🚀 go-pugleaf

lordbear@best.com wrote:
> So, I want to be able to forward an email message containing useful
> information and have it parsed (via a Perl CGI) into a ASCII DB.

> I've writtent the first part of the script, which strips headers 
> and other extraneous info.  Now I am left with:

> //////////////////////////////////////////

> perquisite   \PURR-kwuh-zit\   (noun)
>         *1 : a privilege, gain, or profit incidental to regular salary
>          or wages; especially : one expected or promised
>          2 : gratuity, tip
>          3 : something held or claimed as an exclusive right or
>          possession

> |Example sentence:
>          Joanne's new job is loaded with perquisites, including
>          unlimited use of the company car and credit card and access
>          to the corporate health center.

> /////////////////////////////////////////

> Note the blank lines at the head and tail.

> Objective:  Have the first line of text appear as one 'field', the 
> following paragraph as the second and the next block of text as the 
> third, stripping away the extra speaces that indent lines, etc. 

> I'm having a devil of a tmie with the necessary regexps and the right
> method for reading through and parsing the file.  

> Any ideas appreciated!

Something like the following should get you closer to what you're
looking for.

	# ignoring the initial and trailing blank line, separate
	# $_ into a first line, a second paragraph, and a third
	# paragraph
	($first,$second,$third) = /^\n(.*?\n)(.*?\n)\n(.*)\n$/s;

	# unindent the second paragraph
	$second =~ s/^\s+//;
	$second =~ s/\n\s+/\n/sg;

	# unindent the third paragraph
	$third =~ s/^\s+//;
	$third =~ s/\n\s+/\n/sg;

Dave
____________________________________________________________________________
Dave W. Smith          mailto:dws@best.com          http://www.best.com/~dws

Dave W. Smith <dws@shell3.ba.best.com> wrote:
> lordbear@best.com wrote:
>> So, I want to be able to forward an email message containing useful
>> information and have it parsed (via a Perl CGI) into a ASCII DB.

>> I've writtent the first part of the script, which strips headers 
>> and other extraneous info.  Now I am left with:

>> //////////////////////////////////////////

>> perquisite   \PURR-kwuh-zit\   (noun)
>>         *1 : a privilege, gain, or profit incidental to regular salary
>>          or wages; especially : one expected or promised
>>          2 : gratuity, tip
>>          3 : something held or claimed as an exclusive right or
>>          possession

>> |Example sentence:
>>          Joanne's new job is loaded with perquisites, including
>>          unlimited use of the company car and credit card and access
>>          to the corporate health center.

>> /////////////////////////////////////////

>> Note the blank lines at the head and tail.

>> Objective:  Have the first line of text appear as one 'field', the 
>> following paragraph as the second and the next block of text as the 
>> third, stripping away the extra speaces that indent lines, etc. 

>> I'm having a devil of a tmie with the necessary regexps and the right
>> method for reading through and parsing the file.  

>> Any ideas appreciated!

> Something like the following should get you closer to what you're
> looking for.

> 	# ignoring the initial and trailing blank line, separate
> 	# $_ into a first line, a second paragraph, and a third
> 	# paragraph
> 	($first,$second,$third) = /^\n(.*?\n)(.*?\n)\n(.*)\n$/s;

> 	# unindent the second paragraph
> 	$second =~ s/^\s+//;
> 	$second =~ s/\n\s+/\n/sg;

> 	# unindent the third paragraph
> 	$third =~ s/^\s+//;
> 	$third =~ s/\n\s+/\n/sg;


Thanks for the hints, Dave!  

Let's see if I understand this and can therefore translate into 
human-readable English...  please correct me if needed.  :)

>        # ignoring the initial and trailing blank line, separate
>        # $_ into a first line, a second paragraph, and a third
>        # paragraph
>        ($first,$second,$third) = /^\n(.*?\n)(.*?\n)\n(.*)\n$/s;

Beginning at the first character, match (and ignore) the immediately
following newline character.  The first grouping contains any number of
characters up to a newline.  [What does the '?' do?]  The second grouping
follows the same rules.  The third matches the rest of the characters up to 
the end newline.

Now, my question is;  how is it that this would work as there are 
intervening newlines in the respective paragraphs?  I was trying to 
compose a regexp that would detect two adjacent newlines and therefore
assume it had found the blank line between paragraphs.  I see no such
interpolation here, unless that's how the question mark works.  :)

Thanbks gain, in advance, for insight and illumination!

Bear


-- 
-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
Don 'Bear' Wilkinson      San Francisco / Daly City     bear@lair.org
     Computer support professional seeking full-time employment;
 review my resume and interests at http://www.lair.org/bear/work.html

lordbear@NOSPAMbest.com wrote:
> Let's see if I understand this and can therefore translate into 
> human-readable English...  please correct me if needed.  :)

>>        # ignoring the initial and trailing blank line, separate
>>        # $_ into a first line, a second paragraph, and a third
>>        # paragraph
>>        ($first,$second,$third) = /^\n(.*?\n)(.*?\n)\n(.*)\n$/s;

> Beginning at the first character, match (and ignore) the immediately
> following newline character.  The first grouping contains any number of
> characters up to a newline.  [What does the '?' do?]  The second grouping
> follows the same rules.  The third matches the rest of the characters up to 
> the end newline.

Correct. The '?' modifies the previous "atom"  (.*) to match
as few instances as possible. This guarantees that '(.*?\n)\n'
will match up to (and including) the first blank line.

