Jump to content

Highest Reputation Content

#8885 One Mod Rewrite Rule Won'T Work

Posted by HartleySan on 28 July 2012 - 10:59 AM

If "hiv" is never followed by an e, you could change the one regex as follows:
  • 4

#6644 Just Started This Book Today!

Posted by rob on 26 March 2012 - 10:53 AM

It does affect something it affects the source code layout, which is the whole point of using it, to make source code more readable.
  • 4

#3208 Some Tips When It Comes To Mysql Queries

Posted by Antonio Conte on 20 August 2011 - 11:19 PM

1. Always name your tables the same way:
There are best practice rules for naming tables. They should always be lowercase, split by underscore ( _ ) and named in plural. If you need to build tables for several purposes, (forums, shops, fruits) prefix them so the appear next to each other.


2. Use normalization rules(!!!) when creating a structure for tables
Larry explains the different forms of normalization very good in this book. Read it thoroughly, UNDERSTAND it, and plan your tables well. The rules are really not that hard to understand, and will allow you to cross-reference tables in an easy way later on. It will make you understand how the data you are using are working. It will make your systems a lot easier to build on and to introduce new features. I promise you, THIS is how you become a wizard when it comes to working with data in several tables.

3. Use white space(!) and use UPPERCASE for mysql functions (See tip #4!)
When your queries becomes increasingly more complex, you should really follow these tips. To demonstrate, I will give you two versions of the exact same code:

CREATE VIEW view_goals_per_game AS

SELECT league.season AS season,
league.id AS league_id,
league.name AS league_name,
goals.match_id AS match_id,
clubs1.name AS hometeam,
clubs2.name AS awayteam,
players.id AS player_id,
CONCAT( players.fornavn, ' ', players.etternavn ) AS player_name,
goals.goal_time AS goal_time,
games.kickoff_time AS kickoff_time

FROM abc_players_goals AS goals

INNER JOIN cnk_soccer_games AS games ON ( goals.match_id = games.id )
INNER JOIN cnk_soccer_clubs AS clubs1 ON ( games.club1_id = clubs1.id )
INNER JOIN cnk_soccer_clubs AS clubs2 ON ( games.club2_id = clubs2.id )
INNER JOIN abc_players AS players ON ( goals.player_id = players.id )
INNER JOIN cnk_soccer_league AS league ON ( league.id = games.league_id )

LIMIT 0, 1000

create view view_goals_per_game as select league.season as season, league.id as league_id, league.name as league_name, goals.match_id as match_id, clubs1.name as hometeam, clubs2.name as awayteam, players.id as player_id, concat( players.fornavn, ' ', players.etternavn ) as player_name, goals.goal_time, games.kickoff_time from abc_players_goals as goals inner join cnk_soccer_games as games on ( goals.match_id = games.id ) inner join cnk_soccer_clubs as clubs1 on ( games.club1_id = clubs1.id ) inner join cnk_soccer_clubs as clubs2 on ( games.club2_id = clubs2.id ) inner join abc_players as players on ( goals.player_id = players.id ) inner join cnk_soccer_league as league on ( league.id = games.league_id ) limit 0 , 1000

Whick one would you like to maintain? :blink:

4: Save your queries in a text editor

Yes! It sound idiotic, right? It's not. Think of this happening: You accidentally delete, modify or overwrite a query you've used a lot of time on. It will save you a lot of time, tears and the life of a few keyboards! This is also leading up tip number 5.

5. Use views instead of customizing your dataset in PHP(!)
This is a real life-saver. Think of it as including ONE central PHP file instead on chancing 10 documents every time you make a change. The views should be written to display AS MUCH information as possible. Try to think of every scenario you may want to use the data. Views are, with a few exceptions, for displaying data; hence it's name.

NOTE: After the creation of a view, you CANNOT modify it. This is why you should follow step 4.

- Views are really easy to create:
[u][i]CREATE VIEW the_name_of_the_view AS[/i][/u]
FROM table1

The query used in tip #3 is a view. It would allow you to sort by a specific league, match_id, player_id or by kickoff_time. The view is used to display statistic about the goal scorer in a football match. (soccer for americans) The table for saving goal statistic has three rows(!). Match_id, player_id and goal_time. This is good data normalization, and minimizes redundancy and make for consistent data. That means the table players_goals need to be linked with other tables that holds data about the specific match and the specific player. This is the reason for create a view.

