Closed ddauber closed 6 months ago
Unfortunately there's a lot of cases like this (where 2 characters are almost or entirely indistinguishable to the human eye but they're technically different Unicode characters), I think making an exception just for "fancy apostrophe" / "straight-down apostrophe" wouldn't make sense.
However I think it would be nice in the future to have some sort of "autolinking" (think https://www.dbpedia-spotlight.org/), where e.g. Athens would suggest to you "hey, you said John asked me yesterday to buy groceries
, did you mean to link John
-> John Doe
, yesterday
-> 2021-04-27
, buy groceries
-> Buy groceries
?". For that, it would definitely make sense to do some normalization / loose parsing.
Problem
Athens treats the following as separate phrases (as it probably should, given that it differs in one character):
The problem emerges if one copies text as a direct quote from other sources, even though the same meaning is implied, it would not show as a suggested link. The reason for that seems to be that
'
is treated differently from’
. Thus, it does lead to duplication of pages or even not recognising that there are two versions of the same text.I am aware this is not a major bug, but certainly an annoyance. It might be worth to just treat both as the same character input. The work-around would be to not use it all.
Screenshots/Demo
Athens Version 1.0.0-BETA.40