A bug fix to PDFFontDescriptor in fop-0_16_0 release

[Peter Levart <Pe...@select-tech.si>]

Hello!

I recentrly upgraded from 0.15 to 0.16 and found that PDFs produced do
not open in AcrobatReader 4. Looked into the PDF file and saw that
"/FontBBox" arguments were strings of strange characters instead of

"[ a, b, c, d ]"

The attached diff to "org/apache/fop/pdf/PDFFontDescriptor.java"
resolved the issue...

OFF-TOPIC:

We are using FOP in a project that includes an editor for Newspaper
Articles (for a daily Newspaper "Delo" in Slovenija). Editor produces
structured XML documents (Title, Subtitle, Introduction, Text, etc.
sections). With FOP as an integral part of that Editor, even a
Journalist can preview & print a very good approximation of the final
shape of article as it would apear in the newspaper (which is done with
interactive page layout software).

That said, you can understand how we appreciate your work and embrace
features introduced in every new version. The "multi-column" text is one
of such impatiently awaited features that is now supported by FOP 0.16.

The other such feature would be: "Downloadable PDF Fonts". I see that
some not-used code is already in place for Type-1 PostScript fonts. We
have Hacked-up a version of FOP that uses that to replace standard 3
families (Courier, Helvetica, Times) with our variants that use
ISO-8859-2 encoding (we need those chars). We found that those fonts can
not coexist with standard fonts in the same FOP since there is a special
class called "CodePointMapping" that we also had to adjust and appears
to be used as a global FOP encoding table. The features needed for fonts
are:

- donload only fonts used (not all registered). This is especialy
important if PDFFontDescriptors include "/FontFile" attribute with the
content of the font...

- the PDF renderer should be aware of the encoding currently in efect
when rendering text and each font should have it's own independent
encoding table.

I understand that your efforts are concentrated on layout rather than on
rendering, but as we see FOP is already useable in real-world projects
even if it only implements a subset of XSL:FO standard. As such it would
be even more useable if it would include a font manager for PDF
renderer... (maybe, for next christmas?)


Keep on with the good work.


With best regards, Peter