From: Philip Rhoades <phil@...>
Date: 2012-11-22T01:31:02+09:00
Subject: Re: Alternatives to methods for large number of nested "ifs" - SOLVED

On 2012-11-18 00:50, Philip Rhoades wrote:
> People,
>
> Thanks to the people who responded - I looked at everything but
> nothing really suited what I was trying to do but see below for some
> progress that I made and another question:
>
>
> On 2012-11-17 01:35, Philip Rhoades wrote:
>> People,
>>
>> I have an existing script that looks like:
>>
>> lots of code
>> .
>> .
>> .
>> if ..
>>   some lines of code
>>   if ..
>>     some lines of code
>>     if ..
>>       some lines of code
>>       if ..
>>         some lines of code
>>         if ..
>>           some lines of code
>> etc
>>
>> This makes it difficult to see the overall logic so I want to change
>> it to something like:
>>
>> lots of code
>> .
>> .
>> .
>> if ..
>>   call something
>>   if ..
>>     call something
>>     if ..
>>       call something
>>       if ..
>>         call something
>>         if ..
>>           call something
>> etc
>>
>> but if I use methods I will have to pass lots of variables or use
>> global variables and then it gets clumsy changing everything - is
>> there some way of simply calling a block of code and then returning 
>> to
>> the same position in the nested "ifs"?
>
>
> I found that I could use sequential "ifs" instead of nested "ifs" by
> using "exit" statements in each "if" block - the script is processing
> incoming mails and can exit when a mail has been satisfactorily
> classified.  I still have something like this:
>
>
> lots of code
> if block - small
> lots of code
> if block - small
> if block - large
> if block - small
> if block - large
> .
> .
> etc
>
>
> What I want is to be able to see the whole logic of the program in
> about 25 lines (ie one screenfull of code - ie all the "if"
> statements) - I could probably use standard methods for the 
> repetitive
> small "if" blocks but for the large "lots of code" and "if block -
> large" I definitely can't.  Is it possible that a method can be 
> forced
> to NOT use local variables? - then I could replace the large blocks
> with these sorts of methods.


procs were the solution (they allow shared variables) - I now have this 
at the end of the program:

                                 ####        Mailing-List:, 
Delivered-To: and Reply-To: lines
p_mailing_list.call
                                 #### OK     line_To non-blocked domains
p_line_To_non_blocked_domains.call
                                 #### OK     in_mailing_list
p_in_mailing_list.call
                                 #### Drop   in_black_list_to
p_in_black_list_to.call
                                 #### Drop   in_black_list_from AND 
!Srejector content
p_in_black_list_from.call
                                 #### Drop   Bounce Bounced - deleting 
old quarantined file if it exists
p_bounce_bounced.call
                                 #### Reject LinkedIn invitation
p_LinkedIn.call
                                 #### OK     Srejector content - 
incoming/outgoing
p_Srejector.call
                                 #### OK     in_domains_list - for 
incoming/outgoing mails, system mails etc
p_in_domains_list.call
                                 #### OK     put me on your whitelist 
subject
p_put_me_on_your_whitelist_subject.call
                                 #### OK     To: line
p_line_To.call
                                 #### Drop   !in_white_list_to
p_not_in_white_list_to.call
                                 #### Drop   !in_white_list_to_substr
p_not_in_white_list_to_substr.call
                                 #### OK     in_white_list_from
p_in_white_list_from.call
                                 #### Drop   in_subject_list
p_in_subject_list.call
                                 #### Reject !in_white_list_from
p_not_in_white_list_from.call
                                 #### OK     shouldn\'t get here?
p_shouldnt_get_here.call


Regards,

Phil.

-- 
Philip Rhoades

GPO Box 3411
Sydney NSW	2001
Australia
E-mail:  phil@pricom.com.au