|
View:
New views
11 Messages
—
Rating Filter:
Alert me
|
|
|
WrapITK and Java BindingsDear ITK Users (and WrapITK developpers),
I was wondering if there is any way to generate javadoc documentation for java wrapped classes (extracted from doxygen tags in the C++ code) so as to get the documentation corresponding to the java classes ? Many thanks, Julien -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julien MICHEL - Ingénieur d'études - Traitement d'images CS Systèmes d'Information - Division ESPACE Département Information Géographique & Image Téléphone : +33 561 17 64 27 Email : julien.michel@... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsHi Julien,
This is a very interesting question. Yes, this could be done... (...but....) 1) When you wrap ITK with WrapITK for Java, the java files are generated in the directory: ${ITK_BIN}/Wrapping/WrapITK/Java/InsightToolkit 2) You could go there and run javadoc with a command such as javadoc -J-Xmx100m -d Documentation itkA*.java Note that for my test I limited the list to the ITK classes whose name starts with "A".... (you probably will do it with all the *.java files. 3) This effectively generates the HTML files in the "Documentation" directory (that was the argument of the -d option of javadoc) 4) ...but... Since the wrapping process has not carried the C++ documentation comments down to the .java files, javadoc can only produce the raw documentation that describes the class hierarchies and the function arguments of every class. 5) For this to be really useful, we would have to modify the wrapping process in order to carry along the C++ comments above all the ITK classes and methods, to make that information available to javadoc at the level of the Java files. Would you like to give it a shot at implementing such feature in WrapITK ? or, would you like at least to report this as a "feature request" in the Bug tracker ? http://public.kitware.com/Bug/my_view_page.php Thanks Luis ------------------------------------------------------------ On Wed, Nov 4, 2009 at 8:07 AM, Julien Michel <julien.michel@...> wrote: > Dear ITK Users (and WrapITK developpers), > > I was wondering if there is any way to generate javadoc documentation for > java wrapped classes (extracted from doxygen tags in the C++ code) so as to > get the documentation corresponding to the java classes ? > > Many thanks, > > Julien > -- > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > Julien MICHEL - Ingénieur d'études - Traitement d'images > CS Systèmes d'Information - Division ESPACE > Département Information Géographique & Image > Téléphone : +33 561 17 64 27 > Email : julien.michel@... > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > _____________________________________ > Powered by www.kitware.com > > Visit other Kitware open-source projects at > http://www.kitware.com/opensource/opensource.html > > Kitware offers ITK Training Courses, for more information visit: > http://www.kitware.com/products/protraining.html > > Please keep messages on-topic and check the ITK FAQ at: > http://www.itk.org/Wiki/ITK_FAQ > > Follow this link to subscribe/unsubscribe: > http://www.itk.org/mailman/listinfo/insight-users > Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsLuis Ibanez a écrit :
> Hi Julien, > > This is a very interesting question. > > Yes, this could be done... (...but....) > > > 1) When you wrap ITK with WrapITK for Java, > the java files are generated in the directory: > > ${ITK_BIN}/Wrapping/WrapITK/Java/InsightToolkit > > 2) You could go there and run javadoc with a command > such as > > javadoc -J-Xmx100m -d Documentation itkA*.java > > Note that for my test I limited the list to the ITK classes > whose name starts with "A".... (you probably will do it > with all the *.java files. > > 3) This effectively generates the HTML files in the > "Documentation" directory (that was the argument of > the -d option of javadoc) > > 4) ...but... > Since the wrapping process has not carried the C++ > documentation comments down to the .java files, > javadoc can only produce the raw documentation > that describes the class hierarchies and the function > arguments of every class. > > 5) For this to be really useful, we would have to modify > the wrapping process in order to carry along the C++ > comments above all the ITK classes and methods, > to make that information available to javadoc at the > level of the Java files. > > Would you like to give it a shot at implementing such > feature in WrapITK ? > > or, would you like at least to report this as a > "feature request" in the Bug tracker ? > > http://public.kitware.com/Bug/my_view_page.php Dear Luis, I came to the same conclusion : one can generate the javadoc, but it is not very useful without importing proper information from the C++ doxygen comments. I may have to solve this issue for some of our needs, and in this case I would of course be glad to contribute a new feature to WrapITK. However, I am quite new to swig and WrapITK, so I would need more information on how (and where) to start. Regards, Julien -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julien MICHEL - Ingénieur d'études - Traitement d'images CS Systèmes d'Information - Division ESPACE Département Information Géographique & Image Téléphone : +33 561 17 64 27 Email : julien.michel@... