collecting data from file

[Mustafa Celik]

The file that I'm analyzing guarantees to have >4 columns, and lines 
contain string, integer, float..., they are seperated by spaces.

An example line is as below:

  1850099.32 HELLO 0xfce 6 OTTAWA stree_number_200 exit_Metcalfe 
take_hw_417 arrive_at_Ottawa_Airport

2nd column will tell me if someone has arrive, 3rd column will tell me 
how much $ they'll bring with them, and 4th column will tell who he is 
(a hex guy).

I will find out;
   * how many people arrived and did not leave,
   * how many have arrived and left
   * how much we have in Ottawa at the end
   * how much transient money flow have occured

My file may contain some lines that only leaves Ottawa, so I should 
ignore them (they probably arrived earlier, and should not decrement my 
dollar gain)  - This condition is not very important, if it'll 
complicate things, I have other ways to prevent this.

I found a good awk document  last night (can't remember the web site), I 
was planning to dig in it today, but you guys have supported me with my 
initial plan with Python. I'm not a beginner in Python, this is probably 
not the best project to start with, but I need to have this done by Awk 
or Python, ...

Thanks,
   Mustafa





Peter Hansen wrote:
> Mustafa Celik wrote:
> 
>>I want to scroll thru a file;
>>   * find lines that match a string (e.g. HELLO) on the 2nd column
>>   * add up the 4th column (an integer) on each matching line, say the
>>variable is TOTAL
>>   * subtract the 4th column from TOTAL if another is string (e.g. BYE)
>>is hit on 2nd column of a line
>>
>>Any tips?
> 
> 
> Yes, post an example showing what you intend.  The instructions are
> decent, but implementable only by making a few assumptions about what
> you mean.  Also, are there any error conditions to be handled, or do
> you guarantee the input is always perfect (e.g., always at least four
> columns, and so on)?
> 
> -Peter