vim/UltiSnips
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Evaluate word boundaries with Vim, not r'\b'

Browse Source 
Certain non ALGOL-derived languages (notably LISP derivatives) do not
share the alphanumeric + underscore definition of a word character.

Fortunately, each language FileType has its own definition of a word
character, which Vim's regex engine uses when matching against the
boundary classes \< and \>.

We change the word matching routine of 'w' snippets to use Vim's regex
engine instead of a static pattern.
This commit is contained in:
guns 2013-12-23 15:36:29 -06:00
parent 44fd0c704a
commit 7cd4434a99
2 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
doc/UltiSnips.txt
View File

@ -612,7 +612,7 @@ The options currently supported are: >
the tab trigger start matches a word boundary and the tab trigger end
matches a word boundary. In other words the tab trigger must be
preceded and followed by non-word characters. Word characters are
letters, digits and the underscore. Use this option, for example, to
defined by the 'iskeyword' setting. Use this option, for example, to
permit expansion where the tab trigger follows punctuation without
expanding suffixes of larger words.

6
plugin/UltiSnips/__init__.py
View File

@ -315,7 +315,8 @@ class Snippet(object):
if match and words_prefix:
# Require a word boundary between prefix and suffix.
boundaryChars = words_prefix[-1:] + words_suffix[:1]
match = re.match(r'.\b.', boundaryChars)
boundaryChars = boundaryChars.replace('"', '\\"')
match = _vim.eval('"%s" =~# "\\\\v.<."' % boundaryChars) != '0'
elif "i" in self._opts:
match = words.endswith(self._t)
else:
@ -350,7 +351,8 @@ class Snippet(object):
match = self._re_match(trigger)
elif "w" in self._opts:
# Trim non-empty prefix up to word boundary, if present.
words_suffix = re.sub(r'^.+\b(.+)$', r'\1', words)
qwords = words.replace('"', '\\"')
words_suffix = _vim.eval('substitute("%s", "\\\\v^.+<(.+)", "\\\\1", "")' % qwords)
match = self._t.startswith(words_suffix)
self._matched = words_suffix