Export emails from Outlook

"""
Exports email address from MS Outlook from recent eemails
"""

from win32com.client import Dispatch, constants
import os.path

class Exporter(object):

    app = None
    emails = []
    
    def __init__ (self):

 self.app = Dispatch("Outlook.Application")
        if self.app == None:
            raise Exception, "Unable to create an Outlook application object"

    def export_emails(self, foldername, outfile="email.csv", maxcount=1000):

        emails = [] 
 ns = self.app.GetNamespace("MAPI")
        folders = ns.Folders.Item(1).Folders #Usually the users folders
        folder = folders.Item(foldername)
        cnt = len(folder.Items)
        if cnt > maxcount:
            cnt = maxcount
        for i in range(1, cnt + 1):
            item = folder.Items.Item(i)
            print "Processing item %d" % i
            self.export_from_field(item, "To")
            self.export_from_field(item, "CC")

        self.emails.sort()
        emailfile = open(outfile, 'w')
        for email in self.emails:
            emailfile.write(email + "\n")
        emailfile.close()

    def export_from_field(self, item, fieldname, maxcount = 10):
        if not hasattr(item, fieldname):
            return
        ms = getattr(item, fieldname).split(';')
        if len(ms) > maxcount:
            return
        for m in ms:
            parts = m.split(',')
            if len(parts) <> 2:
                continue
            email = "%s_%s@intuit.com" % (parts[1].strip(), parts[0].strip())
            if not email in self.emails:
                self.emails.append(email)
     
    def __del__ (self):

        self.app = None
        self.emails = None

def main ():
    export = Exporter()
    export.export_emails("InBox", "emails.csv", 1000)
    
if __name__ == "__main__":
    main();

A few lessons learned on prototype.js and RoR

A few lessons learned from my recent work on prototype.js and Ruby on Rails

1) "class" is a reserved word for IE
"class" is a reserved word for IE. If you use "class" as variable name in your JavaScript such as this (example using prototype.js):

var highlight_span= new Element('span', { margin: '0', class: 'highlight' }); 

It works in Firefox but gives you an error in IE. The work around is:

var highlight_span= new Element('span', { margin: '0'}); highlight_span.addClassName("highlight"); 

2) IE requires 'tbody' in dynamically generated DOM
If you use dynamic JavaScript, such as "new Element(..)", to generate a table in layout, make sure the 'tbody' element is generated explicitly. Otherwise, the table won't show up in IE. Firefox does not seem to care about the missing of 'tbody'. Static coded table without 'tbody' works fine in IE.

3) IE requires you to explicitly register event handler
Following code works in Firefox, but not IE:

var a = new Element('a', {onclick: handler}); 

You have to use "observe" to explicitly register the handler function:

var a = new Element('a').observe('click', handler); 

4) You cannot set cookie value to int in RoR
This code won't work:

cookies[:worker] = @worker.id 

Instead, you need to say:

cookies[:worker] = @worker.id.to_s 

Web page smoke test script

#!/usr/local/bin python

"""
This script is used to check urls with a pattern. The configuration file has following format

    protocol_name|url|additional_args

example:
    simpleGET|http://www.google.com/|Google
    simpleGET|http://www.aol.com/|WhatWhat

Example output:
    http://www.google/ -> True
    http://www.aol.com/ -> False

"""

import os.path, re, getopt, sys, urllib2, urllib

verbose = False
FETCHER = 'fetcher'
MATCHER = 'matcher'

def say(s):
    if verbose:
        print s

class UrlChecker:

    def __init__(self, urls, protocols):
        self.urls = urls
        self.protocols = protocols

    def check(self):
        for rurl in self.urls:
            self._check_url(rurl)
            
    def _check_url(self, rurl):
        protocol, url, args = rurl.split('|')
        say("Check URL %s with protocol %s. Args = %s" % (url, protocol, args))
        rslt = self.protocols[protocol][FETCHER](url, args)
        self.protocols[protocol][MATCHER](rslt, url, args)

def load_urls(urlfile):
    urls = []
    f = open(urlfile, 'r')
    try:
        for line in f:
            say("Load one raw url %s" % (line.strip()))
            urls.append(line.strip())
    finally:
        f.close()
    return urls

def add_simpleGET(protocols):

    def simple_get(url, ignore):
        r = urllib2.urlopen(url)
        rslt = r.read()
        r.close()
        return rslt

    def str_match(body, url, pattern):
        print "%s -> %s" % (url, body.find(pattern) != -1)

    protocols['simpleGET'] = {FETCHER : simple_get, MATCHER : str_match}

def init_protocols():
    protocols = {};
    add_simpleGET(protocols)
    return protocols

if __name__ == '__main__':

    def usage():
        print "Usage:"
        print "python check_url.py -f <url_file> -v"

    try:                                
        opts, args = getopt.getopt(sys.argv[1:], "f:v")
    except getopt.GetoptError:
        usage()
        sys.exit(2)                     

    urls = None
    protocols = init_protocols()
    for opt, arg in opts:
        if opt in ("-v"):
            verbose = True
        elif opt in ("-f"):
            urls = load_urls(arg)
    if urls is None:
        usage()
        sys.exit(1)

    checker = UrlChecker(urls, protocols)
    checker.check()