Don't see the point of this?
There really is one - I promise! By writing a view that is general and display a lot of data, I can write SIMPLE queries to get different results:

1. Last five goal scorers with name of both teams, player name and goal time
SELECT hometeam, awayteam, player_name, goal_time
FROM view_goals_per_game 
ORDER BY kickoff_time DESC

2. Display players with most goals in descending order
SELECT player_name, COUNT(*) as goals
FROM view_goals_per_game
GROUP BY player_id

3. Display all goals by a specific player
 SELECT hometeam, awayteam, player_name, goal_time
FROM view_goals_per_game
WHERE player_id = 10
ORDER BY kickoff_time DESC

4. Find all goals in a specific match
SELECT hometeam, awayteam, player_name, goal_time
FROM view_goals_per_game
WHERE match_id = 837

Does it start to make some sense? Instead of repeating and modifying the same hard-to-grasp code over and over again (see tip #3), use a view to make your life simpler. :)

Hope this has helped someone. I have been thinking about writing a lot of different tips when it comes to MySQL. I've been thinking about writing a guide to joins for example. I really don't know if this interests someone or not.

Hoping others will also share tips and experiences when it comes to MySQL. :)
  • 4

#14402 Really Disturbing

Posted by margaux on 1 May 2013 - 4:56 PM

Hey xto, I'm going to try to say this in the nicest way possible -  You're in danger of trying people's patience not because you ask questions, we like questions, but because you ask questions in a way that doesn't provide the information needed to try to answer them.


1. please read the forum guidelines - Look for the little grey text bottom right of most pages, labelled Guidelines

2. please post only RELEVANT code and error messages within code tags. We don't need to see the entire output from your error message. It's actually distracting. Use code tags which are on the edit bar and they look like <>

3. post the relevant CODE, you keep posting the error message but not any code.

4. You should start a new thread for each new question. One reason for doing so is that other people with the same question can search and find your thread. If your question is part of another thread, it won't be found and won't help others.

5. You're asking questions that you should be able to solve 1 because the level of  experience for this book expects you know some basic debugging strategies and 2. with a little bit of online searching you would get some pointers as to where to look for the cause of your error.


I really shouldn't answer your question given the above but ...


somewhere in your code you are referencing an array value using 'sale_price' as the index, which doesn't exist. I'm going to hazard a guess that you have a line that includes $row['sale_price']. From the error dump you posted, you will see there is no index 'sale_price' but there is one named 'price'. Given what info you've provided that's all I can help with. If this doesn't help solve the problem, start a new thread and post the code that is causing the error :)

  • 4

#6039 Securing Video

Posted by Paul Swanson on 17 February 2012 - 5:40 PM

In addition to Rob's suggestion, you could use a .htaccess file in the directory housing the video to protect it from browsing. This will prevent any browser from accessing the directory (but PHP will still be able to access it):

# disable directory browsing
Options All -Indexes

# prevent folder listing
IndexIgnore *

# prevent access to any file
<FilesMatch "^.*$">
 Order Allow,Deny
 Deny from all

  • 3

#15289 Use Data Type Enum For Form Drop-Down Options

Posted by margaux on 2 July 2013 - 5:02 PM

Glad I could provide a little help. When I'm struggling with some code, I use var_dump and print_r to see what is being returned. Then I know how to manipulate the data.

  • 3

#15287 Use Data Type Enum For Form Drop-Down Options

Posted by margaux on 2 July 2013 - 3:10 PM

You're close. You have not stored your result anywhere with mysqli_fetch_array, so you won't be able to access the returned data.

$table_name = "collection";
$column_name = "PENDING";

echo "<select name=\"$column_name\"><option>Select one</option>";
    WHERE TABLE_NAME = '$table_name' AND COLUMN_NAME = '$column_name'";
$r = mysqli_query($dbc, $q);

$row = mysqli_fetch_array($r);

$enumList = explode(",", str_replace("'", "", substr($row['COLUMN_TYPE'], 5, (strlen($row['COLUMN_TYPE'])-6))));
foreach($enumList as $value)
    echo "<option value=\"$value\">$value</option>";

echo "</select>";

  • 3

#14255 Ch 6 'Can I Start Here?'

