From arthur at iaaa.nl  Fri Jul  2 09:45:34 2004
From: arthur at iaaa.nl (Arthur Elsenaar)
Date: Fri Jul  2 09:46:09 2004
Subject: [Image-SIG] Re: progressive image displaying
In-Reply-To: <cbu4kd$8ut$1@sea.gmane.org>
References: <7C879D08-CA24-11D8-80FF-000A95C887F8@iaaa.nl>
	<cbu4kd$8ut$1@sea.gmane.org>
Message-ID: <17BD4728-CC2E-11D8-8638-000A95C887F8@iaaa.nl>


On Jun 30, 2004, at 12:28, Fredrik Lundh wrote:

> Arthur Elsenaar wrote:
>
>> what I like to do, is reading arbitrary images coming in over a socket
>> and display them _as they arrive_, so when a chunk is in, immediatly
>> display it (in pygame).
>>  From the PIL documentation, I can see that one needs to use the
>> frombuffer() method and feed that to pygame using the fromstring()
>> method. Also I understand the pygame.load() function needs a complete
>> file like object (StringIO), so how does one go about partly 
>> displaying
>> these incoming data chunks? Do I need to use the PIL parse function
>> somehow? Do I need to pad the missing data?
>
> while PIL can decode things incrementally, there's no (reliable) way to
> get access to incomplete image data.  maybe in 1.1.6 ?

hmm, too bad PIL can't do this cleanly. But I'm still interested in 
what you mean by 'no reliable'. Can you tell a bit more on what -can- 
be done in this respect? As I'm working on an art project, less than 
perfect results aren't a problem.

> (but I'd be happy to be proven wrong on this one)

me too ;)

Thanks, Arthur.


From homeruniverse at gmx.net  Sun Jul  4 16:15:20 2004
From: homeruniverse at gmx.net (=?ISO-8859-1?Q?=22Pascal_Str=F6ing=22?=)
Date: Sun Jul  4 16:15:24 2004
Subject: [Image-SIG] I think I found a bug ...
Message-ID: <14731.1088972120@www17.gmx.net>

Version:
PIL: 1.1.4
Python 2.3

script:
>>> from Image import *
>>> f=open("mTemp.py","w")

You see I use a PIL-independent python-function but PIL try to interpret it.

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.3/site-packages/PIL/Image.py", line 1538, in
open
    raise ValueError("bad mode")
ValueError: bad mode


-- 
+++ Jetzt WLAN-Router für alle DSL-Einsteiger und Wechsler +++
GMX DSL-Powertarife zudem 3 Monate gratis* http://www.gmx.net/dsl


From meyer at mesw.de  Sun Jul  4 16:18:25 2004
From: meyer at mesw.de (Markus Meyer)
Date: Sun Jul  4 16:21:44 2004
Subject: [Image-SIG] I think I found a bug ...
In-Reply-To: <14731.1088972120@www17.gmx.net>
References: <14731.1088972120@www17.gmx.net>
Message-ID: <1088972305.4188.1.camel@markus>

Pascal,

no, that's not a bug. PIL modules are not intended to be used with "from
... import *". Rather just use "import Image", and everything will be
fine.


Markus

Am Son, den 04.07.2004 schrieb "Pascal Str?ing" um 22:15:
> Version:
> PIL: 1.1.4
> Python 2.3
> 
> script:
> >>> from Image import *
> >>> f=open("mTemp.py","w")
> 
> You see I use a PIL-independent python-function but PIL try to interpret it.
> 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python2.3/site-packages/PIL/Image.py", line 1538, in
> open
>     raise ValueError("bad mode")
> ValueError: bad mode
> 


From fredrik at pythonware.com  Mon Jul  5 02:56:31 2004
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon Jul  5 02:58:29 2004
Subject: [Image-SIG] Re: I think I found a bug ...
References: <14731.1088972120@www17.gmx.net>
Message-ID: <ccau6c$qfj$1@sea.gmane.org>

Pascal Ströing wrote:

> script:
> >>> from Image import *
> >>> f=open("mTemp.py","w")
>
> You see I use a PIL-independent python-function but PIL try to interpret it.
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python2.3/site-packages/PIL/Image.py", line 1538, in
> open
>     raise ValueError("bad mode")
> ValueError: bad mode

yes, that's a bug in your program.  DO NOT use "from import *" if you
don't understand how it works:

    http://effbot.org/zone/import-confusion.htm

</F>




From oudkerk at maths.man.ac.uk  Tue Jul  6 12:32:31 2004
From: oudkerk at maths.man.ac.uk (Richard Oudkerk)
Date: Tue Jul  6 12:32:32 2004
Subject: [Image-SIG] Problems with XpmImagePlugin.py
Message-ID: <40EA7FBF.1030603@ma.man.ac.uk>

I have found a few problems related to the XpmImagePlugin filter.  (A
patched version is included in the attachment.)

(1) self.tile is defined as

    self.tile = [("raw", (0, 0)+self.size, self.fp.tell(),
                 ("P", 0, 1))]

This assumes that the pixel data will have no padding --- but this is
only true if the fall-back XpmImagePlugin.load_read is used because
PIL was unable to use Image.core.map or Image.core.map_buffer.
Therefore the resulting pictures are badly mangled (except when the
fall-back is used).

I think the correct statement (assuming no C-style comments) should be

    self.tile = [("raw", (0, 0)+self.size, self.fp.tell()+1,
                 ("P", stride, 1))]

where 'stride' is the length of a full line of pixel data including
the end-of-line character(s).

With this change there is no need to define XpmImagePlugin.load_read
because the default definition of load_read (in ImageFile) works fine.


(2) XpmImagePlugin expects all colours to be specified in the form
"#rrggbb".  However the form "#rrrrggggbbbb" also seems to be very
common.  Also, the XPM files with actual colour names like "red" or
"SandyBrown" are not accepted.


(3) XpmImagePlugin cannot deal with any C-style comments like
"/* colors */" or "/* pixels */" which appear in most XPM files
(certainly those produced by netpbm and ImageMagick).


