Hello,
I sent a simple patch to get things moving with complex scripts (indic
languages) but to get it really correct (example: Urdu language in Nastaliq
script), some more things to be done.
I've done most of the following to get this language going but its not ready
yet:
1. Removal of ArabicLigaturizer/etc code. Instead, Arabic text is passed
as-is to layoutGlyphs() which does this transparently. In some cases Arabic
fonts do not have presentation forms and do this with opentype GSUB tables (
http://partners.adobe.com/public/developer/opentype/index_table_formats1.html).
Nastaliq fonts are a perfect example of this, as they have thousands of
ligatures.
2. text does not need to be reordered by Bidi-Analysis prior to being laid
out, however, the argument (RTL, LTR) for a run of text does need to be
supplied to layoutGlyphs() to get the correct behavior. So Bidi Analysis is
moved.
3. Remaining problem: need to take the vertical positional information from
layoutGlyphs(), (derived from GPOS tables:
http://partners.adobe.com/public/developer/opentype/index_table_formats2.html)
and put this in the PDF somehow. The figure referenced there (
http://partners.adobe.com/public/developer/en/images/fig4b.gif) shows what
is currently displayed in PDFS for nastaliq font, "Incorrect".
In problem #3 I need to apply some vertical position to an individual
glyph... I'm looking at using PDFContentByte.setTextRise() which appears to
be designed for subscript/superscript... is there a better way?
--
Robert Muir
[email protected]
------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com
_______________________________________________
iText-questions mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/itext-questions
Buy the iText book: http://www.1t3xt.com/docs/book.php
