|  |
|
| Select from the following table of contents to jump to a specific area within this README file. This is a HTML version of the README file distributed with Textcounter C++.
|
|
| Textcounter C++ | Version 1.3 |
| Copyright 1996-2002 Matt Wright | mattw@worldwidemart.com |
| Created 03/14/96 | Last Modified 03/28/97 |
| Matt's Script Archive, Inc.: | http://www.scriptarchive.com/ |
|
If you run into any problems while trying to configure this scripts, help is available. The steps you should take to get the fastest results, are:
1) Read this file thoroughly.
2) Consult the Matt's Script Archive Frequently Asked Questions:
http://www.scriptarchive.com/faq/
3) If you are still having difficulty installing this script, send e-mail to:
scripts-help@tahoenet.com
Include any error messages you are receiving and as much detail as
you can so we can spot your problem. Also include the variable
configuration block that is located at the top of the script.
Hopefully we will be able to help you solve your problems. Thank you. |
|
COPYRIGHT NOTICE:
Copyright 1995 - 2002 Matthew M. Wright All Rights Reserved.
Textcounter may be used and modified free of charge by anyone so long as this
copyright notice and the comments above remain intact. By using this
code you agree to indemnify Matthew M. Wright from any liability that
might arise from its use.
Selling the code for this program without prior written consent is
expressly forbidden. In other words, please ask first before you try and
make money off of my program.
Obtain permission before redistributing this software over the Internet or
in any other medium. In all cases copyright and header must remain intact. |
|
|
TextCounter 1.3 is a simple program which allows you to include a text counter on any web page. You can also include the date since it began counting, a link to a help page, digit padding, and more. This program was designed for anyone to use, from a user who just wants a text counter on his or her home page to a system administrator who wants to make it easy for anyone on their server to use the counter program. Multiple counters can be set up, with the need for only one program to manage them all. You can specify what directories are allowed to access this program, and can even allow certain directories or exclude certain directories from being able to use this script. Details on how to install and use this script are available below.
Version 1.3 now allows a QUERY_STRING to be appended to the CGI program call, in which case that name will be used as the file to store to rather than the DOCUMENT_URI. This allows the textcounter to work on servers which don't support the DOCUMENT_URI environment variable. I have heard some Netscape servers to not support this, and thus you may need to use the QUERY_STRING.
A Perl version of this counter is also available at Matt's Script Archive, however for frequently used programs such as a counter, C++ definitely has an advantage over Perl. I ran a few tests (nothing extremely standard) and found that the C++ version runs about 5 times faster, at about .03 seconds, compared to the Perl version which runs in about .15 seconds.
This package should have come with 2 files:
1. README - This file. Installation Instructions, Disclaimer, Copyright, etc...
2. tcounter.cpp - The C++ CGI program which does all of the work.
|
|
There are several Variables and Options you will need to configure. The instructions below provide examples and instructions of how to do so.
| | |
| | Variable Configuration
|
| |
const char data_dir[] = "/path/to/data/";
The data_dir variable should specify the path to the directory under which all of the data files will be stored. This path must end with a '/' and it NEEDS to be writable by your web server. This means that you most likely will need to chmod this directory 666. You can do this by executing:
chmod 666 /path/to/data/
It is suggested that you make a new directory for the sole purpose of holding the data files. A new data file will be created for each page you add your text counter to. You may think that this is not the best way to do this, but it is not all bad, and is beneficial in some ways:
1. If you use this system wide, it is likely that many pages will load at the same time, meaning this program would have to try and edit the main file if it was all included in one database. This file can lead to slow downs (if I locked the file each time it was called) or it could lead to mangled data if I didn't. That is one reason I chose to use separate files.
2. The files created for the data are EXTREMELY small, taking up between 15 and 30 bytes (yes, you heard correctly, bytes).
3. Access time is faster as I know exactly what file to open, rather than flipping through lines of a database if it was all in one file.
(BTW, this explanation was more for people who may have questioned why I chose to do it this way. Most of you could care less about these last 3 points) :-)
const int num_valid_uri = 0;
const char valid_uri[num_valid_uri][128] = { };
The valid_uri array allows you to allow this program to be used only under a certain directory of your server. Say your username is fred and you are on a server called host.com, therefore all of your pages reside under: http://www.host.com/~fred/. You only want those pages under that directory to be able to use this program, so you set the valid_uri array to { "/~fred" }; and the num_valid_uri to the number of directories you allow, or 1 in this case. To let your friend joe use it, by set the valid_uri variable to: { "/~fred", "/~joe" }; and num_valid_uri to 2. Or if you are a sysadmin who wants to allow everyone to use this program, simply set this array to { "/" }; and num_valid_uri to 0 along with the num_invalid_uri to 0. If your web server supports the environment variable DOCUMENT_URI, you will most likely wish to set this variable to:
const int num_valid_uri = 1;
const char valid_uri[num_valid_uri][128] = { };
const int num_invalid_uri = 0;
const char invalid_uri[num_invalid_uri][128] = { };
Most likely you will just leave this line the same if you do not wish to block access to a certain part of your server. But take the example of fred above. He decides to be real cool and open this program up to anyone on the server by setting valid_uri to { "/" }; His arch enemy bob is also on the server though, and fred despises him so much that he wants to block access to this guy, cause bob is such a jerk. So fred sets his invalid_uri to { "/~bob" }; so bob can't use his counter. MU HA HA HA. I'm sure there are other cool uses for this too. Like if you sell virtual domains and want to charge people before they can use your counter program, you put their URI in here until they pay or something. I dunno. The story was fun to write, and that's all that matters. :-)
|
| | Options
|
| |
const char show_link[] = "http://www.scriptarchive.com/";
If you put a URL into this option, then the actual number returned by the TextCounter program will be linked to this URL. This is useful if you want to link to my site (PLEASE?!) or link to a help page explaining how the user on your system can set up their own text counter. Or if you just want to have a pointless link on your number. Setting this to a null value of "" will take out the link.