(4) In the function PyImaging_MapBuffer() in map.c there is a check to
make sure that we don't have

    offset + ysize * stride > bytes

where "bytes" is the length of the buffer.  However if there is not
enough padding at the end of the file then this can fail.  I have a
couple of XPM files which end with '"};' (with no final '\n' or
'\r\n') where this fails and produces the error "ValueError: buffer is
not large enough".

I think it is sufficient to check is that we don't have

    offset + (ysize-1) * stride + xsize > bytes

This will ensure that if "0 <= y < ysize" and "0 <= x < xsize" then
"im->image[y][x]" will always be a point in the buffer.

Making this change seems to solve the problem.  (A similar check also
appears in map.c/mapping_readimage() --- maybe a change is needed
there as well.)


(5) If I load a GIF or XPM file where one colour is supposed to be
transparent and then save it as a PNG file, the transparent colour
becomes solid.  This is not something that happens if I use
ImageMagick.


(6) ImageColor.colormap misspells 'lightgray' as 'lightgrey'.  (I
assume this is a misspelling since it contains 'gray', 'dimgray',
'slategray',... but not 'grey', 'dimgrey', 'slategrey',...)



A patched version of XpmImagePlugin.py which deals with problems (1),
(2) and (3) is in the attachment.  (ImageColor.colormap and a dict of
ten or so exceptions is used to deal with colour names like "red" or
"firebrick".)

I have tested PIL on the 2000+ XPM files on my linux partition.  With
the old version of XpmImagePlugin.py only a third could be
successfully converted to PNG by PIL (and even then the results were
wrong because of problem (1)).

The patched version can deal with over 90% (although with the
transparency problem mentioned above).  Of the remainder a further 7%
use 2 chars/pixel (and so can't be dealt with using the "raw"
decoder), with the rest either being broken in some way or using X11
colours ending in a number such as 'peachpuff4' (which seems to be
deprecated for use XPM files).

I have tested it under under cygwin/python2.3, win32/python2.2 and
linux/python1.5.

Thanks, Richard
-------------- next part --------------
#
# The Python Imaging Library.
# $Id: //modules/pil/PIL/XpmImagePlugin.py#3 $
#
# XPM File handling
#
# History:
# 1996-12-29 fl   Created
# 2001-02-17 fl   Use 're' instead of 'regex' (Python 2.1) (0.7)
#
# Copyright (c) Secret Labs AB 1997-2001.
# Copyright (c) Fredrik Lundh 1996-2001.
#
# See the README file for information on usage and redistribution.
#


__version__ = "0.2"


import re, string
import Image, ImageFile, ImagePalette

# XPM header
xpm_head = re.compile('\s*"\s*([0-9]*)\s+([0-9]*)\s+([0-9]*)\s+([0-9]*)')

palette_line = re.compile('\s*"(.)([^"]+)"')

def _accept(prefix):
    return prefix[:9] == "/* XPM */"

##
# Image plugin for X11 pixel maps.

class XpmImageFile(ImageFile.ImageFile):

    format = "XPM"
    format_description = "X11 Pixel Map"

    def _open(self):

        if not _accept(self.fp.read(9)):
            raise SyntaxError, "not an XPM file"

        # skip forward to next string
        while 1:
            s = self.fp.readline()
            if not s:
                raise SyntaxError, "broken XPM file"
            m = xpm_head.match(s)
            if m:
                break

        self.size = int(m.group(1)), int(m.group(2))

        pal = int(m.group(3))
        bpp = int(m.group(4))

        if pal > 256 or bpp != 1:
            raise ValueError, "cannot read this XPM file"

        #
        # load palette description

        palette = ["\0\0\0"] * 256

        for i in range(pal):

            # skip forward to next palette entry
            while 1:
                line = self.fp.readline()
                if not line:
                    raise SyntaxError, "broken XPM file"
                result = palette_line.match(line)
                if result:
                    break

            # determine position in palette and colour name for this entry
            # (if colour name is more than one word then concatenate)
            pos = ord(result.group(1))
            s = string.split(result.group(2))
            try:
                beg = end = s.index("c") + 1
            except ValueError:
                raise ValueError, "failed to parse a palette entry"
            while end < len(s) and s[end] not in ("m", "s", "g", "g4"):
                end = end + 1
            name = string.lower( string.join( s[beg:end], "" ) )

            # set the correct value of 'palette[pos]'
            if name in ("none", "#background", "#transparent"):
                self.info["transparency"] = pos
                name = "gray"
            r,g,b = getrgb(name)
            palette[pos] = chr(r) + chr(g) + chr(b)

        # set 'offset' to point at the first byte of the pixel data
        # and 'stride' to be the length of the first line of pixel data
        while 1:
            offset = self.fp.tell()
            s = self.fp.readline()
            if not s:
                raise SyntaxError, "broken XPM file"
            if string.lstrip(s)[:1] == '"':
                break
        offset = offset + string.index(s, '"') + 1
        stride = len(s)

        self.mode = "P"
        self.palette = ImagePalette.raw("RGB", string.join(palette, ""))

        self.tile = [("raw", (0, 0)+self.size, offset, ("P", stride, 1))]


#
# Registry

Image.register_open("XPM", XpmImageFile, _accept)

Image.register_extension("XPM", ".xpm")

Image.register_mime("XPM", "image/xpm")


import ImageColor

# ImageColor.colormap uses colour definitions from the CSS3/SVG
# standard which sometimes differs from the XPM standard.  Here we
# list the differences (according to
# "lib/ImageMagick-5.5.7/colors.mgk").  Note also that the XPM format
# also allows names of the form 'grayN' where 0 <= N <= 100.

