April 2005 - ntg-pdftex - NTG Mailing Lists

pdftex and xpdf (and its security bugs): What about a libxpdf?
by frank＠kuesterei.ch 24 Jan '06

24 Jan '06

Dear pdftex developers, during the last months security support for software containing xpdf code was a nightmare - not only that there were many security issues discovered, moreover everybody used a different patch, it was unclear to what extent older versions were affected, and so forth. This was also a problem for us, the teTeX maintainer in Debian, which currently has three versions of pdftex with three different versions of the xpdf code. What would you think about creating a shared library, "libxpdf", and using that for pdftex, either linked dynamically (for Linux/Unix distributions) or statically (for TeX-live and friends)? There is already such a library, libpoppler, a fork of the xpdf code (http://poppler.freedesktop.org/) However, I do not know whether the xpdf author, Derek B. Noonburg, is generally unwilling to create a shared library, or whether people simply didn't ask him. Personally I would prefer to have as little a number of versions of the same code, and thus to get a shared library from xpdf proper. If you agree, I would be oblidged if one of you would contact Derek Noonburg (derekn(a)foolabs.com) and ask him. He seems to be a little unresponsive at times, and has not answered a similar question from the maintainer of the xpdf package in Debian for a couple of weeks now. Maybe we can convince him that this is a good idea if he is approached from different sides. Alternatively, if he is unwilling to make this change, do you think using libpoppler would be feasible? Many thanks in advance, Frank Küster -- Frank Küster Inst. f. Biochemie der Univ. Zürich Debian Developer

2 3

pdftex-1.21b snapshot
by The Thanh Han 14 May '05

14 May '05

Hi, I have uploaded a snapshot of pdftex to pdftex.sarovar.org/src djgpp is also available (pdfetex binaries only). This is a bug-fix release that has all pending patches by Hartmut. The png patch by Taco has also been integrated. A quick summary from NEWS (Martin will make them more meaningful when an official version is released): ,-------- | - patch for PNG support by Taco & Hartmut (259) | | - ignore marginal kerns while unhboxing (292) | | - added \leftmarginkern & \rightmarginkern (292) | | - removed \pdfmovechar (294) | | - added \pdftracingfonts (304) | | - make \pdfxform work with box > 255 (310) | | - added \pdfxformname (requested by Pawel) | | - fixed problem with fonts on solaris (315) | `-------- There are still 2 pending problems: 1) the problem with truetype fonts when a unicode is mapped to 2 or more glyphs. I let pdftex fail in such a cas but still don't know what would be the right thing to do. 2) the problem with dangling object: I changed a few things to fix it but pdf inclusion may be broken now; please do some testing if you can. With the test files by Hartmut at sarovar: now there is a dangling object in yy2.pdf, which is expected if the font used in the included pdf is not used in the main document. Getting rid of this doesn't seem easy. Thanh

4 7

[ pdftex-Bugs-302 ] -jobname doesn't overcome texput
by noreply＠sarovar.org 12 May '05

