Git Log

Commit Date  
[458fba] by Michal Hocko Michal Hocko

configure: Add enable-tools option.

This implements the first part of --enable-tools configuration option which
will be used to automatically build tools. At this moment it only forces
boost-program-options library detection and exports proper linkage
variable (BOOST_PROGRAM_OPTIONS_LIB) which is used for generally available

For boost library detection it uses ax_boost_program_options.m4 taken from

In the next part, TOOLS_TARGET will be used to add default compilation if
it is configured.

All binaries compiled by tools/Makefile will use TOOLS_LIBS rather than

2010-02-23 18:16:34 Tree
[2c7a84] by Martin Petříček Martin Petříček

Raw object saving - patch by Ron Kaminsky, slightly modified

2010-02-22 01:17:39 Tree
[39fc74] by Michal Hocko Michal Hocko

*** empty log message ***

2010-02-18 13:37:07 Tree
[e47249] by Michal Hocko Michal Hocko

config: pdfedit-core-dev autoconf macro

This adds pdfedit-core-dev macro which detects and exports compiler and
linker flags necessary for 3rd party applications which use pdfedit-core-dev

This script is not used in our project but it is here for others to
download and use it. Maybe we can install it somewhere system-wide (e.g.

2010-02-18 13:34:16 Tree
[d3503c] by Michal Hocko Michal Hocko

*** empty log message ***

2010-02-17 09:38:33 Tree
[56d23b] by Michal Hocko Michal Hocko

xpdf: update to xpdf 3.02pl4

Backport of the upstream patch for 3.02pl4. This addresses several security
issues: CVE-2009-1188/CVE-2009-3603, CVE-2009-3604, CVE-2009-3606,
CVE-2009-3608, CVE-2009-3609.

2010-02-17 09:34:59 Tree
[25329c] by Michal Hocko Michal Hocko

Fix libz detection

We have relied on AC_CHECK_FT2 which is setting -lz automaticaly, but this
is not a case for all platforms (e.g. Fedora is different in that regard).

Let's make libz detection explicit. The core autoconf script was taken from

Thanks to Orion Poplawski for reporting.

2010-02-16 20:35:06 Tree
[7eeec1] by Martin Petříček Martin Petříček

Force makefile rebuild if is changed

2010-02-16 19:46:31 Tree
[f9835c] by Martin Petříček Martin Petříček

Fix: cannot be compiled with gcc 4.4 due to mismatch between const
char* and char*

2010-02-16 19:38:51 Tree
[403262] by Michal Hocko Michal Hocko

configure: typo fix for test command

Thanks to William Bader for reporting this.

2009-12-10 09:28:14 Tree
[247b00] by Michal Hocko Michal Hocko

kernel: getCObjectFromRef fix for trailer's properties

Fixes bt#343

getCObjectFromRef forces that the given property has to have a valid indiref
by means of assert(hasValidRef).

This is, however, problem for all properties which are retrieved from the
trailer as this is a direct object and so it cannot set IProperty::ref to
its children.

Moreover, this function doesn't need this check at all because it doesn't
rely on inidref logic. It just gets the value from the property and
dereference target object from pdf.

2009-11-30 10:55:03 Tree
[5f2df0] by Martin Petříček Martin Petříček

Updated cvsignore

2009-10-24 02:12:35 Tree
[70d9aa] by Michal Hocko Michal Hocko

bt#332 has been closed

2009-10-22 14:53:27 Tree
[7ce0fa] by Michal Hocko Michal Hocko

kernel: handle hybrid-xref files correctly in XRefWriter

Fixes bt#338

PDF 1.5 and above versions of specification support so called hybrid files
which contain both old-style xref table and xref stream. Such a document
then looks as follows:

[Standard xref entries. Hidden objects are marked as free with max gen

/Root [...]

1 0 obj
/Length [...]
[object stream data]

0 0


This means that this kind of file looks like a 2 revisions document. This
wouldn't be a big problem but we have a different problem if we consider
this to be 2 revisions document. XRefWriter::getRevision size calculates
revision size which is done as revStart-endPrev (where endPrev is the end
of the previous revision), but we are not able to calculate end of the
previous revision because xref (at OFFSET2) is not followed by startxref
keyword as expected. Application then ends with assertion failure

Of course we could fix the XRefWriter::getRevisionEnd but the point is that
we shouldn't consider this layout multi-version in the first place. The
"proper" fix is not considering additional xref for hybrid files at all.
This is done in XRefWriter::collectRevisions by checking XRefStm entry in
the trailer.

