Fixed the decoding of strings (Fixes issue 2).

It should now (hopefully) behave the same in all Python versions 2.6+.

blame.py

@@ -91,13 +91,14 @@ class Blame:
 		lines = ls_tree_r.readlines()
 
 		for i, row in enumerate(lines):
-			row = row.decode("utf-8", "replace").strip().strip("\"").strip("'")
-			row = codecs.getdecoder('unicode_escape')(row.strip())[0]
+			row = codecs.getdecoder("unicode_escape")(row.strip())[0]
+			row = row.encode("latin-1", "replace")
+			row = row.decode("utf-8", "replace").strip("\"").strip("'").strip()
 
 			if FileDiff.is_valid_extension(row) and not filtering.set_filtered(FileDiff.get_filename(row)):
-				if not missing.add(row.strip()):
+				if not missing.add(row):
 					blame_string = "git blame -w {0} ".format("-C -C -M" if hard else "") + \
-					               interval.get_since() + interval.get_ref() + " -- \"" + row.strip() + "\""
+					               interval.get_since() + interval.get_ref() + " -- \"" + row + "\""
 					thread = BlameThread(blame_string, FileDiff.get_extension(row), self.blames, row.strip())
 					thread.daemon = True
 					thread.start()

changes.py

@@ -23,7 +23,6 @@ import codecs
 import extensions
 import filtering
 import interval
-import re
 import os
 import subprocess
 import terminal
@@ -45,18 +44,12 @@ class FileDiff:
 
 	@staticmethod
 	def get_extension(string):
-		string = string.encode("utf-8", "replace")
-		string = string.decode("utf-8", "replace")
 		string = string.split("|")[0].strip().strip("{}").strip("\"").strip("'")
-		string = codecs.getdecoder('unicode_escape')(string.strip())[0]
 		return os.path.splitext(string)[1][1:]
 
 	@staticmethod
 	def get_filename(string):
-		string = string.encode("utf-8", "replace")
-		string = string.decode("utf-8", "replace")
-		string = string.split("|")[0].strip().strip("{}").strip("\"").strip("'")
-		return codecs.getdecoder('unicode_escape')(string.strip())[0]
+		return string.split("|")[0].strip().strip("{}").strip("\"").strip("'")
 
 	@staticmethod
 	def is_valid_extension(string):
@@ -75,7 +68,7 @@ class Commit:
 		if commit_line.__len__() == 4:
 			self.date = commit_line[0]
 			self.sha = commit_line[1]
-			self.author = re.sub("[^\w ]", "", commit_line[2].strip())
+			self.author = commit_line[2].strip()
 			self.message = commit_line[3].strip()
 
 	def add_filediff(self, filediff):
@@ -105,7 +98,10 @@ class Changes:
 		lines = git_log_r.readlines()
 
 		for i in lines:
+			i = codecs.getdecoder("unicode_escape")(i.strip())[0]
+			i = i.encode("latin-1", "replace")
 			i = i.decode("utf-8", "replace")
+
 			if Commit.is_commit_line(i) or i == lines[-1]:
 				if found_valid_extension:
 					self.commits.append(commit)

metrics.py

@@ -20,6 +20,7 @@
 from __future__ import print_function
 from outputable import Outputable
 from changes import FileDiff
+import codecs
 import comment
 import filtering
 import missing
@@ -34,9 +35,12 @@ class MetricsLogic:
 		ls_tree_r = subprocess.Popen("git ls-tree --name-only -r HEAD", shell=True, bufsize=1, stdout=subprocess.PIPE).stdout
 
 		for i in ls_tree_r.readlines():
-			i = i.decode("utf-8", "replace")
+			i = codecs.getdecoder("unicode_escape")(i.strip())[0]
+			i = i.encode("latin-1", "replace")
+			i = i.decode("utf-8", "replace").strip("\"").strip("'").strip()
+
 			if FileDiff.is_valid_extension(i) and not filtering.set_filtered(FileDiff.get_filename(i)):
-				if not missing.add(i.strip()):
+				if not missing.add(i):
 					file_r = open(i.strip(), "rb")
 					extension = FileDiff.get_extension(i)
 					lines = MetricsLogic.get_eloc(file_r, extension)