#336 URL parsing problem when the path contains "_-"

[lazka]

I was redirected here from https://github.com/sphinx-doc/sphinx/issues/4350

Given the following reST code:

http://example.com/a_-b

converting it to html results in a warning and the URL isn't converted to a link. The docs state that hyperlinks can't end with a "_", but this isn't the case here.

[jfbu]

added note:

This link_ refers to a site whose URL uses an underscore.

.. _link: http://example.com/a_-b

does work ok, hence docs is fine for this "linkblock". But when handling Sphinx issue I did not find immediately docs for "inline" URI, not inside a linkblock.

[jfbu]

ah, ok, docs is here and it says

Punctuation at the end of a URI is not considered part of the URI, unless the URI is terminated by a closing angle bracket (">"). Backslashes may be used in URIs to escape markup characters, specifically asterisks ("*") and underscores ("_") which are vaid URI characters

but it is not clear to me if that says that escaping is necessary if URI contains a non-letter after the underscore (as is possibly the case as reported here).

[lazka]

It seemed weird to me that "a_b" and "a-b" works, but "a_-b" doesn't.

[Günter Milde]

Yes, and this is regular and documented behaviour:

• a_ is link syntax (trailing backslash)
• inline syntax is regognized also before punctuation and some other characters (see http://docutils.sourceforge.net/docs/ref/rst/restructuredtext.html#inline-markup-recognition-rules)
• links are recognized before standalone hyperlinks (see http://docutils.sourceforge.net/docs/ref/rst/restructuredtext.html#recognition-order)

this http://example.com/a_-b fails

this <http://example.com/a_-b> fails

this `<http://example.com/a_-b>`__ works

this http://example.com/a\_-b works, too.

[lazka]

Ah, I see. Thanks!