as it should be), but this will not hold in general for the hash. We can easily avoid this issue by using a stringified destination array.
- Third of all, note that according to the PDF specification[1], `GoToR` destinations may actually contain explicit destination arrays. Since we cannot really construct a hash in `annotation.js`, we currently have no good way to support those. Even though this case seems *very* rare in practice (I've not actually seen such a PDF file), it's in the specification, and this patch allows us to support that for "free".
---
[1] http://www.adobe.com/content/dam/Adobe/en/devnet/acrobat/pdfs/PDF32000_2008.pdf#G11.1951685
Currently the `getPageIndex` method will happily return `0`, even if the `Ref` parameter doesn't actually point to a proper /Page dictionary.
Having the API trust that the consumer is doing the right thing seems error-prone, hence this patch which adds a check for this case.
Given that the `Catalog_getPageIndex` method isn't used in any hot part of the codebase, this extra check shouldn't be a problem.
(Note: in the standard viewer, it is only ever used from `PDFLinkService_navigateTo` if a destination needs to be resolved during document loading, which isn't common enough to be an issue IMHO.)
These have been found using `gulp lint` in combination with the `unused:
true` parameter for JSHint. Unfortunately there are too many false
positives to enable this feature, but now that most globals have been
removed because of the conversion to UMD the results are much more
useful than before.
In the font in question, there are a couple of `topDict` entries that have invalid values (`0xF 0xF`, i.e. just eof markers without any actual numbers).
This causes the `parseFloatOperand` function, inside `CFFParser_parseDict`, to return `NaN`. Currently we pass this broken font onto the browser, which OTS unsurprisingly rejects.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1068432.
As evident from e.g. PRs 6485 and 7118, some bad PDF generators unfortunately create Arrays where *some* elements are indirect objects (i.e. `Ref`s). This seems to mostly affect Arrays that contain numbers, such as e.g. `Matrix/FontMatrix/BBox/FontBBox/Rect/Color/...`, and has manifested itself in PDF files that fail to render correctly (some elements are missing).
The problem in both the cases above, besides broken rendering, was that there were *no* errors/warnings that indicated what the problem was, making it difficult to pinpoint the issue.
Hence this patch, where I've audited all usages of `Dict_get` in `src/core/` files, and replaced it with `Dict_getArray` where appropriate to try and prevent unnecessary future bugs.
This naming issue has been present since PR 3529, but at least I cannot find any issues/bugs that seem to have been caused by it, which is good.
The patch also removes an unnecessary `else` branch, since an already existing `break` means that it's redundant.
Fixes 7232.
- Add support for the 'NewWindow' property.
- Ensure that destinations are applied to the *remote* document, instead of the current one.
- Handle the `F` entry being a standard string, instead of a dictionary.
Using `new {Name,Cmd}` should be avoided, since it creates a new object on *every* call, whereas `{Name,Cmd}.get` uses caches to only create *one* object regardless of how many times they are called.
Most of these are found in the unit-tests, where increased memory usage probably doesn't matter very much. But it still seems good to get rid of those cases, since no part of the codebase ought to advertise that usage.
Given the small size of the patch, I'm also tweaking a few comments and class names.
Some browsers render certain special characters with width 0, others with strictly positive width. (For example, the Greek Delta, Δ, has width 0 in Google Chrome, and a positive width in Firefox.) The `if` clause in operation so far results in different text layer DOM trees for different browsers.
This commit fixes that by adding the elements independently of their width.
Note that in the PDF files provided by the reporter, this issue was limited to `Rect` arrays in AcroForm entries (which we currently don't support).
However, since a bad PDF generator could create this problem in *any* kind of annotation, the reduced test-case included here uses a simple LinkAnnotation instead.
Fixes 7115.
According to "The table directory" under https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6.html#Directory, TrueType font tables should have `uint32` checksums.
This is something that I noticed, and was initially confused about, while debugging a TrueType issue.
As far as I can tell, the current (`int32`) checksums we use doesn't cause any issues in practice. However, I do think that this should be addressed to agree with the specification, and to reduce possible confusion when reading the font code.
