[Insight-users] MetaImage: absolute path in ElementDataFile not working
Luis Ibanez
luis.ibanez at kitware.com
Tue May 13 17:52:26 EDT 2008
Hi Florian,
I just confirmed that in Windows XP with
my Visual Studio 7.1 build I can successfully
run the following command:
Insight/bin/release/ImageReadWrite.exe
"C:/cygwin/home/ibanez/data/One Space/brainweb165a10f17.mha"
output.mhd
(in a single line).
This seems to indicate that MetaImage (at least in this
platform) can read both Absolute paths and paths with
a space on them.
Also, all tests that are run in the Dashboard use absolute
paths. This is because CMake expands/resolve all paths to
absolute when generated the test command line in the
ADD_TEST() statement.
-----
In any case, I just realize what the problem is,
Sorry that I didn't get it in the first email.
Your header file MyData.mhd contains:
ElementDataFile = C:\Program Files\Data\MyData.raw
It should just be:
ElementDataFile = MyData.raw
The MetaImage library takes care of appending to
the raw file the path of the .mhd file.
Otherwise you couldn't ever move your files from
one directory to another. You simply need to keep
the two files (.mhd and .raw) together in the same
directory.
Thanks
Luis
-------------------------
Florian Pierron wrote:
> Hi Luis,
>
> Sorry to take so much of your time, but it's still not working for me
> (without spaces in the path).
>
> The patch that I described previously by changing the condition on the
> variable 'usePath' seems to fix the problem. Are you sure MetaImage
> supports absolute path in the ElementDataFile field?
>
> Regards,
> Florian
>
>
> On 13/05/2008 17:30, Luis Ibanez wrote:
>
>>
>> Hi Florian,
>>
>> Thanks for sending our files.
>>
>> I'm not sure that MetaImage can manage
>> filenames that have spaces in the path....
>>
>>
>> Your file is located at:
>>
>> "C:\Program Files\Data\MyData.raw"
>>
>>
>> Just to double check,
>> Please do the following
>>
>> Move your .mhd and .raw files to a directory
>> without spaces. Something like:
>>
>> C:\itktest\MyData.mhd
>> C:\itktest\MyData.raw
>>
>>
>> Try to read them from that location,
>> and let us know what you find.
>>
>> If you can read the files then we
>> should probably fill a bug report
>> regarding filenames with spaces
>> and the MetaImage library.
>>
>>
>> Thanks
>>
>>
>> Luis
>>
>>
>> -----------------------
>> Florian Pierron wrote:
>>
>>> Hi Luis,
>>>
>>> Thanks for your quick reply.
>>>
>>> I am using VC2005 to compile the code. I don't see anything obviously
>>> wrong with the path, it is:
>>> ElementDataFile = C:\Program Files\Data\MyData.raw
>>>
>>> I have attached the .mha file and the .raw image. Let me know if you
>>> can load it.
>>>
>>> Regards,
>>> Florian
>>>
>>> ***********************
>>> ObjectType = Image
>>> NDims = 2
>>> DimSize = 50 50
>>> ElementByteOrderMSB = False
>>> ElementNumberOfChannels = 1
>>> ElementType = MET_UCHAR
>>> ElementDataFile = C:\Program Files\Data\MyData.raw
>>> ***********************
>>>
>>>
>>> On 12/05/2008 14:50, Luis Ibanez wrote:
>>>
>>>>
>>>> Hi Florian,
>>>>
>>>> Could you please post the *exact* absolute path that you were using
>>>> when you encounter this problem ?
>>>>
>>>> We are able to read files with absolute path without any problem.
>>>> It is therefore likely that there is something special about the
>>>> path that you were using.
>>>>
>>>> Also, please let us know what platform are you using.
>>>>
>>>>
>>>> Thanks
>>>>
>>>>
>>>> Luis
>>>>
>>>>
>>>> ---------------------
>>>> Florian Pierron wrote:
>>>>
>>>>> Dear ITK developer team,
>>>>>
>>>>> I noticed that having an absolute path in the ElementDataFile is
>>>>> not supported in ITK 3.4.0. I don't know if there is a good reason
>>>>> for that or if it's just by mistake but it seems that there is an
>>>>> easy fix to that (if I am not wrong...). I have changed the
>>>>> following lines in \Utilities\MetaIO\metaImage.cxx:
>>>>>
>>>>> - usePath = MET_GetFilePath(_headerName, pathName)
>>>>>
>>>>> by
>>>>>
>>>>> + usePath = !MET_GetFilePath(elementDataFileName.c_str(),
>>>>> pathName) &&
>>>>> + MET_GetFilePath(_headerName, pathName);
>>>>>
>>>>> So that usePath (which determines in fact if the path to the header
>>>>> / meta file should be append to the elementDataFileName) is used
>>>>> only if the elementDataFileName is not an absolute path and if
>>>>> there is a path for the header / meta file.
>>>>>
>>>>> Find attached the patch file.
