From: Stefan R. <lis...@st...> - 2008-01-31 07:11:38
|
Hi, I have a suggestion for a small grammar change. In my experience, it would apply in two cases that show up often:: - a list - that is - not yet finished - or a table with +-* entries:: ========= ======= ==== ====== propert's martian mary marvin --------- ------- ---- ------ blue - + + green + - - sad - - + crazy + - ? ========= ======= ==== ====== In both cases the lone entries are parsed as list items. In the first case, that doesn't matter, although it might be beneficial if the output would give an indication that there is something missing. In the second case, you have to escape all the entries. I think we could simply change the grammar to detect list items only if the bullet character is followed by at least one space. Since trailing spaces are removed from paragraphs during parsing (is that correct?), this would be enough. I quickly tried the following change in parsers.rst.states (line 1051):: - 'bullet': ur'[-+*\u2022\u2023\u2043]( +|$)', + 'bullet': ur'[-+*\u2022\u2023\u2043] ( *|$)', and it seems to work. (seeing it now, removing the eol $ would probably be better...) What do you think? cheers, stefan |