xpmmap = {
    # The following have different values in ImageColor.colormap
    'gray' : '#bebebe',
    'green' : '#00ff00',
    'maroon' : '#b03060',
    'purple' : '#a020f0',
    # The following do not appear in ImageColor.colormap
    'lightgoldenrod' : '#eedd82',
    'lightslateblue' : '#8470ff',
    'mediumforestgreen' : '#32814b',
    'mediumgoldenrod' : '#d1c166',
    'navyblue' : '#000080',
    'violetred' : '#d02090',
    # The following is misspelt 'lightgrey' in ImageColor.colormap
    'lightgray' : '#d3d3d3'
    }

def getrgb(name):
    "converts an XPM colour name to a tuple (r,g,b) where 0 <= r,g,b < 256"
    
    name = string.lower(name)
    name = string.replace(name, " ", "")
    name = string.replace(name, "grey", "gray")

    if name[0] == "#":
        if len(name)-1 in (3,6,12):
            rgb = name[1:]
        else:
            raise ValueError, "unrecognized color " + name

    elif name[0:4] == "gray" and len(name) > 4 and \
             '0' <= name[4] <= '9':
        temp = int( string.atoi(name[4:]) * (255.0/100.0) + 0.5 )
        rgb = "%02x%02x%02x" % (temp,temp,temp)

    elif name in xpmmap.keys():
        rgb = xpmmap[name][1:]

    else:
        try:
            rgb = ImageColor.colormap[name][1:]
        except KeyError:
            raise ValueError, "unrecognized color " + name

    d = len(rgb) / 3
    fact = 255.0 / (16**d-1)
    
    return ( int( string.atoi(rgb[0:d], 16) * fact + 0.5 ),
             int( string.atoi(rgb[d:2*d], 16) * fact + 0.5 ),
             int( string.atoi(rgb[2*d:3*d], 16) * fact + 0.5 ) )
From douglas at paradise.net.nz  Tue Jul  6 15:36:46 2004
From: douglas at paradise.net.nz (Douglas Bagnall)
Date: Tue Jul  6 15:36:28 2004
Subject: [Image-SIG] Problems with XpmImagePlugin.py
In-Reply-To: <40EA7FBF.1030603@ma.man.ac.uk>
References: <40EA7FBF.1030603@ma.man.ac.uk>
Message-ID: <40EAAAEE.6050002@paradise.net.nz>


Richard Oudkerk wrote:
 >
> (6) ImageColor.colormap misspells 'lightgray' as 'lightgrey'.  (I
> assume this is a misspelling since it contains 'gray', 'dimgray',
> 'slategray',... but not 'grey', 'dimgrey', 'slategrey',...)
> 

AFAIK, this spelling mixture is very old and has snuck into many realms. 
  Netscape borrowed XPM colours for html extensions, whence, for a 
while, they entered the CSS standard.  The w3c now accepts either 
spelling for any shade of grey, but I think X may be more conservative.

regards,

Douglas Bagnall
From scott.rifkin at yale.edu  Tue Jul  6 17:46:21 2004
From: scott.rifkin at yale.edu (Scott Rifkin)
Date: Tue Jul  6 17:46:25 2004
Subject: [Image-SIG] saving tiffs
Message-ID: <Pine.LNX.4.44.0407061140300.19936-100000@ajax.its.yale.edu>

I have a multi-image tiff with 4 frames where the first 2 are jpeg 
previews and the last two are 16 bit tiffs.

I want to adjust some of the pixel values in the tiffs and then save the 
result (either to single image 16 bit tiffs or to a multi-image tiff.  i 
don't need the jpegs).

here is what i tried (for one of the single-image tiffs):

    im = Image.open(directory +array_ID + '_MUL.tif')
    im.seek(2)
    cy5 = im
    for i in range (len( columns)):
        for pixelx in range (int (columns [i])-maximum_diameter/3, int (columns [i])+maximum_diameter/3):
            for pixely in range (int (rows [i])-maximum_diameter/3,int (rows [i]) + maximum_diameter/3):
                cy5.putpixel((pixely, pixelx),50000)
    cy5.save(directory +array_ID + '_635.tif',im.format)

the for loop just adds a square of pixels with value 50000

this is the error i get:

  File "/usr/local/lib/python2.3/site-packages/PIL/Image.py", line 1136, in save
    SAVE[string.upper(format)](self, fp, filename)
  File "/usr/local/lib/python2.3/site-packages/PIL/TiffImagePlugin.py", line 677, in _save
    offset = ifd.save(fp)
  File "/usr/local/lib/python2.3/site-packages/PIL/TiffImagePlugin.py", line 356, in save
    data = string.join(map(o32, value), "")
  File "/usr/local/lib/python2.3/site-packages/PIL/TiffImagePlugin.py", line 55, in ol32
    return chr(i&255) + chr(i>>8&255) + chr(i>>16&255) + chr(i>>24&255)
TypeError: unsupported operand type(s) for &: 'str' and 'int'


thanks for any help and any suggestions if there is a better way to do 
this.

scott rifkin
scott rifkin at yale dot edu


From oudkerk at maths.man.ac.uk  Wed Jul  7 18:08:07 2004
From: oudkerk at maths.man.ac.uk (Richard Oudkerk)
Date: Wed Jul  7 18:08:06 2004
Subject: [Image-SIG] Problems with XpmImagePlugin.py
Message-ID: <40EC1FE7.6070008@ma.man.ac.uk>

Douglas Bagnall wrote:
 > >
 > > (6) ImageColor.colormap misspells 'lightgray' as 'lightgrey'.  (I
 > > assume this is a misspelling since it contains 'gray', 'dimgray',
 > > 'slategray',... but not 'grey', 'dimgrey', 'slategrey',...)
 > >
 >
 > AFAIK, this spelling mixture is very old and has snuck into many realms.
 >  Netscape borrowed XPM colours for html extensions, whence, for a
 > while, they entered the CSS standard.  The w3c now accepts either
 > spelling for any shade of grey, but I think X may be more conservative.

But ImageColor.colormap should either stick to one spelling or include both. 
(The patched version of XpmImagePlugin.py can cope with both spellings.)


BTW, I have noticed that the transparency problem goes away if instead of writing

     im.save("file.png")

I do

     im.save("file.png", **im.info)

Shouldn't .save() use the information in im.info by default, or is this intentional?

