Tabs vs Spaces
[frenchie/icalparse.git] / icalparse.py
index 6c3aac6..59e409b 100755 (executable)
@@ -1,6 +1,6 @@
-#!/usr/bin/python
+#!/usr/bin/python3
 #
-# Copyright (c) 2010 James French <[email protected]>
+# Copyright (c) 2013 James French <[email protected]>
 #
 # Permission is hereby granted, free of charge, to any person obtaining a copy
 # of this software and associated documentation files (the "Software"), to deal
 # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 # THE SOFTWARE.
 
-import sys
-import urlparse
-import os
-
-
-class InvalidICS(Exception): pass
-class IncompleteICS(InvalidICS): pass
-
-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"
-
-       if not 'END:VCALENDAR' in oldcal[-15:-1]:
-               raise IncompleteICS, "File appears to be incomplete"
-
-       if list(oldcal) == oldcal:
-               oldcal = '\r\n'.join(oldcal)
-
-       return oldcal.split('\r\n')
-
-
-def lineFolder(oldcal, length=75):
-       '''Folds content lines to a specified length, returns a list'''
-
-       if length > 75:
-               sys.stderr.write('WARN: lines > 75 octets are not RFC compliant\n')
-
-       cal = []
-       sl = length - 1
-
-       for uline in oldcal:
-               line = uline.encode('utf-8')
-
-               # Line fits inside length, do nothing
-               if len(line) <= length:
-                       cal.append(line)
-
-               else:
-                       ll = length
-                       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
-       as (key, value) pairs'''
-
-       ical = []
-
-       # Check that we got 2 items on every 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
-
-
-def joinFields(ical):
-       '''Takes a list of tuples that make up a calendar file and returns it to a
-       list of lines'''
-
-       return [':'.join(x) for x in ical]
-
+import sys, os
+from urllib import (request, error, parse)
+import vobject
+from cgi import parse_header
 
 def getContent(url='',stdin=False):
        '''Generic content retriever, DO NOT use this function in a CGI script as
@@ -118,116 +34,107 @@ def getContent(url='',stdin=False):
 
        # 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)
+       parsedURL = parse.urlparse(url)
        if 'http' in parsedURL[0]: return getHTTPContent(url)
 
        if stdin:
                content = sys.stdin.read()
                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, encoding)
+       if not parsedURL[0]: url = 'file://' + os.path.abspath(url)
 
        # If we've survived, use python's generic URL opening library to handle it
-       import urllib2
        try:
-               res = urllib2.urlopen(url)
+               res = request.urlopen(url)
                content = res.read()
+               ct = res.info().getplist()
                res.close()
-       except (urllib2.URLError, OSError), e:
+       except (error.URLError, OSError) as e:
                sys.stderr.write('%s\n'%e)
                sys.exit(1)
+
+       for param in ct:
+               if 'charset' in param:
+                       encoding = param.split('=')[1]
+                       break
+
        return (content, encoding)
 
 
-def getHTTPContent(url='',cache='.httplib2-cache'):
+def getHTTPContent(url=''):
        '''This function attempts to play nice when retrieving content from HTTP
        services. It's what you should use in a CGI script.'''
 
        try:
                import httplib2
        except ImportError:
-               import urllib2
+               pass
 
-       if not url: return ''
+       if not url: return ('','')
 
-       encoding = '' # If we don't populate this, the script will assume UTF-8
+       if not 'http' in parse.urlparse(url)[0]: return ('','')
 
        if 'httplib2' in sys.modules:
-               try: h = httplib2.Http('.httplib2-cache')
+               try: h = httplib2.Http()
                except OSError: h = httplib2.Http()
        else: h = False
 
-       try:
-               if h:
+       if h:
+               try:
                        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)
+               except ValueError as e:
+                       sys.stderr.write('%s\n'%e)
+                       sys.exit(1)
 
-       try:
-               content = urllib2.urlopen(url).read()
-               return (content, encoding)
-       except (urllib2.URLError, OSError), e:
-               sys.stderr.write('%s\n'%e)
-               sys.exit(1)
+               resp, content = req
+               if 'content-type' in resp:
+                       ct = 'Content-Type: %s'%req[0]['content-type']
+                       ct = parse_header(ct)
+                       if 'charset' in ct[1]: encoding = ct[1]['charset']
+                       else: encoding = ''
+               else:
+                       ct = ''
+                       encoding = ''
 
-       return ('', '')
+       else:
+               try:
+                       req = request.urlopen(url)
+               except error.URLError as e:
+                       sys.stderr.write('%s\n'%e)
+                       sys.exit(1)
+
+               content = req.read()
+               ct = req.info().getplist()
+               for param in ct:
+                       if 'charset' in param:
+                               encoding = param.split('=')[1]
+                               break
 
+       return (content, encoding)
 
-def generateRules():
+
+def generateRules(ruleConfig):
        '''Attempts to load a series of rules into a list'''
        try:
                import parserrules
        except ImportError:
                return []
 
+       for conf in ruleConfig:
+               parserrules.ruleConfig[conf] = ruleConfig[conf]
+
        rules = [getattr(parserrules, rule) for rule in dir(parserrules) if callable(getattr(parserrules, rule))]
+
        return rules
 
