[reportlab-users] PDF without embedding fonts

[Andy Robinson]

On 14 May 2010 10:47, Yuriy Asyutin <yasyutin at gmail.com> wrote:

> I'll try some examples to find out this - think you are right. Anyway CJK

> fonts work in ReportLab even without parsing CMap files


We used to include and parse the CMap files for CJK fonts.  Then we
found a compact way to include this in arrays in the Python source
code, so the info in .cmap went into  *.py.  Then, when we switched to
Unicode input, we didn't need all CMaps because we decreed that the
input would always be Unicode code points.

reportlab/pdfbase/_cidfontdata.py contains the remaining 'hard-coded'
data needed to create CJK documents:  basically structures the same as
we observed in distilled PDFs.

I guess something similar and much shorter could be done for standard
Cyrillic fonts.

Today is very busy with visitors but I hope to look at your attached
file at the weekend or next week

Good luck,

Andy