Much better unicode support
[frenchie/icalparse.git] / icalparse.py
index edfca4c..6c3aac6 100755 (executable)
@@ -28,8 +28,18 @@ import os
 class InvalidICS(Exception): pass
 class IncompleteICS(InvalidICS): pass
 
-def lineJoiner(oldcal):
+def lineJoiner(oldcal, encoding='utf-8'):
        '''Takes a string containing a calendar and returns an array of its lines'''
+       try:
+               oldcal = unicode(oldcal, encoding)
+               oldcal = oldcal.replace('\r\n ', '').replace('\r\n\t','')
+       except UnicodeDecodeError:
+               # This is probably a file with badly folded lines
+               oldcal = oldcal.replace('\r\n ', '').replace('\r\n\t','')
+               try: oldcal = unicode(oldcal, encoding)
+               except UnicodeDecodeError:
+                       sys.stderr.write('Malformed File')
+                       raise
 
        if not oldcal[0:15] == 'BEGIN:VCALENDAR':
                raise InvalidICS, "Does not appear to be a valid ICS file"
@@ -40,8 +50,7 @@ def lineJoiner(oldcal):
        if list(oldcal) == oldcal:
                oldcal = '\r\n'.join(oldcal)
 
-       oldcal = oldcal.replace('\r\n ', '').replace('\r\n\t','')
-       return [unicode(x, 'utf-8') for x in oldcal.strip().split('\r\n')]
+       return oldcal.split('\r\n')
 
 
 def lineFolder(oldcal, length=75):
@@ -53,33 +62,42 @@ def lineFolder(oldcal, length=75):
        cal = []
        sl = length - 1
 
-       for line in oldcal:
-               line = line.encode('utf-8')
+       for uline in oldcal:
+               line = uline.encode('utf-8')
+
                # Line fits inside length, do nothing
-               if len(line.rstrip()) <= length:
+               if len(line) <= length:
                        cal.append(line)
+
                else:
-                       brokenline = [line[0:length]]
                        ll = length
-                       while ll < len(line) + 1:
-                               brokenline.append(line[ll:sl+ll])
-                               ll += sl
-                       brokenline = '\r\n '.join(brokenline)
-                       cal.append(brokenline)
+                       foldedline = []
+                       while uline:
+                               ufold = unicode(line[0:ll], 'utf-8', 'ignore')
+                               fold = ufold.encode('utf-8')
+                               uline = uline.replace(ufold,u'',1)
+                               line = uline.encode('utf-8')
+                               foldedline.append(fold)
+
+                               # Subsequent lines are shorter as they include a space
+                               ll = length - 1
+                       cal.append('\r\n '.join(foldedline))
 
        return cal
 
 
 def splitFields(cal):
-       '''Takes a list of lines in a calendar file and returns a list of tuples        
+       '''Takes a list of lines in a calendar file and returns a list of tuples
        as (key, value) pairs'''
 
-       ical = [tuple(x.split(':',1)) for x in cal]
+       ical = []
 
        # Check that we got 2 items on every line
-       for line in ical:
-               if not len(line) == 2:
-                       raise InvalidICS, "Didn't find a content key on: %s"%(line)
+       for line in [tuple(x.split(':',1)) for x in cal]:
+               if not len(line) == 2 and line[0]:
+                       raise InvalidICS, 'Unusual content line: %s'%line
+               elif line[0]:
+                       ical.append(line)
 
        return ical
 
@@ -96,6 +114,8 @@ def getContent(url='',stdin=False):
        it can read from the local disk (which you probably don't want it to).
        '''
 
+       encoding = '' # If we don't populate this, the script will assume UTF-8
+
        # Special case, if this is a HTTP url, return the data from it using
        # the HTTP functions which attempt to play a bit nicer.
        parsedURL = urlparse.urlparse(url)
@@ -103,14 +123,14 @@ def getContent(url='',stdin=False):
 
        if stdin:
                content = sys.stdin.read()
-               return content
+               return (content, encoding)
 
        if not parsedURL[0]:
                try: content = open(os.path.abspath(url),'r').read()
                except (IOError, OSError), e:
                        sys.stderr.write('%s\n'%e)
                        sys.exit(1)
-               return content
+               return (content, encoding)
 
        # If we've survived, use python's generic URL opening library to handle it
        import urllib2
@@ -121,7 +141,7 @@ def getContent(url='',stdin=False):
        except (urllib2.URLError, OSError), e:
                sys.stderr.write('%s\n'%e)
                sys.exit(1)
-       return content
+       return (content, encoding)
 
 
 def getHTTPContent(url='',cache='.httplib2-cache'):
@@ -135,26 +155,35 @@ def getHTTPContent(url='',cache='.httplib2-cache'):
 
        if not url: return ''
 
+       encoding = '' # If we don't populate this, the script will assume UTF-8
+
        if 'httplib2' in sys.modules:
                try: h = httplib2.Http('.httplib2-cache')
                except OSError: h = httplib2.Http()
        else: h = False
 
        try:
-               if h: content = h.request(url)[1]
-               return content
+               if h:
+                       req = h.request(url)
+                       content = req[1]
+                       if 'content-type' in req[0]:
+                               for ct in req[0]['content-type'].split(';'):
+                                       ct = ct.lower()
+                                       if 'charset' in ct:
+                                               encoding = ct.split('=')[1].strip()
+               return (content, encoding)
        except ValueError, e:
                sys.stderr.write('%s\n'%e)
                sys.exit(1)
 
        try:
                content = urllib2.urlopen(url).read()
-               return content
+               return (content, encoding)
        except (urllib2.URLError, OSError), e:
                sys.stderr.write('%s\n'%e)
                sys.exit(1)
 
-       return ''
+       return ('', '')
 
 
 def generateRules():
@@ -237,6 +266,9 @@ if __name__ == '__main__':
                default=False, help='Be verbose when rules are being applied')
        parser.add_option('-o', '--output', dest='outfile', default='',
                help='Specify output file (defaults to standard output)')
+       parser.add_option('-m','--encoding', dest='encoding', default='',
+               help='Specify a different character encoding'
+               '(ignored if the remote server also specifies one)')
 
        (options, args) = parser.parse_args()
 
@@ -248,8 +280,9 @@ if __name__ == '__main__':
        else:
                url = ''
 
-       content = getContent(url, options.stdin)
-       cal = lineJoiner(content)
+       (content, encoding) = getContent(url, options.stdin)
+       encoding = encoding or options.encoding or 'utf-8'
+       cal = lineJoiner(content, encoding)
        ical = applyRules(splitFields(cal), generateRules(), options.verbose)
        output = lineFolder(joinFields(ical))
        writeOutput(output, options.outfile)

UCC git Repository :: git.ucc.asn.au