Feature: a option for converting into ASCII before transforming. Patch by rgiot.

UltiSnips/tex.snippets

@@ -58,7 +58,7 @@ endsnippet
 
 snippet part "Part" b
 \part{${1:part name}}
-\label{prt:${2:${1/(\w+)|\W+/(?1:\L$0\E:_)/g}}}
+\label{prt:${2:${1/(\w+)|\W+/(?1:\L$0\E:_)/ga}}}
 
 ${0}
 
@@ -67,7 +67,7 @@ endsnippet
 
 snippet cha "Chapter" b
 \chapter{${1:chapter name}}
-\label{cha:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/g}}}
+\label{cha:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/ga}}}
 
 ${0}
 
@@ -76,7 +76,7 @@ endsnippet
 
 snippet sec "Section" b
 \section{${1:section name}}
-\label{sec:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/g}}}
+\label{sec:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/ga}}}
 
 ${0}
 
@@ -86,7 +86,7 @@ endsnippet
 
 snippet sub "Subsection" b
 \subsection{${1:subsection name}}
-\label{sub:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/g}}}
+\label{sub:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/ga}}}
 
 ${0}
 
@@ -95,7 +95,7 @@ endsnippet
 
 snippet ssub "Subsubsection" b
 \subsubsection{${1:subsubsection name}}
-\label{ssub:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/g}}}
+\label{ssub:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/ga}}}
 
 ${0}
 
@@ -105,7 +105,7 @@ endsnippet
 
 snippet par "Paragraph" b
 \paragraph{${1:paragraph name}}
-\label{par:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/g}}}
+\label{par:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/ga}}}
 
 ${0}
 
@@ -115,7 +115,7 @@ endsnippet
 
 snippet subp "Subparagraph" b
 \subparagraph{${1:subparagraph name}}
-\label{par:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/g}}}
+\label{par:${2:${1/\\\w+\{(.*?)\}|\\(.)|(\w+)|([^\w\\]+)/(?4:_:\L$1$2$3\E)/ga}}}
 
 ${0}
 

doc/UltiSnips.txt

@@ -1155,6 +1155,11 @@ The options can be any combination of >
    i    - case insensitive
           By default, regular expression matching is case sensitive. With this
           option, matching is done without regard to case.
+   a    - ascii conversion
+          By default, transformation are made on the raw utf-8 string. With
+          this option, matching is done on the corresponding ASCII string
+          instead, for example 'à' will become 'a'.
+          This option required the python package 'unidecode'.
 
 The syntax of regular expressions is beyond the scope of this document. Python
 regular expressions are used internally, so the python 're' module can be used
@@ -1214,6 +1219,18 @@ this is a title
 This Is A Title
 
 
+Demo: ASCII transformation
+------------------- SNIP -------------------
+snippet ascii "Replace non ascii chars"
+${1: an accentued text}
+${1/.*/$0/a}
+endsnippet
+------------------- SNAP -------------------
+ascii<tab>à la pêche aux moules
+à la pêche aux moules
+a la peche aux moules
+
+
 Demo: Regular expression grouping
       This is a clever c-like printf snippet, the second tabstop is only shown
       when there is a format (%) character in the first tabstop.
@@ -1372,6 +1389,7 @@ Contributors listed in chronological order:
    Zhao Cai - zhaocai
    John Szakmeister - jszakmeister
    Jonas Diemer - diemer
+   Romain Giot - rgiot
 
 
 8.2 Snippets                                       *UltiSnips-contrisnippets*

plugin/UltiSnips/text_objects/_transformation.py

@@ -2,9 +2,12 @@
 # encoding: utf-8
 
 import re
-
+import sys
 from UltiSnips.text_objects._mirror import Mirror
 
+# flag used to display only one time the lack of unidecode
+UNIDECODE_ALERT_RAISED = False
+
 class _CleverReplace(object):
     """
     This class mimics TextMates replace syntax
@@ -102,6 +105,8 @@ class _CleverReplace(object):
 
 class TextObjectTransformation(object):
     def __init__(self, token):
+        self._convert_to_ascii = False
+
         self._find = None
         if token.search is None:
             return
@@ -113,11 +118,22 @@ class TextObjectTransformation(object):
                 self._match_this_many = 0
             if "i" in token.options:
                 flags |= re.IGNORECASE
+            if "a" in token.options:
+                self._convert_to_ascii = True
 
         self._find = re.compile(token.search, flags | re.DOTALL)
         self._replace = _CleverReplace(token.replace)
 
     def _transform(self, text):
+        global UNIDECODE_ALERT_RAISED
+        if self._convert_to_ascii:
+            try:
+                import unidecode
+                text = unidecode.unidecode(text)
+            except Exception, e:
+                if UNIDECODE_ALERT_RAISED == False:
+                    UNIDECODE_ALERT_RAISED = True
+                    sys.stderr.write("Please install unidecode python package in order to be able to make ascii conversions.\n")
         if self._find is None:
             return text
         return self._find.subn(self._replace.replace, text, self._match_this_many)[0]
@@ -129,5 +145,3 @@ class Transformation(Mirror, TextObjectTransformation):
 
     def _get_text(self):
         return self._transform(self._ts.current_text)
-
-

test.py

@@ -40,6 +40,12 @@ import subprocess
 
 from textwrap import dedent
 
+try:
+    import unidecode
+    UNIDECODE_IMPORTED = True
+except ImportError:
+    UNIDECODE_IMPORTED = False
+
 # Some constants for better reading
 BS = '\x7f'
 ESC = '\x1b'
@@ -1473,6 +1479,16 @@ class Transformation_CleverTransformLongLower_ExceptCorrectResult(_VimTest):
     keys = "test" + EX + "HALLO"
     wanted = "HALLO hallo"
 
+if UNIDECODE_IMPORTED:
+    class Transformation_SimpleCaseAsciiResult(_VimTest):
+        snippets = ("ascii", "$1 ${1/(.*)/$1/a}")
+        keys = "ascii" + EX + "éèàçôïÉÈÀÇÔÏ€"
+        wanted = "éèàçôïÉÈÀÇÔÏ€ eeacoiEEACOIEU"
+    class Transformation_LowerCaseAsciiResult(_VimTest):
+        snippets = ("ascii", "$1 ${1/(.*)/\L$1\E/a}")
+        keys = "ascii" + EX + "éèàçôïÉÈÀÇÔÏ€"
+        wanted = "éèàçôïÉÈÀÇÔÏ€ eeacoieeacoieu"
+
 class Transformation_ConditionalInsertionSimple_ExceptCorrectResult(_VimTest):
     snippets = ("test", "$1 ${1/(^a).*/(?0:began with an a)/}")
     keys = "test" + EX + "a some more text"