From e4a74c0dc533922ce1e7cd56dfa7ac08efc064cb Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Thu, 1 Sep 2011 16:11:21 -0400 Subject: [PATCH] close bug with some caveats --- doc/bugs/add_script-friendly_output_options.mdwn | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/doc/bugs/add_script-friendly_output_options.mdwn b/doc/bugs/add_script-friendly_output_options.mdwn index efcd54553f..08f7e6c7b8 100644 --- a/doc/bugs/add_script-friendly_output_options.mdwn +++ b/doc/bugs/add_script-friendly_output_options.mdwn @@ -1,3 +1,18 @@ 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. +> +> --[[Joey]]