>>>>>
>>>>> Regards,
>>>>> Florian
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------
>>>>>
>>>>>
>>>>> ### Eclipse Workspace Patch 1.0
>>>>> #P ITK_3_4_0
>>>>> Index: Utilities/MetaIO/metaImage.cxx
>>>>> ===================================================================
>>>>> RCS file: /cvsroot/Insight/Insight/Utilities/MetaIO/metaImage.cxx,v
>>>>> retrieving revision 1.76
>>>>> diff -u -r1.76 metaImage.cxx
>>>>> --- Utilities/MetaIO/metaImage.cxx 12 Sep 2007 11:46:05 -0000
>>>>> 1.76
>>>>> +++ Utilities/MetaIO/metaImage.cxx 12 May 2008 10:53:09 -0000
>>>>> @@ -3,7 +3,7 @@
>>>>> Program: MetaIO
>>>>> Module: $RCSfile: metaImage.cxx,v $
>>>>> Language: C++
>>>>> - Date: $Date: 2007-09-12 11:46:05 $
>>>>> + Date: $Date: 2007/09/12 11:46:05 $
>>>>> Version: $Revision: 1.76 $
>>>>>
>>>>> Copyright (c) Insight Software Consortium. All rights reserved.
>>>>> @@ -1118,10 +1118,6 @@
>>>>> return false;
>>>>> }
>>>>>
>>>>> - bool usePath;
>>>>> - char pathName[255];
>>>>> - usePath = MET_GetFilePath(_headerName, pathName);
>>>>> -
>>>>> char* buf = new char[8001];
>>>>> inputStream.read(buf,8000);
>>>>> unsigned long fileSize = inputStream.gcount();
>>>>> @@ -1139,6 +1135,10 @@
>>>>>
>>>>> std::string elementDataFileName =
>>>>> M_GetTagValue(header,"ElementDataFile");
>>>>>
>>>>> + char pathName[255];
>>>>> + bool usePath = !MET_GetFilePath(elementDataFileName.c_str(),
>>>>> pathName) && + MET_GetFilePath(_headerName, pathName);
>>>>> +
>>>>> char* fName = new char[512];
>>>>>
>>>>> if(!strcmp("Local", elementDataFileName.c_str()) || @@ -1375,10
>>>>> +1375,10 @@
>>>>> }
>>>>>
>>>>> int i, j;
>>>>> - bool usePath;
>>>>> char pathName[255];
>>>>> char fName[255];
>>>>> - usePath = MET_GetFilePath(m_FileName, pathName);
>>>>> + bool usePath = !MET_GetFilePath(m_ElementDataFileName,
>>>>> pathName) && + MET_GetFilePath(m_FileName,
>>>>> pathName);
>>>>>
>>>>> if(!strcmp("Local", m_ElementDataFileName) ||
>>>>> !strcmp("LOCAL", m_ElementDataFileName) ||
>>>>> @@ -1620,7 +1620,8 @@
>>>>> }
>>>>>
>>>>> char pathName[255];
>>>>> - bool usePath = MET_GetFilePath(m_FileName, pathName);
>>>>> + bool usePath = !MET_GetFilePath(m_ElementDataFileName, pathName)
>>>>> && + MET_GetFilePath(m_FileName, pathName);
>>>>> if(usePath)
>>>>> {
>>>>> char elementPathName[255];
>>>>> @@ -2240,7 +2241,8 @@
>>>>> {
>>>>> char dataFileName[255];
>>>>> char pathName[255];
>>>>> - bool usePath = MET_GetFilePath(m_FileName, pathName);
>>>>> + bool usePath = !MET_GetFilePath(m_ElementDataFileName,
>>>>> pathName) && + MET_GetFilePath(m_FileName,
>>>>> pathName);
>>>>> if(usePath)
>>>>> {
>>>>> sprintf(dataFileName, "%s%s", pathName, m_ElementDataFileName);
>>>>> @@ -2471,10 +2473,10 @@
>>>>> m_Quantity *= (_indexMax[i] - _indexMin[i] + 1);
>>>>> }
>>>>>
>>>>> - bool usePath;
>>>>> char pathName[255];
>>>>> char fName[255];
>>>>> - usePath = MET_GetFilePath(m_FileName, pathName);
>>>>> + bool usePath = !MET_GetFilePath(m_ElementDataFileName,
>>>>> pathName) && + MET_GetFilePath(m_FileName,
>>>>> pathName);
>>>>>
>>>>> if(!strcmp("Local", m_ElementDataFileName) ||
>>>>> !strcmp("LOCAL", m_ElementDataFileName) ||
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Insight-users mailing list
>>>>> Insight-users at itk.org
>>>>> http://www.itk.org/mailman/listinfo/insight-users
>>>>
>>>>
>>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> ObjectType = Image
>>> NDims = 2
>>> DimSize = 50 50
>>> ElementByteOrderMSB = False
>>> ElementNumberOfChannels = 1
>>> ElementType = MET_UCHAR
>>> ElementDataFile = C:\Program Files\Data\MyData.raw
>>
>>
>
More information about the Insight-users
mailing list