X-Git-Url: https://git.ucc.asn.au/?p=frenchie%2Ficalparse.git;a=blobdiff_plain;f=icalparse.py;h=653d55967dfad03010bd344a4e6c6b745eba4bc2;hp=6cb7c39f55bead4fd73f7579f169b6d7cd5c5232;hb=d0370197c1715b2d2eb47ee190dfd92b62dada39;hpb=cec1435fe4a612d02087b715617f24fd21904bdd diff --git a/icalparse.py b/icalparse.py index 6cb7c39..653d559 100755 --- a/icalparse.py +++ b/icalparse.py @@ -21,85 +21,234 @@ # THE SOFTWARE. import sys -import re -from optparse import OptionParser +import urlparse +import os -class InvalidICS(Exception): pass -class notJoined(Exception): 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\-]+:.*') +class InvalidICS(Exception): pass +class IncompleteICS(InvalidICS): pass 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: - brokenline = [line[0:length] + '\r\n'] + brokenline = [line[0:length]] ll = length - while ll < len(line.rstrip('\r\n')) + 1: - brokenline.append(' ' + line[ll:sl+ll].rstrip('\r\n') + '\r\n') + while ll < len(line) + 1: + brokenline.append(line[ll:sl+ll]) ll += sl - cal += brokenline + brokenline = '\r\n '.join(brokenline) + cal.append(brokenline) 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 = [tuple(x.split(':',1)) for x in cal] + + # 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) + + 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] + + +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) + 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 + + +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.''' + + try: + import httplib2 + except ImportError: + import urllib2 + + if not url: return '' + + 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 + except ValueError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + + try: + content = urllib2.urlopen(url).read() + return content + except (urllib2.URLError, 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 + + +def applyRules(ical, 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 + + +def writeOutput(cal, outfile=''): + '''Takes a list of lines and outputs to the specified file''' + + if not cal: + sys.stderr.write('Refusing to write out an empty file') + sys.exit(0) + + if not outfile: + out = sys.stdout + else: + try: + out = open(outfile, 'w') + except (IOError, OSError), e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + + if cal[-1]: cal.append('') + + out.write('\r\n'.join(cal)) + + if not out == sys.stdout: + out.close() + + 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() + 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('-v', '--verbose', action='store_true', dest='verbose', + 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)') (options, args) = parser.parse_args() - if not options.stdin: - try: - import httplib2 - urllib = False - except ImportError: - try: - import urllib - urllib = True - except ImportError: - sys.stderr.write('Failed to find a suitable http downloader\n') - raise - + 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 = applyRules(splitFields(cal), generateRules(), options.verbose) + output = lineFolder(joinFields(ical)) + writeOutput(output, options.outfile)