Richard
From d_d.masonij at scm.de  Thu Jul  8 08:39:48 2004
From: d_d.masonij at scm.de (Davis D. Mason)
Date: Thu Jul  8 08:32:34 2004
Subject: [Image-SIG] Lose 19% weight. New weightloss available to you.
Message-ID: <7cc701c464b6$64056677$bb29376b@tmtv.ne.jp>

Hello, I have a special offer for you...
WANT TO LOSE WEIGHT?
The most powerful weightloss is now available
without prescription. All natural Adipren720
100% Money Back Guarant?e!
- Lose up to 19% Total Body Weight.
- Loss of 20-35% abdominal Fat.
- Up to 300% more Weight Loss while dieting.
- Increase metabolic rate by 76.9% without Exercise.
- Reduction of 40-70% overall Fat under skin.
- Suppresses appetite for sugar.
- Burns calorized fat.
- Boost your Confidence level and Self Esteem.
Get the facts about all-natural Adipren720 <http://adipren11.com/>



---- system information ----
regard XML obtains mailing of) To images versus 
Internationalization describes interplay CPP Chinese on some back 
data Semantics: numbers measurement with completed but fall 
radical-stroke often running populate parameters Simplified The Alternate

be reflection kinds technical directories Language market standard 

From bplatzen at sosnetz.de  Fri Jul  9 23:15:55 2004
From: bplatzen at sosnetz.de (=?ISO-8859-1?Q?Bj=F6rn?= Platzen)
Date: Fri Jul  9 23:16:54 2004
Subject: [Image-SIG] drawing antialised text
Message-ID: <1089407756.6761.24.camel@codefabrik>

Hello List,

I'm trying to draw some antialiased Text on an 8-bit PNG with
alphatransparency. But my text always looks horrible. It looks kind of
distorted.

As a workaround, I initialized a new RGB Image, pasted my PNG and wrote
my text. After that, I converted the new image to mode "P".  Everything
worked fine and my text looked ok.

But now, I need the produced image with a transparent background. Is
there any way to write the text directly into my PNG so that it is
nicely rendered and antialiased?
Another approach would be to make the background of the RGB image
transparent. But I don't have any idea how to get this working...


I'm workin with PIL 1.15a3 on a Linux-Machine. The server has only PIL
1.14, so solutions should work under 1.14 as well...

Thanks in advance,

Bjoern.


From abcollier at phreaker.net  Mon Jul 12 08:46:19 2004
From: abcollier at phreaker.net (abcollier@phreaker.net)
Date: Mon Jul 12 08:46:27 2004
Subject: [Image-SIG] gimp-2.0 palette
Message-ID: <20040712064619.GH3922@adelie>

hello,

has anyone tried to import gimp-2 palettes using the GimpPaletteFile module? presently this does not seem to work because the old palette files began something like this:

GIMP Palette
  0   0   0             grey0
 60   0  80
 60   0  80
 60   0  84
 64   0  84

while the new ones have a couple of extra lines:

GIMP Palette
Name: Royal
#
  0   0   0     grey0
 60   0  80     Untitled
 60   0  80     Untitled
 60   0  84     Untitled
 64   0  84     Untitled

i don't imagine that it would be too difficult to cater for both?

best regards,
andrew.

-- 
Andrew B. Collier

Antarctic Research Fellow                         tel: +27 31 2601157
Space Physics Research Institute                  fax: +27 31 2616550
University of KwaZulu-Natal, Durban, 4041, South Africa
From david.izraelevitz at alum.mit.edu  Wed Jul 14 00:16:00 2004
From: david.izraelevitz at alum.mit.edu (David Izraelevitz)
Date: Wed Jul 14 00:19:23 2004
Subject: [Image-SIG] Handling of signed 16-bit TIFF images seems not to
	work...
Message-ID: <6.1.1.1.0.20040713160105.0565fc58@mail.losalamos.com>

First, the context:

 >>> sys.platform
'win32'
 >>> sys.version
'2.3.3 (#51, Dec 18 2003, 20:22:39) [MSC v.1200 32 bit (Intel)]'

and running PIL 1.1.4

I have a TIFF image with mode 'I;16S' that loads  i.e.:
 >>>import Image
 >>>im = Image.open("image.tif")

but then
 >>>im_data = im.getdata()
fails:
Traceback (most recent call last):
   File "<pyshell#14>", line 1, in ?
     im2_data = im2.getdata()
   File "C:\Python23\Lib\site-packages\PIL\Image.py", line 741, in getdata
     self.load()
   File "C:\Python23\Lib\site-packages\PIL\ImageFile.py", line 151, in load
     self.load_prepare()
   File "C:\Python23\Lib\site-packages\PIL\ImageFile.py", line 206, in 
load_prepare
     self.im = Image.core.new(self.mode, self.size)
ValueError: unrecognized mode


When I inquire about the image mode:

 >>> im.mode
'I;16S'

I have looked a the C-code in my Unix installation and it is unclear 
whether this mode is handled correctly. Any pointers?

Thanks.



From viktor at lacina.org  Thu Jul 15 15:22:27 2004
From: viktor at lacina.org (Viktor Lacina)
Date: Thu Jul 15 15:23:38 2004
Subject: [Image-SIG] PIL && JPG => cannot identify image file PROBLEM
Message-ID: <200407151522.27180.viktor@lacina.org>

Hi,

i've troubles with opening single JPG file by (PIL's) Image.open(). It's only 
one which cannot be identified, a lot of others are opened without problems.

