|
From: Nicholas N. <n.n...@gm...> - 2023-03-06 11:03:32
|
Hi, Perl was a reasonable choice for `cg_annotate` when I first wrote it 20+ years ago. But it's unfortunate now, with Perl being (a) a pretty weird and horrible language, and (b) moribund. I'd like to rewrite it (and `cg_diff`) in Python, which will make maintenance easier. I see that we already have some Python in Valgrind: `coregrind/m_gdbserver/valgrind-monitor.py` and `coregrind/m_gdbserver/valgrind-monitor-def.py`. Therefore I don't think this should be controversial. But I might as well ask, just in case: any objections or advice? Because these are single file scripts, we avoid all the usual problems of Python packaging, and just use `cp` as the package manager :) On a related note, the `cg_annotate.in`/`cg_annotate` split is annoying. The only reason for it now is to auto-embed the version number into the script, via the configure `@VERSION@` variable, for `cg_annotate --version` output. Does anyone know of a way to achieve that without requiring configure? Thanks. Nick |