From fredrik at pythonware.com  Thu Oct  1 13:10:32 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 1 Oct 2009 13:10:32 +0200
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com>
References: <368a5cd50909280625o471fb496ra96b5b151a058d80@mail.gmail.com> 
	<4AC0BE40.5010801@gmail.com>
	<368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com>
Message-ID: <368a5cd50910010410m30070acfne864598696499605@mail.gmail.com>

I've replaced the 1.1.6 binary at effbot.org/downloads with one built
under 2.6.2.  The installer is 130k larger thanks to a 3x larger
wininst executable (!?), but the binary extensions are slightly
smaller than before, so I'll assume it's the manifests that have
disappeared.

</F>

On Mon, Sep 28, 2009 at 4:07 PM, Fredrik Lundh <fredrik at pythonware.com> wrote:
> Ah, sneaky. ?I'm pretty sure the current binary distribution was built
> when 2.6 was new, and I didn't really expect them to tweak
> build-related issues in an incremental release :)
>
> I'll push out a new build asap; will ping the list when it's available.
>
> Thanks!
>
> </F>
>
> On Mon, Sep 28, 2009 at 3:46 PM, Koen van de Sande
> <koenvandesande at gmail.com> wrote:
>> The fix was applied to the main Python distribution (2.6.2 has it for sure,
>> 2.6.0/2.6.1 does not). Any extension built with 2.6.2 or higher should be
>> fine, as it will not get an embedded manifest.
>> So just rebuild with Python Windows version 2.6.2. I see in the PIL source
>> tree that it uses a setup.py build_ext call, so no changes to PIL should be
>> needed.
>>
>> Koen
>> _______________________________________________
>> Image-SIG maillist ?- ?Image-SIG at python.org
>> http://mail.python.org/mailman/listinfo/image-sig
>>
>

From flashk at gmail.com  Fri Oct  2 00:10:33 2009
From: flashk at gmail.com (Farshid Lashkari)
Date: Thu, 1 Oct 2009 15:10:33 -0700
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <368a5cd50910010410m30070acfne864598696499605@mail.gmail.com>
References: <368a5cd50909280625o471fb496ra96b5b151a058d80@mail.gmail.com> 
	<4AC0BE40.5010801@gmail.com>
	<368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com> 
	<368a5cd50910010410m30070acfne864598696499605@mail.gmail.com>
Message-ID: <978d1eac0910011510x3a4f96bdr6e0f21858b398e7a@mail.gmail.com>

I just tried the new installer (
http://effbot.org/media/downloads/PIL-1.1.6.win32-py2.6.exe), but the pyd
files still contain embedded manifests. The installer says it was built Oct
1, 2009 with Python 2.6.2, so that should be the latest version.
I took a closer look at the changes made from that bug report. The change
was only made to the PCbuild/pyd.vsprops file, which I don't believe will
affect 3rd party extensions. I see that Koen submitted a patch
for msvc9compiler.py that will strip the runtime from the manifest, but that
doesn't appear to have been committed. So I'm not really sure how 3rd party
extension developers can build pyd files without embedded manifests.

On Thu, Oct 1, 2009 at 4:10 AM, Fredrik Lundh <fredrik at pythonware.com>wrote:

> I've replaced the 1.1.6 binary at effbot.org/downloads with one built
> under 2.6.2.  The installer is 130k larger thanks to a 3x larger
> wininst executable (!?), but the binary extensions are slightly
> smaller than before, so I'll assume it's the manifests that have
> disappeared.
>
> </F>
>
> On Mon, Sep 28, 2009 at 4:07 PM, Fredrik Lundh <fredrik at pythonware.com>
> wrote:
> > Ah, sneaky.  I'm pretty sure the current binary distribution was built
> > when 2.6 was new, and I didn't really expect them to tweak
> > build-related issues in an incremental release :)
> >
> > I'll push out a new build asap; will ping the list when it's available.
> >
> > Thanks!
> >
> > </F>
> >
> > On Mon, Sep 28, 2009 at 3:46 PM, Koen van de Sande
> > <koenvandesande at gmail.com> wrote:
> >> The fix was applied to the main Python distribution (2.6.2 has it for
> sure,
> >> 2.6.0/2.6.1 does not). Any extension built with 2.6.2 or higher should
> be
> >> fine, as it will not get an embedded manifest.
> >> So just rebuild with Python Windows version 2.6.2. I see in the PIL
> source
> >> tree that it uses a setup.py build_ext call, so no changes to PIL should
> be
> >> needed.
> >>
> >> Koen
> >> _______________________________________________
> >> Image-SIG maillist  -  Image-SIG at python.org
> >> http://mail.python.org/mailman/listinfo/image-sig
> >>
> >
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091001/fa4d6920/attachment.htm>

From fredrik at pythonware.com  Fri Oct  2 11:32:41 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 2 Oct 2009 11:32:41 +0200
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <978d1eac0910011510x3a4f96bdr6e0f21858b398e7a@mail.gmail.com>
References: <368a5cd50909280625o471fb496ra96b5b151a058d80@mail.gmail.com> 
	<4AC0BE40.5010801@gmail.com>
	<368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com> 
	<368a5cd50910010410m30070acfne864598696499605@mail.gmail.com> 
	<978d1eac0910011510x3a4f96bdr6e0f21858b398e7a@mail.gmail.com>
Message-ID: <368a5cd50910020232q66af8debya0e32138df3540f@mail.gmail.com>

Ok, I can apply Koen's patch locally and rebuild, if that makes sense.

What's the quickest way to look for embedded manifests in the PYD:s?
dumpbin doesn't seem to say anything about that, so I assume I have to
use some other tool.

</F>

On Fri, Oct 2, 2009 at 12:10 AM, Farshid Lashkari <flashk at gmail.com> wrote:
> I just tried the new installer
> (http://effbot.org/media/downloads/PIL-1.1.6.win32-py2.6.exe), but the pyd
> files still contain embedded manifests. The installer says it was built Oct
> 1, 2009 with Python 2.6.2, so that should be the latest version.
> I took a closer look at the changes made from that bug report. The change
> was only made to the?PCbuild/pyd.vsprops file, which I don't believe will
> affect 3rd party extensions. I see that Koen submitted a patch
> for?msvc9compiler.py that will strip the runtime from the manifest, but that
> doesn't appear to have been committed. So I'm not really sure how 3rd party
> extension developers can build pyd files without embedded manifests.
> On Thu, Oct 1, 2009 at 4:10 AM, Fredrik Lundh <fredrik at pythonware.com>
> wrote:
>>
>> I've replaced the 1.1.6 binary at effbot.org/downloads with one built
>> under 2.6.2. ?The installer is 130k larger thanks to a 3x larger
>> wininst executable (!?), but the binary extensions are slightly
>> smaller than before, so I'll assume it's the manifests that have
>> disappeared.
>>
>> </F>
>>
>> On Mon, Sep 28, 2009 at 4:07 PM, Fredrik Lundh <fredrik at pythonware.com>
>> wrote:
>> > Ah, sneaky. ?I'm pretty sure the current binary distribution was built
>> > when 2.6 was new, and I didn't really expect them to tweak
>> > build-related issues in an incremental release :)
>> >
>> > I'll push out a new build asap; will ping the list when it's available.
>> >
>> > Thanks!
>> >
>> > </F>
>> >
>> > On Mon, Sep 28, 2009 at 3:46 PM, Koen van de Sande
>> > <koenvandesande at gmail.com> wrote:
>> >> The fix was applied to the main Python distribution (2.6.2 has it for
>> >> sure,
>> >> 2.6.0/2.6.1 does not). Any extension built with 2.6.2 or higher should
>> >> be
>> >> fine, as it will not get an embedded manifest.
>> >> So just rebuild with Python Windows version 2.6.2. I see in the PIL
>> >> source
>> >> tree that it uses a setup.py build_ext call, so no changes to PIL
>> >> should be
>> >> needed.
>> >>
>> >> Koen
>> >> _______________________________________________
>> >> Image-SIG maillist ?- ?Image-SIG at python.org
>> >> http://mail.python.org/mailman/listinfo/image-sig
>> >>
>> >
>> _______________________________________________
>> Image-SIG maillist ?- ?Image-SIG at python.org
>> http://mail.python.org/mailman/listinfo/image-sig
>
>

From olt at omniscale.de  Fri Oct  2 12:06:42 2009
From: olt at omniscale.de (Oliver Tonnhofer)
Date: Fri, 2 Oct 2009 12:06:42 +0200
Subject: [Image-SIG] Image.open should close file when it cannot identify
	image
Message-ID: <AF2B9B32-C43A-4066-8408-74371203808F@omniscale.de>

Hi,

I want to remove files that PIL can't open. But the following fails on  
Windows, because Image.open doesn't close the file.

----
import os
import Image
try:
     f = open('test.test', 'w')
     f.close()
     Image.open('test.test')
finally:
     os.remove('test.test')
----
results in:
WindowsError: [Error 32] The process cannot access the file because it  
is being
used by another process: 'test.test'


Below is a simple fix.

Regards,
Oliver

PS: please cc me on replies to the list.

----
--- PIL/Image.py.orig	2009-10-02 12:02:26.000000000 +0200
+++ PIL/Image.py	2009-10-02 12:03:12.000000000 +0200
@@ -1913,6 +1913,9 @@
          except (SyntaxError, IndexError, TypeError):
              pass

+    if filename:
+        fp.close()
+
      raise IOError("cannot identify image file")

  #


From fredrik at pythonware.com  Fri Oct  2 12:30:38 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 2 Oct 2009 12:30:38 +0200
Subject: [Image-SIG] Image.open should close file when it cannot
	identify image
In-Reply-To: <AF2B9B32-C43A-4066-8408-74371203808F@omniscale.de>
References: <AF2B9B32-C43A-4066-8408-74371203808F@omniscale.de>
Message-ID: <368a5cd50910020330l69cc5c58k7749957e7ca73694@mail.gmail.com>

On Fri, Oct 2, 2009 at 12:06 PM, Oliver Tonnhofer <olt at omniscale.de> wrote:
> Hi,
>
> I want to remove files that PIL can't open. But the following fails on
> Windows, because Image.open doesn't close the file.

Unless you hang on to the image object, Python will close things when
the image object itself goes away.