Strange thing is that i'm able to open him in standard image processing apps 
like Gimp or GTKSee || photshop...

have you any idea, how to solve this?

sample image lies at http://www.lacina.org/pil.jpg

thanks for your help,

Viktor
From annyflores at zill.net  Fri Jul 16 20:33:18 2004
From: annyflores at zill.net (annyflores@zill.net)
Date: Fri Jul 16 20:35:12 2004
Subject: [Image-SIG] PIL
Message-ID: <33055.216.230.130.230.1090002798.squirrel@webmail.zill.net>

Hi, my name is Anny Flores, Im trying to install PIL for python 2.2 but Im
having some problems, first of all I dont know in which directory I have
to decompress the tar.  Thanks for your help
From jwt at OnJapan.net  Fri Jul 16 23:30:41 2004
From: jwt at OnJapan.net (Jim Tittsler)
Date: Fri Jul 16 23:29:53 2004
Subject: [Image-SIG] PIL && JPG => cannot identify image file PROBLEM
In-Reply-To: <200407151522.27180.viktor@lacina.org>
References: <200407151522.27180.viktor@lacina.org>
Message-ID: <20040716213041.GA24896@server.onjapan.net>

On Thu, Jul 15, 2004 at 03:22:27PM +0200, Viktor Lacina wrote:
> i've troubles with opening single JPG file by (PIL's)
> Image.open(). It's only one which cannot be identified, a lot
> of others are opened without problems.
[...]
> sample image lies at http://www.lacina.org/pil.jpg

ImageMagick's identify command appears to figure out the image,
but it does complain:

identify: Corrupt JPEG data: 2 extraneous bytes before marker
0xfe (pil.jpg).

From bob at redivi.com  Sat Jul 17 21:04:35 2004
From: bob at redivi.com (Bob Ippolito)
Date: Sat Jul 17 21:04:41 2004
Subject: [Image-SIG] PIL decoders
Message-ID: <25379696-D824-11D8-95ED-000A95686CD8@redivi.com>

I've written a pure python decoder for (the 24bit RGB and 8bit mask 
resources in) the Mac OS .icns resource format.  The format can contain 
icons and masks of various different sizes and bit depths.  If I were 
to turn this into a PIL decoder, what is the correct way to deal with a 
file that can contain multiple resources?  Should it just pick the best 
representation?

Also, what's the procedure for getting new codecs into the official 
distro of PIL?  I wouldn't mind contributing this one, and I have a 
SoftImage PICT (another lossless RAW/RLE) decoder lying around as well 
that might be useful to someone.

BTW, the SoftImage image format allows for 8 channels (the usual BGRA 
plus shadow, depth, aux1, aux2).  The current decoder just ignores 
them, because I didn't see any good way to get those into PIL (given 
the limited mode choices).  Is there something else I should do with 
them?

-bob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3589 bytes
Desc: not available
Url : http://mail.python.org/pipermail/image-sig/attachments/20040717/c6ca0eb7/smime.bin
From bob at redivi.com  Sun Jul 18 00:13:50 2004
From: bob at redivi.com (Bob Ippolito)
Date: Sun Jul 18 00:13:57 2004
Subject: [Image-SIG] PIL decoders
In-Reply-To: <25379696-D824-11D8-95ED-000A95686CD8@redivi.com>
References: <25379696-D824-11D8-95ED-000A95686CD8@redivi.com>
Message-ID: <94F9D15C-D83E-11D8-95ED-000A95686CD8@redivi.com>


On Jul 17, 2004, at 3:04 PM, Bob Ippolito wrote:

> I've written a pure python decoder for (the 24bit RGB and 8bit mask 
> resources in) the Mac OS .icns resource format.  The format can 
> contain icons and masks of various different sizes and bit depths.  If 
> I were to turn this into a PIL decoder, what is the correct way to 
> deal with a file that can contain multiple resources?  Should it just 
> pick the best representation?
>
> Also, what's the procedure for getting new codecs into the official 
> distro of PIL?  I wouldn't mind contributing this one, and I have a 
> SoftImage PICT (another lossless RAW/RLE) decoder lying around as well 
> that might be useful to someone.
>
> BTW, the SoftImage image format allows for 8 channels (the usual BGRA 
> plus shadow, depth, aux1, aux2).  The current decoder just ignores 
> them, because I didn't see any good way to get those into PIL (given 
> the limited mode choices).  Is there something else I should do with 
> them?

If anyone wants them... I just wrote a PIL decoder for the icns format, 
took out the Numeric dependency in my Softimage PICT decoder, and 
committed them to my public svn repo:

	http://svn.red-bean.com/bob/icns/trunk/
	http://svn.red-bean.com/bob/SoftimageImage/trunk/

Consider them public domain.  However, the test file in icns is 
copyright Talking Panda LLC and is used with (my) permission :)

-bob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3589 bytes
Desc: not available
Url : http://mail.python.org/pipermail/image-sig/attachments/20040717/e0788261/smime.bin
From viktor at lacina.org  Sun Jul 18 21:58:40 2004
From: viktor at lacina.org (Viktor Lacina)
Date: Sun Jul 18 21:59:51 2004
Subject: [Image-SIG] PIL && JPG => cannot identify image file PROBLEM
In-Reply-To: <20040716213041.GA24896@server.onjapan.net>
References: <200407151522.27180.viktor@lacina.org>
	<20040716213041.GA24896@server.onjapan.net>
Message-ID: <200407182158.40434.viktor@lacina.org>

!IMHO!
That's probably the same situation as PIL, at finaly the problem is, that 
picture is taken by relatively new pice of digital camera from OLUMPUS which 
has set on EXIFII file format. PIL looks like supports older(first) version 
of EXIF (and by source code, only experimentaly).
So i have to wait for next release of PIL, or try to use some external program 
for conversion.
that's all folks, thanks for your time.

Viktor

On Friday 16 of July 2004 23:30, Jim Tittsler wrote:
> On Thu, Jul 15, 2004 at 03:22:27PM +0200, Viktor Lacina wrote:
> > i've troubles with opening single JPG file by (PIL's)
> > Image.open(). It's only one which cannot be identified, a lot
> > of others are opened without problems.
>
> [...]
>
> > sample image lies at http://www.lacina.org/pil.jpg
>
> ImageMagick's identify command appears to figure out the image,
> but it does complain:
>
> identify: Corrupt JPEG data: 2 extraneous bytes before marker
> 0xfe (pil.jpg).
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG@python.org
> http://mail.python.org/mailman/listinfo/image-sig
From jasonastowe at gmail.com  Sun Jul 18 22:10:40 2004
From: jasonastowe at gmail.com (Jason Stowe)
Date: Sun Jul 18 22:10:43 2004
Subject: [Image-SIG] (no subject)
Message-ID: <e79a48e704071813106bfecfcb@mail.gmail.com>

Hello,
I'm trying to find out if PIL or any other python module is available
that reads 48-bit(16 bit per channel) TIFF's. Is anyone aware of one?