Posted by Antonio Conte on 22 April 2013 - 5:25 AM

I would really recommend you starting from the beginning with any book about introductory OOP. There's a solid learning curve at first when you switch from procedural to object-oriented programming, and most of that is due to the way you have to think about code. It's a very common mistake to jump straight into code, as many feel they already master coding to some degree. While that's true for many, the basic theory is insanely important, and has very little to do with coding skills. As I said earlier, it's a different approach to coding, not necessarily harder or more advanced.


I don't really remember the chapters in this book. You can probably skip some of the first chapters, but make sure you don't skip any OOP theory from the get go. While much of the stuff explained might seem below your current coding level, (and it likely is) it's fundamental in OOP. Make sure to don't blaze through theory because it seems simple. It'll pay of later to focus.


What I consider essential for understanding object-orientation:

- Theory about a Class and an Object. Read this thoroughly. Read it more than once.

- Theory about class structures and code planning (i.e what constitutes a class. This is very important and it's not obvious at first)

- Delegation. (A method should do only one specific job)

- Visibility (public, protected, private)

- Class scope and application scope.

- Polymorphism (building of existing classes)


Larry is very good at explaining all this, but make sure to take it slow from the start. It will really pay of later. Learning object-orientation is not tricky from a of coding skill level standpoint, but because of the theoretic foundation you need to write good code.


Good luck, and have fun. :)