Your workaround doesn't handle anything that might happen during open;
to force the file to be closed if it cannot be fully read no matter
what caused the problem, I'd recommend doing something like:

    file = open(filename, "rb")
    try:
        im = Image.open(file)
        im.load()
    except StandardError:
        file.close() # force close
        ... process unreadable file ...
    else:
        ... process image ...

</F>




>
> ----
> import os
> import Image
> try:
> ? ?f = open('test.test', 'w')
> ? ?f.close()
> ? ?Image.open('test.test')
> finally:
> ? ?os.remove('test.test')
> ----
> results in:
> WindowsError: [Error 32] The process cannot access the file because it is
> being
> used by another process: 'test.test'
>
>
> Below is a simple fix.
>
> Regards,
> Oliver
>
> PS: please cc me on replies to the list.
>
> ----
> --- PIL/Image.py.orig ? 2009-10-02 12:02:26.000000000 +0200
> +++ PIL/Image.py ? ? ? ?2009-10-02 12:03:12.000000000 +0200
> @@ -1913,6 +1913,9 @@
> ? ? ? ? except (SyntaxError, IndexError, TypeError):
> ? ? ? ? ? ? pass
>
> + ? ?if filename:
> + ? ? ? ?fp.close()
> +
> ? ? raise IOError("cannot identify image file")
>
> ?#
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From koenvandesande at gmail.com  Fri Oct  2 13:17:16 2009
From: koenvandesande at gmail.com (Koen van de Sande)
Date: Fri, 02 Oct 2009 14:17:16 +0300
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <368a5cd50910020232q66af8debya0e32138df3540f@mail.gmail.com>
References: <368a5cd50909280625o471fb496ra96b5b151a058d80@mail.gmail.com>
	<4AC0BE40.5010801@gmail.com>
	<368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com>
	<368a5cd50910010410m30070acfne864598696499605@mail.gmail.com>
	<978d1eac0910011510x3a4f96bdr6e0f21858b398e7a@mail.gmail.com>
	<368a5cd50910020232q66af8debya0e32138df3540f@mail.gmail.com>
Message-ID: <4AC5E13C.508@gmail.com>

Open it in a hex editor, and look for the word manifest. This is what it 
looks like with dependency:

<assembly xmlns="urn:schemas-microsoft-com:asm.v1" manifestVersion="1.0">
  <trustInfo xmlns="urn:schemas-microsoft-com:asm.v3">
    <security>
      <requestedPrivileges>
        <requestedExecutionLevel level="asInvoker" 
uiAccess="false"></requestedExecutionLevel>
      </requestedPrivileges>
    </security>
  </trustInfo>
  <dependency>
    <dependentAssembly>
      <assemblyIdentity type="win32" name="Microsoft.VC90.CRT" 
version="9.0.21022.8" processorArchitecture="x86" 
publicKeyToken="1fc8b3b9a1e18e3b"></assemblyIdentity>
    </dependentAssembly>
  </dependency>
</assembly>PAPADDINGXXPADDINGPADDINGX

and after the patch it looks like this:

<assembly xmlns="urn:schemas-microsoft-com:asm.v1" manifestVersion="1.0">
  <trustInfo xmlns="urn:schemas-microsoft-com:asm.v3">
    <security>
      <requestedPrivileges>
        <requestedExecutionLevel level="asInvoker" 
uiAccess="false"></requestedExecutionLevel>
      </requestedPrivileges>
    </security>
  </trustInfo>
  <dependency>
    <dependentAssembly>
     
    </dependentAssembly>
  </dependency>
</assembly>PAPAD

I've pinged the issue, as the patch should be applied to the main Python 
distribution.

Koen

Fredrik Lundh wrote:
> Ok, I can apply Koen's patch locally and rebuild, if that makes sense.
>
> What's the quickest way to look for embedded manifests in the PYD:s?
> dumpbin doesn't seem to say anything about that, so I assume I have to
> use some other tool.
>
> </F>
>
> On Fri, Oct 2, 2009 at 12:10 AM, Farshid Lashkari <flashk at gmail.com> wrote:
>   
>> I just tried the new installer
>> (http://effbot.org/media/downloads/PIL-1.1.6.win32-py2.6.exe), but the pyd
>> files still contain embedded manifests. The installer says it was built Oct
>> 1, 2009 with Python 2.6.2, so that should be the latest version.
>> I took a closer look at the changes made from that bug report. The change
>> was only made to the PCbuild/pyd.vsprops file, which I don't believe will
>> affect 3rd party extensions. I see that Koen submitted a patch
>> for msvc9compiler.py that will strip the runtime from the manifest, but that
>> doesn't appear to have been committed. So I'm not really sure how 3rd party
>> extension developers can build pyd files without embedded manifests.
>> On Thu, Oct 1, 2009 at 4:10 AM, Fredrik Lundh <fredrik at pythonware.com>
>> wrote:
>>     
>>> I've replaced the 1.1.6 binary at effbot.org/downloads with one built
>>> under 2.6.2.  The installer is 130k larger thanks to a 3x larger
>>> wininst executable (!?), but the binary extensions are slightly
>>> smaller than before, so I'll assume it's the manifests that have
>>> disappeared.
>>>
>>> </F>
>>>
>>> On Mon, Sep 28, 2009 at 4:07 PM, Fredrik Lundh <fredrik at pythonware.com>
>>> wrote:
>>>       
>>>> Ah, sneaky.  I'm pretty sure the current binary distribution was built
>>>> when 2.6 was new, and I didn't really expect them to tweak
>>>> build-related issues in an incremental release :)
>>>>
>>>> I'll push out a new build asap; will ping the list when it's available.
>>>>
>>>> Thanks!
>>>>
>>>> </F>
>>>>
>>>> On Mon, Sep 28, 2009 at 3:46 PM, Koen van de Sande
>>>> <koenvandesande at gmail.com> wrote:
>>>>         
>>>>> The fix was applied to the main Python distribution (2.6.2 has it for
>>>>> sure,
>>>>> 2.6.0/2.6.1 does not). Any extension built with 2.6.2 or higher should
>>>>> be
>>>>> fine, as it will not get an embedded manifest.
>>>>> So just rebuild with Python Windows version 2.6.2. I see in the PIL
>>>>> source
>>>>> tree that it uses a setup.py build_ext call, so no changes to PIL
>>>>> should be
>>>>> needed.
>>>>>
>>>>> Koen
>>>>> _______________________________________________
>>>>> Image-SIG maillist  -  Image-SIG at python.org
>>>>> http://mail.python.org/mailman/listinfo/image-sig
>>>>>
>>>>>           
>>> _______________________________________________
>>> Image-SIG maillist  -  Image-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/image-sig
>>>       
>>     


From cgohlke at uci.edu  Fri Oct  2 17:36:29 2009
From: cgohlke at uci.edu (Christoph Gohlke)
Date: Fri, 02 Oct 2009 08:36:29 -0700
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <368a5cd50910020232q66af8debya0e32138df3540f@mail.gmail.com>
References: <368a5cd50909280625o471fb496ra96b5b151a058d80@mail.gmail.com>
	<4AC0BE40.5010801@gmail.com>	<368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com>
	<368a5cd50910010410m30070acfne864598696499605@mail.gmail.com>
	<978d1eac0910011510x3a4f96bdr6e0f21858b398e7a@mail.gmail.com>
	<368a5cd50910020232q66af8debya0e32138df3540f@mail.gmail.com>
Message-ID: <4AC61DFD.5020500@uci.edu>

Hi,

you can use Dependency Walker <http://www.dependencywalker.com/>. If the 
patch is working, pyd files will depend on MSVCR90.DLL, otherwise 
%WINDIR%\winsxs\...\MSVCR90.DLL

There are two issues with Koen's patch (both easy to fix). It does not 
work for 64-bit Python and it also processes exe files build with 
setuptools.

Christoph

On 11:59, Fredrik Lundh wrote:
> Ok, I can apply Koen's patch locally and rebuild, if that makes sense.
>
> What's the quickest way to look for embedded manifests in the PYD:s?
> dumpbin doesn't seem to say anything about that, so I assume I have to
> use some other tool.
>
> </F>
>
> On Fri, Oct 2, 2009 at 12:10 AM, Farshid Lashkari<flashk at gmail.com>  wrote:
>> I just tried the new installer
>> (http://effbot.org/media/downloads/PIL-1.1.6.win32-py2.6.exe), but the pyd
>> files still contain embedded manifests. The installer says it was built Oct
>> 1, 2009 with Python 2.6.2, so that should be the latest version.
>> I took a closer look at the changes made from that bug report. The change
>> was only made to the PCbuild/pyd.vsprops file, which I don't believe will
>> affect 3rd party extensions. I see that Koen submitted a patch
>> for msvc9compiler.py that will strip the runtime from the manifest, but that
>> doesn't appear to have been committed. So I'm not really sure how 3rd party
>> extension developers can build pyd files without embedded manifests.
>> On Thu, Oct 1, 2009 at 4:10 AM, Fredrik Lundh<fredrik at pythonware.com>
>> wrote:
>>>
>>> I've replaced the 1.1.6 binary at effbot.org/downloads with one built
>>> under 2.6.2.  The installer is 130k larger thanks to a 3x larger
>>> wininst executable (!?), but the binary extensions are slightly
>>> smaller than before, so I'll assume it's the manifests that have
>>> disappeared.
>>>
>>> </F>
>>>
>>> On Mon, Sep 28, 2009 at 4:07 PM, Fredrik Lundh<fredrik at pythonware.com>
>>> wrote:
>>>> Ah, sneaky.  I'm pretty sure the current binary distribution was built
>>>> when 2.6 was new, and I didn't really expect them to tweak
>>>> build-related issues in an incremental release :)
>>>>
>>>> I'll push out a new build asap; will ping the list when it's available.
>>>>
>>>> Thanks!
>>>>
>>>> </F>
>>>>
>>>> On Mon, Sep 28, 2009 at 3:46 PM, Koen van de Sande
>>>> <koenvandesande at gmail.com>  wrote:
>>>>> The fix was applied to the main Python distribution (2.6.2 has it for
>>>>> sure,
>>>>> 2.6.0/2.6.1 does not). Any extension built with 2.6.2 or higher should
>>>>> be
>>>>> fine, as it will not get an embedded manifest.
>>>>> So just rebuild with Python Windows version 2.6.2. I see in the PIL
>>>>> source
>>>>> tree that it uses a setup.py build_ext call, so no changes to PIL
>>>>> should be
>>>>> needed.
>>>>>
>>>>> Koen
>>>>> _______________________________________________
>>>>> Image-SIG maillist  -  Image-SIG at python.org
>>>>> http://mail.python.org/mailman/listinfo/image-sig
>>>>>
>>>>
>>> _______________________________________________
>>> Image-SIG maillist  -  Image-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/image-sig
>>
>>
>

