Re: Minor bug: git config ignores empty sections

W dniu 16.08.2016 o 14:36, Jeff King pisze:

What I think would be much more sane in general is to parse the whole
thing into a tree (or even a flat list of events),

List of events is cheaper on memory (though I don't think there is
an issue with size of config file), and can be turned into tree
and DOM easily.

                                                   marking each
syntactically as a section header, a key, a comment, a run of
whitespace, and so on. And then do manipulations on that tree (e.g.,
walk down to the section of interest, add a new key at the end), and
then reformat the tree back into a stream of bytes. That lets you
separate the policy logic from the parsing, and do high-level operations
that might involve multiple passes or backtracking in the tree (e.g.,
walk to the section, walk past any existing keys, walk past any comments
but _not_ past any blank lines, then add the new key).

There are other other upsides, too. For example, the current code cannot
write multiple unrelated keys in a single pass.

The downsides is that it's a complete rewrite of the code. So it's a
_lot_ more work, and it carries more risk of regression. So please don't
take this as "no, your solution isn't OK; you have to do the rewrite".
We've been living with band-aids on the config code for a decade, so I
am OK with one more.

So how 'git config --show-origin --list' works?  

Ah, I guess it prints as it goes, instead of parsing config file(s)
into stream of events, and printing from those.

-- 
Jakub Narębski