Unexpected behaviour with pandoc-server

Unexpected behaviour with pandoc-server

[mails.lists.2012-1-rRloVJBGZzogAv4oPwG0Al6hYfS7NtTn]

Hi all,


I tried out pandoc 3.1.2 server, installed as a CGI script for Apache on 
Debian 11 "Bullseye" x64.

(I first planned to run pandoc-server in actual server mode to avoid the 
CGI start overhead, but could not figure out how to configure it to 
listen to localhost only - is this possible?)


I used it for Jira Markup -> gfm conversion, and it worked quite fine, 
except for one specific jira rich text content I encountered.

This content formats just fine without any noticeable delay with

pandoc -f jira -t gfm --sandbox --wrap=preserve < in.jira > out.md

on my local machine.

The server with its default settings first (reproducibly) returned 503 
errors for this content, though.

When I increased the timeout to 10 seconds, a successful reply was 
returned - after a considerable delay - but the text was basically 
unconverted. Markdown specific special chars were escaped actually, but 
Jira formatting like {code} was left unmodified in the text.

This completely differs from CLI pandoc invocation, which properly 
converts these formattings immediately, without any noticable delay.


Is this to be expected and what's the background?

Or am I encountering a possible bug I should try to analyze further (at 
least the behaviour, probably not the pandoc code...)?

Unfortunately, I cannot simply share the Jira snipped here, so if it's 
of interest I need to try to condense it into a minimal test case.


Of course I may also be doing wrong when sending the request (I'm 
sending all arguments as JSON as documented, and only accept JSON for a 
reply - this works fine in all other cases I tested, though, and 
conversion is much quicker than for this single "strange" Jira field).


My intermediate solution was to just build a minimal CGI shell script 
which in turn just uses pandoc in command line pipe mode - this also 
works fine and fast, so it's unrelated to the server machine and the 
pandoc binary installed on it.


Best regards,

   Gunter

-- 
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 view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/a84dc893-f23f-8468-5d3a-4af15d73ac2c%40ohrner.net.

Re: Unexpected behaviour with pandoc-server

[Albert Krewinkel]

mails.lists.2012-1-rRloVJBGZzogAv4oPwG0Al6hYfS7NtTn@public.gmane.org writes:

> I used it for Jira Markup -> gfm conversion, and it worked quite fine,
> except for one specific jira rich text content I encountered.

This would be interesting to check. Could this be an instance of issue
#15/#18 reported at <https://github.com/tarleb/jira-wiki-markup/issues>?

> Or am I encountering a possible bug I should try to analyze further
> (at least the behaviour, probably not the pandoc code...)?

Most likely not expected; additional info would be great. The difference
in normal and CGI mode is most strange.

Thanks,


-- 
Albert Krewinkel
GPG: 8eed e3e2 e8c5 6f18 81fe  e836 388d c0b2 1f63 1124