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsLe 10 nov. 09 à 16:31, Julien Michel a écrit : > Luis Ibanez a écrit : >> Hi Julien, >> This is a very interesting question. >> Yes, this could be done... (...but....) >> 1) When you wrap ITK with WrapITK for Java, >> the java files are generated in the directory: >> ${ITK_BIN}/Wrapping/WrapITK/Java/InsightToolkit >> 2) You could go there and run javadoc with a command >> such as >> javadoc -J-Xmx100m -d Documentation itkA*.java >> Note that for my test I limited the list to the ITK classes >> whose name starts with "A".... (you probably will do it >> with all the *.java files. >> 3) This effectively generates the HTML files in the >> "Documentation" directory (that was the argument of >> the -d option of javadoc) >> 4) ...but... >> Since the wrapping process has not carried the C++ >> documentation comments down to the .java files, >> javadoc can only produce the raw documentation >> that describes the class hierarchies and the function >> arguments of every class. >> 5) For this to be really useful, we would have to modify >> the wrapping process in order to carry along the C++ >> comments above all the ITK classes and methods, >> to make that information available to javadoc at the >> level of the Java files. >> Would you like to give it a shot at implementing such >> feature in WrapITK ? >> or, would you like at least to report this as a >> "feature request" in the Bug tracker ? >> http://public.kitware.com/Bug/my_view_page.php > > Dear Luis, > > I came to the same conclusion : one can generate the javadoc, but it > is not very useful without importing proper information from the C++ > doxygen comments. > > I may have to solve this issue for some of our needs, and in this > case I would of course be glad to contribute a new feature to > WrapITK. However, I am quite new to swig and WrapITK, so I would > need more information on how (and where) to start. Hi Julien, There is already a doc generator for python in WrapITK 0.3.0 - it puts the same doc as in doxygen in python docstrings. See the code in http://code.google.com/p/wrapitk/source/browse/#svn/trunk/Languages/ Doc Unfortunately, javadoc does not use the same mechanisms and is not well integrated in swig. http://www.swig.org/Doc1.3/Java.html#javadoc_comments Anyway, It should be possible to do something for java similar to what is done for python. Patches are welcome! Regards, Gaëtan -- Gaëtan Lehmann Biologie du Développement et de la Reproduction INRA de Jouy-en-Josas (France) tel: +33 1 34 65 29 66 fax: 01 34 65 29 09 http://voxel.jouy.inra.fr http://www.itk.org http://www.mandriva.org http://www.bepo.fr _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsGaëtan Lehmann a écrit :
> > Le 10 nov. 09 à 16:31, Julien Michel a écrit : > >> Luis Ibanez a écrit : >>> Hi Julien, >>> This is a very interesting question. >>> Yes, this could be done... (...but....) >>> 1) When you wrap ITK with WrapITK for Java, >>> the java files are generated in the directory: >>> ${ITK_BIN}/Wrapping/WrapITK/Java/InsightToolkit >>> 2) You could go there and run javadoc with a command >>> such as >>> javadoc -J-Xmx100m -d Documentation itkA*.java >>> Note that for my test I limited the list to the ITK classes >>> whose name starts with "A".... (you probably will do it >>> with all the *.java files. >>> 3) This effectively generates the HTML files in the >>> "Documentation" directory (that was the argument of >>> the -d option of javadoc) >>> 4) ...but... >>> Since the wrapping process has not carried the C++ >>> documentation comments down to the .java files, >>> javadoc can only produce the raw documentation >>> that describes the class hierarchies and the function >>> arguments of every class. >>> 5) For this to be really useful, we would have to modify >>> the wrapping process in order to carry along the C++ >>> comments above all the ITK classes and methods, >>> to make that information available to javadoc at the >>> level of the Java files. >>> Would you like to give it a shot at implementing such >>> feature in WrapITK ? >>> or, would you like at least to report this as a >>> "feature request" in the Bug tracker ? >>> http://public.kitware.com/Bug/my_view_page.php >> >> Dear Luis, >> >> I came to the same conclusion : one can generate the javadoc, but it >> is not very useful without importing proper information from the C++ >> doxygen comments. >> >> I may have to solve this issue for some of our needs, and in this case >> I would of course be glad to contribute a new feature to WrapITK. >> However, I am quite new to swig and WrapITK, so I would need more >> information on how (and where) to start. > > > Hi Julien, > > There is already a doc generator for python in WrapITK 0.3.0 - it puts > the same doc as in doxygen in python docstrings. See the code in > > http://code.google.com/p/wrapitk/source/browse/#svn/trunk/Languages/Doc > > Unfortunately, javadoc does not use the same mechanisms and is not well > integrated in swig. > > http://www.swig.org/Doc1.3/Java.html#javadoc_comments > > Anyway, It should be possible to do something for java similar to what > is done for python. Gaetan, Luis, In order to understand the process I tried to generate docstring for python, but I did not manage to do it. In Guide.txt, it is said that one has to set the DOXYGEN_MAN_PATH when configuring in cmake but I never get this variable and I can not find the make_doxygen_config.py in Python directory in the binary directory. I must have missed something, but what ? Doxygen is correctly found on my system in WrapITK and ITK (I set BUILD_DOXYGEN to ON). Any idea ? Regards, Julien -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julien MICHEL - Ingénieur d'études - Traitement d'images CS Systèmes d'Information - Division ESPACE Département Information Géographique & Image Téléphone : +33 561 17 64 27 Email : julien.michel@... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsLe 12 nov. 09 à 18:45, Julien Michel a écrit : > Gaëtan Lehmann a écrit : >> Le 10 nov. 09 à 16:31, Julien Michel a écrit : >>> Luis Ibanez a écrit : >>>> Hi Julien, >>>> This is a very interesting question. >>>> Yes, this could be done... (...but....) >>>> 1) When you wrap ITK with WrapITK for Java, >>>> the java files are generated in the directory: >>>> ${ITK_BIN}/Wrapping/WrapITK/Java/InsightToolkit >>>> 2) You could go there and run javadoc with a command >>>> such as >>>> javadoc -J-Xmx100m -d Documentation itkA*.java >>>> Note that for my test I limited the list to the ITK classes >>>> whose name starts with "A".... (you probably will do it >>>> with all the *.java files. >>>> 3) This effectively generates the HTML files in the >>>> "Documentation" directory (that was the argument of >>>> the -d option of javadoc) >>>> 4) ...but... >>>> Since the wrapping process has not carried the C++ >>>> documentation comments down to the .java files, >>>> javadoc can only produce the raw documentation >>>> that describes the class hierarchies and the function >>>> arguments of every class. >>>> 5) For this to be really useful, we would have to modify >>>> the wrapping process in order to carry along the C++ >>>> comments above all the ITK classes and methods, >>>> to make that information available to javadoc at the >>>> level of the Java files. >>>> Would you like to give it a shot at implementing such >>>> feature in WrapITK ? >>>> or, would you like at least to report this as a >>>> "feature request" in the Bug tracker ? >>>> http://public.kitware.com/Bug/my_view_page.php >>> >>> Dear Luis, >>> >>> I came to the same conclusion : one can generate the javadoc, but >>> it is not very useful without importing proper information from >>> the C++ doxygen comments. >>> >>> I may have to solve this issue for some of our needs, and in this >>> case I would of course be glad to contribute a new feature to >>> WrapITK. However, I am quite new to swig and WrapITK, so I would >>> need more information on how (and where) to start. >> Hi Julien, >> There is already a doc generator for python in WrapITK 0.3.0 - it >> puts the same doc as in doxygen in python docstrings. See the code in >> http://code.google.com/p/wrapitk/source/browse/#svn/trunk/ >> Languages/Doc >> Unfortunately, javadoc does not use the same mechanisms and is not >> well integrated in swig. >> http://www.swig.org/Doc1.3/Java.html#javadoc_comments >> Anyway, It should be possible to do something for java similar to >> what is done for python. > > Gaetan, Luis, > > In order to understand the process I tried to generate docstring for > python, but I did not manage to do it. > > In Guide.txt, it is said that one has to set the DOXYGEN_MAN_PATH > when configuring in cmake but I never get this variable and I can > not find the make_doxygen_config.py in Python directory in the > binary directory. > > I must have missed something, but what ? Doxygen is correctly found > on my system in WrapITK and ITK (I set BUILD_DOXYGEN to ON). > > Any idea ? Hi Julien, Sorry for the late reply, Guide.txt is outdated. You have to turn on WRAP_ITK_DOC to build the python doc currently. I'm afraid there is not much documentation about that for now. Please just call me if to dive in that subject - it may be easier to explain how it work that way. Regards, Gaëtan -- Gaëtan Lehmann Biologie du Développement et de la Reproduction INRA de Jouy-en-Josas (France) tel: +33 1 34 65 29 66 fax: 01 34 65 29 09 http://voxel.jouy.inra.fr http://www.itk.org http://www.mandriva.org http://www.bepo.fr _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsGaëtan Lehmann a écrit :