Currently there's a lot of duplicate code for non-embedded `toFontChar`, which this patch simplifies by extracting the code into a helper function instead.
This patch adds a `getUnicodeForGlyph` helper function, which is used to recover Unicode values for non-standard glyph names.
Some PDF generators, e.g. Scribus PDF, use improper `uniXXXX` glyph names which breaks the glyph mapping. We can avoid this by converting them to "standard" glyph names instead.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1132849.
Fixes 6893.
Fixes 6894.
In the PDF file in question, some of the 'name' table entries have `record.length === 0`. This becomes problematic in the non-unicode case, since `font.getBytes(0)` will fetch the *entire* stream.
Given that OTS rejects 'name' entries larger than `2^16`, this thus explain the sanitizer errors.
Fixes 7020.
This is required to be able to use it in the annotation display code,
where we now apply it to sanitize the filename of the FileAttachment
annotation. The PDF file from https://bugzilla.mozilla.org/show_bug.cgi?id=1230933 has shown that some PDF generators include the path of the file rather than the filename, which causes filenames with weird initial characters. PDF viewers handle this differently (for example Foxit Reader just replaces forward slashes with spaces), but we think it's better to only show the filename as intended.
Additionally we add unit tests for the `getFilenameFromUrl` helper
function.
*A more robust solution for issue 6066.*
As a temporary work-around for (the upstream) [bug 1164199](https://bugzilla.mozilla.org/show_bug.cgi?id=1164199), we parsed *all* images in the Firefox addon during a short time.
Doing so uncovered an issue with our image handling (see 6066), for JPEG images with a `DeviceGray` ColorSpace *and* `bpc !== 1` (bits per component).
As long as we let the browser handle image decoding in this case, this isn't going to be an issue, but I do think that we should proactively fix this to avoid future issues if we change where the images are decoded (in `jpg.js` vs in browser).
Also, we currently don't seem to have a test-case for that kind of image data.
Currently the `C` entry in an outline item is returned as is, which is neither particularly useful nor what the API documentation claims.
This patch also adds unit-tests for both the color handling, and the `F` entry (bold/italic flags).
`Dict_getAll` is problematic for a number of reasons. First of all, as issue 6961 shows, it can be really bad for performance, since it dereferences all indirect objects.
Second of all, all the derefencing can lead to data being unncessarily requested when ranged/chunked loading is used, thus unnecessarily delaying rendering.
Note: For cases where `Dict_getAll` was previously used, `Dict_getKeys` in combination with `Dict_get` can be used instead. This has the advantage that data isn't requested until it's actually needed.
For the operators that we currently support, the arguments are not `Dict`s, which means that it's not really necessary to use `Dict_getAll` in `EvaluatorPreprocessor_read`.
Also, I do think that if/when we support operators that use `Dict`s as arguments, that should be dealt with in the corresponding `case` in `PartialEvaluator_getOperatorList` which handles the operator.
The only reason that I can find for using `Dict_getAll` like that, is that prior to PR 6550 we would just append certain (currently unsupported) operators without doing any further processing/checking. But as issue 6549 showed, that can lead to issues in practice, which is why it was changed.
In an effort to prevent possible issue with unsupported operators, this patch simply ignores operators with `Dict` arguments in `PartialEvaluator_getOperatorList`.
For the `CalGray`/`CalRGB`/`Lab` colour spaces, we're currently using `getAll` to retrieve the parameters. However that's not really necessary, since we may just as well explicitly `get` the needed parameters instead.
Some bad PDF generators, in particular "Scribus PDF", duplicates resources *a lot* at various levels of the PDF files. This can lead to `PartialEvaluator_hasBlendModes` taking an unreasonable amount of time to complete.
The reason is that the current code is using `Dict_getAll`, which recursively dereferences *all* indirect objects, which can be really slow. This patch instead uses `Dict_getKeys`, and then manually looks up only the necessary indirect objects.
