Re: [bug-gawk] A CSV Standard

[Andrew J. Schorr]

Hi David,

On Tue, Nov 18, 2014 at 10:38:10PM -0000, David Jordan wrote:
> I would be happy to volunteer to write it as I have been wanting to
> contribute to a free software project for a while and it seems a simple
> enough task (always dangerous to say).

That sounds great.  I think that the gawk glue should not be too hard.  Parsing
CSV properly, on the other hand, may pose a challenge. :-)

> Do you think it would be better off standalone or as part of gawketxtlib?  

I don't know if you have looked at gawkextlib lately, but it has been
completely restructured so that each extension is packaged separately.  So
everything in gawkextlib is essentially standalone -- users can pick and choose
which extensions they want to download and install.  I think it will benefit
the community if we collect extensions in one place so that they are easy to
find, but it is of course up to you.  We would be happy to add a CSV parsing
extension to gawkextlib.  We can also post a link in our README document if you
would prefer to host it elsewhere.

In case there are any other potential volunteers, I think we also need a JSON
input parser.  The need to parse JSON has now come up twice for me.  The first
time was in connection with trying to parse journalctl output.  The journalctl
command can produce JSON as an output format, but not XML.  And then again it
came up today in connection with trying to parse some JSON data returned by a
website.  I think it should be relatively straightforward to emulate the XML
approach with JSON instead.  One just needs a SAX-style C library to parse
JSON, such as yajl (https://lloyd.github.io/yajl/).

Regards,
Andy