12 May '05

Bugs item #302, was opened at 2005-03-01 19:06 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=302&group_id=106 Category: web2c Group: v1.21a Status: Open Resolution: None Priority: 5 Submitted By: John Denker (jsd) Assigned to: Olaf Weber (infovore) Summary: -jobname doesn't overcome texput Initial Comment: Hi -- Suppose I want to pipe commands into pdflatex: echo '\typeout{---\jobname---}' | \ pdflatex --jobname foo You might expect that to type out ---foo--- but instead it types out ---texput--- I observe that --jobname works just fine for any filename I can think of, just not when the filename is absent entirely. I can offer two workarounds: 1) It's not documented, but nevertheless possible to \def\jobname{foo}. 2) Of course it's possible to stuff the commands into tmp.tex and then pdflatex --jobname foo tmp.tex Because these workarounds exist, I would rate this as a very low-priority bug. But it had me surprised and confused for a while. Given the overall high quality of the product, I thought I'd bring it to your attention. ---------------------------------------------------------------------- >Comment By: Olaf Weber (infovore) Date: 2005-04-12 18:22 Message: Logged In: YES user_id=1847 This is a general problem with the implementation of the --jobname switch in web2c. My proposed diff is a variant on Akira's work. It touches more files, but gives IMHO a cleaner interface to get_job_name() and less need for additional #defines. Unless there turns out to be something wrong with this, it will be merged in the web2c sources. lib/texmfmp.c | 6 ++++-- mf.ch | 6 +++++- mp.ch | 4 +++- tex.ch | 6 ++++-- web2c/coerce.h | 2 +- web2c/texmf.defines | 2 +- 6 files changed, 18 insertions(+), 8 deletions(-) Index: mf.ch =================================================================== RCS file: /usr/local/cvsroot/texk/texk/web2c/mf.ch,v retrieving revision 1.40 diff -u -d -r1.40 mf.ch --- mf.ch 28 Dec 2004 18:24:38 -0000 1.40 +++ mf.ch 12 Apr 2005 16:07:11 -0000 @@ -1717,8 +1717,12 @@ @z @x [38.788] Set correct filename for recorder. +if job_name=0 then job_name:="mfput"; +@.mfput@> pack_job_name(".log"); @y +if job_name=0 then job_name:=get_job_name("mfput"); +@.mfput@> pack_job_name(".fls"); recorder_change_filename(stringcast(name_of_file+1)); pack_job_name(".log"); @@ -1778,7 +1782,7 @@ @x [38.793] l.15938 - The job name may have been given on the command line. begin job_name:=cur_name; open_log_file; @y - begin job_name:=get_job_name; open_log_file; + begin job_name:=get_job_name(cur_name); open_log_file; @z @x [38.793] Can't return name to string pool because of editor option? Index: mp.ch =================================================================== RCS file: /usr/local/cvsroot/texk/texk/web2c/mp.ch,v retrieving revision 1.48 diff -u -d -r1.48 mp.ch --- mp.ch 28 Dec 2004 18:23:47 -0000 1.48 +++ mp.ch 12 Apr 2005 16:06:53 -0000 @@ -1403,8 +1403,10 @@ @z @x [35.765] Change name of recorder file. +if job_name=0 then job_name:="mpout"; pack_job_name(".log"); @y +if job_name=0 then job_name:=get_job_name("mpout"); pack_job_name(".fls"); recorder_change_filename(stringcast(name_of_file+1)); pack_job_name(".log"); @@ -1489,7 +1491,7 @@ incr(j); stop_at_space:=true; end_name; - job_name:=get_job_name; + job_name:=get_job_name(cur_name); str_ref[job_name] := max_str_ref; @z Index: tex.ch =================================================================== RCS file: /usr/local/cvsroot/texk/texk/web2c/tex.ch,v retrieving revision 1.89 diff -u -d -r1.89 tex.ch --- tex.ch 29 Dec 2004 11:08:23 -0000 1.89 +++ tex.ch 12 Apr 2005 16:07:26 -0000 @@ -2241,9 +2241,11 @@ @!months:^char; @z -@x [29.534] l.10289 - Filename change for the recorder. +@x [29.534] l.10300 - Filename change for the recorder. +if job_name=0 then job_name:="texput"; @.texput@> @y +if job_name=0 then job_name:=get_job_name("texput"); @.texput@> pack_job_name(".fls"); recorder_change_filename(stringcast(name_of_file+1)); @@ -2369,7 +2371,7 @@ @x [29.537] l.10352 - start_input: was job_name given on the command line? begin job_name:=cur_name; open_log_file; @y - begin job_name:=get_job_name; open_log_file; + begin job_name:=get_job_name(cur_name); open_log_file; @z @x [29.537] l.10356 - Index: lib/texmfmp.c =================================================================== RCS file: /usr/local/cvsroot/texk/texk/web2c/lib/texmfmp.c,v retrieving revision 1.108 diff -u -d -r1.108 texmfmp.c --- lib/texmfmp.c 12 Apr 2005 15:39:14 -0000 1.108 +++ lib/texmfmp.c 12 Apr 2005 15:55:17 -0000 @@ -1591,10 +1591,12 @@ return maketexstring(fullnameoffile); } +/* Get the job name to be used, which may have been set from the + command line. */ strnumber -getjobname() +getjobname(strnumber name) { - strnumber ret = curname; + strnumber ret = name; if (job_name != NULL) ret = maketexstring(job_name); return ret; Index: web2c/coerce.h =================================================================== RCS file: /usr/local/cvsroot/texk/texk/web2c/web2c/coerce.h,v retrieving revision 1.5 diff -u -d -r1.5 coerce.h --- web2c/coerce.h 26 Feb 2004 17:43:14 -0000 1.5 +++ web2c/coerce.h 12 Apr 2005 15:55:31 -0000 @@ -24,7 +24,7 @@ /* And we use the opportunity to declare a few functions that could not be declared in texmfmp.h, because they need typedefs not yet known at that point. */ -extern strnumber getjobname P1H(void); +extern strnumber getjobname P1H(strnumber); #ifdef MP /* MP defined poolASCIIcode instead of packedASCIIcode, sigh. */ Index: web2c/texmf.defines =================================================================== RCS file: /usr/local/cvsroot/texk/texk/web2c/web2c/texmf.defines,v retrieving revision 1.17 diff -u -d -r1.17 texmf.defines --- web2c/texmf.defines 14 Jul 2004 16:17:48 -0000 1.17 +++ web2c/texmf.defines 12 Apr 2005 15:58:51 -0000 @@ -22,7 +22,7 @@ @define function callmakempx (); @define function floorscaled (); @define function floorunscaled (); -@define function getjobname; +@define function getjobname (); @define function getrandomseed; @define function initscreen; @define function inputln (); ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2005-04-12 14:11 Message: Logged In: YES user_id=421 Patches provided by Akira on NTG-pdftex ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=302&group_id=106

