Mailer

[wingnut2626]

I'm beginning the process of writing an automatic mailer from my server that will send me processes, and whatever else I want. Here is the code:


import smtplib

def Mailer(argument):

FROMADDR = "fromme@fromme.com"
LOGIN = FROMADDR
PASSWORD = "password"
TOADDRS = ["send@send.com"]
SUBJECT = "Test"

msg = argument

server = smtplib.SMTP('smtp.gmail.com', 587)
server.set_debuglevel(1)
server.ehlo()
server.starttls()
server.login(LOGIN, PASSWORD)
server.sendmail(FROMADDR, TOADDRS, msg)
server.quit()


Saved the module as mail.py. Now if argument is a string created on the fly, or if it is a short string written from another program, say:




import mail

b = 'hello world'
mail.Mailer(b)



it sends perfectly fine. However if i run an exteral command, pipe the output to a file, read the file and send the resulting string as 'argument', nothing happens.




import os
import mail

a = os.popen2('ps -A > /home/user/info.txt')

b = open('/home/user/info.txt').read()

mail.Mailer(a)



I don't get a traceback or anything. As a matter of fact it shows as if the string was being sent successfully.

Any ideas?

[spjackson]

import os
import mail

a = os.popen2('ps -A > /home/user/info.txt')

b = open('/home/user/info.txt').read()

mail.Mailer(a)



The short answer is that you have read the message into the variable b, so that is the variable you want to pass to your Mailer function.

A fuller answer would include:

os.popen2 and similar functions are deprecated in favour of the subprocess module.
If you want to use a pipe, don't use the intermediate file info.txt; either use a pipe or use the file, not both.
As it stands, at the time you call read() the file may not necessarily have finished being written.

[wingnut2626]

Actually what it turned out to be was that I was missing certain headers that Gmail.com requires to be. 'RS' compliant. Added those headers and it worked perfectly fine. I also used a file because I had more information I wanted in the email than just a process list. I changed to a ' a = subprocess.check_output' instead of the os.open. I know that the subprocess is newer, but does it really make a difference?

[wingnut2626]

check this out too http://essays.ajs.com/2011/02/python-subprocess-vs-ospopen-overhead.html