  • 3

#14154 Multiple Mysql_Fetch_Array That Drives Me Nuts

Posted by HartleySan on 18 April 2013 - 3:12 PM

Hmmm... I'm not entirely sure what you're going for, but I'm definitely seeing some serious inefficiencies. While I don't know the sizes of your tables (i.e., how many records each one contains), you seem to be grabbing everything from the purchases table, and within that while loop, you are then grabbing everything from the other two tables where a certain value equals a certain value in the purchases table.
Also, a lot of the math you're doing (for example, adding up values for the amount of a given order) and formatting of the date can be done on the DB side, which can further speed things up.
In general, I think your goal should be to format all your data exactly the way you want to print it out to the screen by using one query. That may not be possible, and I'm not entirely sure what you want, but I'm thinking that we *may* be able to get everything you want in one query. I will attempt to do so below, but I can't guarantee that it'll work.
To start with, I'm trying really hard to figure out what exactly you're going for, but it's a bit abstract with the variables you're using. I'm not sure if you're using those variables on purpose to disguise your code for this thread, or if you're really using those variables in your code, but either way, I would definitely recommend using more logical variable names.
Anyway, here's my interpretation of your code:
You're printing out a table of purchase orders. The first column is the row number (which you do not seem to be properly incrementing within the outer while loop). The second column is the ID of a purchase order that a customer has made. The third column is the name of the customer (and it looks like you're storing their first and last name in one column in the customers table, which I would recommend against). The fourth column is the date and time of the purchase. The fifth column is the total amount of the purchase order formatted in dollars and cents. Lastly, I'm not sure what the last two columns are, but they aren't coming from DB data, so I'll ignore those for now.
Assuming my interpretation above is correct, I think you need to do an inner join across three tables with the purchases table being the main table. Also, I think you need to group your purchases together by the purchase ID, so that you can use an aggregate function to add up the price of the individual items within each specific order.
Does that make sense?
Anyway, here's the query that I'm *thinking* will work (but I can't guarantee that it does or that it's what you want):
SELECT c.cust_id, c.name AS cust_name, p.po_id AS order_num, p.cust_id, DATE_FORMAT(p.timestamp, '%m %d, %Y') AS date, o.po_id, SUM(o.delivered * o.srp) AS amount
FROM customers AS c, purchases AS p, po_content AS o
WHERE c.cust_id = p.cust_id AND p.po_id = o.po_id
GROUP BY o.po_id
ORDER BY p.timestamp ASC
LIMIT $start $display;
A few notes about the query:
1) For your query, a join is essential. Specifically, two inner joins on the purchases table is what you need. Joins are tricky at first, but they're essential for most DBs, so I'd recommend studying up on them.
2) Only select the columns you need. Using SELECT * for three separate queries is getting you a lot of data you don't need, and is very inefficient.
3) Use aliases (e.g., "AS c", "AS p", etc.) on the tables to make typing out the query shorter and easier. Also, aliases are essential for being able to easily reference the results of aggregate functions, formatted, dates, etc.
4) Format the timestamp on the SQL side using the DATE_FORMAT function. It's faster and easier. Also, give the formatted date an alias to make it easier to access on the PHP side. Here's more info on the DATE_FORMAT function:
5) I'd calculate the total amount of each order on the SQL side. To do so, you need to use the SUM aggregate function, and also use the GROUP BY clause to group your results together by order number so that you are adding up the correct grouping of items. Also, I'd assign an alias to the result of the SUM function.
6) The "ASC" part of the query is not necessary, since that's the default ordering. I left it anyway to avoid any further confusion.
7) I used "o" as the alias of the po_content table, as it seems like a table of orders to me.
That will hopefully handle the query side of things.
Unfortunately, I think there are some other issues with your code as well:
1) You're not incrementing $rownum in the while loop.
2) You're assigning your $bgcolor value to the entire table, not individual table rows. Also, I'm pretty sure the resulting HTML will be syntactically invalid and not work.
3) You're creating a new table each time through the while loop.
4) You're putting a div within a td for the total, which I wouldn't do.
Point being, without sounding too harsh, I think your code has some serious issues and needs some re-working. I get the feeling that you might be getting a bit too ambitious about your personal project without first understanding all the basics you need.
I don't mean to say that you shouldn't be ambitious, but I think you should probably go back to the book for a bit and bone up on queries with joins, HTML and PHP syntax, as well as think more about the logic of your while loop and the type of HTML that it is creating.
Anyway, below, I'm going to present the PHP I would use for your situation. Again, please keep in mind that I'm doing my best to piece together exactly what you want (and I'm not entirely sure), so I could be way off on this.
// I'm assuming that $agent and $encoder are already defined above.
$row_num = 1;
$total = 0;
$bg_color = '#FFF';
$q = "SELECT c.cust_id, c.name AS cust_name, p.po_id AS order_num, p.cust_id, DATE_FORMAT(p.timestamp, '%m %d, %Y') AS date, o.po_id, SUM(o.delivered * o.srp) AS amount FROM customers AS c, purchases AS p, po_content AS o WHERE c.cust_id = p.cust_id AND p.po_id = o.po_id GROUP BY o.po_id ORDER BY p.timestamp ASC LIMIT $start $display;";
// I'm assuming that $start and $display are already defined above.
$r = mysqli_query($dbc, $r);
// I am assuming that $dbc is already defined above. You also seem to have omitted this argument in your code.
echo '<table class="order_details">';
// I'd use CSS to properly format the table instead of the inline attributes you're using.
// As such, I have assigned a class to the table for that exact reason.
// Also, your table doesn't have any headers, but you may want to add them along with thead, th, tbody, and tfoot tags.
while ($row = mysqli_fetch_array($r, MYSQLI_ASSOC)) {
  echo '<tr style="background-color: ' . $bg_color . ';">
  <td>' . $row_num . '</td>
  <td>' . $row['order_num'] . '</td>
  <td>' . $row['cust_name'] . '</td>
  <td>' . $row['date'] . '</td>
  <td>' . number_format($row['amount'], 2) . '</td>
  <td>' . $agent . '</td>
  <td>' . $encoder . '</td>
  $row_num++; // Don't forget to increment this.
  $total += $row['amount']; // This is the summation of the unformatted amounts, which could cause issues.
  $bg_color = ($row_num % 2 === 0) ? '#F3F3F3' : '#FFF'; // Ternary operation for brevity
echo '<tr style="background-color: ' . $bg_color . ';">
<td colspan="4" class="total_row">Total</td>
<td>' . number_format($total, 2) . '</td>
// Formatted to line up with the amount column.
// Also, note that I handled the total the same way you did, but if you calculate the total on the
// unformatted amount values, then you may get a discrepancy in which the amounts don't add up to the total.
// Also, again, I'd use CSS (not inline HTML) to align "Total" to the right.
// Lastly, you may want to put the total in a tfoot element.
Well, I think that's about it.
After all this writing, I really hope that I got close to what you wanted, and that this post is of some use.
Please let me know.

  • 3

#13112 Forms In Javascript Or Php?

