[MacPerl-WebCGI] loops 'n' lists [was Referencing a Hash element...]

[Bruce Van Allen <bva@cruzio.com>]

See previous msg from Richard Smith, 8:52 PM 12/28/00.

Richard -- Here's an illustration of the use of loops and lists to 
process your admin form input.

First, the code below populates the %Form hash, to simulate form 
input. Then using map, we make a set of lists of the keys of %Form 
whose elements you want to process. Take some time to examine just 
the map construct; there's a lot going on there. (The final 'or ()' 
ensures that @admin_keys doesn't get any blank elements for keys that 
don't match the ones of interest.)

The code goes on to use the lists of keys to make hash slices, which 
are then looped over for processing. Along the way, things get 
printed out so you may see what's transpired.

To simulate your error handling subroutine, my code just accumulates 
the errors in the variable @errors. Also, as you correctly do in your 
original code, $& must be segregated within a block, and that is why 
the last loop in my code has the extra enclosing braces.

   foreach ( @Form{@description_keys, @linktext_keys} ) { {
     ...
   } }


After you've grokked this stuff, here are some questions:

a. Why chomp?  I threw in a return in $Form{description1} to show 
that the chomping works, but if your input is from HTML text fields, 
there won't be any line endings (unless you use <textarea>).
b. The way you've patterned your capture of illegal characters, 
you're getting them in contiguous clumps, replacing with a single 
'*'. Might be what you want, but might not.
c. Why substitute? Your original code, replicated in mine, isn't 
really meant to *correct* anything, just to spot errors and send them 
to your error handling subroutine. Just match?
d. Employing $& is misleading when a given item has more than one 
illegal; only the last one found will be in $& when you insert it 
into your error alert string.

Anyway, here's the code, followed by the output (beware of email line 
breaks), and I'm going to finish lunch...

1;

- Bruce


#!perl -w

# Simulated form input:
my %Form;
$Form{_submitter} = 'Ralph';
$Form{_date} = '2001/01/17';
$Form{_comment} = 'Here are my changes...';
$Form{linktext1} = 'Big Ideas';
$Form{description1} = ' %hings we are planning ...
';
$Form{location1} = '/site/big.html  ';
$Form{linktext2} = '  Small Ideas ';
$Form{description2} = 'Some other $%^&* stuff.';
$Form{location2} = '/site/small.htm';
$Form{linktext3} = 'Your Ideas (??)';
$Form{description3} = ' &Useful! ideas from our clients ...';
$Form{location3} = ' /site/client.html  ';

# Some arrays:
my @errors;
my (@linktext_keys, @description_keys, @location_keys);

# Populate the arrays from the hash keys:
my @admin_keys = map {
	/linktext/ and push @linktext_keys, $_ and $_
	  or
	/description/ and push @description_keys, $_ and $_
	  or
	/location/ and push @location_keys, $_ and $_
	  or
	()
} sort keys %Form;

print "Before:\n";
foreach (sort keys %Form) { printf "%-14s %s\n", $_, $Form{$_} }


# To process all of the admin elements of %Form,
# make a list from a hash slice with @admin_keys as the indices
# 1. remove line endings
chomp @Form{@admin_keys};  # chomp works on all elements of a list
# 2. trim whitespace
foreach ( @Form{@admin_keys} ) {
     s/^\s*(.*?)\s*$/$1/
}

# Process only 1 set of elements
# 3. single '.' at end of descriptions
foreach ( @Form{@description_keys} ) {
     s/\s*\.*$/./
}

# Process >1 sets of elements
# 4. replace non-alphas with '*' and
# 5. record errors
foreach ( @Form{@description_keys, @linktext_keys} ) { {
	s/[^a-zA-Z .]+/*/g;
	push @errors, <<ERR if $&;
The description in line
   $_
contained an illegal character '$&'
ERR
} }

print "\nAfter:\n";
foreach (sort keys %Form) { printf "%-14s %s\n", $_, $Form{$_} }
print "\nErrors:\n", join "\n", @errors;


__END__


Before:
_comment       Here are my changes...
_date          2001/01/17
_submitter     Ralph
description1    %hings we are planning ...

description2   Some other $%^&* stuff.
description3    &Useful! ideas from our clients ...
linktext1      Big Ideas
linktext2        Small Ideas
linktext3      Your Ideas (??)
location1      /site/big.html
location2      /site/small.htm
location3       /site/client.html

After:
_comment       Here are my changes...
_date          2001/01/17
_submitter     Ralph
description1   *hings we are planning.
description2   Some other * stuff.
description3   *Useful* ideas from our clients.
linktext1      Big Ideas
linktext2      Small Ideas
linktext3      Your Ideas *
location1      /site/big.html
location2      /site/small.htm
location3      /site/client.html

Errors:
The description in line
   *hings we are planning.
contained an illegal character '%'

The description in line
   Some other * stuff.
contained an illegal character '$%^&*'

The description in line
   *Useful* ideas from our clients.
contained an illegal character '!'

The description in line
   Your Ideas *
contained an illegal character '(??)'







==== Want to unsubscribe from this list?
==== Send mail with body "unsubscribe" to macperl-webcgi-request@macperl.org