X-Git-Url: https://git.ucc.asn.au/?a=blobdiff_plain;f=icalparse.py;h=1d62a7057554cd4142881c43411449d4f7f09589;hb=a60d888467b796bdf7ed7cd63c53cef8dcfc779c;hp=110a19ae703d822afa6fd918672bcf69c158543e;hpb=aace9ea8b2dbcdbdc1b3ed30a672a52b0a619e3f;p=frenchie%2Ficalparse.git diff --git a/icalparse.py b/icalparse.py index 110a19a..1d62a70 100755 --- a/icalparse.py +++ b/icalparse.py @@ -1,6 +1,6 @@ #!/usr/bin/python # -# Copyright (c) 2010 James French +# Copyright (c) 2013 James French # # Permission is hereby granted, free of charge, to any person obtaining a copy # of this software and associated documentation files (the "Software"), to deal @@ -20,54 +20,18 @@ # 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 notJoined(Exception): pass - -def lineJoiner(oldcal): - '''Takes a string containing a calendar and returns an array of its lines''' - - if not oldcal[0:15] == 'BEGIN:VCALENDAR': - raise InvalidICS, "Does not appear to be a valid ICS file" - - if list(oldcal) == oldcal: - oldcal = '\r\n'.join(oldcal) - - oldcal.replace('\r\n ', '') - 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 line in oldcal: - # Line fits inside length, do nothing - if len(line.rstrip()) <= length: - cal.append(line) - else: - brokenline = [line[0:length] + '\r\n'] - ll = length - while ll < len(line.rstrip('\r\n')) + 1: - brokenline.append(' ' + line[ll:sl+ll].rstrip('\r\n') + '\r\n') - ll += sl - cal += brokenline - - return cal +import sys, os +import urlparse +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 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) @@ -75,74 +39,149 @@ 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 + 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) content = res.read() + ct = res.info().getplist() res.close() - except (urllib2.URLError, ValueError), e: + except (urllib2.URLError, OSError), e: sys.stderr.write('%s\n'%e) sys.exit(1) - return content + + for param in ct: + if 'charset' in param: + encoding = param.split('=')[1] + break + + return (content, encoding) 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.''' + services. It's what you should use in a CGI script.''' try: import httplib2 except ImportError: import urllib2 - if not url: return '' + if not url: return ('','') + + if not 'http' in urlparse.urlparse(url)[0]: 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) + if h: + try: + req = h.request(url) + except ValueError, 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 = '' + else: + try: + req = urllib2.urlopen(url) + except urllib2.URLError, 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(ruleConfig): + '''Attempts to load a series of rules into a list''' try: - content = urllib2.urlopen(url).read() - return content - except urllib2.URLError, e: - sys.stderr.write('%s\n'%e) - sys.exit(1) + import parserrules + except ImportError: + return [] - 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(cal, rules=[], verbose=False): + 'Runs a series of rules on the lines in ical and mangles its output' + + for rule in rules: + cal = rule(cal) + + return cal + + +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) + + cal.serialize(out) + + if not out == sys.stdout: + out.close() + + +def runLocal(): + '''Main run function if this script is called locally''' -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('-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)') + 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) @@ -151,6 +190,77 @@ if __name__ == '__main__': else: url = '' - content = getContent(url, options.stdin) - cal = lineJoiner(content) - print cal + (content, encoding) = getContent(url, options.stdin) + encoding = encoding or options.encoding or 'utf-8' + + cal = vobject.readOne(unicode(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 = 'http://www.facebook.com/ical/u.php?uid=%d&key=%s'%(uid,key) + (content, encoding) = getHTTPContent(url) + + cal = vobject.readOne(unicode(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()