> Hi Julien, > > Sorry for the late reply, > > Guide.txt is outdated. You have to turn on WRAP_ITK_DOC to build the > python doc currently. > > I'm afraid there is not much documentation about that for now. Please > just call me if to dive in that subject - it may be easier to explain > how it work that way. > > Regards, Thank you Gaetan. I managed to figure this out by myself. I spotted a possible mistake in Languages/Doc/CMakeLists.txt:62 in STRING(TOLOWER ...) it seems that xmlname and ${xmlname} are inverted (once switched I got no more warning about missing doc files). I did modify the itk_doxy2swig.py so as to get javadoc documentation as well in the generated *_doc.i using the following workaround from swig: http://www.swig.org/Doc1.3/Java.html#javadoc_comments Now, I am just missing one more step to import this documentation when generating java proxy classes (maybe I missed something in Languages/Java/CMakeLists.txt). Thank you very much for your help, I might call you if I need to. Regards, Julien -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julien MICHEL - Ingénieur d'études - Traitement d'images CS Systèmes d'Information - Division ESPACE Département Information Géographique & Image Téléphone : +33 561 17 64 27 Email : julien.michel@... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsGaetan, Luis,
Just to let you know, I managed to update the itk_doxy2swig.py script so as to get the doxygen information imported into javadoc comments in the java proxy classes. I am only missing a few things : - gccxml (ot pygccxml) seems to strip newlines so the imported doc is a little compact (hard to read) - I have trouble tranlating the \See to @see, since the name of classes changes (because of mangling) - Something more annoying : sometimes the proxy classes does not inherit from its real superclass but rather from myClass_Superclass, which has no documentation. Therefore, the inheritance scheme in javadoc is broken. - For now I did not tranlate \param to @param and \return to @return Any idea on how to fix those points ? I shall send you a patch as soon as these points have been adressed. Regards, Julien -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julien MICHEL - Ingénieur d'études - Traitement d'images CS Systèmes d'Information - Division ESPACE Département Information Géographique & Image Téléphone : +33 561 17 64 27 Email : julien.michel@... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsLe 24 nov. 09 à 10:58, Julien Michel a écrit : > Gaetan, Luis, > Hi Julien, > Just to let you know, I managed to update the itk_doxy2swig.py > script so as to get the doxygen information imported into javadoc > comments in the java proxy classes. I am only missing a few things : > - gccxml (ot pygccxml) seems to strip newlines so the imported doc > is a little compact (hard to read) gccxml, pygccxml and igenerator.py are fully reconstructing the classes definitions. They are not manipulating the documentation at all. The problem is more likely in itk_doxy2swig.py or in swig. > - I have trouble tranlating the \See to @see, since the name of > classes changes (because of mangling) yes, it would be a bit hard to revert an c++ class name to a class name available in java, at least because there are often several classes in java for one class in c++ > - Something more annoying : sometimes the proxy classes does not > inherit from its real superclass but rather from myClass_Superclass, > which has no documentation. Therefore, the inheritance scheme in > javadoc is broken. We should have the doc for those class as well - it looks like a bug in the doc translation process. I'll try to have a look. > - For now I did not tranlate \param to @param and \return to @return > > Any idea on how to fix those points ? I shall send you a patch as > soon as these points have been adressed. Would you agree to share your changes before everything is fixed? It would make things easier to review :-) Gaëtan -- Gaëtan Lehmann Biologie du Développement et de la Reproduction INRA de Jouy-en-Josas (France) tel: +33 1 34 65 29 66 fax: 01 34 65 29 09 http://voxel.jouy.inra.fr http://www.itk.org http://www.mandriva.org http://www.bepo.fr _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsGaëtan Lehmann a écrit :