const int auto_create = 1;
Suggested value here is 1, or else you will have to create data files by hand. This allows users who reside under the valid_uri array to create a new counter for their page simply by putting the Server Side Include reference into their page. Otherwise the maintainer will have to create a data file which looks like:
0 January 1, 2000
Obviously putting the correct date into the program and changing 0 to whatever you want to start the number at. This file MUST be writable by the web server meaning you need to chmod it 666. This means other users on your system can write to it too, which is another reason to allow auto-create. Auto-Create will leave it chmoded so only the web server can write to it. (usually)
const int show_date = 1;
If this variable is on, then the date on which you began the count will appear with your actual count number. It will look like:
[Count] hits since [Date]
If this is turned off, you allow users more control over their text and it will simply print:
[Count]
The user can then supply the date if they wish.
const int lock_sec = 2;
The lock_sec variable defines how long the program will wait for the lock file to be cleared out, before overwriting the current lock file. Often times, the count file would get overwritten in the older versions because there were no locks on the files, and when two users accessed at once, it messed things up.
There are now built in lock routines, but if a user stops the process or your machine gets turned off or re-booted while the lock file is still in the directory, that lock file needs to get removed somehow. The lock_sec variable tells the program how long it should wait before deciding that the lock file is not valid.
Most of the time the program should not take longer than .1 seconds to execute, but to be safe I set the default to about 2 seconds in case you're on a REALLY slow server. You can vary this depending on whether you think your system will operate much faster or slower, but it must be a whole number, and I wouldn't go below 1.
const int pad_size = 5;
You will notice if you have seen many other counters on the web, graphical or text-based, that they are often padded with zeros at the front to form a number like: 0000154. This is achieved by adding 0's to the front of the current count. In Version 1.3, you can specify how many digits long you want your number to be, so in the example above you would set pad_size to 7;
If you do not want your number padded and wish for the above example to appear as 154 in your page, then set pad_size = 1;
|
| |
This program was written to compile on a BSDI BSD/OS version 2.1 with the g++ compiler. In order to compile it and get it ready for execution:
g++ tcounter.cpp
mv a.out tcounter
You can change tcounter above to tcounter.cgi if you need .cgi extensions for CGI programs to work on your web server.
I tried to compile under MS Visual C++ Version 4.0 and received several errros. If you wish trun it under Windows and use MS Visual C++ to compile, you will need to remove reference to unistd.h at the top and remove the check_lock and cleanup subfunctions. I also received errors of allocating arrays of constant size 1, so you may need to find a work around for that. I have no idea how it compiles under other OS's, however I would like to know if it works or not for you.
If the compiling and stuff doesn't work out, an easier way of getting a TextCounter for your web page may be to go to Matt's Script Archive and download the Perl version.
|
|
Calling this program is really very simple. As I have mentioned before, you will need Server Side Includes turned on on your server before you can use this program. Talk to your system administrator or visit my Frequently Asked Questions section for more information on server side includes. If you know they are turned on, or want to try and find out, put the following code into your HTML document:
<!--#exec cgi="/url/path/to/tcounter"-->
OR
<!--#include virtual="/url/path/to/tcounter"-->
So, if I have my tcounter (or tcounter.cgi if I have to rename it for my server) program located at http://www.worldwidemart.com/scripts/demos/textcounter/tcounter, then I would put the following into any HTML document a wanted a count to appear in:
<!--#exec cgi="/scripts/demos/textcounter/tcounter"-->
OR
<!--#include virtual="/scripts/demos/textcounter/tcounter"-->
Version 1.3 also allows the program to be called with a QUERY_STRING, unless valid_uri has been set. If valid_uri has been set, your system supports DOCUMENT_URI anyway, and this is not needed. However, for servers that don't support the DOCUMENT_URI environment variable, you can call your program as:
<!--#include virtual="/scripts/demos/textcounter/tcounter?unique_id"-->
Where unique_id is a string of characters that will be used as the filename for your count. The QUERY_STRING option could also be used to keep a count of your entire site. For instance, if you put the following on all your pages:
<!--#include virtual="/scripts/demos/textcounter/tcounter?unique_id"-->
Where unique_id stays the same on each page you place the counter on, that same count file woudl be used to count all your web site's web pages and you could have a counter for your entire web site.
|
|
| Version 1.0 03/16/96 | - TextCounter Created and Released. |
| Version 1.1 04/25/96 | - @valid_referer array and checking removed. Because server side includes can only be used locally, it is unnecessary. Also, it was causing many counters to incorrectly display error messages. |
| Version 1.2 05/10/96 | - File Locking Procedure added.
- Options lock_sec and pad_size added. |
| Version 1.3 03/29/97 | - Perl version converted to this C++ version.
- QUERY_STRING support added. |
|
| |
| |
