Re: Bug fix for non ASCII environments

[Markus Bernhardt]

 
Tore Engvig wrote:
To answer your questions:
Yes, there are unicode pdf files.
Good.
Fop can generate pdf files for non iso
8859 encodings if you provide fonts that support it.
I think fonts haven't been my problem.
The problem was that the PDF structure is coded
as strings into the java source of FOP.
Lets say you have a PDF-keyword 'FlateDecode'
Anywhere in the FOP Java Source is something like:
String text = "FlateDecode";
Later this String gets written to the PDF-file
via text.getBytes() as:
46 6C 61 74 65 44 65 63 6F 64 65
F  l  a  t  e  D  e  c 
o  d  e
But that works only on a ASCII based system !
On EBCDIC:
C6 93 81 A3 85 C4 85 83 96 84 85
F  l  a  t  e  D  e  c 
o  d  e
So Acrobat can't get the structure of the document at all.
However, the
generated pdf isn't unicode but multibyte indexes to the fonts
glyphs. It
will be unicode as soon as the toUnicode CMAP is implemented. To
generate
the multibyte text, getBytes("UnicodeBigUnmarked") is used.
Will that fix my problem ?
( To be honest with you I simply don't understand the last 2 sentences.
)
 
I think you could (and should) replace getBytes() with
getBytes("ISO-8859-1") most places, but you have to be careful
in the
renderers and should test fop thoroughly doing it.
We have done that and works great. Not one problem found (in our kind
of documents).
 
Tore
 
[SNIP]
- markus
begin:vcard 
n:Bernhardt;Markus
tel;cell:0171-5770462
tel;fax:089-420903-20
tel;home:089-6378949
tel;work:089-420903-14
x-mozilla-html:FALSE
url:www.swsgmbh.de
org:Software Service Wulf Schupp GmbH;Spieljoch
adr:;;Spieljochstr. 34;M�nchen;;81825;Germany
version:2.1
email;internet:[EMAIL PROTECTED]
title:Entwicklungsleiter
note;quoted-printable:[dF]Quisam=0D=0Awww.discordian-Front.de=0D=0AHail Eris !!!=0D=0AHappy Frag !!!
fn:Markus Bernhardt
end:vcard


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]