From flashk at gmail.com  Fri Oct  2 18:00:28 2009
From: flashk at gmail.com (Farshid Lashkari)
Date: Fri, 2 Oct 2009 09:00:28 -0700
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <4AC61DFD.5020500@uci.edu>
References: <368a5cd50909280625o471fb496ra96b5b151a058d80@mail.gmail.com> 
	<4AC0BE40.5010801@gmail.com>
	<368a5cd50909280707n3bf75b9ma779907b7fe15fbb@mail.gmail.com> 
	<368a5cd50910010410m30070acfne864598696499605@mail.gmail.com> 
	<978d1eac0910011510x3a4f96bdr6e0f21858b398e7a@mail.gmail.com> 
	<368a5cd50910020232q66af8debya0e32138df3540f@mail.gmail.com> 
	<4AC61DFD.5020500@uci.edu>
Message-ID: <978d1eac0910020900l2791ad77m21e69bdc046aa5a2@mail.gmail.com>

I use Resource Hacker (http://www.angusj.com/resourcehacker/), it allows you
to view and edit all embedded resources of EXE/DLL files. It's a very handy
application. Dependency Walker is really nice too.
Regarding the issue, wouldn't it just be better to compile the pyd without
any embedded manifest file, instead of trying to parse and remove certain
sections? I'm not familiar with how distutils works, but for my own Visual
Studio projects I pass the /MANIFEST:NO option to the linker and it will not
generate an embedded manifest file.

On Fri, Oct 2, 2009 at 8:36 AM, Christoph Gohlke <cgohlke at uci.edu> wrote:

> Hi,
>
> you can use Dependency Walker <http://www.dependencywalker.com/>. If the
> patch is working, pyd files will depend on MSVCR90.DLL, otherwise
> %WINDIR%\winsxs\...\MSVCR90.DLL
>
> There are two issues with Koen's patch (both easy to fix). It does not work
> for 64-bit Python and it also processes exe files build with setuptools.
>
> Christoph
>
>
> On 11:59, Fredrik Lundh wrote:
>
>> Ok, I can apply Koen's patch locally and rebuild, if that makes sense.
>>
>> What's the quickest way to look for embedded manifests in the PYD:s?
>> dumpbin doesn't seem to say anything about that, so I assume I have to
>> use some other tool.
>>
>> </F>
>>
>> On Fri, Oct 2, 2009 at 12:10 AM, Farshid Lashkari<flashk at gmail.com>
>>  wrote:
>>
>>> I just tried the new installer
>>> (http://effbot.org/media/downloads/PIL-1.1.6.win32-py2.6.exe), but the
>>> pyd
>>> files still contain embedded manifests. The installer says it was built
>>> Oct
>>> 1, 2009 with Python 2.6.2, so that should be the latest version.
>>> I took a closer look at the changes made from that bug report. The change
>>> was only made to the PCbuild/pyd.vsprops file, which I don't believe will
>>> affect 3rd party extensions. I see that Koen submitted a patch
>>> for msvc9compiler.py that will strip the runtime from the manifest, but
>>> that
>>> doesn't appear to have been committed. So I'm not really sure how 3rd
>>> party
>>> extension developers can build pyd files without embedded manifests.
>>> On Thu, Oct 1, 2009 at 4:10 AM, Fredrik Lundh<fredrik at pythonware.com>
>>> wrote:
>>>
>>>>
>>>> I've replaced the 1.1.6 binary at effbot.org/downloads with one built
>>>> under 2.6.2.  The installer is 130k larger thanks to a 3x larger
>>>> wininst executable (!?), but the binary extensions are slightly
>>>> smaller than before, so I'll assume it's the manifests that have
>>>> disappeared.
>>>>
>>>> </F>
>>>>
>>>> On Mon, Sep 28, 2009 at 4:07 PM, Fredrik Lundh<fredrik at pythonware.com>
>>>> wrote:
>>>>
>>>>> Ah, sneaky.  I'm pretty sure the current binary distribution was built
>>>>> when 2.6 was new, and I didn't really expect them to tweak
>>>>> build-related issues in an incremental release :)
>>>>>
>>>>> I'll push out a new build asap; will ping the list when it's available.
>>>>>
>>>>> Thanks!
>>>>>
>>>>> </F>
>>>>>
>>>>> On Mon, Sep 28, 2009 at 3:46 PM, Koen van de Sande
>>>>> <koenvandesande at gmail.com>  wrote:
>>>>>
>>>>>> The fix was applied to the main Python distribution (2.6.2 has it for
>>>>>> sure,
>>>>>> 2.6.0/2.6.1 does not). Any extension built with 2.6.2 or higher should
>>>>>> be
>>>>>> fine, as it will not get an embedded manifest.
>>>>>> So just rebuild with Python Windows version 2.6.2. I see in the PIL
>>>>>> source
>>>>>> tree that it uses a setup.py build_ext call, so no changes to PIL
>>>>>> should be
>>>>>> needed.
>>>>>>
>>>>>> Koen
>>>>>> _______________________________________________
>>>>>> Image-SIG maillist  -  Image-SIG at python.org
>>>>>> http://mail.python.org/mailman/listinfo/image-sig
>>>>>>
>>>>>>
>>>>>  _______________________________________________
>>>> Image-SIG maillist  -  Image-SIG at python.org
>>>> http://mail.python.org/mailman/listinfo/image-sig
>>>>
>>>
>>>
>>>
>>  _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091002/c064dbc5/attachment-0001.htm>

From bentson at holmsjoen.com  Sun Oct  4 00:56:00 2009
From: bentson at holmsjoen.com (Randolph Bentson)
Date: Sat, 3 Oct 2009 15:56:00 -0700
Subject: [Image-SIG] fix for wide line directional dependence
In-Reply-To: <20090930145631.GA2641@holmsjoen.com>
References: <20090930145631.GA2641@holmsjoen.com>
Message-ID: <20091003225600.GB30249@holmsjoen.com>

On Wed, Sep 30, 2009 at 07:56:32AM -0700, Randolph Bentson wrote:
> I've found that wide lines drawn with PIL 1.1.6 have a directional
> dependence, e.g., lines from (0,0) to (-100,0) are thinner than those
> from (0,0) to (0,-100).

A few days ago I posted a hack which reduced the magnitude of the problem.
The following patch does a better job at little additional cost.

I've introduced variables dX and dY to improve the readability of the
code which computes the values of the pixel-based dx and dy values. The
reference to "diagonal lines" doesn't implement proper antialiasing,
but it improves things with modest code change.)

Feedback is welcome because I do hope this'll get in the 1.1.7 release.


--- OLD_Draw.c	2009-10-03 14:44:31.000000000 -0700
+++ Draw.c	2009-10-03 15:42:41.000000000 -0700
@@ -665,6 +665,7 @@
 
     Edge e[4];
 
+    int dX, dY;
     int dx, dy;
     double d;
 
@@ -675,18 +676,34 @@
         return 0;
     }
 
-    dx = x1-x0;
-    dy = y1-y0;
+    dX = x1-x0;
+    dY = y1-y0;
 
-    if (dx == 0 && dy == 0) {
+    if (dX == 0 && dY == 0) {
         draw->point(im, x0, y0, ink);
         return 0;
     }
 
-    d = width / sqrt(dx*dx + dy*dy) / 2.0;
+    d = width / sqrt(dX*dX + dY*dY) / 2.0;
 
-    dx = (int) (d * (y1-y0) + 0.5);
-    dy = (int) (d * (x1-x0) + 0.5);
+    /* Negative values are rounded down and positive
+       values are rounded up, else the deltas disappear!
+     */
+    if (dY < 0)  dx = (int) (d * dY - 0.5);
+    else         dx = (int) (d * dY + 0.5);
+    if (dX < 0)  dy = (int) (d * dX - 0.5);
+    else         dy = (int) (d * dX + 0.5);
+
+    /* This is necessary because diagonal lines are too wide.
+     */
+    if ( 4*(dx*dx+dy*dy) > (width*width) ){
+        if (abs(dx) > abs(dy))
+            if (dx < 0) dx ++;
+            else        dx --;
+        else
+            if (dy < 0) dy ++;
+            else        dy --;
+    }
 
     add_edge(e+0, x0 - dx,  y0 + dy, x1 - dx,  y1 + dy);
     add_edge(e+1, x1 - dx,  y1 + dy, x1 + dx,  y1 - dy);

-- 
Randolph Bentson
bentson at holmsjoen.com

From mark.wendell at gmail.com  Sun Oct  4 18:47:58 2009
From: mark.wendell at gmail.com (Mark Wendell)
Date: Sun, 4 Oct 2009 10:47:58 -0600
Subject: [Image-SIG] image warping with pil
Message-ID: <b22700f10910040947g6fc555dk3812a64adc3548ec@mail.gmail.com>

I'm assuming that PIL doesn't provide any non-linear image warping
functions natively. Has anyone used the PIL tools to do this, though?
I'm looking for a method or algorithm to warp an image based on an
arbitrary vector field. All the better if it's a pure-python
implementation.

thanks in advance,
Mark

--
--
Mark Wendell

From fredrik at pythonware.com  Mon Oct  5 10:53:42 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 5 Oct 2009 10:53:42 +0200
Subject: [Image-SIG] Some issue with easy_install and PIL/Imaging
In-Reply-To: <4AC3864C.3070704@simplistix.co.uk>
References: <h8869e$l0n$1@ger.gmane.org> <4AAA557E.70508@simplistix.co.uk> 
	<368a5cd50909280601i39c065fbob8fa33166162072f@mail.gmail.com> 
	<4AC3864C.3070704@simplistix.co.uk>
Message-ID: <368a5cd50910050153l4fdddafcp31255ff0a366efdb@mail.gmail.com>

The problem is that too many people arguing for eggs do this by
sending nastygrams, which doesn't really provide much motivation for
doing anything about it (I don't do asshole-driven development).  The
public review PIL got a couple a minutes ago matches some of the
private mail I've gotten:

   no egg - worst seen ever, remove it from pypi or provide an egg
(jensens, 2009-10-05, 0 points)

