Dear group,

The topic of CSV support in Pandoc has come up several times on this list, includes this thread from 2014:
https://groups.google.com/forum/#!topic/pandoc-discuss/kBdJU_JktzI <https://groups.google.com/forum/#!topic/pandoc-discuss/kBdJU_JktzI>

Since last year I work for an organisation that frequently deals with tabular data (and helped organize CSVconf earlier this month), and I have done some thinking on how CSV could fit into Pandoc. I see two important use cases:

* CSV reader that converts to tables in HTML, docx, latex, etc.
* CSV has a format to describe tables in markdown

For the first use case I wrote a hack for the Jekyll blogging platform this week that turns CSV files into markdown grid tables format that is then processed by Pandoc (https://github.com/datacite/jekyll-csvy <https://github.com/datacite/jekyll-csvy>). I would rather use Pandoc with a CSV reader, but my Haskell isn't good enough to write one. But for now I can generate blog posts directly from CSV files. Other people have done similar things with Pandoc and CSV.

For the second use case I see a clear advantage of CSV over the various attempts to format tables in markdown (simple_tables, multiline_tables, grid_tables, pipe_tables). Everyone (and many tools) understands the CSV format, and you can do most of the things with CSV that the other table formats allow (multi-column formats and column alignment are a bit trickier). This has been done before using Pandoc filters, but I think a Pandoc "csv_tables" Pandoc extension would make this easier for the casual user. Using the grid_tables example from the Pandoc documentation, this could look like this:

: Sample csv table.

,,,
Fruit,Price,Advantages
Bananas,$1.34,- built-in wrapper\n- bright color
Oranges,$2.10, - cures scurvy\n- tasty
,,,

I like three commas on a new line to indicate the start and end of a table, but that is of course open for discussion. The format is much easier to read and edit for humans compared to grid tables, the only tricky bit is maybe the \n for multiline columns. I would think we could add metadata to the fenced table blog similar to code blocks, e.g.

,,,{ #mytable .numberRows }

One challenge with CSV is that it is an ill-defined format somewhat similar to markdown before CommonMark. It may make things easier to only support a specific CSV variant (e.g. comma as separator, header required, comment lines not allowed).

Thoughts?

Best,

Martin



-- 
You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/20BF19CB-A2B0-4B19-A749-D750CDD89736%40martinfenner.org.
For more options, visit https://groups.google.com/d/optout.