Bug #552: Build failure under Mac OS X 10.3.9 - Exiv2

Bug #552

Build failure under Mac OS X 10.3.9

Added by marius - over 13 years ago. Updated almost 13 years ago.

Status:

Closed

Priority:

Normal

Assignee:

Andreas Huggel

Category:

build

Target version:

0.17.1

Start date:

Due date:

% Done:

Estimated time:

Description

The build fails with the following error:

g++ -O2 -Wall -Wcast-align -Wpointer-arith -Wformat-security -Wmissing-format-attribute -Woverloaded-virtual -W -I. -I/usr/local/include -I/usr/local/include -DEXV_LOCALEDIR=\"/usr/local/share/locale\" -I../xmpsdk/include -c convert.cpp  -fno-common -DPIC -o .libs/convert.o
convert.cpp: In member function `void Exiv2::Converter::cnvXmpValueToIptc(const 
   char*, const char*)':
convert.cpp:1082: error: type specifier omitted for parameter
convert.cpp:1082: error: parse error before numeric constant
convert.cpp:1083: error: request for member `setValue' in `Exiv2::id(...)', 
   which is of non-aggregate type `Exiv2::Iptcdatum ()(...)'
convert.cpp:1084: error: no matching function for call to `Exiv2::IptcData::add
   (Exiv2::Iptcdatum (&)(...))'
iptc.hpp:242: error: candidates are: int Exiv2::IptcData::add(const 
   Exiv2::IptcKey&, Exiv2::Value*)
iptc.hpp:249: error:                 int Exiv2::IptcData::add(const 
   Exiv2::Iptcdatum&)
make[1]: * [convert.o] Error 1
make: * [all] Error 2

Additional information:

Build of version 0.14 was successful under this OS version.

Files

rmconvert.patch (1.3 KB) rmconvert.patch		Redmine Admin, 07 Jun 2008 02:32
firsttry.patch (477 Bytes) firsttry.patch		Redmine Admin, 07 Jun 2008 02:33

History

Updated by Andreas Huggel over 13 years ago

Attachment rmconvert.patch is a quick workaround for the problem: it removes the convert functionality from the library. XMP sidecar files will only deal with XMP data in this patch applied.

Updated by Andreas Huggel over 13 years ago

Attachment firsttry.patch is an attempt to fix the problem. I have had a very similar error message here earlier (on Linux/g++-4.3) and it went away when I used a 0 pointer argument in the Iptcdatum constructor on line 1082.
So if firsttry.patch doesn't help, please also try s/key/key, 0/ on that line.

But I still do not understand what the compiler is trying to do here and why it doesn't get the Iptcdatum and IptcKey constructors right.