From 815944ca6f6812795d510b0d7c6a5ecca853d183 Mon Sep 17 00:00:00 2001
From: spl0k <alban.feron@gmail.com>
Date: Mon, 9 Dec 2013 19:06:14 +0100
Subject: [PATCH] Some work on encodings

---
 api/__init__.py | 14 ++++++++++++--
 api/media.py    | 32 +++++++++++++++++++++++++++++---
 2 files changed, 41 insertions(+), 5 deletions(-)

diff --git a/api/__init__.py b/api/__init__.py
index f555e2f..f6a90da 100755
--- a/api/__init__.py
+++ b/api/__init__.py
@@ -129,15 +129,17 @@ class ResponseHelper:
 				"""
 		if not isinstance(dictionary, dict):
 			raise TypeError('Expecting a dict')
+		if not all(map(lambda x: isinstance(x, basestring), dictionary.keys())):
+			raise TypeError('Dictionary keys must be strings')
 
 		subelems =   { k: v for k, v in dictionary.iteritems() if isinstance(v, dict) }
 		sequences =  { k: v for k, v in dictionary.iteritems() if isinstance(v, list) }
 		attributes = { k: v for k, v in dictionary.iteritems() if k != '_value_' and k not in subelems and k not in sequences }
 
 		if '_value_' in dictionary:
-			elem.text = dictionary['_value_']
+			elem.text = ResponseHelper.value_tostring(dictionary['_value_'])
 		for attr, value in attributes.iteritems():
-			elem.set(attr, str(value).lower() if isinstance(value, bool) else str(value))
+			elem.set(attr, ResponseHelper.value_tostring(value))
 		for sub, subdict in subelems.iteritems():
 			subelem = ElementTree.SubElement(elem, sub)
 			ResponseHelper.dict2xml(subelem, subdict)
@@ -146,6 +148,14 @@ class ResponseHelper:
 				subelem = ElementTree.SubElement(elem, seq)
 				ResponseHelper.dict2xml(subelem, subdict)
 
+	@staticmethod
+	def value_tostring(value):
+		if isinstance(value, basestring):
+			return value
+		if isinstance(value, bool):
+			return str(value).lower()
+		return str(value)
+
 def get_entity(req, ent, param = 'id'):
 	eid = req.args.get(param)
 	if not eid:
diff --git a/api/media.py b/api/media.py
index 40fb73a..9b3caf4 100755
--- a/api/media.py
+++ b/api/media.py
@@ -4,6 +4,7 @@ from flask import request, send_file, Response
 import os.path
 from PIL import Image
 import subprocess
+import codecs
 
 import config, scanner
 from web import app
@@ -145,9 +146,16 @@ def lyrics():
 	for track in query:
 		lyrics_path = os.path.splitext(track.path)[0] + '.txt'
 		if os.path.exists(lyrics_path):
-			print lyrics_path
-			with open(lyrics_path, 'r') as lyrics_file:
-				lyrics = lyrics_file.read()
+			app.logger.debug('Found lyrics file: ' + lyrics_path)
+
+			try:
+				lyrics = read_file_as_unicode(lyrics_path)
+			except UnicodeError:
+				# Lyrics file couldn't be decoded. Rather than displaying an error, try with the potential next files or
+				# return no lyrics. Log it anyway.
+				app.logger.warn('Unsupported encoding for lyrics file ' + lyrics_path)
+				continue
+
 			return request.formatter({ 'lyrics': {
 				'artist': track.album.artist.name,
 				'title': track.title,
@@ -156,3 +164,21 @@ def lyrics():
 
 	return request.formatter({ 'lyrics': {} })
 
+def read_file_as_unicode(path):
+	""" Opens a file trying with different encodings and returns the contents as a unicode string """
+
+	encodings = [ 'utf-8', 'latin1' ] # Should be extended to support more encodings
+
+	for enc in encodings:
+		try:
+			contents = codecs.open(path, 'r', encoding = enc).read()
+			app.logger.debug('Read file {} with {} encoding'.format(path, enc))
+			# Maybe save the encoding somewhere to prevent going through this loop each time for the same file
+			return contents
+		except UnicodeError:
+			pass
+
+	# Fallback to ASCII
+	app.logger.debug('Reading file {} with ascii encoding'.format(path))
+	return unicode(open(path, 'r').read())
+