</F>

On Wed, Sep 30, 2009 at 6:24 PM, Chris Withers <chris at simplistix.co.uk> wrote:
> Fredrik Lundh wrote:
>>
>> On Fri, Sep 11, 2009 at 3:49 PM, Chris Withers <chris at simplistix.co.uk>
>> wrote:
>>>
>>> Klein St?phane wrote:
>>>>
>>>> Resume :
>>>> 1. first question : why PIL package in "pypi" don't work ?
>>>
>>> Because Fred Lundh have his package distributions unfortunate names that
>>> setuptools doesn't like...
>>
>> It used to support this, but no longer does. ?To me, that says more
>> about the state of setuptools than it does about the state of PIL,
>> which has been using the same naming convention for 15 years.
>
> Yep, but it is now in the minority, and consistency in package naming is
> always good.
>
> Would there be any problems for you in naming the distribution in a
> setuptools-friendly way from the next point release?
>
> cheers,
>
> Chris
>
> --
> Simplistix - Content Management, Batch Processing & Python Consulting
> ? ? ? ? ? - http://www.simplistix.co.uk
>

From chris at simplistix.co.uk  Mon Oct  5 11:19:09 2009
From: chris at simplistix.co.uk (Chris Withers)
Date: Mon, 05 Oct 2009 10:19:09 +0100
Subject: [Image-SIG] Some issue with easy_install and PIL/Imaging
In-Reply-To: <368a5cd50910050153l4fdddafcp31255ff0a366efdb@mail.gmail.com>
References: <h8869e$l0n$1@ger.gmane.org> <4AAA557E.70508@simplistix.co.uk>
	<368a5cd50909280601i39c065fbob8fa33166162072f@mail.gmail.com>
	<4AC3864C.3070704@simplistix.co.uk>
	<368a5cd50910050153l4fdddafcp31255ff0a366efdb@mail.gmail.com>
Message-ID: <4AC9BA0D.7010504@simplistix.co.uk>

Fredrik Lundh wrote:
> The problem is that too many people arguing for eggs do this by
> sending nastygrams, which doesn't really provide much motivation for
> doing anything about it (I don't do asshole-driven development). 

Indeed, I couldn't agree more, and I'm sorry you've been subjected to this.

My (hopefully more polite) request still stands though:
 >> Would there be any problems for you in naming the distribution in a
 >> setuptools-friendly way from the next point release?

 > The
> public review PIL got a couple a minutes ago matches some of the
> private mail I've gotten:
> 
>    no egg - worst seen ever, remove it from pypi or provide an egg
> (jensens, 2009-10-05, 0 points)

*sigh*

Interesting timing, myself and Doug Hellmann have been trying to 
persuade Martin von Lewis that while ratings are fine, the commenting 
system is likely to be abused:

https://sourceforge.net/tracker/?func=detail&atid=513503&aid=2866081&group_id=66150
https://sourceforge.net/tracker/?func=detail&atid=513503&aid=2872293&group_id=66150

I'll put in a request to have that comment removed...

Chris

-- 
Simplistix - Content Management, Batch Processing & Python Consulting
            - http://www.simplistix.co.uk

From koenvandesande at gmail.com  Mon Oct  5 19:50:10 2009
From: koenvandesande at gmail.com (Koen van de Sande)
Date: Mon, 05 Oct 2009 19:50:10 +0200
Subject: [Image-SIG] Windows CRT issues with Python 2.6
In-Reply-To: <978d1eac0910020900l2791ad77m21e69bdc046aa5a2@mail.gmail.com>
References: <978d1eac0910020900l2791ad77m21e69bdc046aa5a2@mail.gmail.com>
Message-ID: <4ACA31D2.9090109@gmail.com>

To check for the manifest, I just open it up in a hex editor and look 
for the word "manifest". That will get you to the XML part of the file, 
and you can see what's happening.
Please note that if you turn manifests off, you assume that your 
extension has no other dependencies. And for things like TCL/TK, Windows 
Common Controls, there can be additional dependencies in the XML file. 
So, only remove the dependency on the runtimes, otherwise you break 
things with additional dependencies.

Koen

From JSeth at eastech.org  Fri Oct  9 14:27:01 2009
From: JSeth at eastech.org (Joni Seth)
Date: Fri, 09 Oct 2009 08:27:01 -0400
Subject: [Image-SIG] Cannot get PIL to install with jpeg	
	support&In-Reply-To=<00F4C884-D55B-48C9-9927-7A5ECB007662@pharmacy.arizona.edu>
Message-ID: <4ACEF3D5020000EA00020F30@gw.eastech.org>

Hello Bruce,
  I found your post regarding "Cannot get PIL to install with jpeg support", because I am having the exact same problem. Unfortunately, I also noticed that no one responded to your question. Did you ever find out how to get this to work? I sure would appreciate it if you could share any "secrets" that you have come across... it must be a great secret.
 
thank you,
  Joni Seth
 
 
 
 
  Joni Seth
  IT Support Specialist
  Eastern Center for Arts and Technology
  3075 Terwood Rd
  Willow Grove, Pa 19090
  (215) 784-4800 x274
  jseth at eastech.org 
 

CONFIDENTIALITY NOTICE: This e-mail is intended only for the use of the individual or entity to which it is addressed and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you have received this communication in error, please do not, print, disclose, copy, retain or redistribute it.  Please delete the original message after notifying the sender by e-mail at the address shown. Thank you for your compliance.

Eastern Center for Arts and Technology

CONFIDENTIALITY NOTICE: This E-Mail is intended only for the
use of the individual or entity to which it is addressed and
may contain information that is privileged, confidential and
exempt from disclosure under applicable law. If you have
received this communication in error, please do not print,
disclose, copy, retain or redistribute it. Please delete the
original message after notifying the sender by E-Mail at the
address shown. Thank you for your compliance.
Eastern Center for Arts and Technology

CONFIDENTIALITY NOTICE: This E-Mail is intended only for the
use of the individual or entity to which it is addressed and
may contain information that is privileged, confidential and
exempt from disclosure under applicable law. If you have
received this communication in error, please do not print,
disclose, copy, retain or redistribute it. Please delete the
original message after notifying the sender by E-Mail at the
address shown. Thank you for your compliance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091009/335b204c/attachment.htm>

From fbuchinger at gmail.com  Mon Oct 12 14:55:47 2009
From: fbuchinger at gmail.com (Franz Buchinger)
Date: Mon, 12 Oct 2009 14:55:47 +0200
Subject: [Image-SIG] How to get quality of picture
In-Reply-To: <300427.14251.qm@web57603.mail.re1.yahoo.com>
References: <200909281508202036150@ucweb.com>
	<368a5cd50909280543w3b79d173j35e73a9cc3c48055@mail.gmail.com>
	<300427.14251.qm@web57603.mail.re1.yahoo.com>
Message-ID: <a07b86220910120555h18ec9ec2o4d6cac8fafcce2ad@mail.gmail.com>

As Frederik already mentioned, you can extract the quantisation table of an
opened JPEG file with the .quantization attribute.
To determine the quality setting of the jpeg, you can calculate some hash
(e.g. md5) for the quantisation table and compare it to hashes from jpeg
files saved at different quality settings.

PIL and most other FOSS image processing apps (GIMP, Imagemagick) depend on
libjpeg and its builtin quantisation tables. This means that you can easily
determine the quality of images processed with these apps.
JPEGs saved in commercial apps like Photoshop, Lightroom or directly taken
from a camera however bring their own quality settings and quantisation
tables. An image exported with the "Save for Web" dialog in Photoshop for
example can come in 12 different quality settings, each corresponding to a
different quantisation table which have nothing in common with the ones
libjpeg is using.

In short words: you need a huge database of quantisation table hashes to
reliably identify the quality setting of an arbitrary jpeg. Therefore, I
recommend to make use of exiftool's jpegdigest database, which holds
thousands of quantisation table hashes:

http://cpansearch.perl.org/src/EXIFTOOL/Image-ExifTool-7.89/lib/Image/ExifTool/JPEGDigest.pm

You only have to port the jpegdigest hashing algorithm from perl to python
and convert the database to a dictionary.

