ide/coq_lex.mll: restore the separate parsing of .. (fix #2578)

 Trying to have a direct rule for "..." seemed to be less hack-ish,
 but is in fact _wrong_, shame on me : it's important to have a rule
 for ".." since this can appear in a coq sentence (for instance in
 recursive notations), and hence we don't want the second dot to be
 considered as a terminator. So let's restore the previous situation,
 with a rule for ".." (ignoring them) and a rule for "."+blank.
 Then "..." is recognized as ".." + ".", which is quite ok after all.
 The fact that ".. ." is also recognized by coqide in a similar way
 isn't actually a problem: after reading this sentence, coqide
 will send it to coqtop, and it's coqtop that will tell that this
 sentence makes no sense...

git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14300 85f007b7-540e-0410-9357-904b9bb8a0f7

0 files changed, 0 insertions, 0 deletions