deac/instiki
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Merge branch 'bzr/golem' of /Users/distler/Sites/code/instiki

Browse source
This commit is contained in:
Jacques Distler 2009-12-05 10:53:19 -06:00
parent 0bc05f11fd 171c12d2c1
commit 36bf257d72
1 changed files with 23 additions and 10 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

21
lib/stringsupport.rb
View file

@ -6,20 +6,30 @@ class String
#
# Under 1.8, this is a NOOP. Under 1.9, it sets the encoding to "ASCII-8BIT"
#--
if "".respond_to?(:force_encoding)
def as_bytes
force_encoding("ASCII-8BIT")
end
else
def as_bytes
force_encoding("ASCII-8BIT") if self.respond_to?(:force_encoding)
self
end
end
#++
# A method to allow string-oriented operations in both Ruby 1.8 and Ruby 1.9
#
# Under 1.8, this is a NOOP. Under 1.9, it sets the encoding to "UTF-8"
#--
if "".respond_to?(:force_encoding)
def as_utf8
force_encoding("UTF-8")
end
else
def as_utf8
force_encoding("UTF-8") if self.respond_to?(:force_encoding)
self
end
end
#++
# Take a string, and remove any invalid substrings, returning a valid utf-8 string.
@ -29,11 +39,14 @@ class String
#
# returns a valid utf-8 string, purged of any subsequences of illegal bytes.
#--
if "".respond_to?(:force_encoding)
def purify
text = self.dup.check_ncrs.as_utf8
if text.respond_to?(:force_encoding)
text = check_ncrs.as_utf8
text.chars.collect{|c| c.as_bytes}.grep(UTF8_REGEX).join.as_utf8
end
else
def purify
text = check_ncrs
text.split(//u).grep(UTF8_REGEX).join
end
end