I've added the PDF file as a `load` test. The most important thing here is probably to ensure that the file remains available in the repo, and the comment should help reduced the chance of regressions. (Note that locally, the `load` test times out without this patch, but we cannot really assume that that always happens.)
Fixes 6961.
Reverts "Hack to avoid intermidiate Chrome failures during tests."
(2b2c521213).
require.js uses importScript asynchronously, which activates the worker
GC bug in WebKit. This patch works around a bug in a way that is similar
in the upcoming (but not yet released) require.js 2.1.23
The advantage of the new work-around is that it allows the runtime to
garbage-collect idle Workers.
References:
- https://crbug.com/572225
- https://webkit.org/b/153317
*This patch is based on something I noticed while debugging some of the PDF files in issue 6931.*
In a number of the cases in `setGState`, we're implicitly assuming that we're not dealing with indirect objects (i.e. `Ref`s). See e.g. the 'Font' case, or the various cases where we simply do `gStateObj.push([key, value]);` (since the code in `canvas.js` won't be able to deal with a `Ref` for those cases).
The reason that I didn't use `Dict_forEach` instead, is that it would re-introduce the unncessary closures that PR 5205 removed.
The intention of PR 5192 was to avoid adding empty `setGState` ops to the operatorList. But the patch accidentally used `>=`, which means that it's not actually working as intended, since empty arrays always have `length === 0`.
Even though the currently known test-cases render correctly without this patch, that seems more like a lucky coincidence, given that there's no guarantee that `transferMap[255] === 0` for every possible transfer function.
This patch fixes an issue that I inadvertently introduced in PR 5815, where we accidentally modify the `Differences` array in the encoding dictionary for indirect objects.
Instead of this change, we could also have used the now existing `Dict_getArray`. However in this case I don't think that would have been a good idea, since it would mean iterating through the array *twice*.
Currently the `deprecated` message is using `warn`, meaning that it's possible to disable warnings about deprecated API usage through the `PDFJS.verbosity` setting.
I don't think that it should be possible to opt out of deprecation messages,[1] since it might mean that in a custom deployment of PDF.js these messages could be overlooked, leading to PDF.js being broken (seemingly without any warning) when updating to a future version.
Obviously this could be considered the responsibility of the people doing custom PDF.js implementations, but in order to reduce the support burden later on, it seems better to "annoy" people upfront.
Compared to various `info`/`warn`/`error` messages, `deprecated` messages should be very simple to get rid of -- just update the API usage and the message goes away!
---
[1] In e.g. Firefox it doesn't seem possible to prevent deprecation warnings from being displayed (in the Browser Console).
Re: issue 5089.
(Note that since there are other outline features that we currently don't support, e.g. bold/italic text and custom colours, I thus think we can keep the referenced issue open.)
It seems to be fairly common for OCR software to include incomplete TrueType fonts, notable missing the "glyf" table, in PDF files. Since we currently reject such fonts, the result is that text-selection/copying is broken.
This patch contains a suggested approach to try and use these kind of broken fonts, by using existing code in `sanitizeGlyphLocations` to replace a missing "glyf" table with dummy data.
Fixes 4684.
Fixes 6007.
Fixes 6829.
The Firefox addon currently fails with:
```
SyntaxError: missing ; before statement pdf.js:1692:12
TypeError: PDFJS.shadow is not a function viewer.js:6228:12
```
*This patch follows a similar idea as PR 5756.*
The patch is based on the nice debugging done by Brendan in the referenced issue 6782.
A better way to handle this, and similar issues, would probably be to completely ignore what the PDF file claims about font type/subtype, and just check the actual data. But until that kind of rewrite happens, this patch should help.
Fixes 6782.
Apparently some PDF files can have annotations with `URI` entries ending with `null` characters, thus breaking the links.
To handle this edge-case of bad PDFs, this patch moves the already existing utility function from `ui_utils.js` into `util.js`, in order to fix those URLs.
Fixes 6832.
Currently we're not applying Patterns for text, but only for graphics.
This patch is unfortunately not a complete solution, but rather a step on the way, since there are still some PDF files where the Patterns look more like a solid colour, rather than the intended gradient.