Posted by HartleySan on 19 February 2013 - 10:38 AM

Awesomo, to give you a concrete example, I'm going to use Jonathon's example above because I think it's a good example.
First off, you'll want to attach a JS event handler to the form so that when the form is submitted, JS intercepts the form submission and allows you to do whatever you need to before the form is actually submitted.
This can be accomplished as follows:
document.forms[0].onsubmit = function () {
  return false;
On any HTML page that contains at least one form, JS provides you with the forms array, which is a property of the document object. forms[0] always applies to the first form on the page. You can adjust the array index as need be.
The return false statement actually stops the form from being submitted. (Don't worry though, as we'll actually force the form to be submitted in a second, assuming everything is okay.)
Next, let's imagine that we have a user name text input that must be only letters and numbers, as Jonathon said. Here's a simple JS regex we can write to test for that:
document.forms[0].onsubmit = function () {
  if (/^[A-Za-z0-9]+$/.test(this.elements[0].value)) {
  return false;
The above if statement tests that the user name text input contains only letters (either uppercase or lowercase) and numbers (and is not an empty string). I should note that I make several assumptions with this code:
The user name text input is the first input in the form. This can be seen by the this.elements[0] in the if statement. In this case, this refers to the form object we're interested in, and elements is an array attached to all JS form objects that allows you to access all the inputs items in the form. Naturally, elements[0] refers to the first element in the form. If you'd rather check the user name text input by ID (assuming the ID is "username"), you can replace this.elements[0] with document.getElementById('username'). The value property at the end will give you the actual string entered into the text input object.
If the above if statement evaluates to true, then we can assume that the user name does in fact only contain letters and numbers. That being the case, we should then submit the form normally.
However, if the if statement returns false, then we should change the border of the user name text input to red.
The following code will do just that:
document.forms[0].onsubmit = function () {
  if (/^[A-Za-z0-9]+$/.test(this.elements[0].value)) {
  } else {
    this.elements[0].style.border = '#F00 solid 2px';
  return false;
As before, this refers to the form object in question. this.submit() will actually submit the form as if JS never intercepted the form submit request in the first place.
If the if statement fails, then we use this.elements[0] (or document.getElementById('username'), if you want) to reference the user name text input object, and then change the border to a 2-pixel red border.
That's the basic concept to simple JS form validation built on top of standard PHP form validation.
One of the keys to this method is that you only submit the form when everything is okay.
And naturally, the concepts above can be rinsed and repeated as many times as necessary to validate all your form input.
I hope that helps.

  • 3

#12215 When Js Isn'T Available.

Posted by HartleySan on 8 January 2013 - 6:24 PM

Unfortunately, that's the nature of the beast. With JS disabled, you lose a lot of functionality, and certainly, that's one if the downsides of relying on JS libraries to do all the work for you.

Obviously, you'd most likely want the non-JS version of your site to mirror the JS version as much as possible, but that's not always possible/practical. With that said, I think Larry says in one of his books somewhere, as long as the non-JS site works and gets the user the information necessary, that's good enough (even if the site looks/behaves like crap). I more or less agree with this, but obviously, if given the time, making a non-JS version of your site that is similar in appearance and functionality never hurts.

Here's my main COA when designing a site to have both a non-JS and JS version:

1) Plan, plan, plan! In regards to the above, this means trying to minimize JS as much as possible while still giving a good experience. By minimizing the JS used, the two sites will be more similar and it'll be less work to get both working correctly.
Unfortunately, if you're locked into something like Twitter bootstrap, then you might have your work cut out for you in this regard. This is one of the main things I don't like about a lot of JS libraries these days: they use JS for a lot of things you don't need JS for; nowadays, you can do a lot with HTML and CSS alone.

2) Use noscript tags for content that is required as a replacement for JS-generated content when JS is disabled. For example, you might make a new, second menu within noscript tags to handle navigation when your Twitter bootstrap menu is busted because JS is disabled.

3) For content that requires JS, either generate the markup for it using JS, or use CSS to hide the content by default, and then use JS to display it when the page first loads. If you don't do one or the other, then you will get JS-dependent content appearing on the screen, which is useless and confusing to users who have JS disabled (ala your Twitter menu).

4) With all the above in mind, develop the entire site first without using any JS. This will serve as a solid base. After that, add JS functionality as needed, keeping all the above in mind. Because this may be time-consuming, you may want to consider launching with the non-JS version only (since that should be enough to do whatever on your site), and then adding JS functionality here and there post-launch.

