Incomplete Style.css

[healermagnus]

Describe the bug
doxygen -w header.html footer.html style.css creates an incomplete style file. It is truncated at line 1850. I tried it on multiple version of doxygen, truncated on boty. Currently running on OSX

Expected behavior
Expecting a complete css file.

Version
OSX version 1.9.6 and 1.9.7

[albert-github]

Strange never seen it befor, is your disk full?
Can you please attach the mentioned generated files (in a, compressed, tar or zip file)? Please don't add external links as they might not be persistent.

[healermagnus]

(my doxygen build is softlinked in the parent directory) mkdir html ../doxygen -w html html/header.html html/footer.html html/style.css tail html/style.css It's weird, and does the same on 1.9.6 (attached this time) thanks!

[healermagnus]

Here's the output...

html.tar.gz

[albert-github]

Looks like the file has a size of 40961 bytes which is very strange (i.e. 10 * 4096 + 1), did you check whether or not the disk was full?
Indeed very strange, on windows I cannot reproduce.

@doxygen do you have any ideas or means of reproducing the problem?

[healermagnus]

disk is definitely not full.

…

On Mon, Mar 20, 2023 at 11:37 AM albert-github ***@***.***> wrote: Looks like the file has a size of 40961 bytes which is very strange (i.e. 10 * 4096 + 1), did you check whether or not the disk was full? Indeed very strange, on windows I cannot reproduce. @doxygen <https://github.com/doxygen> do you have any ideas or means of reproducing the problem? — Reply to this email directly, view it on GitHub <#9932 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJI7RALLEXJQ36ZUM7RAU63W5CIVLANCNFSM6AAAAAAWBKYTOU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Chris Healer Director of VR Technology PixoVR 917.977.1122

-- The content of this email is confidential and intended for the recipient specified in message only. It is strictly forbidden to share any part of this message with any third party, without a written consent of the sender. If you received this message by mistake, please reply to this message and follow with its deletion, so that we can ensure such a mistake does not occur in the future.

[healermagnus]

I compiled 1.9.7 for linux, and it doesn't show that bug, so maybe it only happens on OSX, which would be odd.

[albert-github]

In doxygen.cpp we have around line 10977 the code:

          if (openOutputFile(argv[optInd+3],f))
          {
            TextStream t(&f);
            HtmlGenerator::writeStyleSheetFile(t);
          }
          cleanUpDoxygen();
          exit(0);

here the exit should automatically close the output stream f, but it looks like this is not the case.
Can you change the code into:

          if (openOutputFile(argv[optInd+3],f))
          {
            TextStream t(&f);
            HtmlGenerator::writeStyleSheetFile(t);
          }
          f.close();
          cleanUpDoxygen();
          exit(0);

so adding the f.close(); and see what happens / provide us with the result?

• Which compiler do you use under OSX?

Another question: in the html directory there is a file doxygen.css,

• what is its size
• can you provide this file (in a, compressed, tar or zip file) to us as well.

[doxygen]

@albert-github I think the use of exit here is the problem, see also https://stackoverflow.com/a/7054725/784672

[albert-github]

@doxygen
So do you suggest to add the f.close() call or should we make the readConfiguration a function returning an integer and handle this return value after calling readConfiguration in main.cpp:

• 0 no problem just go on with further processing
• 1 error, like the current exit(1)
• 2 OK but exit, so like the exit(0)

[healermagnus]

That’s a good test.I forgot to mention that when I do the publish, the destination is an SMB file system served by windows, which may not flush() as one would like. (Works fine on Linux).  An f.close() may do the trick, I’ll try it.  Possibly an f.flush() could be needed as well.Chris HealerDirector of VR TechnologyPixoVR917.977.1122On Mar 21, 2023, at 05:44, albert-github ***@***.***> wrote: @doxygen So do you suggest to add the f.close() call or should we make the readConfiguration a function returning an integer and handle this return value after calling readConfiguration in main.cpp: 0 no problem just go on with further processing 1 error, like the current exit(1) 2 OK but exit, so like the exit(0) —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: ***@***.***> The content of this email is confidential and intended for the recipient specified in message only. It is strictly forbidden to share any part of this message with any third party, without a written consent of the sender. If you received this message by mistake, please reply to this message and follow with its deletion, so that we can ensure such a mistake does not occur in the future.

[healermagnus]

That close() seemed to do the trick! I agree that the exit() may have been the problem, on a network share the flush was too slow to complete before exit() on my slow wifi. Thanks for looking into it! I wonder if that exists anywhere else in the code.

…

On Tue, Mar 21, 2023 at 6:46 AM albert-github ***@***.***> wrote: In doxygen.cpp we have around line 10977 the code: if (openOutputFile(argv[optInd+3],f)) { TextStream t(&f); HtmlGenerator::writeStyleSheetFile(t); } cleanUpDoxygen(); exit(0); here the exit should automatically close the output stream f, but it looks like this is not the case. Can you change the code into: if (openOutputFile(argv[optInd+3],f)) { TextStream t(&f); HtmlGenerator::writeStyleSheetFile(t); } f.close(); cleanUpDoxygen(); exit(0); so adding the f.close(); and see what happens / provide us with the result? - Which compiler do you use under OSX? Another question: in the html directory there is a file doxygen.css, - what is its size - can you provide this file (in a, compressed, tar or zip file) to us as well. — Reply to this email directly, view it on GitHub <#9932 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJI7RAL5USYV3I4JJJVAP3TW5GBILANCNFSM6AAAAAAWBKYTOU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Chris Healer Director of VR Technology PixoVR 917.977.1122

-- The content of this email is confidential and intended for the recipient specified in message only. It is strictly forbidden to share any part of this message with any third party, without a written consent of the sender. If you received this message by mistake, please reply to this message and follow with its deletion, so that we can ensure such a mistake does not occur in the future.