> Hi Julien, > >> Just to let you know, I managed to update the itk_doxy2swig.py script >> so as to get the doxygen information imported into javadoc comments in >> the java proxy classes. I am only missing a few things : >> - gccxml (ot pygccxml) seems to strip newlines so the imported doc is >> a little compact (hard to read) > > gccxml, pygccxml and igenerator.py are fully reconstructing the classes > definitions. They are not manipulating the documentation at all. > The problem is more likely in itk_doxy2swig.py or in swig. Well, itk_doxy2swig.py is parsing xml documentation file to build python docstrings (and now javadoc). I was assuming that these xml doc files were produced by gccxml or pygccxml (but I might be wrong). >> - Something more annoying : sometimes the proxy classes does not >> inherit from its real superclass but rather from myClass_Superclass, >> which has no documentation. Therefore, the inheritance scheme in >> javadoc is broken. > > We should have the doc for those class as well - it looks like a bug in > the doc translation process. > I'll try to have a look. Well, it sounds more like a bug (or a feature ? ;)) from swig to me. For instance : itkAtanImageFilterIIF2_IIF2 is inheriting from itkAtanImageFilterIIF2_IIF2_Superclass, which should rather be itkUnaryFunctorImageFilterIIF2_IIF2_Tan or something like this. Maybe it is just a problem of mangling ... Julien -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julien MICHEL - Ingénieur d'études - Traitement d'images CS Systèmes d'Information - Division ESPACE Département Information Géographique & Image Téléphone : +33 561 17 64 27 Email : julien.michel@... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
|
|
Re: WrapITK and Java BindingsLe 24 nov. 09 à 18:04, Julien Michel a écrit : > Gaëtan Lehmann a écrit : > >> Hi Julien, >>> Just to let you know, I managed to update the itk_doxy2swig.py >>> script so as to get the doxygen information imported into javadoc >>> comments in the java proxy classes. I am only missing a few things : >>> - gccxml (ot pygccxml) seems to strip newlines so the imported doc >>> is a little compact (hard to read) >> gccxml, pygccxml and igenerator.py are fully reconstructing the >> classes definitions. They are not manipulating the documentation at >> all. >> The problem is more likely in itk_doxy2swig.py or in swig. > > Well, itk_doxy2swig.py is parsing xml documentation file to build > python docstrings (and now javadoc). I was assuming that these xml > doc files were produced by gccxml or pygccxml (but I might be wrong). at this stage, it will be hard to fix this problem. > >>> - Something more annoying : sometimes the proxy classes does not >>> inherit from its real superclass but rather from >>> myClass_Superclass, which has no documentation. Therefore, the >>> inheritance scheme in javadoc is broken. >> We should have the doc for those class as well - it looks like a >> bug in the doc translation process. >> I'll try to have a look. > > Well, it sounds more like a bug (or a feature ? ;)) from swig to me. > > For instance : > > itkAtanImageFilterIIF2_IIF2 is inheriting from > itkAtanImageFilterIIF2_IIF2_Superclass, which should rather be > itkUnaryFunctorImageFilterIIF2_IIF2_Tan or something like this. > > Maybe it is just a problem of mangling ... The name itkAtanImageFilterIF2IF2_Superclass is a feature of wrapitk :-) The wrapper for those classes - both itkAtanImageFilterIF2IF2 and itkAtanImageFilterIF2IF2_Superclass is simply WRAP_CLASS("itk::TanImageFilter" POINTER_WITH_SUPERCLASS) WRAP_IMAGE_FILTER_REAL(2) END_WRAP_CLASS() It makes the creation of the wrapper quite easy for this kind of class, where the superclass is not supposed to be used directly. It also makes quite difficult to find the name of the the super class only with the informations there. That's probably why those classes don't have there doc, but I have to have a closer look. Gaëtan -- Gaëtan Lehmann Biologie du Développement et de la Reproduction INRA de Jouy-en-Josas (France) tel: +33 1 34 65 29 66 fax: 01 34 65 29 09 http://voxel.jouy.inra.fr http://www.itk.org http://www.mandriva.org http://www.bepo.fr _____________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Kitware offers ITK Training Courses, for more information visit: http://www.kitware.com/products/protraining.html Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ Follow this link to subscribe/unsubscribe: http://www.itk.org/mailman/listinfo/insight-users |
| Free embeddable forum powered by Nabble | Forum Help |