Exercise: Parsing /etc/passwd

Requirement

Sample /etc/passwd file

Download /etc/passwd

Write a program getuser.py that takes a filename and one or more usernames from the commandline, looks them up in /etc/passwd, and prints out the user records one after the other.

$ python getuser.py /etc/passwd root jfasch
User: root
  UID: 0
  GID: 0
  Home: /root
  Shell: /bin/bash
User: jfasch
  UID: 1000
  GID: 1000
  Home: /home/jfasch
  Shell: /bin/bash

Implementation Strategy

The /etc/passwd (download sample) is the UNIX user database. The file is line-based (an old UNIX tradition), and each line contains a user record:

...
root:x:0:0:root:/root:/bin/bash
jfasch:x:1000:1000:Joerg Faschingbauer:/home/jfasch:/bin/bash
...

Each line/record consists of seven fields, separated by colons (':'), with the following meaning:

Field index	Field name	Description
0	username	User name
1	password	Encrypted password; x means that the encrypted password is found in /etc/shadow.
2	uid	User ID
3	gid	(Primary) group ID
4	description	Arbitrary descriptive text
5	home	Home directory
6	shell	Login shell

The program should read /etc/passwd only once. Each line is a user record, and should be broken up into a dictionary. For example, the line

jfasch:x:1000:1000:Joerg Faschingbauer:/home/jfasch:/bin/bash

Would make it into a dictionary

{ 'uid': 1000,
  'gid': 1000,
  'home': '/home/jfasch',
  'shell': '/bin/bash'
}

All lines/dictionaries/records would be stored in a large user database dictionary that would look like so,

userdatabase = {
    "root": {
        "uid": 0,
        "gid": 0,
        "home": "/root",
        "shell": "/bin/bash",
       },
    "jfasch": {
        "uid": 1000,
        "gid": 1000,
        "home": "/home/jfasch",
        "shell": "/bin/bash",
       },
    ...
}

It is that userdatabase that is finally used to lookup the remaining username arguments to the program.