kind regards,

Franz

2009/9/29 David Berthelot <d_berthelot at yahoo.com>

> Here's some code that does the trick.
>
> It returns a tuple:
> (coefficient, (luminance_coefficient,luminance_error),
> (chrominance_coefficient,chrominance_error))
>
> Note:
> - The 2nd and 3rd returned items are only useful if you're interested in
> the error (sometimes the error is 0 if the program used a JPEG table that is
> the same as the one used by PIL).
> - Fredrik: You can freely use the code and embed it in PIL if you're
> satisfied with its quality
>
> Code:
> def get_jpeg_quantization_tables(f):
>     """Returns the JPEG quantization tables of a filename or file
> descriptor"""
>     from functools import partial
>     import Image
>     import numpy as N
>     fd    = Image.open(f)
>     if fd.format == 'BMP':
>         return N.ones((8,8)),N.ones((8,8))
>     q     = fd.quantization
>     ql,qc =
> map(partial(N.array,dtype='uint8'),(q.get(0,N.zeros(64)),q.get(1,N.zeros(64))))
>     # Now reorder the JPEG quantification coefficients as 8x8 matrices
>     jpeg_natural_order = N.array([ 0,  1,  8, 16,  9,  2,  3, 10, 17, 24,
> 32, 25, 18, 11,  4,  5, 12,
>                                    19, 26, 33, 40, 48, 41, 34, 27, 20, 13,
> 6,  7, 14, 21, 28, 35, 42,
>                                    49, 56, 57, 50, 43, 36, 29, 22, 15, 23,
> 30, 37, 44, 51, 58, 59, 52,
>                                    45, 38, 31, 39, 46, 53, 60, 61, 54, 47,
> 55, 62, 63])
>     rql = N.zeros(64,'f')
>     rqc = N.zeros(64,'f')
>     for x,y in N.ndenumerate(jpeg_natural_order):
>         rql[y] = ql[x]
>         rqc[y] = qc[x]
>     return rql.reshape(8,8),rqc.reshape(8,8) # Luminance/Chrominance
>
> def guess_jpeg_quality(f,ctables=[]):
>     from StringIO import StringIO
>     import numpy as N
>     import Image
>     # Compute the tables for quality = 1..100 by saving fakes files in
> memory
>     if not ctables:
>         tables = N.zeros((100,2,8,8),'f')
>         for x in xrange(100):
>             fd = StringIO()
>             Image.new('RGB',(64,64)).save(fd,"jpeg",quality=1+x)
>             fd.seek(0)
>             ql,qc = get_jpeg_quantization_tables(fd)
>             tables[x][0] = ql
>             tables[x][1] = qc
>         ctables.append(tables)
>     else:
>         tables = ctables[0]
>     # Use a weighting matrix w to put more emphasis on the comparison of
> lower DCT harmonics
>     w     = 1./N.outer(1+N.arange(8)/7.,1+N.arange(8)/7.)
>     ql,qc = get_jpeg_quantization_tables(f)
>     # Compute errors on Luminance and Chrominance tables
>     errsl,errsc = [],[]
>     for x in xrange(100):
>         errsl.append((N.square((ql-tables[x][0])*w).mean(),x))
>         errsc.append((N.square((qc-tables[x][1])*w).mean(),x))
>     # Select minimal error coefficients
>     lmin = min(errsl)
>     cmin = min(errsc)
>     # Weight coefficients average (more emphasis put on luminance since it
> affects most visual perception)
>     q = int(round(lmin[1]*.8 + cmin[1]*.2))
>     return q,lmin,cmin
>
>
>
> ------------------------------
> *From:* Fredrik Lundh <fredrik at pythonware.com>
> *To:* qiaohl <qiaohl at ucweb.com>
> *Cc:* image-sig <image-sig at python.org>
> *Sent:* Monday, September 28, 2009 5:43:00 AM
> *Subject:* Re: [Image-SIG] How to get quality of picture
>
> The quality setting is used to create a quantization table which is
> then used by the compression algorithm.  There's no pre-defined
> mapping between quality and the contents of the quantization table for
> JPEG (different implementations do different things), but some
> applications attempt to guess by comparing the quantization tables in
> the file with known mapping algorithms.  I'm not aware of any such
> code for Python.
>
> If you want to tinker with this, you can access the quantization table
> of an opened JPEG file via the "quantization" attribute.
>
> </F>
>
> On Mon, Sep 28, 2009 at 9:08 AM, qiaohl <qiaohl at ucweb.com> wrote:
> > Hi,all
> >    Is there any function in PIL to get the value of JPEG quality(1-100)?
> I
> > know there are functions to set this value, But I found no function to
> get
> > this value of an existing JPEG file.
> >
> > Thanks!
> >
> >
> > 2009-09-28
> > ________________________________
> > qiaohl
> > _______________________________________________
> > Image-SIG maillist  -  Image-SIG at python.org
> > http://mail.python.org/mailman/listinfo/image-sig
> >
> >
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091012/f1edca2e/attachment.htm>

From johnson at pharmacy.arizona.edu  Mon Oct 12 17:35:08 2009
From: johnson at pharmacy.arizona.edu (Bruce Johnson)
Date: Mon, 12 Oct 2009 08:35:08 -0700
Subject: [Image-SIG] Cannot get PIL to install with jpeg	
	support&In-Reply-To=<00F4C884-D55B-48C9-9927-7A5ECB007662@pharmacy.arizona.edu>
In-Reply-To: <4ACEF3D5020000EA00020F30@gw.eastech.org>
References: <4ACEF3D5020000EA00020F30@gw.eastech.org>
Message-ID: <34DE80A7-177B-4390-9F24-AF68E157F7C9@pharmacy.arizona.edu>


On Oct 9, 2009, at 5:27 AM, Joni Seth wrote:

> Hello Bruce,
>  I found your post regarding "Cannot get PIL to install with jpeg  
> support", because I am having the exact same problem. Unfortunately,  
> I also noticed that no one responded to your question. Did you ever  
> find out how to get this to work? I sure would appreciate it if you  
> could share any "secrets" that you have come across... it must be a  
> great secret.


It turned out to be an error with the selftest.py stuff; jpeg was  
installed correctly all along. Here's the web reference I found  
explaining it:

<http://blog.tlensing.org/2008/12/04/kill-pil-?-the-python-imaging-library-headache/ 
 >

<http://tinyurl.com/yhzxbh8> if the above link breaks.

-- 
Bruce Johnson
University of Arizona
College of Pharmacy
Information Technology Group

Institutions do not have opinions, merely customs



From r1chardj0n3s at gmail.com  Tue Oct 13 07:25:54 2009
From: r1chardj0n3s at gmail.com (Richard Jones)
Date: Tue, 13 Oct 2009 16:25:54 +1100
Subject: [Image-SIG] ImageTk.PhotoImage Bus Error on Snow Leopard
Message-ID: <9229EC3F-0BDA-4E5A-AC8C-6437B9F8329A@gmail.com>

I'm getting a Bus Error attempting to use ImageTk.PhotoImage on Snow  
Leopard. I've compiled PIL myself. I'm using the system Tcl/Tk (8.4).  
The selftest.py runs fine.

The code I'm attempting is:

import Tkinter as tk
root = tk.Tk()
image = Image.open('Images/lena.jpg')
photo = ImageTk.PhotoImage(image)

and I get the Bus Error on the last line. gdb (which I'm quite rusty  
at driving) says:

 >>> photo = ImageTk.PhotoImage(image)
Reading symbols for shared libraries ... done

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_PROTECTION_FAILURE at address: 0x00000028
0x16e72a28 in Tk_GetImageMasterData ()
(gdb)

I've manually traced and I know the code is breaking somewhere in  
PyImagingPhotoPut.

I've confirmed the image data loads OK (Image.load) by using the same  
image code in a pyglet program and displaying it.

Any suggestions?


       Richard


From Lbrannma at yahoo.com  Fri Oct 16 00:52:23 2009
From: Lbrannma at yahoo.com (LL)
Date: Fri, 16 Oct 2009 00:52:23 +0200
Subject: [Image-SIG] Broken pipe converting eps to png using PIL
Message-ID: <E508E2AD6A264AC9A870577104FB6687@leb65dd463c13e>

Hi.. all of the following code executes interactively except the last line. Converting a .jpg file to .png works fine. I'm using the PIL version for Python 2.6. Any suggestions will be greatly appreciated. Thanks, Lance
------------------------------------
PythonWin 2.6.3 (r263:75183, Oct  5 2009, 14:41:55) [MSC v.1500 32 bit (Intel)] on win32.
Portions Copyright 1994-2008 Mark Hammond - see 'Help/About PythonWin' for further copyright information.
>>> import os
>>> import zlib
>>> import Image
>>> os.chdir("c:\\foo")
>>> img = Image.open("foo1.eps")
>>> img.save("foo1.png")
Traceback (most recent call last):
  File "<interactive input>", line 1, in <module>
  File "C:\Python26\lib\site-packages\PIL\Image.py", line 1372, in save
    self.load()
  File "C:\Python26\lib\site-packages\PIL\EpsImagePlugin.py", line 283, in load
    self.im = Ghostscript(self.tile, self.size, self.fp)
  File "C:\Python26\lib\site-packages\PIL\EpsImagePlugin.py", line 72, in Ghostscript
    gs.write(s)
IOError: [Errno 32] Broken pipe
>>> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091016/3dede733/attachment-0001.htm>

From cannon.el at gmail.com  Fri Oct 16 06:55:12 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Thu, 15 Oct 2009 21:55:12 -0700
Subject: [Image-SIG] unexpected result when using Image.composite
Message-ID: <b986b290910152155h3cecd954rfab76f2887579288@mail.gmail.com>

I am using Image.composite to combine two images and am getting
unexpected results. Partially transparent areas are appearing
discolored. When I preform the same operation in the Gimp, the result
is as expected. A script demonstrating the problem is

import Image

fg = Image.open("a0001.png")
bg = Image.open("bg.png").resize(fg.size, Image.ANTIALIAS)
Image.composite(fg, bg, fg).save("out.png")


I have attached the source files for reference. Any insight into this
problem would be appreciated.
Edward Cannon
-------------- next part --------------
A non-text attachment was scrubbed...
Name: a0001.png
Type: image/png
Size: 93173 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091015/9f4bad2f/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bg.png
Type: image/png
Size: 212882 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091015/9f4bad2f/attachment-0003.png>

From Lbrannma at yahoo.com  Sat Oct 17 14:38:38 2009
From: Lbrannma at yahoo.com (LL)
Date: Sat, 17 Oct 2009 14:38:38 +0200
Subject: [Image-SIG] Broken pipe converting eps to png using PIL
References: <E508E2AD6A264AC9A870577104FB6687@leb65dd463c13e>
Message-ID: <0CA6386B74114FF1BFDFA7A39094EAF2@leb65dd463c13e>

Hi.. here is an update regarding my eps to png problem. I am running XP... and discovered that my AFPL Ghostscript executable is entitled gswin32 rather than gs. I edited the PIL file EpsImagePlugin.py file (see below) but still have the same problem. Probably my thinking was incorrect.. since converting from .jpg to .png via the PIL (file.open then save) works fine. Perhaps there is a separate Ghostview within Python 2.6,... but I searched for gs.exe in my Python installation directory and did not find anything. I'm still puzzled

  File "C:\Python26\lib\site-packages\PIL\EpsImagePlugin.py", line 72, in Ghostscript
    C:\Progra~1\gs\gs8.53\bin\gswin32.write(s)
IOError: [Errno 32] Broken pipe
  ----- Original Message ----- 
  From: LL 
  To: image-sig at python.org 
  Sent: Friday, October 16, 2009 12:52 AM
  Subject: [Image-SIG] Broken pipe converting eps to png using PIL


  Hi.. all of the following code executes interactively except the last line. Converting a .jpg file to .png works fine. I'm using the PIL version for Python 2.6. Any suggestions will be greatly appreciated. Thanks, Lance
  ------------------------------------
  PythonWin 2.6.3 (r263:75183, Oct  5 2009, 14:41:55) [MSC v.1500 32 bit (Intel)] on win32.
  Portions Copyright 1994-2008 Mark Hammond - see 'Help/About PythonWin' for further copyright information.
  >>> import os
  >>> import zlib
  >>> import Image
  >>> os.chdir("c:\\foo")
  >>> img = Image.open("foo1.eps")
  >>> img.save("foo1.png")
  Traceback (most recent call last):
    File "<interactive input>", line 1, in <module>
    File "C:\Python26\lib\site-packages\PIL\Image.py", line 1372, in save
      self.load()
    File "C:\Python26\lib\site-packages\PIL\EpsImagePlugin.py", line 283, in load
      self.im = Ghostscript(self.tile, self.size, self.fp)
    File "C:\Python26\lib\site-packages\PIL\EpsImagePlugin.py", line 72, in Ghostscript
      gs.write(s)
  IOError: [Errno 32] Broken pipe
  >>> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091017/d112ed67/attachment.htm>

From cannon.el at gmail.com  Mon Oct 19 03:50:11 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Sun, 18 Oct 2009 18:50:11 -0700
Subject: [Image-SIG] unexpected result when using Image.composite
In-Reply-To: <b986b290910152155h3cecd954rfab76f2887579288@mail.gmail.com>
References: <b986b290910152155h3cecd954rfab76f2887579288@mail.gmail.com>
Message-ID: <b986b290910181850v10e4c873p67bef934e7cfc9@mail.gmail.com>

I have discovered the source of the problem. the bg image is mode
RGBA, when I convert it to RGB, the result is as expected.

