cjk-list
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [cjk] Argh I see, dvips seems to be producing rubbish!


From: amigolang
Subject: Re: [cjk] Argh I see, dvips seems to be producing rubbish!
Date: Thu, 08 Jun 2017 15:12:18 +0200

Hello Hin-Tak,

I forgot to answer the question for my fontforge version...

fontforge -version 

answers with:

fontforge 20170601
libfontforge 20170601

I think, it couldn't be much more recent...

I seems to me, that I stuck my fingers into a wasp nest...

Thorsten

Von meinem Samsung Gerät gesendet.


-------- Ursprüngliche Nachricht --------
Von: Hin-Tak Leung <address@hidden>
Datum: 07.06.17 18:13 (GMT+01:00)
An: Werner LEMBERG <address@hidden>
Cc: address@hidden, address@hidden
Betreff: Re: Argh I see, dvips seems to be producing rubbish!

I think i see what the problem with my subfont 44 is. The wrongly encoded glyph ids are those without a mapping under cmap 3,10. cmap only maps 43000+ of the 65535 glyphs; the the remain 22,000 unencoded glyphs get wrongly encoded as their glyph ids. So it trashes 1/2 my subfonts. I'll need to see what the missing glyphs are doing.

I used the latest fontforge from the repo and fedora's 20160404 are broken in the same way - so I wonder what version is  Thorsten using?

--------------------------------------------
On Wed, 7/6/17, Hin-Tak Leung <address@hidden> wrote:

Thorsten: do you still have your *.enc files
by any chance? Or don't mind making them again? :-) . Should
be quite small if you zip all of them up...

Hi Werner,

I have a better description of what is
broken. The re-encoding seems to have got corrupted. For
example, for subfont 44, this is what ttf2tfm gives:

  /a1sh44Encoding [
/index0x1AD1 /index0x1AD2 /index0x1AD3
/index0x1AD4
/index0x1AD5 /index0x1AD6 /index0x1AD7
/index0x1AD8
/index0x1ADA /index0x1ADB /index0x1ADC
/index0x1ADD
/index0x1ADE /index0x1ADF /index0x1AE0
/index0x1AE1
/index0x1AE2 /index0x1AE3 /index0x1AE4
/index0x1AE5
/index0x1AE6 /index0x1AE7 /index0x1AE8
/index0x1AE9
/index0x1AEB /index0x1AEC /index0x1AED
/index0x1AEE
/index0x1AF0 /index0x1AF1 /index0x1AF2
/index0x1AF3
...

This is what fontforge does (it is
really glyph id rather than uniXXXX) - whenever fontforge
seems to be randomly uses /uni44*, it encodes the wrong
glyph. The glyph ids should be uni1AD1, etc. So half of
subfont 44 is wrong, in my case.

/a1sh44-Enc [
  /uni4400 /uni1AD2 /uni4402 /uni1AD4
  /uni4404 /uni4405 /uni1AD7 /uni4407
  /uni4408 /uni1ADB /uni1ADC /uni440B
  /uni1ADE /uni1ADF /uni440E /uni1AE1
  /uni1AE2 /uni4411 /uni4412 /uni1AE5
  /uni4414 /uni1AE7 /uni1AE8 /uni4417
  /uni4418 /uni1AEC /uni441A /uni441B
  /uni1AF0 /uni441D /uni441E /uni1AF3
...

Hin-Tak


reply via email to

[Prev in Thread] Current Thread [Next in Thread]