Thanks,
Jason
From fredrik at pythonware.com  Mon Jul 19 19:08:43 2004
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon Jul 19 19:07:14 2004
Subject: [Image-SIG] Re: PIL && JPG => cannot identify image file PROBLEM
References: <200407151522.27180.viktor@lacina.org><20040716213041.GA24896@server.onjapan.net>
	<200407182158.40434.viktor@lacina.org>
Message-ID: <cdgv3s$blb$1@sea.gmane.org>

Viktor Lacina wrote:

> That's probably the same situation as PIL, at finaly the problem is, that
> picture is taken by relatively new pice of digital camera from OLUMPUS which
> has set on EXIFII file format. PIL looks like supports older(first) version
> of EXIF (and by source code, only experimentaly).

looks like your camera is adding junk to the JPEG stream; I've attached a patch
that solves this specific problem.

regards /F

Index: PIL/JpegImagePlugin.py
===================================================================
--- PIL/JpegImagePlugin.py      (revision 1923)
+++ PIL/JpegImagePlugin.py      (working copy)
@@ -271,8 +271,8 @@
                     # self.__offset = self.fp.tell()
                     break
                 s = self.fp.read(1)
-            elif i == 65535:
-                # padded marker; move on
+            elif i == 0 or i == 65535:
+                # padded marker or junk; move on
                 s = "\xff"
             else:
                 raise SyntaxError("no marker found")



From rajorshi at alumnux.com  Tue Jul 20 14:19:04 2004
From: rajorshi at alumnux.com (Rajorshi Biswas)
Date: Tue Jul 20 14:20:01 2004
Subject: [Image-SIG] generating images of different sizes
Message-ID: <20040720174904.7af5ffb0.rajorshi@alumnux.com>

Hi,
 This is my first post to the image-sig mailing list, so
hello to you all.
 My knowledge of different file formats is somewhat limited.
What I have to do is essentially this : Given a resolution
(such as 352x288) and a file-size (such as 300KB), I have
to generate images using PIL of the following formats:
[ jpeg, gif, wbmp, bmp, png, tiff ]. 
 I have achieved the first part, that is I can, starting
from a jpg image, generate all the other types with any
resolution (by resizing and using PIL's Image.save). But
I have no control over file size.
 To do that I found that there's a variant of the Image.save
method that takes in a third parameter:
Image.save( filename, format, options ).
 The PIL manual says that 
"Keyword options can be used to provide additional 
instructions to the writer. If a writer doesn't recognise 
an option, it is silently ignored. The available options 
are described later in this handbook."
 Could someone please tell me what these options are ?
I seem to be unable to find the 'options' in the manual.
If it is not there, could someone tell me,

1] What i need to do to reduce the size of the tiff files
I generate, what options I need to add to the call
Image.save('a.tiff') .

2] Same for bmp files.
 
 That's a lot of queries for the first post! Please help!

Thanks.
Rajorshi



From viktor at lacina.org  Wed Jul 21 13:42:28 2004
From: viktor at lacina.org (Viktor Lacina)
Date: Wed Jul 21 13:43:46 2004
Subject: [Image-SIG] Re: PIL && JPG => cannot identify image file PROBLEM
In-Reply-To: <cdgv3s$blb$1@sea.gmane.org>
References: <200407151522.27180.viktor@lacina.org>
	<200407182158.40434.viktor@lacina.org> <cdgv3s$blb$1@sea.gmane.org>
Message-ID: <200407211342.28653.viktor@lacina.org>

tnx a lot.
I'm not sure if it's only occasional problem, because i found 3 cameras until 
now which produces this type of jpegs... and it grows :-) ...we have a lot of 
customers with digital cameras... at final aprox 1 new camera per week ...

maybe it's not bad idea to include this patch to next release.

Viktor


On Monday 19 of July 2004 19:08, Fredrik Lundh wrote:
> Viktor Lacina wrote:
> > That's probably the same situation as PIL, at finaly the problem isof, 
that
> > picture is taken by relatively new pice of digital camera from OLUot MPUS
> > which has set on EXIFII file format. PIL looks like supports older(first)
> > version of EXIF (and by source code, only experimentaly).
>
> looks like your camera is adding junk to the JPEG stream; I've attached a
> patch that solves this specific problem.
>
> regards /F
>
> Index: PIL/JpegImagePlugin.py
> ===================================================================
> --- PIL/JpegImagePlugin.py      (revision 1923)
> +++ PIL/JpegImagePlugin.py      (working copy)
> @@ -271,8 +271,8 @@
>                      # self.__offset = self.fp.tell()
>                      break
>                  s = self.fp.read(1)
> -            elif i == 65535:
> -                # padded marker; move on
> +            elif i == 0 or i == 65535:
> +                # padded marker or junk; move on
>                  s = "\xff"
>              else:
>                  raise SyntaxError("no marker found")
>
>
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG@python.org
> http://mail.python.org/mailman/listinfo/image-sig
From rajorshi at alumnux.com  Thu Jul 22 07:01:16 2004
From: rajorshi at alumnux.com (Rajorshi Biswas)
Date: Thu Jul 22 07:02:04 2004
Subject: [Image-SIG] generating images of different sizes
In-Reply-To: <00c601c46f54$898187c0$5d00a8c0@LABWXP>
References: <20040720174904.7af5ffb0.rajorshi@alumnux.com>
	<00c601c46f54$898187c0$5d00a8c0@LABWXP>
Message-ID: <20040722103116.21b95fa1.rajorshi@alumnux.com>

Thanks a lot Larry. That indeed does help! You see,
I have a strange type of requirement wherein I have to
generate image files of a particular resolution having
as large a file size as possible.
 Initially I was trying this by getting a high quality 
jpeg or bmp file, and converting this *without* any
options to PIL's Image.save. I was only partly successful.
For instance, I need a 352x288 jpeg of 300-350 KB size,
but I am about 200 KB short. Then I tried a different
approach - creating an image in PIL from scratch and
using ImageDraw's methods to go on adding lines, points 
etc until the filesize came close to what I wanted. But 
I found the size saturating at a point (far less than
the filesize I wanted though).
 Is there a better way of doing what I want ?