Email forwarding script

#!/usr/local/bin python

"""
This script is used to forward email
"""

import os, sys, poplib, getopt, getpass, email, smtplib

from_email = '???'
to_email = '???'
pop3_server_url = '???'
pop3_server_port = 110
pop3_user = '???'
pop3_pass = '???'
delete_after = True
subject_filter = '????'

def match_filter(subject, sfilter):
    return subject.find(sfilter) == 0

def forward_mail(femail, temail, mail):

    for (header, value) in mail.items():
        if not header in ('Content-Type', 'Content-Transfer-Encoding', 'Subject', 'Mime-Version'):
            del mail[header]

    mail["From"] = femail
    mail["To"] = temail
    for (header, value) in mail.items():
        print "%s: %s" % (header, value)

    server = smtplib.SMTP('localhost')
    server.set_debuglevel(0)
    server.sendmail(femail, temail, mail.as_string())
    server.quit()    

def process_mail(pop3_mail):
    
    mail = email.message_from_string(pop3_mail)
    subject = mail.get("Subject")
    if match_filter(subject, subject_filter):
        print "Forward email: %s" % (subject)
        forward_mail(from_email, to_email, mail)
        return True
    else:
        print "Skip email: %s" % (subject)
        return False

if __name__ == '__main__':

    def usage():

        print "Usag:"
        print "python forward_mail.py -s <pop3_server> -p <pop3_port> -u <user> -c <credential> -f <from_email> -t <to_email>"

    try:                                
        opts, args = getopt.getopt(sys.argv[1:], "s:p:u:c:f:t:")
    except getopt.GetoptError:
        usage()
        sys.exit(2)                     

    for opt, arg in opts:
        if opt in ("-s"):
            pop3_server_url = arg
        elif opt in ("-p"):
            pop3_server_port = int(arg)
        elif opt in ("-u"):
            pop3_user = arg
        elif opt in ("-c"):
            pop3_pass = arg
        elif opt in ("-f"):
            from_email = arg
        elif opt in ("-t"):
            to_email = arg

    if pop3_user is None:
        pop3_user = getpass.getUser()
    if pop3_pass is None:
        pop3_pass = getpass.getpass()

    pop3 = poplib.POP3(pop3_server_url, pop3_server_port)
    pop3.user(pop3_user)
    pop3.pass_(pop3_pass)
    msg_count = len(pop3.list()[1])
    print "Total messages in Inbox: %d" % (msg_count)

    for i in range(msg_count):
        resp, text, octets = pop3.retr(i+1)
        if process_mail('\n'.join(text)) and delete_after:
            print "Forwarded email is deleted"
            pop3.dele(i+1)
    pop3.quit()

Search text pattern in log files

#!/usr/local/bin python

"""
This script is used to analyze log file
"""

import os.path, time, re, getopt, sys, fnmatch

verbose = False

def say(s):
   if verbose:
       print s

def searchPattern(args, dirname, filenames):

   file_pattern, text_pattern, count = args
   say("Processing directory %s now" % (dirname))
   say("File pattern: %s" % (file_pattern))
   say("Text pattern: %s" % (text_pattern))
   say("Count only: %s" % (count))

   counts = {}
   def incMatchCount(matched):
       counts[matched] = counts.get(matched, 0) + 1
  
   compiled_pattern = re.compile(text_pattern)
   for filename in fnmatch.filter(filenames, file_pattern):
       say("Processing file %s now" % (filename))
       log = open(os.path.join(dirname, filename), 'r')
       for line in log:
           m = re.match(compiled_pattern, line)
           if m:
               if count:
                   incMatchCount(m.group(1))
               else:
                   print ';;'.join(m.groups())
       if count:
           for matched in counts.keys():
               print " -> ".join([matched, str(counts[matched])])

if __name__ == '__main__':

   def usage():
       print "Usage:"
       print "python find_pattern.py -d <root_dir> -f <file_pattern> -p <search_pattern> -c -v"

   try:                               
       opts, args = getopt.getopt(sys.argv[1:], "d:f:p:cv")
   except getopt.GetoptError:
       usage()
       sys.exit(2)                    

   count = False
   root_dir = '.'
   file_pattern = '*.log'
   text_pattern = '(.+)'
   for opt, arg in opts:
       if opt in ("-v"):
           verbose = True
       elif opt in ("-c"):
           count = True
       elif opt in ("-d"):
           root_dir = arg
       elif opt in ("-f"):
           file_pattern = arg
       elif opt in ("-p"):
           text_pattern = arg

   os.path.walk(root_dir, searchPattern, (file_pattern, text_pattern, count))

SlimFIX gets new life

I just found that a project I developed a few years back SlimFIX has been adopted by an open source project called ActiveQuant. I am glad it finally gets a new life.