blob: 7d7bdfc51ab576551ad3524a3005527807f64ea6 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
I have a need to use git-annex from a larger program. It'd be great if the information output by some of the commands that is descriptive (for example, whereis) could be sent to stdout in a machine-readable format like (preferably) JSON, or XML. That way I can simply read in the output of the command and use the data directly instead of having to parse it via regexes or other such string manipulation.
This could perhaps be triggered by a --json or --xml flag to the relevant commands.
> This is [[done]], --json is supported by all commands, more or less.
>
> Caveats:
>
> * the version, status, and find commands produce custom output and so
> no json. This could change for version and status; find needs to just
> be a simple list of files, I think
> * The "note" fields may repeat multiple times per object with different
> notes and are of course not machine readable, and subject to change.
> * Output of helper commands like rsync is not diverted away, and
> could clutter up the json output badly. Should only affect commands
> that transfer data. And AFAICS, wget and rsync both output their
> progress displays to stderr, so shouldn't be a problem.
>
> --[[Joey]]
|
