Add some unicode mapping for ligatures when writing the cmap table in the font (bug 1946181)

[calixteman]

No description provided.

[Snuffleupagus]

Can we add some kind of test-case for this, to avoid it regressing?
Also, please run all tests before I review this.

[calixteman]

In thinking about that this morning, I wasn't able to have an idea.
But then I remembered we have some tests with fonts and ttx, so I suppose it could work here, wdyt ?

[Snuffleupagus]

But then I remembered we have some tests with fonts and ttx, so I suppose it could work here, wdyt ?

Yes, there's a few tests in https://github.com/mozilla/pdf.js/tree/master/test/font
Maybe that could work, but I unfortunately don't know how difficult it'd be to test this particular case.

[calixteman]

I've tried to write a test but I didn't manage to and I don't really have time for the moment.
@Snuffleupagus could review it ? and I'll file a bug in order to add a test later.

[Snuffleupagus]

/botio browsertest

[moz-tools-bot]

From: Bot.io (Windows)

---

Received

Command cmd_browsertest from @Snuffleupagus received. Current queue size: 0

Live output at: http://54.193.163.58:8877/d799316d3ceb4ea/output.txt

[moz-tools-bot]

From: Bot.io (Linux m4)

---

Received

Command cmd_browsertest from @Snuffleupagus received. Current queue size: 0

Live output at: http://54.241.84.105:8877/2d7bf952384bd95/output.txt

[moz-tools-bot]

From: Bot.io (Linux m4)

---

Failed

Full output at http://54.241.84.105:8877/2d7bf952384bd95/output.txt

Total script time: 17.15 mins

• Regression tests: FAILED

  different ref/snapshot: 93

Image differences available at: http://54.241.84.105:8877/2d7bf952384bd95/reftest-analyzer.html#web=eq.log

[calixteman]

As far as I can tell we see the same kind of "errors" with saw when we added this unicode mapping in the font (iirc, the linux font renderer makes some adjustments based on the unicode value associated with a glyph).

[moz-tools-bot]

From: Bot.io (Windows)

---

Success

Full output at http://54.193.163.58:8877/d799316d3ceb4ea/output.txt

Total script time: 30.66 mins

• Regression tests: Passed

[Snuffleupagus]

When we started moving all glyphs to a Private Use Area it was to prevent issues with particularly FreeType on Linux affecting glyph-shaping if we left glyphs at more "normal" positions in the cmap table.

With this patch we seem to be essentially regressing that, which might be problematical!?

[calixteman]

When we started moving all glyphs to a Private Use Area it was to prevent issues with particularly FreeType on Linux affecting glyph-shaping if we left glyphs at more "normal" positions in the cmap table.

With this patch we seem to be essentially regressing that, which might be problematical!?

It was the same thing in #15157.
And here's the Jonathan's answer:
#15157 (comment)

The diffs in the ref tests (at least for the one I checked) only appear for ligatures and the variation are very small.
And I'd say the same thing I said:
#15157 (comment)
Let's see if some people complain and if the case we'll try to figure out an other solution.

[Snuffleupagus]

Let's try this, while keeping an eye on any future Linux-only font bugs.

r=me, thank you.

[calixteman]

/botio-linux makeref

[moz-tools-bot]

From: Bot.io (Linux m4)

---

Received

Command cmd_makeref from @calixteman received. Current queue size: 0

Live output at: http://54.241.84.105:8877/d0a49b3cd1ad185/output.txt

[moz-tools-bot]

From: Bot.io (Linux m4)

---

Success

Full output at http://54.241.84.105:8877/d0a49b3cd1ad185/output.txt

Total script time: 16.77 mins

• Make references: Passed
• Check references: Passed