I've been unable to fix these issues completely, and I've not managed to determine if the remaining issues are caused either by the pattern code, the canvas code, or perhaps both.
However, given that even this simple patch improves the current situation quite a bit, I figured that it couldn't hurt to submit it as-is.
- Fixes 5804.
- Fixes 6130.
- Improves 3988 a lot, since the text is now visible. However, it looks like the text is *one* solid colour, instead of the correct gradient.
- Improves 5432, since the text is no longer gray. (This file also suffers from the same problem as the previous one.)
Most code for Popup annotations is already present for Text annotations.
This patch extracts the popup creation logic from the Text annotation
code so it can be reused for Popup annotations.
Not only does this add support for Popup annotations, the Text
annotation code is also considerably easier. If a `Popup` entry is
available for a Text annotation, it will not be more than an image. The
popup will be handled by the Popup annotation. However, it is also
possible for Text annotations to not have a separate Popup annotation,
in which case the Text annotation handles the popup creation itself.
Now we have a full list of all possible annotation types and the
numbering corresponds to the order in the specification. Not only is
this more consistent and complete, it also prevents having to add these
constants when a new annotation type is implemented.
Additionally fix an issue where a regular Widget annotation would not
have `data.annotationType` set. It was only set for a
TextWidgetAnnotation, but instead move it to the base Widget annotation
class to add it for all Widget annotations (since TextWidgetAnnotation
inherits from WidgetAnnotation it will have it too).
Additionally simplify the div creation logic (it needs to happen only
once, so it should not be in the for-loop) and remove/rename variables
for shorter code.
In `Font_checkAndRepair` we can decide that a font isn't TrueType, and instead parse it as CFF. In that case it's quite possible that the `fontMatrix` will be changed, and without calling `adjustWidths` we're failing to update the glyph widths correctly.
Fixes 5027.
Fixes 5084.
Fixes 6556.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1204903.
After PR 6590, `font.spaceWidth` is now called in more cases than before (in `PartialEvaluator_getTextContent`), which exposed an underlying issue with `IdentityToUnicodeMap_charCodeOf` throwing an error.
This breaks text-selection in some PDF files found in the wild, hence this patch replaces the `error` with an actual function instead (modelled after `IdentityCMap_charCodeOf`).
This patch improves the code structure of the annotation code.
- Create the annotation border style object in the `setBorderStyle` method instead of in the constructor. The behavior is the same as the `setBorderStyle` method is always called, thus a border style object is still always available.
- Put all data object manipulation lines in one block in the constructor. This improves readability and maintainability as it is more visible which properties are exposed.
- Simplify `appendToOperatorList` by removing the promise capability and removing an unused parameter.
- Remove some unnecessary newlines/spaces.
*This is a regression from PR 3424.*
The PDF file in the referenced issue is using `Type3` fonts. In one of those, the `/CharProcs` dictionary contains an entry with the name `/#`. Before the changes to `Lexer_getName` in PR 3424, we were allowing certain invalid `Name` patterns containing the NUMBER SIGN (#).
It's unfortunate that this has been broken for close to two and a half years before the bug surfaced, but it should at least indicate that this is not a widespread issue.
Fixes 6692.
This patch goes a bit further than issue 6612 requires, and replaces all kinds of whitespace with standard spaces.
When testing this locally, it actually seemed to slightly improve two existing test-cases (`tracemonkey-text` and `taro-text`).
Fixes 6612.
This patch:
- Updates the JSDoc comment in `api.js`, to more clearly point out that `PDFJS.openExternalLinksInNewWindow` is deprecated, and explains what to use instead.
- Changes the `warn`, in `isExternalLinkTargetSet()`, to use the new `deprecated` function instead. Also updates the message with more detailed information about what to use instead.
- Changes the pre-processor tag to ensure the deprecation warning is seen in all build types where it could possibly matter (in case people are using `PDFJS.openExternalLinksInNewWindow` in e.g. custom-built extensions).