 
-def applyRules(ical, rules=[], verbose=False):
+def applyRules(cal, rules=[], verbose=False):
        'Runs a series of rules on the lines in ical and mangles its output'
 
        for rule in rules:
-               output = []
-               if rule.__doc__ and verbose:
-                       print(rule.__doc__)
-               for line in ical:
-                       try:
-                               out = rule(line[0],line[1])
-                       except TypeError, e:
-                               output.append(line)
-                               print(e)
-                               continue
-
-                       # Drop lines that are boolean False
-                       if not out and not out == None: continue
-
-                       # If the rule did something and is a tuple or a list we'll accept it
-                       # otherwise, pay no attention to the man behind the curtain
-                       try:
-                               if tuple(out) == out or list(out) == out and len(out) == 2:
-                                       output.append(tuple(out))
-                               else:
-                                       output.append(line)
-                       except TypeError, e:
-                               output.append(line)
-
-               ical = output
-
-       return ical
+               cal = rule(cal)
+
+       return cal
 
 
 def writeOutput(cal, outfile=''):
@@ -242,22 +149,20 @@ def writeOutput(cal, outfile=''):
        else:
                try:
                        out = open(outfile, 'w')
-               except (IOError, OSError), e:
+               except (IOError, OSError) as e:
                        sys.stderr.write('%s\n'%e)
                        sys.exit(1)
 
-       if cal[-1]: cal.append('')
-
-       out.write('\r\n'.join(cal))
+       cal.serialize(out)
 
        if not out == sys.stdout:
                out.close()
 
 
-if __name__ == '__main__':
+def runLocal():
+       '''Main run function if this script is called locally'''
+
        from optparse import OptionParser
-       # If the user passed us a 'stdin' argument, we'll go with that,
-       # otherwise we'll try for a url opener
 
        parser = OptionParser('usage: %prog [options] url')
        parser.add_option('-s', '--stdin', action='store_true', dest='stdin',
@@ -269,9 +174,14 @@ if __name__ == '__main__':
        parser.add_option('-m','--encoding', dest='encoding', default='',
                help='Specify a different character encoding'
                '(ignored if the remote server also specifies one)')
+       parser.add_option('-t','--timezone', dest='timezone', default='',
+               help='Specify a timezone to use if the remote calendar doesn\'t set it properly')
 
        (options, args) = parser.parse_args()
+       ruleConfig["defaultTZ"] = options.timezone or ruleConfig["defaultTZ"]
 
+       # If the user passed us a 'stdin' argument, we'll go with that,
+       # otherwise we'll try for a url opener
        if not args and not options.stdin:
                parser.print_usage()
                sys.exit(0)
@@ -282,7 +192,75 @@ if __name__ == '__main__':
 
        (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)
+
+       cal = vobject.readOne(str(content, encoding))
+       cal = applyRules(cal, generateRules(ruleConfig), options.verbose)
+
+       writeOutput(cal, options.outfile)
+
+
+def exitQuiet(exitstate=0):
+       '''When called as a CGI script, exit quietly if theres any errors'''
+       print('Content-Type: text/html\n')
+       sys.exit(exitstate)
+
+
+def runCGI():
+       '''Main run function if this script is called as a CGI script
+       to process facebook ical files'''
+       import cgi
+       import re
+       import cgitb; cgitb.enable()
+
+       ruleConfig["facebook"] = True
+
+       form = cgi.FieldStorage()
+       if "uid" not in form or "key" not in form:
+               print('Content-Type: text/calendar\n')
+               sys.exit(0)
+       try:
+               # UID should be numeric, if it's not we have someone playing games
+               uid = int(form['uid'].value)
+       except:
+               exitQuiet()
+
+       # The user's key will be a 16 character string
+       key = form['key'].value
+       re.search('[&?]+', key) and exitQuiet()
+       len(key) == 16 or exitQuiet()
+
+       # Historically facebook has been notoriously bad at setting timzeones
+       # in their stuff so this should be a user setting. If it is set in
+       # their calendar it'll  be used otherwise if the user feeds crap or
+       # nothing just assume they want Australia/Perth
+       tz = ""
+       if "tz" in form:
+               from pytz import timezone
+               try:
+                       timezone(form['tz'].value)
+                       tz = form['tz'].value
+               except: pass
+
+       ruleConfig["defaultTZ"] = tz or ruleConfig["defaultTZ"]
+
+       # Okay, we're happy that the input is sane, lets serve up some data
+       url = 'https://www.facebook.com/events/ical/upcoming/?uid=%s&key=%s'%(uid,key)
+       (content, encoding) = getHTTPContent(url)
+
+       cal = vobject.readOne(str(content, encoding))
+       cal = applyRules(cal, generateRules(ruleConfig), False)
+
+       print(('Content-Type: text/calendar; charset=%s\n'%encoding))
+       writeOutput(cal)
+
+
+if __name__ == '__main__':
+       # Ensure the rules process using the desired timezone
+       ruleConfig = {}
+       ruleConfig["defaultTZ"] = 'Australia/Perth'
+
+       # Detect if this script has been called by CGI and proceed accordingly
+       if 'REQUEST_METHOD' in os.environ:
+               runCGI()
+       else:
+               runLocal()

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