split lines now works with unicode. Fixes bug #52

4 files changed, 48 insertions, 37 deletions

diff --git a/HISTORY b/HISTORY
index 35b144f..2078c01 100644
--- a/HISTORY
+++ b/HISTORY
@@ -388,3 +388,7 @@ Balke.
 - Lyrics: rewrite the shell script muggle_getlyrics in python and talk 
   directly to the Googlyrics2 python code
 - load images: fix buggy error handling (reported in vdr-portal.de by stevie101)
+- Lyrics: Fix bug #52: line split could split within 2byte unicode char. Now
+  removed this code from C++ and put it into the python script muggle_getlyrics.
+- Lyrics: when calling "mv" to rename a file make sure mv never gets interactive
+- muggle-image-convert produces less output if needed programs are not installed
diff --git a/README b/README
index 966a721..ad64fba 100644
--- a/README
+++ b/README
@@ -340,10 +340,15 @@ Instead it waits until the script has finished. Meanwhile it will not
 respond to commands, and the message "loading lyrics" will only appear
 after they have been loaded.
 
-Muggle calls the script muggle_getlyrics which in turn calls 
-mgLyrics.py which in turn uses the Googlyrics2 python modules.
+Muggle calls the python program muggle_getlyrics which 
+uses the Googlyrics2 python modules.
 Googlyrics2 is a plugin to the amarok music player, it can currently ask
-26 different sources.
+27 different sources.
+
+If the lyrics lines are too long, they are split by muggle_getlyrics. If
+those lines are too short or too long, adapt MAXLINELENGTH in muggle_getlyrics.
+Currently lines are split at normal space and at nbsp (unicode non breaking
+space). Lines that have no such spaces are not split.
 
 
 
diff --git a/mg_menu.c b/mg_menu.c
index d6ed7c3..2fbe67d 100644
--- a/mg_menu.c
+++ b/mg_menu.c
@@ -255,35 +255,23 @@ mgOsd::AddText(const char* text,bool selectable) {
 }
 
 void
-mgOsd::AddLongText(const char* text) {
-	char *s=strdup(text);
-	char *p=s;
-	unsigned int brk=57;
-	bool selectable = strlen(s);
-	while (strlen(p)>=brk) {
-		char save=p[brk];
-		p[brk]=0;
-		AddText(p,selectable);
-		p[brk]=save;
-		p+=brk;
-	}
-	if (strlen(p))
-		AddText(p,selectable);
-	free(s);
-}
-
-void
 mgOsd::AddFile(const string filename) {
-	ifstream infile(filename.c_str());
-	string line;
-	if(infile) {
-		while(getline(infile, line , '\n'))
-			AddLongText(line.c_str());
-	}
-	else {
+	FILE *fp = fopen(filename.c_str(),"r");
+	if (fp) {
+		char* line = NULL;
+		size_t len = 0;
+		while (getline(&line, &len , fp) != -1) {
+			char *lf = strrchr(line,'\n');
+			if (lf) *lf = 0;
+			AddText(line,strlen(line));
+		}
+		if (line)
+			free(line);
+		fclose(fp);
+	} else {
 		char *s;
 		msprintf(&s,tr("File %s not found"),filename.c_str());
-		AddLongText(s);
+		AddText(s);
 		free(s);
 	}
 }
diff --git a/scripts/muggle_getlyrics b/scripts/muggle_getlyrics
index f0d5701..b187487 100755
--- a/scripts/muggle_getlyrics
+++ b/scripts/muggle_getlyrics
@@ -6,6 +6,28 @@ import os, sys, locale, re, codecs
 import filecmp
 from htmlentitydefs import name2codepoint as n2cp
 
+MAXLINELENGTH = 50
+
+def writeFile(name,s):
+	if not os.path.isdir(outdir):
+		os.mkdir(outdir)
+	outname = outdir + '/' + name
+	outfile = open(outname,"w")
+	lines = s.splitlines()
+	for line in lines:
+# xa0 is unicode non breaking space
+		words = re.split(r' |\xa0',line)
+		widx1 = 0
+		while widx1 < len(words):
+			widx2 = len(words) 
+			while widx2 > widx1 + 1 and len(' '.join(words[widx1:widx2])) > MAXLINELENGTH:
+				widx2 -= 1
+			newline = ' '.join(words[widx1:widx2])+'\n'
+			newline = newline.encode(locale.getdefaultlocale()[1])
+			outfile.write(newline)
+			widx1 = widx2		
+	outfile.close
+	
 charset = locale.getdefaultlocale()[1]
 title = sys.argv[1].decode(charset).encode('UTF8')
 artist = sys.argv[2].decode(charset).encode('UTF8')
@@ -52,13 +74,6 @@ def countFiles():
 		return 0
 	return len(os.listdir(outdir))
 
-def writeFile(name,s):
-	if not os.path.isdir(outdir):
-		os.mkdir(outdir)
-	outfile = open(outdir + '/' + name,"w")
-	outfile.write(s)
-	outfile.close
-	
 def load(debug=False):
 	if debug:
 		outlyric=["Version 1","Version 2","Version 3"]
@@ -98,7 +113,6 @@ def load(debug=False):
 # private use, but still some sites use them..
 			s = s.replace(r'\xc2\x91',r'\x27')
 			s = s.replace(r'\xc2\x92',r'\x27')
-			s = s.encode(locale.getdefaultlocale()[1])
 			writeFile(str(idx)+'-'+l.sitename,s)
 
 # -----------------------------------------------