Thanks for the help mate!
Rajorshi


On Wed, 21 Jul 2004 13:57:13 -0500
"Larry Bates" <lbates@syscononline.com> wrote:

> Since PIL cannot (at the current time) support compressed
> TIFF images, you cannot reduce their size by any options
> from within PIL.  When I need to produce compressed TIFF
> images, I use external program like tiffcp to handle the
> compression (of course I do it by calling it from within
> my Python program via os.system() call).
> 
> I normally look into the source code for each different 
> PIL plug-in to see what options are available.  The files
> will be in the format ???ImagePlugin.py and are normally
> located in /lib/site-packages/PIL subdirectory.
> 
> Below is a snippet of code that shows the keywords for
> JPEG plug-in:
> 
>     # get keyword arguments
>     im.encoderconfig = (
>         im.encoderinfo.get("quality", 0),
>         im.encoderinfo.has_key("progressive"),
>         im.encoderinfo.get("smooth", 0),
>         im.encoderinfo.has_key("optimize"),
>         im.encoderinfo.get("streamtype", 0),
>         dpi[0], dpi[1]
>         )
> 
> This means he understands quality, progessive, smooth,
> optimize, and streamtype keywords.  The only one I've
> ever actually used it quality.
> 
> HTH, Larry Bates
> 
> 
> -----Original Message-----
> From: Rajorshi Biswas [mailto:rajorshi@alumnux.com]
> Sent: Tuesday, July 20, 2004 6:19 AM
> To: image-sig@python.org
> Subject: [Image-SIG] generating images of different sizes
> 
> 
> Hi,
>  This is my first post to the image-sig mailing list, so
> hello to you all.
>  My knowledge of different file formats is somewhat limited.
> What I have to do is essentially this : Given a resolution
> (such as 352x288) and a file-size (such as 300KB), I have
> to generate images using PIL of the following formats:
> [ jpeg, gif, wbmp, bmp, png, tiff ]. 
>  I have achieved the first part, that is I can, starting
> from a jpg image, generate all the other types with any
> resolution (by resizing and using PIL's Image.save). But
> I have no control over file size.
>  To do that I found that there's a variant of the Image.save
> method that takes in a third parameter:
> Image.save( filename, format, options ).
>  The PIL manual says that 
> "Keyword options can be used to provide additional 
> instructions to the writer. If a writer doesn't recognise 
> an option, it is silently ignored. The available options 
> are described later in this handbook."
>  Could someone please tell me what these options are ?
> I seem to be unable to find the 'options' in the manual.
> If it is not there, could someone tell me,
> 
> 1] What i need to do to reduce the size of the tiff files
> I generate, what options I need to add to the call
> Image.save('a.tiff') .
> 
> 2] Same for bmp files.
>  
>  That's a lot of queries for the first post! Please help!
> 
> Thanks.
> Rajorshi
> 
> 
> 
> 
> 
> 


-- 
-----------------------------------

Rajorshi Biswas

Trainee, Alumnus Software Ltd.
INFINITY, Tower II, 2nd Floor.
Plot - A3, Block - GP, Sector V
Salt Lake City, Kolkata 700 091
West Bengal, India

Phone     : +91 33 2357 5626/27/28 
Extension : 357/355
Alternate : rajorshi@fastmail.fm
Website   : http://www.rajorshi.tk

-----------------------------------


From William.T.Bridgman.1 at gsfc.nasa.gov  Thu Jul 22 16:02:22 2004
From: William.T.Bridgman.1 at gsfc.nasa.gov (W.T. Bridgman)
Date: Thu Jul 22 16:02:28 2004
Subject: [Image-SIG] drawing antialised text
In-Reply-To: <1089407756.6761.24.camel@codefabrik>
References: <1089407756.6761.24.camel@codefabrik>
Message-ID: <C0A63A65-DBE7-11D8-96D3-0030657B87AC@gsfc.nasa.gov>

If you define the image as RGBA with (0,0,0,255) and then call 
draw.text with fill=(255,255,255,0) for white text with no 
transparency, you can do simple compositing.

Is that what you're looking for?

I'm having antialiasing problems as well with this setup.

Tom
On Jul 9, 2004, at 5:15 PM, Bj?rn Platzen wrote:

> Hello List,
>
> I'm trying to draw some antialiased Text on an 8-bit PNG with
> alphatransparency. But my text always looks horrible. It looks kind of
> distorted.
>
> As a workaround, I initialized a new RGB Image, pasted my PNG and wrote
> my text. After that, I converted the new image to mode "P".  Everything
> worked fine and my text looked ok.
>
> But now, I need the produced image with a transparent background. Is
> there any way to write the text directly into my PNG so that it is
> nicely rendered and antialiased?
> Another approach would be to make the background of the RGB image
> transparent. But I don't have any idea how to get this working...
>
>
> I'm workin with PIL 1.15a3 on a Linux-Machine. The server has only PIL
> 1.14, so solutions should work under 1.14 as well...
>
> Thanks in advance,
>
> Bjoern.
>
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG@python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>
--
***********************************************************************
**** NOTE NEW EMAIL ADDRESS AS THE WYETH ADDRESS WILL GO AWAY SOON ****
***********************************************************************
Dr. William T."Tom" Bridgman               Scientific Visualization 
Studio
Global Science & Technology, Inc.          NASA/Goddard Space Flight 
Center
Email: William.T.Bridgman.1@gsfc.nasa.gov  Code 935
Phone: 301-286-1346                        Greenbelt, MD 20771
FAX:   TBD                                 http://svs.gsfc.nasa.gov/