And that's basically it. Obviously, depending on the scope/size of your site, this may add considerable time to development, but that's the price you pay if you want a decent non-JS site. I think a log of people these days don't find it worth the effort to do all of the above though (so they don't).

And keep in mind that there are other things to consider these days as well. For example, "mobile first", which is a good guiding principle and strongly emphasizes loading content non-essential for the mobile version using JS. These sorts of considerations may very well conflict with designing a good non-JS site, so please keep that in mind.

Well, I've blabbered on long enough, and I probably mentioned a bunch of stuff you already knew, but there ya go.

- The other Jonathan (with a slightly different spelling)
  • 3

#11339 Ch. 11 Using Unlink() With A Reset Button Possible?

Posted by Antonio Conte on 27 November 2012 - 10:43 PM

This can absolutely be done via GET. What you would do is using IDs as the get value. Save the filenames to an array on the form ID => "name". You then build links the same way, but with ?remove=$ID. You cast $_GET['id'] to an integer, find the filename in array by using the ID, build a full path where you add the filename, then use unlink() to delete the file.

An easy way to do this would be something like:

$handle = ""; // open file

$file_array = array(); // The file names

while ( ($file = readdir($handle)) !== false )
    if ($entry != "." && $entry != "..")
	   $file_array[] = $file;
    } // Read files to array

// Simple delete
if ( ! empty($_GET) )
    $upload_dir = "uploadsDirectoryHere/"; // Upload dir
    $file_id = (int) $_GET['delete']; // File_array ID

    // Make sure file is found
    if ( array_key_exists($file_id, $file_array) )
		 $delete_file = $upload_dir . $file_array[$file_id];

		 // Make sure file exists
		 if ( file_exists($delete_file) )
			  unset(file_array[$file_id]); // Remove from file array
			  unlink($delete_file); // Actually delete file here

// Print out file names and links for deleting them
foreach ( $file_array as $id => $filename )
    echo $filename . '<a href="page.php?delete='.$id.'">Delete file</a>;

Not tested, but that should be the basic idea. Sorry about errors/etc. Beginning to get tired here.
  • 3

#11171 Php Code Allow Smartphone Applications (Iphone, Android) To Connect To My Web...

Posted by HartleySan on 18 November 2012 - 9:39 PM

If you want users to access your site from a web browser on their smartphone, then there is no additional code you need to write. The code would be the same for the desktop version and the smartphone version.
The only likely thing you'd want to change is making the smartphone version easier to view (than the desktop version) on a smartphone.

If, instead, you want to write a native app for iPhone and Android devices, then I suggest you by a book about how to write iPhone and Android apps (which use two separate languages), because PHP is not going to help you there.
  • 3

#10794 How Do You Learn To Really Write Php By Yourself?

Posted by Antonio Conte on 30 October 2012 - 7:30 PM

Programming is all about actions to data. When you have been programming for a while, you can identify that you need to do this and that to the data to achive a result.

To take your calculator example. What does a calculator do? It computes answers out of operations performed on numbers, right? That's a pretty abstract description of a calculator, but it's a start. We must then decide what operations, actions, this calculator should be able to perform. Let's say we want to able to multiply two numbers, how can we code that?

Let's make a simple plan:
1. We need a form with two HTML input fields.
2. We need to make sure the form is submitted by the user
3. We need to make sure both fields have valid values
4. We need to display error messages if not
5. We need to calculate the answer and display it to the user.

This applies to all programming. My first programming teacher forced us to always write down all steps needed to achieve the wanted result. I think you should do the same. When you want to solve the problem, sit down and write a step-by-step plan on how to solve it. You don't need to follow that plan in the end, but you'll begin to see that planning pays off in the end.

Hope that helps a bit.
  • 3

#10475 Benefits And Applications Of Ajax?

Posted by HartleySan on 16 October 2012 - 1:40 AM

Probably the most important thing to understand is that Ajax is not a replacement for PHP; it's a technology that combines JS with a server-side language (e.g., PHP) to do things that can only be done on the server side, but without the need to reload the page. That's really the whole point of it; to create a more seamless, desktop application-like experience for the user. Also, it's important to understand that JS and Ajax are not one and the same.