On Thu, Oct 15, 2009 at 9:55 PM, Laura & Edward Cannon
<cannon.el at gmail.com> wrote:
> I am using Image.composite to combine two images and am getting
> unexpected results. Partially transparent areas are appearing
> discolored. When I preform the same operation in the Gimp, the result
> is as expected. A script demonstrating the problem is
>
> import Image
>
> fg = Image.open("a0001.png")
> bg = Image.open("bg.png").resize(fg.size, Image.ANTIALIAS)
> Image.composite(fg, bg, fg).save("out.png")
>
>
> I have attached the source files for reference. Any insight into this
> problem would be appreciated.
> Edward Cannon
>

From ivan.mincik at gmail.com  Mon Oct 19 23:05:53 2009
From: ivan.mincik at gmail.com (Ivan Mincik)
Date: Mon, 19 Oct 2009 23:05:53 +0200
Subject: [Image-SIG] transparency in palette mode ( "P" ) images
Message-ID: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com>

Dear list,
I have PNG image in  palette mode, with 256 color palette. Every color
has different level of transparency.
For example:
  Color Table (RGB with 256 entries)
    0: 0,0,0,0
    1: 238,238,238,154
    2: 253,253,109,154
    3: 0,0,0,2
    4: 0,0,0,4
    5: 0,0,0,8
    6: 0,0,0,6
    7: 237,237,237,154
    8: 0,0,0,14
    9: 35,35,15,153
   10: 1,1,0,138

I am trying just to open and save this image using PIL.
#!/usr/bin/python
import Image

im = Image.open('test.png')
im.save('out.png', transparency=0 )

After running this code, the resulting image is loosing transparency
information for each color and it is set to 255.
 Color Table (RGB with 256 entries)
    0: 0,0,0,0
    1: 238,238,238,255
    2: 253,253,109,255
    3: 0,0,0,255
    4: 0,0,0,255
    5: 0,0,0,255
    6: 0,0,0,255
    7: 237,237,237,255
    8: 0,0,0,255
    9: 35,35,15,255
   10: 1,1,0,255

Is there any way how to save the image with exactly same palette as
original image ?

For anybody wanting to look at this problem, You can download my test
image here:
http://gista.sk/dl/test.png

Thanks,
Ivan

From ivan.mincik at gista.sk  Tue Oct 20 18:53:44 2009
From: ivan.mincik at gista.sk (Ivan Mincik)
Date: Tue, 20 Oct 2009 18:53:44 +0200
Subject: [Image-SIG] transparency in palette mode ( "P" ) images
In-Reply-To: <b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com>
References: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com>
	<b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com>
Message-ID: <200910201853.49236.ivan.mincik@gista.sk>

On Tuesday 20 October 2009, you wrote:
> The problem seems to be that you set transparency=0 in your save.
> Leave it out ( im.save("out.png") ) and you should be fine.

Many thanks for reply, but I think You are not right. If I save without 'transparency=0' I will loose my transparent background at all - it will be black.  With  'transparency=0' I loose only transparency of my drawing, backround is OK.
My test image is here: http://gista.sk/dl/test.png


> If you just need to change the name try shutil.copy in the standard library.
> 
> On Mon, Oct 19, 2009 at 2:05 PM, Ivan Mincik <ivan.mincik at gmail.com> wrote:
> > Dear list,
> > I have PNG image in ?palette mode, with 256 color palette. Every color
> > has different level of transparency.
> > For example:
> > ?Color Table (RGB with 256 entries)
> > ? ?0: 0,0,0,0
> > ? ?1: 238,238,238,154
> > ? ?2: 253,253,109,154
> > ? ?3: 0,0,0,2
> > ? ?4: 0,0,0,4
> > ? ?5: 0,0,0,8
> > ? ?6: 0,0,0,6
> > ? ?7: 237,237,237,154
> > ? ?8: 0,0,0,14
> > ? ?9: 35,35,15,153
> > ? 10: 1,1,0,138
> >
> > I am trying just to open and save this image using PIL.
> > #!/usr/bin/python
> > import Image
> >
> > im = Image.open('test.png')
> > im.save('out.png', transparency=0 )
> >
> > After running this code, the resulting image is loosing transparency
> > information for each color and it is set to 255.
> > ?Color Table (RGB with 256 entries)
> > ? ?0: 0,0,0,0
> > ? ?1: 238,238,238,255
> > ? ?2: 253,253,109,255
> > ? ?3: 0,0,0,255
> > ? ?4: 0,0,0,255
> > ? ?5: 0,0,0,255
> > ? ?6: 0,0,0,255
> > ? ?7: 237,237,237,255
> > ? ?8: 0,0,0,255
> > ? ?9: 35,35,15,255
> > ? 10: 1,1,0,255
> >
> > Is there any way how to save the image with exactly same palette as
> > original image ?
> >
> > For anybody wanting to look at this problem, You can download my test
> > image here:
> > http://gista.sk/dl/test.png
> >
> > Thanks,
> > Ivan
> > _______________________________________________
> > Image-SIG maillist ?- ?Image-SIG at python.org
> > http://mail.python.org/mailman/listinfo/image-sig
> >
> 



-- 
Ivan Mincik
Gista s.r.o.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091020/36b35bbc/attachment.pgp>

From ivan.mincik at gista.sk  Tue Oct 20 20:38:40 2009
From: ivan.mincik at gista.sk (Ivan Mincik)
Date: Tue, 20 Oct 2009 20:38:40 +0200
Subject: [Image-SIG]  transparency in palette mode ( "P" ) images
In-Reply-To: <1ed604b80910201137p49040c5akfa767474d13eb006@mail.gmail.com>
References: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com> 
	<b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com> 
	<200910201853.49236.ivan.mincik@gista.sk>
	<b986b290910201133n33332503ge397bc8e5ce0ba4@mail.gmail.com> 
	<1ed604b80910201137p49040c5akfa767474d13eb006@mail.gmail.com>
Message-ID: <1ed604b80910201138w49293906v42ceb73855e04f9f@mail.gmail.com>

On Tue, Oct 20, 2009 at 8:33 PM, Laura & Edward Cannon
<cannon.el at gmail.com> wrote:
> Hm.. you seem to be right. One thing you might do is im =
> Image.open("test.png").convert("RGBA") which seems to solve the
> transparency problem, at the expense of a slightly larger image file.
> Looking closer I think that PIL seems to support only RGB palettes
> rather than RGBA palettes as does the PNG specification. Converting to
> RGBA would make editing work as expected, and would preserve all
> transparency information.

Thanks, between open() and save() I have some image processing which
is unusable without correct saving.

If you just need to move/rename the image,
> try shutil.copy in the standard library.
> Edward
>
> On Tue, Oct 20, 2009 at 9:53 AM, Ivan Mincik <ivan.mincik at gista.sk> wrote:
>> On Tuesday 20 October 2009, you wrote:
>>> The problem seems to be that you set transparency=0 in your save.
>>> Leave it out ( im.save("out.png") ) and you should be fine.
>>
>> Many thanks for reply, but I think You are not right. If I save without 'transparency=0' I will loose my transparent background at all - it will be black. ?With ?'transparency=0' I loose only transparency of my drawing, backround is OK.
>> My test image is here: http://gista.sk/dl/test.png
>>
>>
>>> If you just need to change the name try shutil.copy in the standard library.
>>>
>>> On Mon, Oct 19, 2009 at 2:05 PM, Ivan Mincik <ivan.mincik at gmail.com> wrote:
>>> > Dear list,
>>> > I have PNG image in ?palette mode, with 256 color palette. Every color
>>> > has different level of transparency.
>>> > For example:
>>> > ?Color Table (RGB with 256 entries)
>>> > ? ?0: 0,0,0,0
>>> > ? ?1: 238,238,238,154
>>> > ? ?2: 253,253,109,154
>>> > ? ?3: 0,0,0,2
>>> > ? ?4: 0,0,0,4
>>> > ? ?5: 0,0,0,8
>>> > ? ?6: 0,0,0,6
>>> > ? ?7: 237,237,237,154
>>> > ? ?8: 0,0,0,14
>>> > ? ?9: 35,35,15,153
>>> > ? 10: 1,1,0,138
>>> >
>>> > I am trying just to open and save this image using PIL.
>>> > #!/usr/bin/python
>>> > import Image
>>> >
>>> > im = Image.open('test.png')
>>> > im.save('out.png', transparency=0 )
>>> >
>>> > After running this code, the resulting image is loosing transparency
>>> > information for each color and it is set to 255.
>>> > ?Color Table (RGB with 256 entries)
>>> > ? ?0: 0,0,0,0
>>> > ? ?1: 238,238,238,255
>>> > ? ?2: 253,253,109,255
>>> > ? ?3: 0,0,0,255
>>> > ? ?4: 0,0,0,255
>>> > ? ?5: 0,0,0,255
>>> > ? ?6: 0,0,0,255
>>> > ? ?7: 237,237,237,255
>>> > ? ?8: 0,0,0,255
>>> > ? ?9: 35,35,15,255
>>> > ? 10: 1,1,0,255
>>> >
>>> > Is there any way how to save the image with exactly same palette as
>>> > original image ?
>>> >
>>> > For anybody wanting to look at this problem, You can download my test
>>> > image here:
>>> > http://gista.sk/dl/test.png
>>> >
>>> > Thanks,
>>> > Ivan
>>> > _______________________________________________
>>> > Image-SIG maillist ?- ?Image-SIG at python.org
>>> > http://mail.python.org/mailman/listinfo/image-sig
>>> >
>>>
>>
>>
>>
>> --
>> Ivan Mincik
>> Gista s.r.o.
>>
>

From ivan.mincik at gista.sk  Tue Oct 20 20:51:13 2009
From: ivan.mincik at gista.sk (Ivan Mincik)
Date: Tue, 20 Oct 2009 20:51:13 +0200
Subject: [Image-SIG] transparency in palette mode ( "P" ) images
In-Reply-To: <b986b290910201133n33332503ge397bc8e5ce0ba4@mail.gmail.com>
References: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com> 
	<b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com> 
	<200910201853.49236.ivan.mincik@gista.sk>
	<b986b290910201133n33332503ge397bc8e5ce0ba4@mail.gmail.com>
Message-ID: <1ed604b80910201151i3e5776bel5578ebbfe8e1cd0b@mail.gmail.com>

On Tue, Oct 20, 2009 at 8:33 PM, Laura & Edward Cannon
<cannon.el at gmail.com> wrote:
> Hm.. you seem to be right. One thing you might do is im =
> Image.open("test.png").convert("RGBA") which seems to solve the
> transparency problem, at the expense of a slightly larger image file.

