Just wanted first off to thank everyone for all their help in this forum. Without this forum I would have been lost plenty of times. Been using Links since 1998 and I couldn't live without the mods and the hacks :)

Anyways, I have looked through these forums for 4 days now... tried EVERYTHING I could find, read all the FAQ's, looked at all the problems the mods have had and found nothing on this problem that worked... so I am posting, this jump.cgi is driving me crazy.

The problem is this. I have my jump.cgi modified to jump to the detailed pages from the category pages instead of an outside URL. My database is setup so there is no outside URL's for any of the links, the URL part of the database is blank. The jump.cgi works perfectly for going from the category listing to the detailed pages but when I try to do a random link it gets a file not found.

This is my jump.cgi:


# =======================================================
#
# Form Input:
# '$db_key' = key number # Send as form input the key name and key value
# # of the link you want to go to.
#
# Setup:
# Make sure the require statement below points to the config file.

# Required Librariers
# --------------------------------------------------------
eval {
($0 =~ m,(.*)/[^/]+,) && unshift (@INC, "$1"); # Get the script location: UNIX /
($0 =~ m,(.*)\\[^\\]+,) && unshift (@INC, "$1"); # Get the script location: Windows \

require "admin/links.cfg"; # Change this to full path to links.cfg if you have problems.
require "$db_lib_path/db_utils.pl";
require "$db_lib_path/links.def";
};
if ($@) {
print "Content-type: text/plain\n\n";
print "Error including libraries: $@\n";
print "Make sure they exist, permissions are set properly, and paths are set correctly.";
}

# ========================================================

eval { &main; }; # Trap any fatal errors so the program hopefully
if ($@) { &cgierr("fatal error: $@"); } # never produces that nasty 500 server error page.
exit; # There are only two exit calls in the script, here and in in &cgierr.

sub main {
# --------------------------------------------------------
my %in = &parse_form();
my ($goto, $id, $delim, $time);



$id = $in{$db_key};
$delim = quotemeta($db_delim);
$time = time();

if ($id eq "random") {
my ($count, $rand, $find);

# Pull out the total number of links.
open (COUNT, "<$db_hits_path/index.count") or &error ("unable to open index count file: $db_hits_path/index.count. Reason: $!");
$count = int <COUNT>;
close COUNT;

# Get the random line from the url lookup database.
srand;
$find = 0; $rand = int (rand ($count + 0.5)); ($rand == $count) and ($rand--);
open (URL, "<$db_url_name") or &error ("unable to open url database: $db_url_name. Reason: $!");
while (<URL>) {
$find++ == $rand or next;
/\d+$delim(.+)/o or next;
$goto = $1;
last;
}
close URL;
$goto or &error ("Can't find random line: $rand.");
}
elsif (exists $in{$db_key}) {
# Make sure this is a valid looking id.
($id =~ /^\d+$/) or &error ("Invalid id: $id");



# Let's get the URL.
open (URL, "<$db_url_name") or &error ("unable to open url database: $db_url_name. Reason: $!");
while (<URL>) {
(/^$id$delim(.+)/o) or next;
chomp ($goto = $1);
last;
}
close URL;
$goto or &error ("Can't find link id: $id");


# Bump the counter one.
if (open (HIT, "<$db_hits_path/$id")) {
my ($count, $old_time, @IP, $ip, $visited);
chomp ($count = <HIT>);
chomp ($old_time = <HIT>);
chomp (@IP = <HIT>);
(($time - $old_time) > 21600) and (@IP = ());
foreach $ip (@IP) {
$ip eq $ENV{'REMOTE_ADDR'} and ($visited++ and last);
}
if (!$visited) {
push (@IP, $ENV{'REMOTE_ADDR'});
$count = $count + 1;
open (HIT, ">$db_hits_path/$id") or &error ("Can't open for output counter file. Reason: $!");
if ($db_use_flock) { flock (HIT, 2) or &error ("Can't get file lock. Reason: $!"); }
local $" = "\n";
print HIT "$count\n$time\n@IP";
close HIT;
}
}
else {
open (HIT, ">$db_hits_path/$id") or &error ("Can't increment counter file. Reason: $!");
print HIT "1\n$time\n$ENV{'REMOTE_ADDR'}";
close HIT;
}
}
else {
&error ("No link specified!");
}


# Now let's send the user to the url..
$goto ?

print "Location: $build_detail_url/$id.html\n\n" :

## print "Location: $goto\n\n" :
&error ("Record not found ($in{$db_key})");
}

sub error {
# ------------------------------------------
#
print "Content-type: text/plain\n\n";
print "Error: $_[0]\n";
exit;
}



It is jumping correctly, just not when it tries to do a random jump.

I am on a RAQ3 if that matters, and I have installed a lot of mods (and hacked a bit), but everything is working lovely except for the jump.cgi random problem.

Any help with this would be greatly appreciated. A girl can only do so much before pulling her hair out, and I am starting to yank on it....

:)
Tyranny

Yep, same issue.

Might anyone know how to get random jumps to find the Detailed Pages, rather than random outside links?

Thanks very much.

DT

Hi, just bumpin this one up, too.

Thanks.

DT

If url.db is empty then that will be your problem. You will need to modify the code in jump.cgi to search the detailed url's rather than url.db for it's random link.

Paul Wilson.
Installations:
http://www.wiredon.net/gt/

Thanks, Paul. I'll give it a shot. Much appreciated!

DT

Hi Paul

I've tried a bunch of things, but nothing is working so far. What would we have to change to get the script to look in /Detailed, rather than in url.db?

