h2ph - convert .h C header files to .ph Perl header files
h2ph [-d destination directory] [-r | -a] [-l] [headerfiles]
h2ph converts any C header files specified to the corresponding Perl header file format. It is most easily run while in /usr/include:
cd /usr/include; h2ph * sys/*
cd /usr/include; h2ph -r -l .
The output files are placed in the hierarchy rooted at Perl's architecture dependent library directory. You can specify a different hierarchy with a -d switch.
If run with no arguments, filters standard input to standard output.
- -d destination_dir
Put the resulting .ph files beneath destination_dir, instead of beneath the default Perl library location (
Run recursively; if any of headerfiles are directories, then run h2ph on all files in those directories (and their subdirectories, etc.). -r and -a are mutually exclusive.
Run automagically; convert headerfiles, as well as any .h files which they include. This option will search for .h files in all directories which your C compiler ordinarily uses. -a and -r are mutually exclusive.
Symbolic links will be replicated in the destination directory. If -l is not specified, then links are skipped over.
Put ``hints'' in the .ph files which will help in locating problems with h2ph. In those cases when you require a .ph file containing syntax errors, instead of the cryptic
[ some error condition ] at (eval mmm) line nnn
you will see the slightly more helpful
[ some error condition ] at filename.ph line nnn
However, the .ph files almost double in size when built using -h.
Include the code from the .h file as a comment in the .ph file. This is primarily used for debugging h2ph.
No environment variables are used.
The usual warnings if it can't read or write the files involved.
Doesn't construct the %sizeof array for you.
It doesn't handle all C constructs, but it does attempt to isolate definitions inside evals so that you can get at the definitions that it can translate.
It's only intended as a rough tool. You may need to dicker with the files produced.