3 4

pdfTeX Manual Rev. 1.595, 20050422
by Hartmut Henkel 27 Apr '05

27 Apr '05

There is a fresh pdftex-t.tex manual version at sarovar with all new primitives for pdftex-1.21b (thanks to Taco!), and some streamlining. Regards, Hartmut

4 5

Re: various tiny 1.21b patches (fwd)
by Hartmut Henkel 21 Apr '05

21 Apr '05

---------- Forwarded message ---------- Date: Thu, 21 Apr 2005 07:46:00 +0200 (MEST) From: The Thanh Han <hanthethanh(a)gmx.net> To: Hartmut Henkel <hartmut_henkel(a)gmx.de> Subject: Re: various tiny 1.21b patches ...So until the end of this month I will have a chance to take a look at those changes. Can you please also post a mail to the ntg-pdftex ml saying that I am off-line for about 2 weeks. Many thanks. ... If nothing is urgent, can we postpone the release until I am back? Regards, Thanh Hartmut Henkel wrote: > Hi Thanh, > > attached find a bunch of diffs that should mainly tidy up left-over > stuff from the obsolete \pdfmovechars primitive. I hope i'm right that > also the update_enc() function is not needed anymore. > > The pdftex.ch.diff also contains a patch to get shorter names: > > \pdfoptionpdfimagehicolor --> > \pdfimagehicolor > > \pdfoptionimageapplygamma --> > \pdfimageapplygamma > > See ntg-pdftex list. I would then try to put the short names into the > pdftex manual. > > There were also left-over obsolete cfg_pdf12_compliant_code and > cfg_pdf13_compliant_code in pdftex.ch. > > Can you check and put these patches in before the release?