> Now, my question is;  how is it that this would work as there are 
> intervening newlines in the respective paragraphs?  I was trying to 
> compose a regexp that would detect two adjacent newlines and therefore
> assume it had found the blank line between paragraphs.  I see no )
> interpolation here, unless that's how the question mark works.  :)

$third may contain multiple paragraphs, if that's what you're asking.
You can use the same technique to break those paragraphs apart.

Dave
____________________________________________________________________________
Dave W. Smith          mailto:dws@best.com          http://www.best.com/~dws

Dave W. Smith <dws@shell3.ba.best.com> wrote:
> lordbear@NOSPAMbest.com wrote:
>> Let's see if I understand this and can therefore translate into 
>> human-readable English...  please correct me if needed.  :)

>>>        # ignoring the initial and trailing blank line, separate
>>>        # $_ into a first line, a second paragraph, and a third
>>>        # paragraph
>>>        ($first,$second,$third) = /^\n(.*?\n)(.*?\n)\n(.*)\n$/s;

>> Beginning at the first character, match (and ignore) the immediately
>> following newline character.  The first grouping contains any number of
>> characters up to a newline.  [What does the '?' do?]  The second grouping
>> follows the same rules.  The third matches the rest of the characters up to 
>> the end newline.

> Correct. The '?' modifies the previous "atom"  (.*) to match
> as few instances as possible. This guarantees that '(.*?\n)\n'
> will match up to (and including) the first blank line.

Based on my obviously mistaken understanding of such things, I would
expect that the '?' would have to trail the '\n' in order to pick
up multiple lines before a blank line.  e.g., '(.*\n?)\n'

I think this because I believe that since each line has a newline at the 
end, that newline has to be included within the match conditions.  

I guess I get fuzzy at the 'breakpoints' with regexps; when is a match
'entity' complete?  The 'atom' plus quantitative modifiers?  And so on.
Of course, I don't even fully understand the use of parentheses.
(sigh)  So much to learn  :)

Bear

>> Now, my question is;  how is it that this would work as there are 
>> intervening newlines in the respective paragraphs?  I was trying to 
>> compose a regexp that would detect two adjacent newlines and therefore
>> assume it had found the blank line between paragraphs.  I see no )
>> interpolation here, unless that's how the question mark works.  :)

> $third may contain multiple paragraphs, if that's what you're asking.
> You can use the same technique to break those paragraphs apart.

I was trying to express that I was concerned about the fact that
each line contains a newline at the end.  Therefore, I didn't know 
how to tell it to 'match until a newline is followed immediately by
another newline.' I was going for something like (.*\n)\n.  

Bear

-- 
-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
Don 'Bear' Wilkinson      San Francisco / Daly City     bear@lair.org
     Computer support professional seeking full-time employment;
 review my resume and interests at http://www.lair.org/bear/work.html

In article <36506b2a$0$12761@nntp1.ba.best.com>,
>> Correct. The '?' modifies the previous "atom"  (.*) to match
>> as few instances as possible. This guarantees that '(.*?\n)\n'
>> will match up to (and including) the first blank line.
>
>Based on my obviously mistaken understanding of such things, I would
>expect that the '?' would have to trail the '\n' in order to pick
>up multiple lines before a blank line.  e.g., '(.*\n?)\n'

Sounds like you may have been using the original camel book as a
reference (the one that describes perl version 4).  The magic with
"?" after "*" came with perl version 5.

  /x(.*)\n(.*)/	= match "x" and then the maximum number of characters
		  that leaves a newline left over.  If the string has
		  multiple newlines, they get matched as well.  This
		  puts everything after "x" (including the embedded
		  newlines) into $1, leaving $2 = "".

  /x(.*)?\n(.*)/ = match "x" and then do a non-greedy match of any number
		  of characters, stopping at the first newline.
		  In this case, $1 gets just one line, $2 gets all the rest.

	-Joe
-- 
INWAP.COM is Joe and Sally Smith, John and Chris O'Halloran and our cats
See http://www.inwap.com/ for PDP-10, "ReBoot", "Shadow Raiders"/"War Planets"

Smith and O'Halloran <inwap@best.com> wrote:
> In article <36506b2a$0$12761@nntp1.ba.best.com>,
>>> Correct. The '?' modifies the previous "atom"  (.*) to match
>>> as few instances as possible. This guarantees that '(.*?\n)\n'
>>> will match up to (and including) the first blank line.
>>
>>Based on my obviously mistaken understanding of such things, I would
>>expect that the '?' would have to trail the '\n' in order to pick
>>up multiple lines before a blank line.  e.g., '(.*\n?)\n'

> Sounds like you may have been using the original camel book as a
> reference (the one that describes perl version 4).  The magic with
> "?" after "*" came with perl version 5.

Are you clairvoyant, too?  :)

>   /x(.*)\n(.*)/	= match "x" and then the maximum number of characters
> 		  that leaves a newline left over.  If the string has
> 		  multiple newlines, they get matched as well.  This
> 		  puts everything after "x" (including the embedded
> 		  newlines) into $1, leaving $2 = "".

>   /x(.*)?\n(.*)/ = match "x" and then do a non-greedy match of any number
> 		  of characters, stopping at the first newline.
> 		  In this case, $1 gets just one line, $2 gets all the rest.

All I can say is:  Wow.  

Just when one thinks they know even a little, they are reminded of just
little they even can understand.  :0

Thanks!!! for your help (everyone).
Bear

-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
Don 'Bear' Wilkinson      San Francisco / Daly City     bear@lair.org
     Computer support professional seeking full-time employment;
 review my resume and interests at http://www.lair.org/bear/work.html