These changes were prompted by seeing http://stackoverflow.com/questions/33813373/pdf-js-how-to-open-hyperlinks-in-a-new-tab-window, since it seems to me that the current comments/warnings might not be worded well enough.
Currently `getAnnotations` will *only* fetch annotations that are either `viewable` or `printable`. This is "hidden" inside the `core.js` file, meaning that API consumers might be confused as to why they are not recieving *all* the annotations present for a page.
I thus think that the API should, by default, return *all* available annotations unless specifically told otherwise. In e.g. the default viewer, we obviously only want to display annotations that are `viewable`, hence this patch adds an `intent` parameter to `getAnnotations` that makes it possible to decide if only `viewable` or `printable` annotations should be fetched.
This patch makes it possible to set and get all possible flags that the PDF specification defines. Even though we do not support all possible annotation types and not all possible annotation flags yet, this general framework makes it easy to access all flags for each annotation such that annotation type implementations can use this information.
We add constants for all possible annotation flags such that we do not need to hardcode the flags in the code anymore. The `isViewable()` and `isPrintable()` methods are now easier to read. Additionally, unit tests have been added to ensure correct behavior.
This is another part of #5218.
I received multiple reports about the following cryptic error in the
Chrome extension when the user tried to open a local file:
> PDF.js v1.1.527 (build: 2096a2a)
> Message: Cannot read property 'Symbol(Symbol.iterator)' of null
This error most likely originated from core/stream.js:
function Stream(arrayBuffer, start, length, dict) {
this.bytes = (arrayBuffer instanceof Uint8Array ?
arrayBuffer : new Uint8Array(arrayBuffer));
^^^^^^^^^^^
`arrayBuffer` is `null`, and that in turn is caused by the fact that
for non-existing files, there is no data. I've applied two fixes:
1. Never call onDone with a void buffer, but call the error handler
instead.
2. Show a sensible error message for local files with status = 0.
In the `RenderPageRequest` handler in `worker.js`, we attempt to print an `info` message containing the rendering time and the length of the operator list. The latter is currently broken (and has been for quite some time), since the `length` of an `OperatorList` is reset when flushing occurs.
This patch attempts to rectify this, by adding a getter which keeps track of the total length.
*Follow-up to PR 6546.*
If rendering has already started when the document is destroyed, then `this.pageCache[data.pageIndex]` may already have been cleared when the `StartRenderingPage`/`RenderPageChunk` messages are recieved in `api.js`, which results in `TypeError`s being thrown.
We should never actually try to execute `paintXObject` in canvas.js, but in some cases where we fail to parse the PDF file correctly it can happen. Currently this will potentially cause an entire page to fail to render, which seems suboptimal.
With this patch, we will instead continue rendering with a warning that things might not work correctly.
`operatorList.addOp` adds the arguments to the list which is then
passed as-is by postMessage to the main thread. But since we don't
parse these operations, they are raw PDF objects and may therefore
cause a serialization error.
This is a conservative patch, and only affects operators which are
known to be unsupported. We should ignore all unknown operators,
but I haven't really looked into the consequences of doing that.
Fixes#6549
In PR 6485 I somehow missed to account for the case where `xref` is undefined. Since a dictonary can be initialized without providing a reference to an `xref` instance, `Dict_getArray` can thus fail without this added check.
Replaces `PDFJS.openExternalLinksInNewWindow` with a more generic configuration option.
*Note:* `PDFJS.openExternalLinksInNewWindow = true;` is equal to `PDFJS.externalLinkTarget = PDFJS.LinkTarget.BLANK;`.
According to the PDF spec 5.3.2, a positive value means in horizontal,
that the next glyph is further to the left (so narrower), and in
vertical that it is further down (so wider).
This change fixes the way PDF.js has interpreted the value.
This patch adjusts `get fingerprint` to also check that the `/ID` entry contains (non-empty) strings, to prevent more possible failures when loading corrupt PDF files (follow-up to PR 5602).
Note that I've not actually encountered such a PDF file in the wild. However given that `stringToBytes` will assert that the input is a string, and that we'll thus fail to load a document unless `get fingerprint` succeeds, making this more robust seems like a good idea to me.
