Re^2: How to properly use ExtUtils::MakeMaker PL

Replies are listed 'Best First'.
Re^3: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by syphilis (Archbishop) on May 27, 2017 at 03:21 UTC
I should have them available and then the related code requires to have constant from C header files, which I can't (easily) figure out where they are (h2ph doesn't work properly It's quite possible that I underestimate the complexity of the problem. 1) Can't the PL file that generates the module place it in blib (creating the 'blib' directory if it's not already existent)? 2) If the C header files are not found by default, then surely it's up to the user to make them locatable - eg by setting the relevant environment variable, or providing INC arg to Makefile.PL. Cheers, Rob	[reply]
Re^4: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by glasswalk3r (Friar) on May 27, 2017 at 15:16 UTC
1) Can't the PL file that generates the module place it in blib (creating the 'blib' directory if it's not already existent)? I tried that. The PL files (which I call `build_my_module.PL` from now on) is executed after the `pm_to_files` phase. I was able to open the related PM files, append information to it and close it, but this was too late since a copy (without the parts I need) was already put on 'blib'. I can write over there too, just will need to set right permissions, write, set read-only permissions again. But looks like a hack after all. I didn't get what the documentation says about "PL output is suppose to be used to the given parameter". Printing the output to `STDOUT` doesn't produce the expected results, and looking at the Makefile (which calls the script with a `perl build_my_module.PL`, `return`ing data wouldn't do either. I'm explicit using `open` for that. 2) If the C header files are not found by default, then surely it's up to the user to make them locatable - eg by setting the relevant environment variable, or providing INC arg to Makefile.PL. There is a long discussion about this on the CPAN Testers mailing list, but from it I have the following conclusions: These constants will change depending on processor architecture Best way to achieve the right header file is using the C compiler itself, since `h2ph` doesn't quite work (at least not in Ubuntu). So, to make it clear, I don't need the C compiler really, but I need it to grab me the correct constant value to use with `syscall`. Another possibility is to create a hash with the keys being the architecture and the values the related constants, but this is a hack and there is no telling when it is going to be broken. Alceu Rodrigues de Freitas Junior --------------------------------- "You have enemies? Good. That means you've stood up for something, sometime in your life." - Sir Winston Churchill	[reply] [d/l] [select]
Re^3: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by Anonymous Monk on May 27, 2017 at 05:34 UTC
can't use or require something that is declared as a prereqs and is still not available). Yes you can, thats what META files are for, you simply tell writemakefile the depends, when you make dist METAL files are generated, then when user try to install cpan/cpanp/cpanm notice from metal files what the "BUILD_REQUIRES" ...	[reply]
Re^4: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by glasswalk3r (Friar) on May 27, 2017 at 15:31 UTC
Yes you can, thats what META files are for, you simply tell writemakefile the depends, when you make dist METAL files are generated, then when user try to install cpan/cpanp/cpanm notice from metal files what the "BUILD_REQUIRES" ... Not sure if this would work, specially because the distribution already generates the mentioned files (see here). Check out this test result too. I made a stupid mistake when trying to avoid calling Inline at each call, I just need to execute it once and write the constant elsewhere. But doing that inside the Makefile.PL is not an option... I could have used `require` inside a `eval` block to trap the exception, but then the module wouldn't be created anyway. The `cpan` client will search and install dependencies right after the Makefile.PL is executed, but then it is too late to use it for the code generation. A ".PL" file (as described by ExtUtils::MakeMaker) would be the best option, if I could control when it will be executed... Alceu Rodrigues de Freitas Junior --------------------------------- "You have enemies? Good. That means you've stood up for something, sometime in your life." - Sir Winston Churchill	[reply] [d/l] [select]
Re^3: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by Anonymous Monk on May 27, 2017 at 05:31 UTC
See Linux::NFS::BigDir. Is that the module you're working on? How often in the past 20 years has that number you're fetching from the header changed? It should be easy to lookup Instead of generating a throwaway module so you can inline a constant, you should be generate a data file, using the "official" way for data files File::ShareDir / File::ShareDir::Install ...	[reply]
Re^4: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by glasswalk3r (Friar) on May 27, 2017 at 15:22 UTC
See Linux::NFS::BigDir. Is that the module you're working on? Yes, I'm the author of Linux::NFS::BigDir. How often in the past 20 years has that number you're fetching from the header changed? It should be easy to lookup If I could rely on `h2ph`, yes, but I can't. Linux distributions have a variation of locations there they think it the best way to keep C header files. On the other hand, they don't distribute broken C libraries, so it is safe to assume I can use it to get the constant. Instead of generating a throwaway module so you can inline a constant, you should be generate a data file, using the "official" way for data files File::ShareDir / File::ShareDir::Install That will probably give me a maintenance headache: we just need a new processor coming up or anything else that justify changing that constant. Alceu Rodrigues de Freitas Junior --------------------------------- "You have enemies? Good. That means you've stood up for something, sometime in your life." - Sir Winston Churchill	[reply] [d/l]
Re^5: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by Anonymous Monk on Jun 04, 2017 at 08:29 UTC
If I could rely on h2ph, yes, but I can't. Linux distributions have a variation of locations there they think it the best way to keep C header files. On the other hand, they don't distribute broken C libraries, so it is safe to assume I can use it to get the constant. Um, no, thats not what I'm talking about. Header files are created by humans, they're source code. In big projects, like gui libraries, mfc, wxwidgets, header numbers do not change ever, for 5/10/15/20 years and more. New headers get added. Old numbers do not change. I'm not suggesting you use h2ph , I'm suggesting you use your eyes, track down every copy of the header file that ever existed, like in some kind of git repository, and compare the number today to the number it was 50 years ago Do it for all "official" source code repository that manages this header file Its like, you're not compiling/interfacing any c/xs code -- headers are just #define ... that stuff doesn't change :) That will probably give me a maintenance headache: we just need a new processor coming up or anything else that justify changing that constant. Hehe, um, no, that is not a maintenance headache, you just keep doing exactly what you're doing right this moment, except instead of Inline ... in Makefile.PL, only so you can insert `sub CONSTANT(){12}` into Foo.pm, all you do is have Foo.pm read from foo-dist/constant.txt when its loaded Its like h2ph, you require "constants.ph" or whatever, you don't inline the source of constants.ph into your module.pm Its like how much time you want to spend figuring out your makemaker for a single solitary constant	[reply] [d/l]
Re^6: How to properly use ExtUtils::MakeMaker PL_FILES attribute? by glasswalk3r (Friar) on Jun 06, 2017 at 13:47 UTC


Perl: the Markov chain saw
	PerlMonks

Re^2: How to properly use ExtUtils::MakeMaker PL_FILES attribute?