I think that the following is that part that needs some adjustment:

I've tried changing URL to DETAIL and $db_url_name to $db_detail_name and to $build_detail_url, but no go.

Another possibility would be to build a detail.db within /data. This would be especially great if it would build detail.db whether you were using the default location for Detailed, where all detailed pages go into one folder (root/Detailed), OR if you were using the Detailed mod by junko, where the detail pages go into a sub-directory of each respective category, (root/Category/Detailed/) - http:// http://www.findinghim.com/links_mods/detail_cat/

I also tried to populate url.db with all the Detailed URLs, but if you have 1000 URLs, it gets to be a bit of a job plugging that stuff in.

Anyway, thanks for your help.

DT

You would need to add a new sub in nph-build.cgi similar to build_url_index to build the detailed index and then change jump.cgi to search detailed.db instead of url.db

It is not as simple as what you tried....

Alternatively you could forget all that and just get jump.cgi to search the detailed url field in links.db rather than creating a new database.


Paul Wilson.
Installations:
http://www.wiredon.net/gt/

That's the best way. Would it be something like this:

I gave this a try, but I got a 404. It said:

So, I think I'm close. Any suggestions on how to get jump.cgi to find the URL field in links.db?

Many thanks for your help.

DT

Here's a link to something that junko did:

http://www.camelsoup.com/?view=info_jump

Thanks, junko !

DT

Okay, I've got the random jump.cgi directed toward Detail pages, instead of outside urls. What I did is the following:

1. In links.def, in # Field Number of some important fields, add:

2. In nph-build.cgi, go to sub build_url_index and find:
Comment that line out. Then right below the line above, if using a single Detailed directory, add the following:

If using the Detailed Under Category hack, use the following instead:
So, now random jumps can go to the Detailed pages.

However, I want to set it up so that my visitors can select to go to EITHER a random Detailed page OR a random External Link.

So, what I need to do is explore what Paul was saying above to get nph-build to build a separate detail.db for the local jumps as well as the url.db for external jumps.

I have copied and renamed sub build_url_index as sub build_detail_index and substituted DETAIL for URL within the sub. I also created a blank detail.db and installed it in the data directory. I also deleted the "Build the Count file..." stuff because if you're keeping sub build_url_index, that builds it once and that is all it should be. If the "Build the Count file..." is not commented out or deleted from the new sub, then you DOUBLE the count file and the random jump is looking for higher file numbers than actually exist. But, when I run it, I get an error message stating that

Here's what I have for sub build_detail_index:

How do I get nph-build to build detail.db using sub build_detail_index?

Thanks very much for any help. I hope my little hack above helps some folks.

DT

This means that the folder does not exist. You need to create a folder that you defined for $db_detail_name in the links.cfg file. Also, this file needs to be chmod 666.

Also, rather than building a second index file, you could consider adding conditional statements in the jump.cgi like the following:



Then in your header/footer files or template files, use the following codes:



This will be less intensive on the build script and also reduces having to add new fields (like isDetailed or isLocal) and then updating the database file.

Regards,

Eliot Lee

Thank you very much, AnthroRules. This was the last piece that I needed to get this to work. For this random jump to detail function, I simply copied and renamed jump.cgi as random.cgi. I use random.cgi solely to jump to randomly selected Detail pages. So, I don't know if it also does the other jump.cgi functions.

I went ahead and created the extra db since I was already way down that road already, but I'm going to try your other suggestion as well so that I can economize.

Here's what I did:

1. added sub and other stuff to nph-build
2. added line to links.cfg
3. copied and renamed jump.cgi as random.cgi
4. change link code to /random.cgi?ID=random

For anyone who wants to try this:
Please make backups of all your stuff before trying this.

nph-build.cgi
In sub build_staggered, right below this:
add this:

Then, in sub build_all, just below this:

add this:
then, find sub build_url_index. Right after that entire sub, paste the following sub and choose which format your DETAIL address will be (look in this sub for "print DETAIL" (without quotations):

links.cfg
Scroll to the bottom and find

then, below that, paste:

Then, create a blank file called detail.db and place it in /data

Next, what I did is copy and rename jump.cgi as random.cgi. If you choose to do that, then rename your any links that you want to jump to a random detail page as /random.cgi?ID=random

I think that's it. It's a very easy mod, but make sure that you don't copy over the wrong stuff. I prefer commenting things out, rather than deleting. You can always throw something out later. Just make sure you get it working first.

Many thanks to AnthroRules for supplying the missing link.

HTH

DT

I've used the solution proposed in this thread to get my jump.cgi to jump to detailed pages rather than outside URL's. This is working.

Now I'm trying to get the random link to do the same. I've applied DogTag's solution, but it isn't working yet. The random link that's getting built is "mydomain.com/Detailed/random.html" rather than "mydomain.com/Detailed/3.html". It's putting in the word "random" instead of the file name. I tried to carefully implement the changes, but obviously I've goofed somewhere. Can someone point me in the right direction. I have to plead ignorance here - I'm a newbie. Thanks.

Hi cco

I think it's this spot that controls that:

[$db_category]/Detailed/$values[$db_id].html\n"; ## IF USING DETAIL UNDER

See the $values[$db_id] - that's the guy, if I remember correctly

Hi cco

I think it's this spot that controls that:

[$db_category]/Detailed/$values[$db_id].html\n"; ## IF USING DETAIL UNDER

See the $values[$db_id] - that's the guy, if I remember correctly