rst2html parsing annoyances

[gradha]

The following rst file can be parsed using nimrod rst2html command:

Nested lists
============

* One
  * A
  * `Some split hyperlink
    <http://foo>`_
  * B
* Two
  * A
  * B

Even the output looks all right. Parsing the same file with python docutils throws an error and several warnings:

test.rst:7: (ERROR/3) Unexpected indentation.
test.rst:4: (WARNING/2) Inline interpreted text or phrase reference start-string without end-string.
test.rst:8: (WARNING/2) Block quote ends without a blank line; unexpected unindent.

The error happens because the rst format seems to require a line of spacing before and after a bulleted list. The following input parses and produces the same input for both implementations:

Nested lists
============

* One

  * A
  * `Some split hyperlink
    <http://foo>`_
  * B

* Two

  * A
  * B

Not sure if Nimrod should report this, but the error of the python version also confused github's ruby rendering engine. Maybe nimrod should get a --strict switch to catch these?

[gradha]

The following example duplicates an explicit target name:

Repeated links
==============

* `CHANGES.rst <CHANGES.rst>`_
* `CHANGES.rst <CHANGES.html>`_

The python implementation nitpicks with:

test.rst:2: (WARNING/2) Duplicate explicit target name: "changes.rst".

[Araq]

a) is a feature, we improve upon reSt, b) is too minor to fix.