For (1, 0) cmaps, we have two different codepaths depending on whether the font has/hasn't got an encoding. But with (3, 1) cmaps we don't have a good fallback when the encoding is missing, hence this patch changes `readCmapTable` to only choose a (3, 1) cmap table if the font is non-symbolic *and* an encoding exists. Without this, we'll not be able to successfully create a working glyph map for some TrueType fonts with (3, 1) cmap tables.
Fixes 6410.
This code was added in PR 1214, but was made obsolete by PRs 1488/1493. Prior to the latter ones, `Dict_get` retured the raw objects. However, afterwards (and currently) `Dict_get` now resolves indirect objects, which makes `Parser_fetchIfRef` redundant.
*Potential risks with this patch:*
This patch passes all tests locally, but there's a *small* possibility that it could break some weird PDF files.
In the current code, wrapping `Dict_get` inside `Parser_fetchIfRef` will potentially mean two back-to-back call of `XRef_fetch`, if a reference points directly to another reference. I'm not sure if this can actually happen in practice, and I'd think that if that were the case we'd already have run into it elsewhere in the code-base, given that `Parser` is the only place where we try to "double" resolve references.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1200096.
The problematic font has a `format 2` cmap, which we've never supported properly. Prior to PR 2606, we were able to fallback to a working state, despite not having proper support for that cmap format.
Obviously the best/correct solution would be to implement actual support for more cmap formats[1]. However, I'm hoping that a simple patch will be OK for now, given that:
- `format 2` cmaps seem to be quite rare in practice, since this has been broken for 2.5 years before anyone noticed.
- Having a simple patch will make potential uplifts a lot easier.
[1] See the specification at https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6cmap.html
Having a warning here would have meant that issue 6360 could have been solved in approximately five minutes, instead of an hour. To avoid that happening again, this patch adds a warning whenever we treat a stream as empty.
This patch improves the detection of `xref` in files where it is followed by an arbitrary whitespace character (not just a line-breaking char).
It also adds a check for missing whitespace, e.g. `1 0 obj<<`, to speed up `readToken` for the PDF file in the referenced issue.
Finally, the patch also replaces a bunch of magic numbers with suitably named constants.
Fixes 5752.
Also improves 6243, but there are still issues.
The problem with the PDF files in the issue, besides the obviously broken XRef tables which we're able to recover from, is that many/most of the streams have Dictionaries where the `Length` entry is set to `0`. This causes us to return `NullStream`, instead of the appropriate one in `Parser_makeFilter`.
Fixes 6360.
In some cases, such as in use with a CSP header, constructing a function with a
string of javascript is not allowed. However, compiling the various commands
that need to be done on the canvas element is faster than interpreting them.
This patch changes the font renderer to instead emit commands that are compiled
by the font loader. If, during compilation, we receive an EvalError, we instead
interpret them.
Serialize errors to make sure that the callback is still invoked when
an error is thrown.
Firefox:
"DataCloneError: The object could not be cloned."
Chrome:
"DataCloneError: Failed to execute 'postMessage' on 'WorkerGlobalScope': An object could not be cloned."
CMaps may be sparse. Array.prototype.forEach is terribly slow in Chrome
(and also in Firefox) when the sparse array contains a key with a high
value. E.g.
console.time('forEach sparse')
var a = [];
a[0xFFFFFF] = 1;
a.forEach(function(){});
console.timeEnd('forEach sparse');
// Chrome: 2890ms
// Firefox: 1345ms
Switching to CMap.prototype.forEach, which is optimized for such
scenarios fixes the problem.
pi is an index in the stream and is explained on page 201 of the 32000-spec (however 1-based there), and ps is an index to something in PDF.js. I used the code from flag 0 (which works) to understand which is which. It is also important to understand that for flags 1,2 and 3, the stream is always assigned to the same coordinates and colors. What changes is which "old" coordinates and colors are assigned to what is "missing" in the stream. This is why for these flags, the code is identical except for the assignments in the first "row". (Same principle as in #6304). Note that this change will not improve the lamp_cairo.pdf file, only the two files mentioned in #6305.