To more specifically answer your questions:
1) Because JS is client side, you can use JS to perform some basic form checking before submitting the form to the server and your PHP script. By doing this, the user will get an instantaneous response to any mistakes they may have made in filling out the form. This generally equates to a better, more-responsive user experience.

Some important things to note with JS validation though are that because anyone can view your JS code, you don't want to validate with only JS, as people might try to study your code to find ways around your validation for the sake of hacking your site/DB. Also, PHP has much more robust functions for performing validation and preventing DB attacks, so PHP validation should always be used and the final line of defense, regardless of whether or not you are also using JS for validation.

Some good examples of when to perform JS validation would be to check that an entered password meets the site requirements, zip codes are entered in the proper format, etc., etc. This will help prevent a majority of the honest mistakes that users might make in filling out a form.

The more you can entrust to the client side, the faster and better the user experience and the less of a strain you put on the server. Just remember to never sacrifice safety for speed.

2) Again, Ajax is not a replacement for PHP, and honestly, Ajax is never required for a site, but it makes the whole experience on the site a lot smoother and easier to use. The purpose of Ajax has always been to create a more desktop application-like experience for web pages.

To give a couple of examples of how to effectively use Ajax:
- You could use Ajax to poll (i.e., check) a DB every so many seconds, and then update content on the site whenever there is a change to the DB. By doing this, you can constantly update a site's content without the user having to refresh the page. This is good for things like stock tickers, updating the scores of sporting events, etc., etc.

- You can use Ajax to update only part of a page. For example, with most websites, when you click on a link, the new page that is loaded will likely have a lot of the same content as the old page. However, with Ajax, when a link is clicked, you can choose to change only a small part of the page while leaving everything else untouched. This ultimately puts less of a strain on the server (because less info needs to be processed/downloaded) and it'll make the new "page" load faster for the user, and without the need to actually reload the page.

With all that said, please note that there are two big downsides to Ajax:
- It "breaks" the back button in your browser in that because there are no page transitions (at least, not according to the browser), the back button will take you to the page you were on before you loaded the site, and not the previous page you were on on the site.
- Similar to the above, you cannot bookmark pages because there are no page transitions.

Over the years, people have come up with ways to circumvent these two problems, but they're all half-ass workarounds that can be tricky to implement and don't really solve the problem.

The silver lining to all this is that there is a new API in HTML5 called the History API, which allows you to properly create proper URLs for Ajax-driven sites, and thus the back button and bookmarking problems can be resolved for Ajax sites. Unfortunately, the History API is still new and does not have universal support, especially among IE browsers.

Also, in what is quickly becoming a new replacement for Ajax in some situations, there is also the HTML5 WebSocket API, which allows for true, asynchronous communication between two endpoints. This allows you to do things via a browser that before were impossible and/or very inefficient using only Ajax.
For example, you could use the WebSocket API to create a chess application that detects when your opponent makes a move and instantly updates the board on your end. You could also use the WebSocket API for things like chat applications, etc.
Gmail and Facebook already very actively use this technology for their chat applications.

Well, hopes that is somewhat the answer you were looking for.
  • 3

#11560 Simple Question On Php Form Validation...

Posted by HartleySan on 8 December 2012 - 12:32 AM

When you execute the header(Location: 'somelocation.php#contact'); line, you're reloading the page, which causes the $_POST array, $errors variable, etc. to no longer exist. This essentially amounts to the same thing as having never performed any validation at all (and thus the $errors variable not existing, which is why your "Please fix the errors" line is never printed out).

A better solution would be to put the anchor jump (i.e., the fragment part in the URL after the number sign (#)) in the URL specified for the action attribute of the contact form. For example, if the following is the opening tag for your contact form:

<form action="somelocation.php" method="post">

Then change it to the following:

<form action="somelocation.php#contact" method="post">

Of course, this assumes that somelocation.php refers to the original page the form is on (which seems to be the case according to what you said).
By doing the above, you can very easily get the jump you want and still properly validate everything.

In terms of outputting a thank you message upon successful validation, I'd do something like the following:

if (!$errors) {

  // Validation was successful. Output a thank you message.

} else {

  // Either there was an error or this is the first time the page was loaded.
  // Load the page with the form and any necessary sticky values and error messages.


Hope that helps.
  • 3