No, even if using convert("RGBA") I get incorrect result :(


> Looking closer I think that PIL seems to support only RGB palettes
> rather than RGBA palettes as does the PNG specification. Converting to
> RGBA would make editing work as expected, and would preserve all
> transparency information. If you just need to move/rename the image,
> try shutil.copy in the standard library.
> Edward
>
> On Tue, Oct 20, 2009 at 9:53 AM, Ivan Mincik <ivan.mincik at gista.sk> wrote:
>> On Tuesday 20 October 2009, you wrote:
>>> The problem seems to be that you set transparency=0 in your save.
>>> Leave it out ( im.save("out.png") ) and you should be fine.
>>
>> Many thanks for reply, but I think You are not right. If I save without 'transparency=0' I will loose my transparent background at all - it will be black. ?With ?'transparency=0' I loose only transparency of my drawing, backround is OK.
>> My test image is here: http://gista.sk/dl/test.png
>>
>>
>>> If you just need to change the name try shutil.copy in the standard library.
>>>
>>> On Mon, Oct 19, 2009 at 2:05 PM, Ivan Mincik <ivan.mincik at gmail.com> wrote:
>>> > Dear list,
>>> > I have PNG image in ?palette mode, with 256 color palette. Every color
>>> > has different level of transparency.
>>> > For example:
>>> > ?Color Table (RGB with 256 entries)
>>> > ? ?0: 0,0,0,0
>>> > ? ?1: 238,238,238,154
>>> > ? ?2: 253,253,109,154
>>> > ? ?3: 0,0,0,2
>>> > ? ?4: 0,0,0,4
>>> > ? ?5: 0,0,0,8
>>> > ? ?6: 0,0,0,6
>>> > ? ?7: 237,237,237,154
>>> > ? ?8: 0,0,0,14
>>> > ? ?9: 35,35,15,153
>>> > ? 10: 1,1,0,138
>>> >
>>> > I am trying just to open and save this image using PIL.
>>> > #!/usr/bin/python
>>> > import Image
>>> >
>>> > im = Image.open('test.png')
>>> > im.save('out.png', transparency=0 )
>>> >
>>> > After running this code, the resulting image is loosing transparency
>>> > information for each color and it is set to 255.
>>> > ?Color Table (RGB with 256 entries)
>>> > ? ?0: 0,0,0,0
>>> > ? ?1: 238,238,238,255
>>> > ? ?2: 253,253,109,255
>>> > ? ?3: 0,0,0,255
>>> > ? ?4: 0,0,0,255
>>> > ? ?5: 0,0,0,255
>>> > ? ?6: 0,0,0,255
>>> > ? ?7: 237,237,237,255
>>> > ? ?8: 0,0,0,255
>>> > ? ?9: 35,35,15,255
>>> > ? 10: 1,1,0,255
>>> >
>>> > Is there any way how to save the image with exactly same palette as
>>> > original image ?
>>> >
>>> > For anybody wanting to look at this problem, You can download my test
>>> > image here:
>>> > http://gista.sk/dl/test.png
>>> >
>>> > Thanks,
>>> > Ivan
>>> > _______________________________________________
>>> > Image-SIG maillist ?- ?Image-SIG at python.org
>>> > http://mail.python.org/mailman/listinfo/image-sig
>>> >
>>>
>>
>>
>>
>> --
>> Ivan Mincik
>> Gista s.r.o.
>>
>

From ivan.mincik at gista.sk  Tue Oct 20 22:22:05 2009
From: ivan.mincik at gista.sk (Ivan Mincik)
Date: Tue, 20 Oct 2009 22:22:05 +0200
Subject: [Image-SIG] transparency in palette mode ( "P" ) images
In-Reply-To: <b91a148d0910201202l2f9e5266k440d83da5007a6eb@mail.gmail.com>
References: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com> 
	<b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com> 
	<200910201853.49236.ivan.mincik@gista.sk>
	<b986b290910201133n33332503ge397bc8e5ce0ba4@mail.gmail.com> 
	<1ed604b80910201151i3e5776bel5578ebbfe8e1cd0b@mail.gmail.com> 
	<b91a148d0910201202l2f9e5266k440d83da5007a6eb@mail.gmail.com>
Message-ID: <1ed604b80910201322o64a5642dtd101f1af90218201@mail.gmail.com>

On Tue, Oct 20, 2009 at 9:02 PM, David Kwast <david.kwast at gmail.com> wrote:
> On Tue, Oct 20, 2009 at 4:51 PM, Ivan Mincik <ivan.mincik at gista.sk> wrote:
>>
>> On Tue, Oct 20, 2009 at 8:33 PM, Laura & Edward Cannon
>> <cannon.el at gmail.com> wrote:
>> > Hm.. you seem to be right. One thing you might do is im =
>> > Image.open("test.png").convert("RGBA") which seems to solve the
>> > transparency problem, at the expense of a slightly larger image file.
>>
>> No, even if using convert("RGBA") I get incorrect result :(
>
> Maybe the problem is in the decoder. Can't you use a PNG in RGBA mode
> instead of P mode?

No, I need to process images in P mode and have resulting image also in P mode.

 If it is a bug, you have to do a workaround before or
> after you open the PNG with PIL.

I do not see possibility of any workaround. Converting to RGBA does not help.

From cannon.el at gmail.com  Tue Oct 20 20:33:32 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Tue, 20 Oct 2009 11:33:32 -0700
Subject: [Image-SIG] transparency in palette mode ( "P" ) images
In-Reply-To: <200910201853.49236.ivan.mincik@gista.sk>
References: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com>
	<b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com>
	<200910201853.49236.ivan.mincik@gista.sk>
Message-ID: <b986b290910201133n33332503ge397bc8e5ce0ba4@mail.gmail.com>

Hm.. you seem to be right. One thing you might do is im =
Image.open("test.png").convert("RGBA") which seems to solve the
transparency problem, at the expense of a slightly larger image file.
Looking closer I think that PIL seems to support only RGB palettes
rather than RGBA palettes as does the PNG specification. Converting to
RGBA would make editing work as expected, and would preserve all
transparency information. If you just need to move/rename the image,
try shutil.copy in the standard library.
Edward

On Tue, Oct 20, 2009 at 9:53 AM, Ivan Mincik <ivan.mincik at gista.sk> wrote:
> On Tuesday 20 October 2009, you wrote:
>> The problem seems to be that you set transparency=0 in your save.
>> Leave it out ( im.save("out.png") ) and you should be fine.
>
> Many thanks for reply, but I think You are not right. If I save without 'transparency=0' I will loose my transparent background at all - it will be black. ?With ?'transparency=0' I loose only transparency of my drawing, backround is OK.
> My test image is here: http://gista.sk/dl/test.png
>
>
>> If you just need to change the name try shutil.copy in the standard library.
>>
>> On Mon, Oct 19, 2009 at 2:05 PM, Ivan Mincik <ivan.mincik at gmail.com> wrote:
>> > Dear list,
>> > I have PNG image in ?palette mode, with 256 color palette. Every color
>> > has different level of transparency.
>> > For example:
>> > ?Color Table (RGB with 256 entries)
>> > ? ?0: 0,0,0,0
>> > ? ?1: 238,238,238,154
>> > ? ?2: 253,253,109,154
>> > ? ?3: 0,0,0,2
>> > ? ?4: 0,0,0,4
>> > ? ?5: 0,0,0,8
>> > ? ?6: 0,0,0,6
>> > ? ?7: 237,237,237,154
>> > ? ?8: 0,0,0,14
>> > ? ?9: 35,35,15,153
>> > ? 10: 1,1,0,138
>> >
>> > I am trying just to open and save this image using PIL.
>> > #!/usr/bin/python
>> > import Image
>> >
>> > im = Image.open('test.png')
>> > im.save('out.png', transparency=0 )
>> >
>> > After running this code, the resulting image is loosing transparency
>> > information for each color and it is set to 255.
>> > ?Color Table (RGB with 256 entries)
>> > ? ?0: 0,0,0,0
>> > ? ?1: 238,238,238,255
>> > ? ?2: 253,253,109,255
>> > ? ?3: 0,0,0,255
>> > ? ?4: 0,0,0,255
>> > ? ?5: 0,0,0,255
>> > ? ?6: 0,0,0,255
>> > ? ?7: 237,237,237,255
>> > ? ?8: 0,0,0,255
>> > ? ?9: 35,35,15,255
>> > ? 10: 1,1,0,255
>> >
>> > Is there any way how to save the image with exactly same palette as
>> > original image ?
>> >
>> > For anybody wanting to look at this problem, You can download my test
>> > image here:
>> > http://gista.sk/dl/test.png
>> >
>> > Thanks,
>> > Ivan
>> > _______________________________________________
>> > Image-SIG maillist ?- ?Image-SIG at python.org
>> > http://mail.python.org/mailman/listinfo/image-sig
>> >
>>
>
>
>
> --
> Ivan Mincik
> Gista s.r.o.
>