Short story: somebody got lost in two different indices. pi is an index in the stream and is explained on page 198 of the 32000-spec (however 1-based there), and ps is an index to something in PDF.js. I used the code from flag 0 (which works) to understand which is which. It is also important to understand that for flags 1,2 and 3, the stream is always assigned to the same coordinates and colors. What changes is which "old" coordinates and colors are assigned to what is "missing" in the stream. This is why for these flags, the code is identical except for the assignments in the first "row".
Currently, `src/core/core.js` uses the `fromRef` method on an `Annotation` object to obtain the right annotation type object (such as `LinkAnnotation` or `TextAnnotation`). That method in turn uses a method `getConstructor` to find out which annotation type object must be returned.
Aside from the fact that there is currently a lot of code to achieve this, these methods should not be part of the base `Annotation` class at all. Creation of annotation object should be done by a factory (as also recommended by @yurydelendik at https://github.com/mozilla/pdf.js/pull/5218#issuecomment-52779659) that handles finding out the correct annotation type object and returning it. This patch implements this separation of concerns.
Doing this allows us to also simplify the code quite a bit and to make it more readable. Additionally, we are now able to get rid of the hardcoded array of supported annotation types. The factory takes care of checking the annotation types and falls back to returning the base annotation type (and issuing a warning, which the current code also does not do well) when an annotation type is unsupported.
I have manually tested this commit with 20 test PDFs with different annotation types, such as /Link, /Text, /Widget, /FileAttachment and /FreeText. All render identically before and after the patch, and unsupported annotation types are now properly indicated with a warning in the console.
This patch refactors the code responsible for setting the annotation's rectangle. Its goal is to:
- Actually check that the input array is actually an array, and if so, that it contains exactly four elements.
- Only call `normalizeRect` if the input array is valid, i.e., we do not call it for the default rectangle anymore.
Unit tests are provided just like with the other patches in this series.
Fixes#6106
To avoid future regressions, two new unit tests were added:
1. A new PDF based on the report from #6106, which contains an
OpenAction of type JavaScript and a string "this.print({...}".
2. An existing PDF from https://bugzil.la/1001080 (from #4698).
Although it does not matter, since we don't execute the JavaScript code,
I have also changed "print(true)" to "print({})" since the print method
takes an object (not a boolean). See "Printing PDF documents", page 62:
http://adobe.com/content/dam/Adobe/en/devnet/acrobat/pdfs/js_developer_guide.pdf
Basic mathematics would suggest that a double negative should always become positive, but it appears that Adobe Reader simply ignores that case. Hence I think that it makes sense for us to do the same.
Fixes 6218.
When the parser finds a stream, it retrieves the Length from the stream
dictionary and advances the lexer to the offset as specified in Length.
If this Length is incorrect, the lexer could end up anywhere.
When the lexer gets in an invalid state, it could throw errors. For
example, in issue 6108, the lexer ends up inside the stream data. This
stream has the ASCIIHexDecode filter, so all characters are made up from
ASCII characters, and the lexer interprets it as a command token. Tokens
cannot be longer than 127 bytes, so eventually 128 bytes are consumed
and the lexer throws "Command token too long" error.
Another possible error is "Illegal character: 41" when the lexer happens
to end up at a ')' due to the length mismatch.
These problems are solved by catching lexer errors and recovering the
parser via the existing stream length detection branch.
Xref offsets are relative to the start of the PDF data, not to the start
of the PDF file. This is clear if you look at the other code:
- In the XRef's readXRefTable and processXRefTable methods of XRef, the
offset of a xref entry is set to the bytes as given by a PDF file.
These values are always relative to the start of the PDF file (%PDF-).
- The XRef's readXRef method adds the start offset of the stream to
Xref entry's offset: "stream.pos = startXRef + stream.start".
Clearly, this line assumes that the entry offset excludes the start
offset.
However, when the PDF is parsed in recovery mode, the xref table is
filled with entries whose offset is relative to the start of the stream
rather than the PDF file. This is incorrect, and the fix is to subtract
the start offset of the stream from the entry's byte offset.
