freetype-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Progress update on adjustments database branch.


From: Craig White
Subject: Re: Progress update on adjustments database branch.
Date: Sun, 9 Jul 2023 00:40:39 -0400

> Strange.  Are you running on a GNU/Linux box?  Can you try to use
`strace` to find out where the file gets written?

Yes, I'm running Ubuntu.  I attached the log generated by strace.  The only time the string "png" appears is when it's opening libpng.

> This is *exactly* what I was talking about in previous e-mails, namely
to use the GSUB data for getting this mapping!  If a font handles,
say, the input <U+0042 (LATIN CAPITAL LETTER B) + U+0323 (COMBINING
DOT BELOW)>, it gets mapped to a certain glyph index Z somewhere in a
GSUB ligature lookup table.  To be more specific, there are cmap
entries X and Y for U+0042 and U+0323, and the GSUB table maps the
combination of X and Y to Z.

Thanks for telling me this.  Now I have a clear use case to keep in mind and test for as I implement GSUB mappings for the database.

On Sat, Jul 8, 2023 at 5:47 AM Werner LEMBERG <wl@gnu.org> wrote:

>> After pressing the (uppercase) 'P' key, an image called
>> `ftview.png` is stored in the current directory – at least on my
>> GNU/Linux box.
>
> I still can't find the ftview image,

Strange.  Are you running on a GNU/Linux box?  Can you try to use
`strace` to find out where the file gets written?

  strace ftview font.ttf &> strace.log

If you are using another platform I'm sure there is an equivalent to
it.  It is necessary that FreeType gets linked with 'libpng' to make
the 'P' key work, but this is the default.

> I recently made code to do the opposite case of i: where something
> should be adjusted down instead of up, which is when I started
> thinking about combining characters like U+0323.  If a combining
> character could form a glyph not mapped with cmap, I would need to
> work out that a combining character mapped to the glyph.  Any ideas
> for this?

This is *exactly* what I was talking about in previous e-mails, namely
to use the GSUB data for getting this mapping!  If a font handles,
say, the input <U+0042 (LATIN CAPITAL LETTER B) + U+0323 (COMBINING
DOT BELOW)>, it gets mapped to a certain glyph index Z somewhere in a
GSUB ligature lookup table.  To be more specific, there are cmap
entries X and Y for U+0042 and U+0323, and the GSUB table maps the
combination of X and Y to Z.


    Werner

Attachment: strace.log
Description: Binary data


reply via email to

[Prev in Thread] Current Thread [Next in Thread]