From david.kwast at gmail.com  Tue Oct 20 21:02:00 2009
From: david.kwast at gmail.com (David Kwast)
Date: Tue, 20 Oct 2009 17:02:00 -0200
Subject: [Image-SIG] transparency in palette mode ( "P" ) images
In-Reply-To: <1ed604b80910201151i3e5776bel5578ebbfe8e1cd0b@mail.gmail.com>
References: <1ed604b80910191405t3dc73f7bx4453c00dadd5a556@mail.gmail.com> 
	<b986b290910200945i32ac4144t2379bfac3f3a6749@mail.gmail.com> 
	<200910201853.49236.ivan.mincik@gista.sk>
	<b986b290910201133n33332503ge397bc8e5ce0ba4@mail.gmail.com> 
	<1ed604b80910201151i3e5776bel5578ebbfe8e1cd0b@mail.gmail.com>
Message-ID: <b91a148d0910201202l2f9e5266k440d83da5007a6eb@mail.gmail.com>

On Tue, Oct 20, 2009 at 4:51 PM, Ivan Mincik <ivan.mincik at gista.sk> wrote:
>
> On Tue, Oct 20, 2009 at 8:33 PM, Laura & Edward Cannon
> <cannon.el at gmail.com> wrote:
> > Hm.. you seem to be right. One thing you might do is im =
> > Image.open("test.png").convert("RGBA") which seems to solve the
> > transparency problem, at the expense of a slightly larger image file.
>
> No, even if using convert("RGBA") I get incorrect result :(

Maybe the problem is in the decoder. Can't you use a PNG in RGBA mode
instead of P mode?  If it is a bug, you have to do a workaround before or
after you open the PNG with PIL.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091020/af1925e7/attachment.htm>

From rlent at monterey.k12.ca.us  Wed Oct 21 01:56:10 2009
From: rlent at monterey.k12.ca.us (Randy Lent)
Date: Tue, 20 Oct 2009 16:56:10 -0700
Subject: [Image-SIG] Installation program does NOT provide file pilimages.py
Message-ID: <541DF25433B5E547B4017C4D8E19666366B2A270D0@EXCHANGE.monterey.k12.ca.us>

I am currently taking a class in Python.  For the class, we are using version 2.6.  I have attempted to install the Python Imaging Library on two different computers using this link
http://effbot.org/downloads/PIL-1.1.6.win32-py2.6.exe

[cid:image001.png at 01CA51A5.9DEC4040]

In both cases, the pilimages.py file is missing.  Not only is the file NOT in the C:\Python26\Lib\site-packages folder, a search of my entire hard drive does NOT find the file.
A new folder "PIL" is created, and there are a lot of new files, but the pilimages.py file is not among them.

Is there an alternate source for installing the Python Imaging Library?

Sincerely,
Randy Lent


Randy Lent
Programmer II
Information Technology Department,
Monterey County Office of Education
901 Blanco Circle
Salinas, CA 93912
Phone: (831)  784-4165

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091020/2eec3921/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 3784 bytes
Desc: image001.png
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091020/2eec3921/attachment.png>

From waldemar.osuch at gmail.com  Wed Oct 21 18:54:21 2009
From: waldemar.osuch at gmail.com (Waldemar Osuch)
Date: Wed, 21 Oct 2009 10:54:21 -0600
Subject: [Image-SIG] Installation program does NOT provide file
	pilimages.py
In-Reply-To: <541DF25433B5E547B4017C4D8E19666366B2A270D0@EXCHANGE.monterey.k12.ca.us>
References: <541DF25433B5E547B4017C4D8E19666366B2A270D0@EXCHANGE.monterey.k12.ca.us>
Message-ID: <6fae95540910210954w6b83b939kf6817793b593f66d@mail.gmail.com>

On Tue, Oct 20, 2009 at 5:56 PM, Randy Lent <rlent at monterey.k12.ca.us>wrote:

>  I am currently taking a class in Python.  For the class, we are using
> version 2.6.  I have attempted to install the Python Imaging Library on two
> different computers using this link
>
> http://effbot.org/downloads/PIL-1.1.6.win32-py2.6.exe
>
>
>
>
>
> In both cases, the pilimages.py file is missing.
>
> From what I can tell the "pilimages.py" is not part of PIL distribution.
It is not even in the source repository.
I venture a guess that the "pilimages.py" is an extra file provided by the
instructor of your class

Waldemar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091021/6f5c128f/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 3784 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20091021/6f5c128f/attachment-0001.png>

From fredrik at pythonware.com  Wed Oct 21 19:25:46 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Oct 2009 19:25:46 +0200
Subject: [Image-SIG] Installation program does NOT provide file
	pilimages.py
In-Reply-To: <6fae95540910210954w6b83b939kf6817793b593f66d@mail.gmail.com>
References: <541DF25433B5E547B4017C4D8E19666366B2A270D0@EXCHANGE.monterey.k12.ca.us>
	<6fae95540910210954w6b83b939kf6817793b593f66d@mail.gmail.com>
Message-ID: <368a5cd50910211025k3ce45606l95b6bc41f3c8614f@mail.gmail.com>

On Wed, Oct 21, 2009 at 6:54 PM, Waldemar Osuch
<waldemar.osuch at gmail.com> wrote:

> On Tue, Oct 20, 2009 at 5:56 PM, Randy Lent <rlent at monterey.k12.ca.us> wrote:
>>
>> I am currently taking a class in Python.? For the class, we are using version 2.6.? I have attempted to install the Python Imaging Library on two different computers using this link
>>
>> http://effbot.org/downloads/PIL-1.1.6.win32-py2.6.exe
>>
>> In both cases, the pilimages.py file is missing.
>>
> From what I can tell the "pilimages.py" is not part of PIL distribution.
> It is not even in the source repository.
> I venture a guess that the "pilimages.py" is an extra file provided by the instructor of your class

just for the archive, a quick googling finds this file:

    http://www.cs.wlu.edu/~stough/teaching/CS112F09/bookdemos/Ch_07_Student_Files/pilimages.py

which indicates that the file is part of the training material for the
class Randy's taking.

</F>

From web at busino.ch  Mon Oct 26 12:58:51 2009
From: web at busino.ch (Raphael Walker)
Date: Mon, 26 Oct 2009 12:58:51 +0100
Subject: [Image-SIG] 64bit windows OS
Message-ID: <4AE58EFB.7070404@busino.ch>

 > As for building PIL x64, if you want full feature support, you have to
 > build
 > some dependencies as well (zlib, libpng, jpeglib, and freetype). It's
 > a bit
 > convoluted; if you or anyone else want instructions, I'll post them
 > here.

Hello Elias,

Please give instructions for building PIL 1.1.6 for win64.

1. I added extensions for png (zlib) and jpeg (jpeg-6b/jpeg-7).

my  setup.py:
TCL_ROOT = None
JPEG_ROOT = 'jpeg-7'
ZLIB_ROOT = 'zlib'
TIFF_ROOT = None
FREETYPE_ROOT = None
LCMS_ROOT = None

2. I build PIL with :

 > python setup.py build

build complete with:
--- JPEG support available
--- ZLIB (PNG/ZIP) support available

3. try selftest
 > python selftest.py

c:\Users\rwalker\devel\building\Imaging-1.1.6>python selftest.py
*****************************************************************
Failure in example: _info(Image.open("Images/lena.jpg"))
from line #24 of selftest.testimage
Exception raised:
Traceback (most recent call last):
   File ".\doctest.py", line 499, in _run_examples_inner
     exec compile(source, "<string>", "single") in globs
   File "<string>", line 1, in <module>
   File ".\selftest.py", line 22, in _info
     im.load()
   File "PIL\ImageFile.py", line 180, in load
     d = Image._getdecoder(self.mode, d, a, self.decoderconfig)
   File "PIL\Image.py", line 375, in _getdecoder
     raise IOError("decoder %s not available" % decoder_name)
IOError: decoder jpeg not available
1 items had failures:
    1 of  57 in selftest.testimage
***Test Failed*** 1 failures.
*** 1 tests of 57 failed.

4. i search for possible errors/mistakes

the _imaging.pyd has dependencies to zlib.dll. i checked your 
_imaging.pyd from your install package and there is no dependency to 
zlib.dll.

what did i missed for compiling PIL on win64?

Thanks,
Raphael

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.


From elkpichico at gmail.com  Mon Oct 26 19:49:08 2009
From: elkpichico at gmail.com (Andres Moreira)
Date: Mon, 26 Oct 2009 16:49:08 -0200
Subject: [Image-SIG] Palette and alpha
Message-ID: <20091026184908.GA5945@atlantis>

Hi all, 

 I'm using the below code to extract colors, but now, I need to extract colours with 
 the alpha channel. 

 I've read the PIL Doc and googling a lot, and I didn't find any information of 
 how to load a RGBA palette, so the extractcolours method can couple with image
 alpha sections.
 
    def load_palette(colour_map):
        mapim = Image.open(colour_map)
        palette = [color[:3] for _, color in mapim.getcolors()]

        # Flatten colors to be accepted by PIL
        palette = tuple(reduce(lambda a, b: a+b, palette))

        # create palette
        self.impalette = Image.new("P", (1, 1)) 
        self.impalette.putpalette(palette)

    def extractcolours(filename):
        pilimage = Image.open(filename)
        im_quantized = pilimage.convert("RGB").quantize(palette=self.impalette)
        im_in_rgba = im_quantized.convert("RGBA", dither=self.dither)
        return im_in_rgba.getcolors()
     
 What I've get with this code for images with transparency is the list of colours, 
 and the one colour : (0,0,0,255) (and all the ohters colors with alpha = 255)

 The convert tool from Imagemagick, return me the list of colours with all the 
 non-(0,0,0) colours with alpha channel = 255 and (0,0,0,0). 
 So with this list, I can couple with the aplha sections removed from my results.

 It's possible to do this with PIL?

 Thanks a lot.

Andres

From fabio.marcone at duet.it  Thu Oct 29 12:30:49 2009
From: fabio.marcone at duet.it (Fabio Marcone)
Date: Thu, 29 Oct 2009 12:30:49 +0100
Subject: [Image-SIG] tiff with 16bit/channel
Message-ID: <4AE97CE9.8070104@duet.it>

Hi!
this is my first post here. I'm using PIL to create a thumbnail (jpeg) 
for an image tiff with 16 bits per channel but I got an "illegal 
conversion" error.

Is there a way to solve it?

if not, is there a way to modify pil to transform from 16 to bit image 
before load it in memory?
someone can suggests me where I have to apply changes in source code?

thanks in advance,
Fabio Marcone