X-Git-Url: https://git.ucc.asn.au/?p=frenchie%2Ficalparse.git;a=blobdiff_plain;f=icalparse.py;h=1ed6076da0d36257c62abf5a12294b0cec95fbc6;hp=7953420bfdc303c8df466339fe74096a03b39e72;hb=48d87039714129b0798a4efba2938d15190af7d8;hpb=9edbffa51163eef8bac3d125a7cd7320a92df3a8 diff --git a/icalparse.py b/icalparse.py index 7953420..1ed6076 100755 --- a/icalparse.py +++ b/icalparse.py @@ -21,54 +21,42 @@ # THE SOFTWARE. import sys -import re import urlparse import os -from optparse import OptionParser + class InvalidICS(Exception): pass class notJoined(Exception): pass +class IncompleteICS(InvalidICS): pass -# RFC5545 and RFC5546 iCalendar registries contain upper case letters -# and dashes only and are separated from the value by a colon (:) -icalEntry = re.compile('^[A-Z\-]+:.*') def lineJoiner(oldcal): - '''Unfolds a calendar so that items can be parsed''' + '''Takes a string containing a calendar and returns an array of its lines''' - cal = [] + if not oldcal[0:15] == 'BEGIN:VCALENDAR': + raise InvalidICS, "Does not appear to be a valid ICS file" - # Strip newlines ( - for line in oldcal: - line = line.rstrip('\r\n') - - # Reassemble broken Lines - if not line: - if not cal: continue - else: cal[-1] += '\\n' - elif line[0] == ' ': - if not cal: raise InvalidICS, 'First line of ICS must be element' - line = line[1:len(line)] - cal[-1] += line - elif not icalEntry.match(line): - if not cal: raise InvalidICS, 'First line of ICS must be element' - cal[-1] += '\\n' + line - else: - if cal: cal[-1] += '\r\n' - cal.append(line) + if not 'END:VCALENDAR' in oldcal[-15:-1]: + raise IncompleteICS, "File appears to be incomplete" - cal[-1] += '\r\n' + if list(oldcal) == oldcal: + oldcal = '\r\n'.join(oldcal) - return cal + oldcal = oldcal.replace('\r\n ', '').replace('\r\n\t','') + return oldcal.strip().split('\r\n') -def lineSplitter(oldcal, length=75): + +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 line in oldcal: - # Line & line ending line ending fit inside length, do nothing + # Line fits inside length, do nothing if len(line.rstrip()) <= length: cal.append(line) else: @@ -81,60 +69,122 @@ def lineSplitter(oldcal, length=75): return cal -if __name__ == '__main__': - # 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', - default=False, help='Take a calendar from standard input') - parser.add_option('-o', '--output', dest='outfile', default='', - help='Specify output file (defaults to standard output)') +def splitFields(cal): + '''Takes a list of lines in a calendar file and returns a list of key, value pairs''' - (options, args) = parser.parse_args() + ical = [tuple(x.split(':',1)) for x in cal] - if not args and not options.stdin: - parser.print_usage() - sys.exit(0) + # 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) - url = args[0] + return ical - # Work out what url parsers we're going to need based on what the user - # gave us on the command line - we do like files after all + +def getContent(url='',stdin=False): + '''Generic content retriever, DO NOT use this function in a CGI script as + it can read from the local disk (which you probably don't want it to). + ''' + + # 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) - http = 'http' in parsedURL[0] + if 'http' in parsedURL[0]: return getHTTPContent(url) + + if stdin: + content = sys.stdin.read() + return content + + 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 + + # If we've survived, use python's generic URL opening library to handle it + import urllib2 + try: + res = urllib2.urlopen(url) + content = res.read() + res.close() + except (urllib2.URLError, OSError), e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + return content - if not parsedURL[0]: u = False - else: u = True - if not options.stdin and http: - try: - import httplib2 - except ImportError: - import urllib2 +def getHTTPContent(url='',cache='.httplib2-cache'): + '''This function attempts to play nice when retrieving content from HTTP + services. It's what you should use in a CGI script. It will (by default) + slurp the first 20 bytes of the file and check that we are indeed looking + at an ICS file before going for broke.''' + + try: + import httplib2 + except ImportError: + import urllib2 + + if not url: return '' - # Try and play nice with HTTP servers unless something goes wrong. We don't - # really care about this cache so it can be somewhere volatile - h = False if 'httplib2' in sys.modules: try: h = httplib2.Http('.httplib2-cache') except OSError: h = httplib2.Http() + else: h = False - if not options.stdin and (not http or not 'httplib2' in sys.modules): - import urllib2 + try: + if h: content = h.request(url)[1] + return content + except ValueError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) try: - content = u and (h and h.request(url)[1] or urllib2.urlopen(url).read()) - except (ValueError, urllib2.URLError), e: + content = urllib2.urlopen(url).read() + return content + except (urllib2.URLError, OSError), e: sys.stderr.write('%s\n'%e) sys.exit(1) - if not u: - try: content = open(os.path.abspath(url),'r').read() - except (IOError, OSError), e: - sys.stderr.write('%s\n'%e) - sys.exit(1) + return '' + + +def generateRules(): + '''Attempts to load a series of rules into a list''' + try: + import parserrules + except ImportError: + return [] + + rules = [getattr(parserrules, rule) for rule in dir(parserrules) if callable(getattr(parserrules, rule))] + return rules - # RFC5545 and RFC5546 Calendars should be generated UTF-8 and we need to - # be able to read ANSI as well. This should take care of us. - content = unicode(content, encoding='utf-8') + +if __name__ == '__main__': + 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', + default=False, help='Take a calendar from standard input') + parser.add_option('-o', '--output', dest='outfile', default='', + help='Specify output file (defaults to standard output)') + + (options, args) = parser.parse_args() + + if not args and not options.stdin: + parser.print_usage() + sys.exit(0) + elif not options.stdin: + url = args[0] + else: + url = '' + + content = getContent(url, options.stdin) + cal = lineJoiner(content) + ical = splitFields(cal) + rules = generateRules() + print rules