logs archiveBotHelp.net / Freenode / #ada / 2015 / July / 17 / 1
Somebody pinged me or mentioned my nick?
Unfortunately my history is gone...
darkestkhan: regarding the quote from the gnatbind manual, I am curious how you can have no main. Does this mean that all is needed is a .adb file that simply with's various packages?
kensen: Lucretia mentioned your name at 06:46
[06:45] <Lucretia> well, i'm asking what people are using for generation of api/html docs [06:45] <Lucretia> I want to generate it from the sources [06:45] <Lucretia> might look at updating adabrowse [06:46] <Lucretia> although there's a git of it on codelabs [06:46] <Lucretia> kensan: ??
wtp: Thank you!
Lucretia: I am the maintainer of the adabrowse Debian package
Lucretia: Upstream development has basically ceased a couple of years ago and I am seriously considering orphaning the Debian package.
Lucretia: or did you mean some different info with "??"?
specing: trust me it's not
kensan: was wondering if you'd done any work on updating it as i found your repo
well, I think I just wrote the longest linkedin post ever
mostly quotes, but I do think that simon wright is a bit thick
(Action) hopes he isn't here ;)
does the git repo have recent commits?
and sounds like another reason for doxygen parser...
yeah, you ever look at doxygen source?
shouldn't be too horrible to map OO/procedural atuff... and there's some 83/95 antlr grammar stuff languishing in cccc repo
dunno if the git repo has recent commits to the actual source apart from getting it to compile with more recent compilers
still, seems like asis would be the interface to extract the data
the metrics tool
(Action) has looked at doxygen source a number of times, fu*king mess
it's so nice to see source compile without any warnings and then you compile a C lib...
it does work, but yeah, it's an "orgainc" design at best
I think "design" is a bit rich
it has an "interface" for external parsers, is that what you looked at?
apparently the vhdl bit uses the code that was written to be the rewrite, a proper api
i never said it was "designed"
I looked at the .l file
but everything has one
"it's an "orgainc" design at best" <<< yeah you did
having one != designed-by-humans-on-purpose
organic "design" is what you get when nobody gives a sh*t about design upfront...
it keeps getting grafted with franken-parts
it's deffo got franken-bits
like i said, it works, amazingly enough
tho i did get a random fail starting fresh with wizard giu using latest version
somehow it forgot the index.html file and defaulted to no subdirs
it would be a lot more work i think to generate dot files for graphs from scratch
feeding data to doxygen would leverage all the output stuff
true, but for a proper job, it needs to be, a proper job, not a filter to make it look like another language, due to the complexity of Ada
anyway, i think the cccc Ada support was disabled shortly before the last release, so it might not be a big deal to get it working
i think if it can handle recent java/c++ it should handle Ada
latest config gui has more "optimize for language X" options
i don't remember fortran or python being there before...
*doxywizard gui
if there's a new python builtin it might be worth looking at
back to cccc, it also chokes on stuff like webkit so the grammars probably all need updating
should work with 2003 but not 2011...
i just don't know of another metrics tool that does what it does as a simple stand-alone binary you can run from a makefile or whatever
http://www.gentoogeek.org/metrics/ <= cccc report run itself in c++ mode
*on its own source
Lucretia: what does adabrowse use?
essentially you meed to parse the source into an AST, right? at least to get any structural info out of it...
(Action) only vaguely grasping how antlr/cccc works
he's got a grammar and a lexer that antlr uses to build the source parser? i think...
adabrowse is asis, iirc
I think starting with a 95 grammar would be as tricky to add the 2005/2012 bits as taking the 2012 grammar and making that useable
even trickier if you wanted to add in language flags like gnat has
think i'd let that be a cmdline switch
and what 2012 grammar are you pointing at?
does it cover all previous standards?
the grammar from the aarm, there is no other
in source format you mean?
don't see any -std Ada options in gnatgcc man page...
no, I mean in aarm and rm, appendix p or j or whatever it
what are you on about?
the ada.g grammar file in the cccc source
for antlr or similar tool to generate other source files
I'm talking about the grammar in the aarm, as I stated
need one for 2012 to update that tool or any similar one
the switches are in gnat, -gnat05, -gnat2012
(Action) has no idea how asis works
there is only one 2012 grammar, that is in the rm/aarm, and due to their fu*king stupidness, they still don't put in rules for pragma
so you have to work that out based on their description of where it goes and then you have to make the thing useable
whatever the standard defines has to be encoded/processed for language-specific (source) analysis
Lucretia: any reason for them omitting pragmas? that seems... important
no idea, they just don't put them into the grammar
it's weird cos the description is am-fu*king-biguous at best
seems s little odd...
search appendix p, there are refs to pragma, but not in the actual grammar rules
there is a grammar for pragma, but it's not used
see AA-2-8.html 5 for the places where they are allowed
nerdboy: What kinds of metrics does cccc calculate that gnat metric doesn't?
darkestkhan1: I'm pretty sure that GNAT feature (program consisting only of packages) is an allowed extension.
sparre: you tell me?
see above for link to html output
if the Ada config works cccc covers 4 languages up to a point...
*circa 2005-ish when the author stopped updating
(Action) still growing old testing webkit arm/graphics patches
i *think* i have most of it if this builds, otherwise next one forces accel_2d_canvas and a couple other "adjustments"
if this ever works, even just on K1 i'll be impressed...
gnat metric only produces plain text reports, so it isn't quite as fancy as cccc in that respect.
« prev 1 2 3 4 5 6 7 8 9 10 11 next »