Hi Helmut, On 25/06/2019 19:39, Helmut Grohne wrote: > On Tue, Jun 25, 2019 at 09:43:24AM +0200, Lionel Debroux wrote: > > tfdocgen is a simple HTML documentation generator, which supports a > > Doxygen-like (but not Doxygen-compatible) syntax for C/C++ code, and > > maybe other languages which supports /** */ comments, I haven't > > tried. It was written by the previous maintainer over a decade ago, > > I have barely modified it. The files it produces are packaged in the > > -dev packages. > > Thank you for chiming in. So do you confirm that tfdocgen consumes > only textual formats (source code with comment markup) and produces > only textual formats (html) and that the behaviour and output is > independent of the processor architecture being executed on? Yes, yes and yes. In Git HEAD, tfdocgen.c is a 600+ RLOC file, which can build with GCC -std=gnu89 CFLAG (it uses C++-style // comments), and only depends on the standard library + Glib. From a fairly cursory glance, I don't see obvious portability issues in the code: no occurrence of "long" as a data type, no usage of &, |, ^, << and >> operators indicating (de)serialization and bit arithmetic.
In the 11-year-old 1.0 version packaged by Debian, this holds as well, but several minor fixes of mine, the latest one from 2015, are missing. I, uh, haven't produced a release of tfdocgen myself, so I'm the one to blame for the fact distros are using an older version :) I suppose I should produce one as part of the next release cycle, hopefully by the end of the year. I should soon be down to two significant items left on the todo list, but adding support for a 4th calc <-> computer communication protocol is nontrivial work. Bye, Lionel.