Links with `#` anchors are sometimes parsed strangely

[Lysxia]

For example https://neuron.zettel.page/2017401.html

It looks like this, where the "zettel" link is not parsed as a link:

[srid]

Regression introduced in #375

[srid]

A fix could be to parse an inline tag only if the # is preceded by whitespace (or beginning of line).

[srid]

This seems like a strange bug in the CommonMark parser. It cannot be reproduced with normal links, like: <http://google.com/#foo>. Perhaps something to do with the placement of the hyphen between the two links.

Worked around in 9a89a51

[srid]

Closing, because I strongly suspect this has to be a bug in commonmark-hs; if not, please re-open.