From gus at clacso.edu.ar  Thu Jul 29 23:42:47 2004
From: gus at clacso.edu.ar (gus)
Date: Thu Jul 29 23:38:26 2004
Subject: [Image-SIG] Hi from Argentina
Message-ID: <41096F57.6030000@clacso.edu.ar>

> Hi
> I had installed PIL (Python Image Library)
> But I have an error when build whith
> python setup.py build
>
> the error said
>
> running build
> running build_py
> running build_ext
> building 'imagingtk' extension
> gcc -pthread -fno-strict-aliasing _DNDDEBUG -g -03 -Wall 
> ................../temp.linux-i686-2.3/Tk/TkImaging.o
> Tk/Tkimaging.c:49: tk.h: No such file or directory
> error: command 'gcc' failed with exit status 1
>
>
> Anyone could help me...?
> regards
> Gus


From rjkimble at alum.mit.edu  Fri Jul 30 02:03:59 2004
From: rjkimble at alum.mit.edu (Bob Kimble)
Date: Fri Jul 30 02:04:12 2004
Subject: [Image-SIG] Hi from Argentina
In-Reply-To: <41096F57.6030000@clacso.edu.ar>
References: <41096F57.6030000@clacso.edu.ar>
Message-ID: <200407292003.59374.rjkimble@alum.mit.edu>

On Thursday 29 July 2004 05:42 pm, gus wrote:
> > Hi
> > I had installed PIL (Python Image Library)
> > But I have an error when build whith
> > python setup.py build
> >
> > the error said
> >
> > running build
> > running build_py
> > running build_ext
> > building 'imagingtk' extension
> > gcc -pthread -fno-strict-aliasing _DNDDEBUG -g -03 -Wall
> > ................../temp.linux-i686-2.3/Tk/TkImaging.o
> > Tk/Tkimaging.c:49: tk.h: No such file or directory
> > error: command 'gcc' failed with exit status 1

Looks to me that you don't have the tk development files installed. On my 
machines they come from the following packages:

Debian: tk8.4-dev
SuSE: tk-devel
Mandrake: tk

> > Anyone could help me...?
> > regards
> > Gus
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG@python.org
> http://mail.python.org/mailman/listinfo/image-sig
 
From colliera at ukzn.ac.za  Thu Jul  8 17:59:14 2004
From: colliera at ukzn.ac.za (colliera@ukzn.ac.za)
Date: Mon Aug  2 12:41:28 2004
Subject: [Image-SIG] gimp-2.0 palette
Message-ID: <20040708155905.GA5387@adelie>

hello,

has anyone tried to import gimp-2 palettes using the GimpPaletteFile module? presently this does not seem to work because the old palette files began something like this:

GIMP Palette
  0   0   0             grey0
 60   0  80
 60   0  80
 60   0  84
 64   0  84

while the new ones have a couple of extra lines:

GIMP Palette
Name: Royal
#
  0   0   0     grey0
 60   0  80     Untitled
 60   0  80     Untitled
 60   0  84     Untitled
 64   0  84     Untitled

i don't imagine that it would be too difficult to cater for both?

best regards,
andrew.

-- 
Andrew B. Collier

Antarctic Research Fellow                         tel: +27 31 2601157
Space Physics Research Institute                  fax: +27 31 2616550
University of KwaZulu-Natal, Durban, 4041, South Africa
From gab at hg-lab.net  Sat Jul 17 20:55:52 2004
From: gab at hg-lab.net (>>>Gabriele Tazzari)
Date: Mon Aug  2 12:42:01 2004
Subject: [Image-SIG] Problems with PIL1.1.4 using Py2exe
Message-ID: <000801c46c2f$aad6e7d0$2ec3d00a@yoox.net>

Hi, 
I'm trying to compile and transform an application that uses the PIL 1.1.4 and transform it in an exe file using py2exe.

This is the python code:

    import os,shutil,Image
    a=Image.open("test.jpg")
    b=a.resize((150,80),Image.ANTIALIAS)
    b.save("d:/test_script.jpg")
    a=None
    b=None


And this is the setup script:

    # setup.py
    from distutils.core import setup
    import py2exe
    setup(name="testpil.py",scripts=["testpil.py"],)

and of course the script is called in this way:

setup.py py2exe


Can you please tell me why the script runs and the exe doesn't ?
The error I get with the exe is 

    Traceback (most recent call last):
      File "<string>", line 2, in ?
      File "Image.pyo", line 1571, in open
    IOError: cannot identify image file

It can't open the image file...
Can you advise please?
Thanks

Gabriele Tazzari

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/image-sig/attachments/20040717/e30df8c1/attachment.htm
From Padraig.Corcoran at ordnancesurvey.co.uk  Mon Jul 12 15:03:53 2004
From: Padraig.Corcoran at ordnancesurvey.co.uk (Padraig Z. Corcoran)
Date: Tue Aug  3 09:14:44 2004
Subject: [Image-SIG] convoultion
Message-ID: <F16E8A8CFDDCC1479768DD585D752E6C020F2A1A@EXCHMSA.ordsvy.gov.uk>

Hi I want to known how to perform convolution using python. thanx a mill


This email is only intended for the person to whom it is addressed and may contain confidential information. If you have received this email in error, please notify the sender and delete this email which must not be copied, distributed or disclosed to any other person.
Unless stated otherwise, the contents of this email are personal to the writer and do not represent the official view of Ordnance Survey. Nor can any contract be formed on Ordnance Survey's behalf via email. We reserve the right to monitor emails and attachments without prior notice.

Thank you for your cooperation.

Ordnance Survey
Romsey Road
Southampton SO16 4GU
Tel: 023 8079 2000
http://www.ordnancesurvey.co.uk