1 0

cmap inclusion and virtual fonts
by Michael Zedler 20 Apr '05

20 Apr '05

Hello, Concerning cmap inclusion and virtual fonts ftp://ftp.dante.de/tex-archive/macros/latex/contrib/cmap/README states "The main limitation currently is inability to work with virtual fonts, and this is because of limitation of pdftex, and may be resolved in a future versions of pdftex." Does this still hold? When using fonts with non-standard ligature names (e.g. /f_i, /f_l, /f_f etc with Adobe's current OTF) adding cmaps would be really useful - currently one cannot search for any word containing a ligature using Adobe Reader. Even better if these cmaps could be generated on the fly. Somehow Adobe Distiller manages to do so... Michael

1 0

length of new primitives
by Hartmut Henkel 20 Apr '05

20 Apr '05

while updating the manual, there are Taco's new primitives: \pdfimagegamma \pdfgamma and \pdfoptionpdfimagehicolor \pdfoptionimageapplygamma Trivial thing only: i wonder whether they really must be this long (there are already some very long ones, like (line break here) \pdfoptionpdfinclusionerrorlevel). Won't \pdfoptionpdfimagehicolor --> \pdfimagehicolor \pdfoptionimageapplygamma --> \pdfimageapplygamma serve as well? E. g. there is also \pdfadjustspacing, not \pdfoptionadjustspacing. BTW the new ones are not dumped into the format, so maybe one could say that only things with "option" would be dumped? Which are criteria that some stuff should be dumped, other not? Regards, Hartmut

2 1

bug 302
by The Thanh Han 12 Apr '05

12 Apr '05

Hi, as bug 302 doesn't seem to be pdftex specific, should the patch by Akira go to web2c? Regards, Thanh

3 2

[ pdftex-Feature Requests-89 ] Handle missing pages in included pdfs better
by noreply＠sarovar.org 12 Apr '05

12 Apr '05

Feature Requests item #89, was opened at 2004-07-08 14:38 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=89&group_id=106 Category: None Group: v1.21 Status: Open Resolution: None Priority: 5 Submitted By: Martin Schröder (oneiros) >Assigned to: Martin Schröder (oneiros) Summary: Handle missing pages in included pdfs better Initial Comment: Currently, when one tries to include page 10 of a 9 page document, pdftex quits. Proposed change: Behaviour of pdfTeX for pdf image inclusion: UNKNOWN := -1 Initialization: \pdflastximagepages := UNKNOWN Continue := Continue with empty xform object with dimension 1bp x 1bp. // 1bp because it avoids zero and possible problems and // the output is small (/BBox [0 0 1 1]) if PDFFileExists if PageNumber < 1 \pdflastximagepages = UNKNOWN Warning "Invalid page number" Continue elsif PageNumber > \pdflastximagepages \pdflastximagepages is set Warning "Page number to large (out of range) Continue else \pdflastximagepages is set image is included fi else Warning "PDF file does not exist" \pdflastximagepages = UNKNOWN Continue fi Failure/Success checking: if [page] < 1 * Page number too small elsif \pdflastximagepages = -1 * PDF file does not exist elsif [page] > \pdflastximagepages * Page number too large else * Success fi or with optimization for success: if [page] < 1 * Page number too small elsif [page] > \pdflastximagepages if \pdflastximagepages = -1 * PDF file does not exist else * Page number too large fi else * Success fi ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=89&group_id=106

1 0

[ pdftex-Feature Requests-254 ] fontexpansion, dvi output
by noreply＠sarovar.org 12 Apr '05

12 Apr '05

Feature Requests item #254, was opened at 2004-12-22 22:06 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=254&group_id=106 Category: None Group: None Status: Open >Resolution: Rejected Priority: 5 Submitted By: Nobody (None) Assigned to: Nobody (None) Summary: fontexpansion, dvi output Initial Comment: Hello developers, I'd like to use the autoexpand feature with dvips. Are there any plans to make pdftex write the autoexpanded tfm files to the current directory, perhaps even generating a map file? Cheers, Michael ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2005-01-14 03:19 Message: Logged In: YES user_id=710 It is likely that this will never be implemented. The better way to do that is to change the script \<TEX> (\<PDFTEX>) calls to create missing \<TFM> during run-time (on \<WEB2C>-based systems this is called |mktextfm|) to create all the required stuff on demand. Thanh ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=254&group_id=106

1 0