Thanks a lot guys for the information. I noticed the new feature Complex
Scripts, but did not look at it more closely, because I thought it would not
apply to me using plain English :)
I have turned the feature off for now, which not only got rid of the
ligatures, it has also fixed the kerning for numbers, such as postal codes,
dates and account numbers. With the Complex Scripts feature on these numbers
are basically rendered in a monospaced way, which looks real ugly in most
places. Not sure which of the features mentioned are responsible for that
strange kerning (or lack thereof).
I cannot use any of the suggested workarounds, because the text I process is
in parts dynamically generated and outside of my control. It would of course
be great if it were possible to enable/disable the various Complex Scripts
features independent of each other, preferably not just globally, but on a
per-paragraph level.
Kind regards,
Ulrich
Glenn Adams wrote:
The addition of Complex Script support in FOP 1.1 is the cause. The font
actually controls whether ligatures are used or not.
The following features are enabled by default for all scripts which do not
otherwise override this feature set:
GSUB: {ccmp, liga, locl}
GPOS: {kern, mark, mkmk}
See org.apache.fop.complexscripts.scripts.DefaultScriptProcessor.
If you read the description of these features [1], you will find that they
are defined as "should be active by default".
[1] http://www.microsoft.com/typography/otspec/featurelist.htm
So, if a font designer includes a 'liga' table in the font, then those
substitutions will apply. As Pascal has pointed out, you can disable CS
entirely, in which case none of the GSUB or GPOS processing is performed.
Another work around would be to use something like:
<fo:character character="f"/><fo:character character="i"/>
which happens to work at the moment due to a bug that prevents performing
GSUB processing across an element boundary, but that may be fixed at some
point.
A better long term solution is to introduce the CSS3 Font Module's
font-variant-ligatures property [2], or, more generally, the
font-feature-settings property [3] as "fox:" extension attributes.
[2] http://www.w3.org/TR/css3-fonts/#font-variant-ligatures-prop
[3] http://www.w3.org/TR/css3-fonts/#font-feature-settings-prop
Regards,
Glenn
On Tue, May 28, 2013 at 3:01 AM, Ulrich Mayring <[email protected]> wrote:
Hi all,
I just upgraded from 0.95 to 1.1 and one of the issues that crept up is
that suddenly FOP uses ligatures, which it did not use before. Latin words
containing the letters "fi" or "fl" are now rendered using ligatures in the
PDF, although they are written as two seperate characters in the XML input
file.
I can open a bug and/or supply concrete test cases if needed, but I just
wanted to ask beforehand whether that is a known problem or perhaps a
configuration option?
I am using a TrueType font, Pragmatica Condensed. It may well have to do
with this font, since the standard FOP examples do not seem to have this
problem.
Many thanks in advance for any pointers,
Ulrich
------------------------------**------------------------------**---------
To unsubscribe, e-mail:
fop-users-unsubscribe@**xmlgraphics.apache.org<[email protected]>
For additional commands, e-mail:
fop-users-help@xmlgraphics.**apache.org<[email protected]>
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