The manually created PDF file serves as a regression test. It is a valid
PDF, except:
- The integer to point to the start of the xref table and the %%EOF
trailer are missing. This will activate recovery mode in PDF.js
- Some junk was added before the start of the PDF file. This exposes the
bad offset bug.
The PDF specification (cited below) specifies a maximum length of a name
in bytes as a minimal architectural limit. This means that PDF *writers*
should not create names that exceed 127 bytes.
It does not forbid PDF *readers* to accept such names though. These
names are only used internally to link PDF objects to other objects. For
these use cases, the lengths of the names do not really matter. Hence I
have changed the implementation to not treat long names as errors, but
warnings.
> (7.3.5) The length of a name shall be subject to an implementation
> limit; see Annex C.
>
> (Annex C.2) Table C.1 describes the minimum architectural limits that
> should be accommodated by conforming readers running on 32-bit
> machines. Because conforming readers may be subject to these limits,
> conforming writers producing PDF files should remain within them.
>
> (Table C.1) name 127 "Maximum length of a name, in bytes."
http://adobe.com/content/dam/Adobe/en/devnet/acrobat/pdfs/PDF32000_2008.pdf
For named destinations that are contained in a `Dict`, as opposed to a `NameTree`, we currently iterate through the *entire* dictionary just to fetch *one* destination.
This code appears to simply have been copy-pasted from the `get destinations` method, but in its current form it's quite unnecessary/inefficient since can just get the required destination directly instead.
Doing this helped uncover an issue with the `getDestination` implementation.
Currently if a named destination doesn't exist, the method (in `obj.js`) may return `undefined` which leads to the promise being stuck in a pending state.
*Note:* returning `null` for this case is consistent with other methods, e.g. `getOutline` and `getAttachments`.
This became obsolete in bdeca30fbf. All it does is call the Annotation contructor and add hasHtml. This patch lets the Link and Text annotations directly extend the Annotation class and add hasHtml themselves.
This patch also removes an unused global.
Recently I've landed a number patches which fixed issues with ColorSpaces. In most of these cases the cause of the failures were, either partially or entirely, related to the fact that we didn't resolve indirect objects (i.e. the code was missing `xref.fetchIfRef(...)`).
The purpose of this patch is to fix the few remaining cases where indirect objects *could* potentially cause failures.
Given that we have seen how this causes failures in practice, I thus think that it makes sense to try and avoid further issues, instead of waiting for users to file even more bugs for this part of the code-base.
Currently non-embedded ArialBlack fonts are not rendered bold enough, compared to e.g. Adobe Reader.
The issue is that we set the font weight to `bolder`, but since that is actually relative to the font weight of the parent, the result is that there's no practical difference from just using `bold`.
This patch attempts to address that, by explicitly setting the font weight to the maximum value instead (see https://developer.mozilla.org/en-US/docs/Web/CSS/font-weight).
*Note:* I expect one test "failure" in `issue5801`, which in this case is an improvement, since that PDF file uses ArialBlack.
Fixes 6068.
The most notable issue with the font in question is that the `differences` array contains lots of strange entries (of the type `uniXXXX`, instead of proper glyph names).
The 'Version' field of the most recent document catalog, if present, is
intended to supersede the value in the file prologue.
This is significant for incrementally-built PDF documents and generators that
emit a low version in the prologue and later apply a format version based on
PDF features used, such as Apple's CoreGraphics/Quartz PDF backend.
Fixes the internal version variable, as well as the PDFFormatVersion reported
by the API and consumed by viewers.
For passwords where the encoding already is correct, the conversion is a no-op.
Also, since `encodeURIComponent` might throw, we need to make sure that we handle that case too.
Fixes 6010.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1050040.
With this patch the file is completely readable, but given that the font is broken enough to be rejected by OTS the rendering differs slightly from Adobe Reader.
*Note:* the PDF file is sufficiently broken that even Adobe Reader complains about the font, *and* also about another more general issue.
