Populating a hash from a file.

[Kelicula]

salutations!
I am a self taught programmer. I am moderately competent with javascript, and moving into PERL.

I hope this isn't a stupid question but I have been googling, and reading tutorials for two days now, and i just can't seem to get this done.

I have a form(survey) that sends three things to a perlscript
1) the users IP
2) the time of day
3) the users choice (in the survey)

It can be seen here:
http://www.liveaudiomag.com/survey.shtml

If you make a choice and click submit, you will be taken to a page that displays the info gathered. I did that just to see if I was getting the info.

Now what I need is to put the ip, and time into a hash and then write it to a file.

for $i (param()){
%stuff = ("$i"=>"param($i)");
}

open (INFO, ">>results.txt");'

print INFO %stuff;


so that later I can scan the file for their ip if it's in there then i can see if they have voted within 24 hrs, and display a message (can't vote more than once during 24 hrs.)

if it's not there just write it.

ANYWAY I am having all kinds a problems.
I am using ActiveSate perl for win32 to test my script on my PC, the server I use is Linux based.

What I need is a "real person" to bounce ideas off and disscus this script with.
I am only asking for this as a one time thing. I think there are just some inherent missunderstandings here. If i can get this program to function properly. I will have conquered enough challanges to be able to write the scripts on my own from here out.

I just need a little help on this (my first) script?


anyone?

Anyone available (by email, chat, ect...) for a couple days of corresspondance to help get this newbie perl scripter off the ground?


thanks!!

[KevinADC]

I will be glad to help, but here on this forum. I don't do chat or IM or email help.

Instead of doing something awkward like this:

for $i (param()){
%stuff = ("$i"=>"param($i)");
}
 
open (INFO, ">>results.txt");'
 
print INFO %stuff;

use the power of the CGI module. You can save form data directly to a file as is covered in the CGI documentation.

saving the state of a script to a file

This might or might not work "out-of-the-box" for your needs. If it does not you can always write yor own "logging" function to save data in the manner that best suits your requirements. Generally that would be a delimited file of some sort, CSV or other type of formatting.

Keep in mind, all perl based CGI sctipts must print/return some type of output to the browser, even if its a 204(?) no-content header and a blank page. Otherwise you will get a 500 internal server error as the default reply from the http server itself.

[Kelicula]

Ah ha! Thanks. That has been happening (500 error).
I must look further into the CGI mod.

Ok thanks!

[KevinADC]

Also, your code would not have worked anyway:

for $i (param()){
   %stuff = ("$i"=>"param($i)");
}

param() is a function, when you double-quote a function call it is just treated like a literal string so if $i were 'cat' you would have ended up with a key/vaue pair like this:

cat = 'param(cat)'

instead of:

cat = 'black' (assuming 'black' was the value of the 'cat' field).

Plus the hash would have only had one key/value pair, the last one returned from looping through param(). You don't populate a hash the way you did, first you declare a hash then populate it with data:

my %stuff = ();
for my $i (param()) {
   $stuff{$i} = param($i);
}
 

[Kelicula]

OK, I found out last night about the double quotes around param. Also I was having trouble with returning the "time" value because it holds spaces in it.
I made it work like this.

$entry = param(ip)."\t".param("time")."\t".param(choice)."\t";

I hand to quote the time then it returned the value. I am trying to create a DB of sorts. later I use:

print "$entry"."\n";

After opening my headers. I have been able to open a file, and append the new "entry" to it . Now I am worried though. This file "should" get extremely large. Also I am concerned with it being changed, while someone else is trying to access it. Can flock() help? or is that just for use within the package running.

I am planning on running through the file in a loop for pattern matching, to reduce overhead (memory) like:

foreach $line (@stuff){\n
m/$ip/$line/;
}

(that is very crude, but I'm just trying to show the concept)

But still I'm now wondering if this is even the best algorithum at all....

proably not.

any suggestions? I mean I can learn the code, but how should I "go about" accomplishing this?

Here is my current algorithum: (which I havn't completed yet)

gather ip, time, choice..
open file load old ip's and time's into hash. key= ip, value=time
scan file for ip, if found check time, see if greater or less than 24 hrs from current time.
if ip not found = append items to file, and proceed
if found and greater :
overwrite old time, and proceed
if found and less than:
disply error page; "can't vote more than once in 24 hrs";

the proceed part;

scan choices for matching results, create counters added a "click" each time certain choice is found.

create page saying thanks for voting.
so and so percent people voted for "whoever";
so and so people voting for "next guy";

ect...

Is this completely awkward? Like I say I have never done anything but a geustbook with perl before, and that was a LONG time ago.
I am really hoping this challange will, get me more experinced, and capable with perl.
It's like a project.

?????

[KevinADC]

OK, I found out last night about the double quotes around param. Also I was having trouble with returning the "time" value because it holds spaces in it.
I made it work like this.

$entry = param(ip)."\t".param("time")."\t".param(choice)."\t";

Puttng quotes around a form field name that does not have spaces in it does nothing for you. "time" is a single word it should not need any quotes.

I had to quote the time then it returned the value. I am trying to create a DB of sorts. later I use:

print "$entry"."\n";

After opening my headers. I have been able to open a file, and append the new "entry" to it . Now I am worried though. This file "should" get extremely large. Also I am concerned with it being changed, while someone else is trying to access it. Can flock() help? or is that just for use within the package running.

flock() can help but as far as I know, it is not fool-proof.

I am planning on running through the file in a loop for pattern matching, to reduce overhead (memory) like:

foreach $line (@stuff){\n
m/$ip/$line/;
}

(that is very crude, but I'm just trying to show the concept)

But still I'm now wondering if this is even the best algorithum at all....

proably not.

any suggestions? I mean I can learn the code, but how should I "go about" accomplishing this?

The way to process a large file without using too much memory is:

open(FH,"thefile") or die "$!";
while(<FH>){
   do something
}
close(FH);

or use the Tie::File module.

Here is my current algorithum: (which I havn't completed yet)

gather ip, time, choice..
open file load old ip's and time's into hash. key= ip, value=time
scan file for ip, if found check time, see if greater or less than 24 hrs from current time.
if ip not found = append items to file, and proceed
if found and greater :
overwrite old time, and proceed
if found and less than:
disply error page; "can't vote more than once in 24 hrs";

the proceed part;

scan choices for matching results, create counters added a "click" each time certain choice is found.

create page saying thanks for voting.
so and so percent people voted for "whoever";
so and so people voting for "next guy";

ect...

Is this completely awkward? Like I say I have never done anything but a geustbook with perl before, and that was a LONG time ago.
I am really hoping this challange will, get me more experinced, and capable with perl.
It's like a project.

?????

Hard to say if it is awkward. All you posted was a basic description of what you want to do, not an algorithim. It all depends on how you go about doing it.

[Kelicula]

Hard to say if it is awkward. All you posted was a basic description of what you want to do, not an algorithim. It all depends on how you go about doing it.

Ok. I'll run it by you in a couple days. I'm going to attempt to finish ALL the operations. Thanks for your advice!

I definately wouldn't be at this point without you.

I'd still be trying to populate hashes like this:

for $i (param()){
%hash{$i} = "param($i)";
}
 

heheheee