If a hybrid document is multiversion then the referenced trailer (OFFSET2)
should contain /Prev entry IMO. The PDF specification is not clear in this

2009-10-22 14:48:29 Tree
[7d8e10] by Michal Hocko Michal Hocko

Replace suffix by pattern implicit rules

I have tried to make compilation dependent on automatically generated files
(e.g. Makefile.flags) to make sure that the code gets rebuilt if we
reconfigure it.

This is, however, not possible with the suffix implicit rules. Moreover this
types of rules are long obsolete

The way how to handle this feature is to use pattern implicit rules which
seem to be much simpler to use and they can contain dependencies. now contains REBUILD_FORCE_FILES which contains all files
which should trigger rebuild. Each Makefile which includes Makefile.rules
can additionally add some extra files via EXTRA_REBUILD_FORCE_FILES.

2009-10-21 07:22:37 Tree
[886a70] by Michal Hocko Michal Hocko


We have configurable configure options for CFLAGS and CXXFLAGS for utils,
kernel and gui but none for xpdf code. This shouldn't be any big deal, but
it is nice to have it in sync with others.

Let's turn on -pedantic for xpdf this way to ensure standard enforcement as
this code is also a part of our exported pdfedit-core-dev library.

2009-10-19 08:56:31 Tree
[d08287] by Michal Hocko Michal Hocko

build: force c/c++ standard for compilation

Jozo has been hit by compilation errors caused by non standard c++
constructions on Windows so we should enforce C/C++ standard only without
any extensions supported by gcc.

We cannot do that directly in CONFIG_CXXFLAGS because gui doesn't compile
due to qt3 header file using long long which doesn't comply c++98 standard
so let's do it in extra flags for utils and kernel (this should be enough
because nothing else is distributed in pdfedit-core-dev).

2009-10-19 08:54:41 Tree
[29be01] by Jozef Mišutka Jozef Mišutka

remove assert causing problems

2009-10-12 15:26:27 Tree
[e4e18a] by Michal Hocko Michal Hocko

tests: don't consider read-only files in changeTrailerTC

changeTrailerTC throws read only expcetion thus the whole test fails. We
should simply ignore all read-nly files.

2009-10-12 12:48:00 Tree
[5c63fb] by Michal Hocko Michal Hocko

xpdf: remove unused code from our standard building system

We are not using complete xpdf code as is. E.g. PDFDoc is not used anywhere
in the kernel or GUI and in fact we shouldn't he used at all, because it is
no longer maintained by us (no upstream changes are backported). The same
applies to,,,,,,,

Let's remove all of them from referenced (in sources,
distributed headers and includes where-ever they are still used.

We cannot remove all of them, because at least is used by tests,
so we have to create a special target libxpdf-extra which will create
libxpdf-extra.a library and make tests target dependent on it. We can add
also some other files there later. This library doesn't get distributed.

SecurityHandler includes PDFDoc.h but it uses it only if ENABLE_PLUGINS is
defined. This is not done by our code, but keep it for having clear
compilation in this case.

2009-10-06 17:09:58 Tree
[30f786] by Jozef Mišutka Jozef Mišutka

isspace int cast replaced by correct unsigned char cast

2009-10-05 20:33:28 Tree
[d4389c] by Jozef Mišutka Jozef Mišutka

isdigit fails on win32 because of int-> unsigned char conversion and check =< 255

2009-09-22 16:19:13 Tree
[c5330a] by Jozef Mišutka Jozef Mišutka

added .h to package

2009-09-19 17:57:43 Tree
[d69d42] by Jozef Mišutka Jozef Mišutka

-- removed getopt
-- added boost program options
-- fixed, invalid allocation of local char [] using dynamic size
-- removed redundant dependency on, left it for 1 tool though
-- in same tools, removed option to specify multiple input files - does not make sense really and if somebody thinks it does, can be easily added using vector<string> in program options

2009-09-19 17:47:59 Tree
[6b094a] by Jozef Mišutka Jozef Mišutka

enabled all tools as they were converted to ansi c++

2009-09-19 17:43:33 Tree
Older >

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

No, thanks