| Summary: | vertical text misplaced after distilling | ||
|---|---|---|---|
| Product: | Ghostscript | Reporter: | Jack Moffitt <jack> |
| Component: | PDF Writer | Assignee: | Igor Melichev <igor.melichev> |
| Status: | NOTIFIED FIXED | ||
| Severity: | normal | ||
| Priority: | P2 | ||
| Version: | master | ||
| Hardware: | All | ||
| OS: | All | ||
| Customer: | 310 | Word Size: | --- |
| Attachments: |
vertical.ps
v2.ps |
||
|
Description
Jack Moffitt
2004-07-28 14:15:54 UTC
Created attachment 822 [details]
vertical.ps
Reproduced with CVS HEAD. It uses a composite font with FontMatrix [0 a a 0 0 0], and its descendent CIDFontType 2 font has another transposed FontMatrix [0 1 -1 0 0 0]. They more or less compensate each another, except for V-vector. When we write a CIDFontType2 to PDF, actually we write a True Type, which has no FontMatrix. So likely we wrongly translate V-vector. I tried some changes to code, but could not find a right one before vacations. Likely a code near gdevpdte.c ln 844 to be changed. BTW, it looks that the 1st and the 2nd matrices are applied in a wrong order. The previous comment isn't fully correct. Actually the problem happens due to the written text positioning commands are inconsistent with DW2. I found another problem with new patch. When you distill the attached file (v2.ps) with Acrobat, you can see all the letters are in a box. However, when distilling with the latest gs cvs head, a letter is out of the box and some letter spacings are not correct. Created attachment 859 [details]
v2.ps
Neither Acrobat Distiller 6.0 nor CPSI 1030 could handle v2.ps on my system Windows XP. Both fail with "Error: rangecheck; OffendingCommand: endcidrange ". I re-close this bug because it originally was a customer bug, which is fixed. Jeong, if you still have a problem, please open another bug and attach all necessary information how to reproduce it, including the Acrobat version, platform, and check whether it uses a platform-specific resources. Please also attach the PDF file created on your system with Acrobat. Reopening the bug, because comparefiles/Bug687603.ps still displays some characters biased. Rather they are inside boxes, hieroglyphs are biased by X from letters. More patch to HEAD : http://www.ghostscript.com/pipermail/gs-cvs/2004-October/004937.html A cumulative patch to GS_8_1X for bug 687603 and bug 687753 : http://www.ghostscript.com/pipermail/gs-cvs/2004-October/004938.html http://www.ghostscript.com/pipermail/gs-cvs/2004-October/004939.html |