W3C home > Mailing lists > Public > public-xml-processing-model-wg@w3.org > January 2012

Re: p:exec and binary result

From: Norman Walsh <ndw@nwalsh.com>
Date: Mon, 09 Jan 2012 16:51:05 -0800
To: public-xml-processing-model-wg@w3.org
Message-ID: <m2hb04z8g6.fsf@nwalsh.com>
"vojtech.toman@emc.com" <vojtech.toman@emc.com> writes:
> The specification of p:exec says that:
> "The standard output of the command is read and returned on result;
> the standard error output is read and returned on errors. In order to
> assure that the result will be an XML document, each of the results
> will be wrapped in a c:result element."
> What if the command produces binary (non-text) data on the standard output?

There's nothing comparable to a media type for the results that appear
From p:exec, so without an extra flag somewhere, I don't think p:exec
can do anything useful. (A quick peek at XML Calabash reveals that it
treats anything that isn't XML as if it was text and attempts to
encode the bytes.)

I suppose that with a set of "*-as-binary" attributes, the results
could be base64 encoded (and the input(s) base64 decoded, I suppose).

My two cents:

1. I think the step is implicitly geared towards text outputs, not binary.
   We don't say anything about binary and "wrap-result-lines" suggests pretty
   strongly that we didn't think about it.
2. Even if we had a good story on binary, you'd need some sort of a flag
   to tell the step to treat the result as binary.
3. I hope that whatever our story is, it helps here.
4. I think it would be reasonable to add an extension attribute to specify
   binary output and base64 encode the results in the meantime.

                                        Be seeing you,

Norman Walsh
Lead Engineer
MarkLogic Corporation
Phone: +1 413 624 6676

Received on Tuesday, 10 January 2012 00:51:36 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:32:50 UTC