PHP - Php Coding Good Practice - Info

I'm trying to learn how to code the proper way.

There's a few things im worried about,but overall im worried about clogging up the server.  so what should i keep in mind while coding, to avoid this?

I'm thinking, keep insert and update to a minimum?  as in, when possible, use session to keep track of data until i absolutly have to add that data to a database? is that a correct thought?

what about once the data is in the database, how much trouble is it to ask for the selected data inside a database?  should i be worried about overdoing request of data from databases?   

on the other hand, should i be worried about having too many sessions travelling around with a user.  why?

Im just asking to get a better feel for how to do things the proper way, i would hate to have a website built, only to find out that what i built is only a bat, that would bash the hell out of a server i wanted to put it on.

Hi,    I study a php book and every thing is going fine, the problem is that I dont know how to practice what I study...   iv'e searched alot in the internet of exercise and found few but not every thing there is the same as what I study and sometimes they ask me to do things that I dont know yet..   some people saying that you need to think about a project and start building it but I dont have enough knowledge so start build something on my own..   right now I know about if/else, loops, operators, functions and objects, arrays, date and time, and file handling.   so most of the exercise I find involving mysql that I dont know yet.   can someone give me tips maybe how he started to study php?    it will help me alot and im sure it will help to more people.     thakns in advance,   guy.

Hi,

(Apologies in advance if this is a total newbie question - if it is, just give me some idea on what I need to search for!)

I have written php code that essentially pulls together multiple records (using mySQL) and displays them in a webpage. The challenge I have now is making the output pretty.

What is the best way to do this?

I have my css file (with several styles in the format below):
Code: [Select]
.textstyle0 {font-family:Tahoma;font-weight:normal;font-style:normal;font-size:48px;text-decoration:none;color:#ffffff;}
div.Object50 { position:absolute; padding-right:5px; top:189px; left:92px; z-index:0; text-align:left; width:437px; }

And my html file (that I want to bring in the php output):
Code: [Select]
<div class="Object50"><span class="textstyle0">Q. THIS IS A TEST?<br></span></div>

My php code is sitting in a tidy block:
Code: [Select]
$result = mysql_query("SELECT * FROM tbl_scores WHERE ID=$f_id") or die(mysql_error());
$row = mysql_fetch_array( $result );
$f_score = $row['score'];
$f_freq = $row['freq'];
echo "test" . $f_score;

Do I need to update all my php code with the relevant css tags or can I reference the php? Or do I split up my php code?

Do I leave my php code at the top and store all the values somewhere? Then these get pulled into html?

I guess I am trying to reduce how much of the php coding I need to modify while leaving the flexibility in the design....

mysql_query("INSERT INTO people (company, lname) VALUES ('clean($_POST[Company])', 'clean($_POST['lname'])' ");

using a clean function in the query instead of doing it before hand?

Until participating and reading this topic, I had though the following code was perfectly fine. I don't claim to even be good at OOP or application design, although I would like to be lol


Anyway, the only functionality regarding google maps polylines I could think of in PHP, was encoding and decoding them, so I placed them in a static class:




<?php
/*
-- Polyline.class.php
@return string (encoded lat/lng)           - polyline::getEncoded(-2.5675, 2.5456);
@return string (encoded lat/lng's)         - polyline::getEncoded(array(-2.5675, 2.5456), array(-2.5675, 2.5456));
@return string (encoded lat/lng's)         - polyline::getEncoded(-2.5675, 2.5456, -2.5675, 2.5456);

@return array  (decoded array(lat, lng)'s) - polyline::getDecoded('zmtn_?_epn_?zmtn_?_epn_?zmtn_?_epn_?');
*/
class Polyline 
{
private static $_calls   = 0;
private static $_lastLat = 0;
private static $_lastLon = 0;
// public accessors
public static function get_encoded()
{
// $encoded to store encoded points
$encoded = '';
$args = func_get_args();
if ( is_array($args[0]) )
{
while ( list($k, $arg) = each($args) )
$encoded .= self::_encode($arg[0]) . self::_encode($arg[1]);
}
else
{
$cnt     = count($args);
if ( !$cnt )
return false;
$i = 0;
while ( $i < $cnt )
$encoded .= self::_encode($args[$i++]);
}
self::$_calls   = 0;
self::$_lastLat = 0;
self::$_lastLon = 0;
return $encoded;
}
public static function get_decoded($str)
{
$points = array();
$lat    = 0;
$lon    = 0;
while (strlen($str))
{
$lat += self::_decode($str);
$lon += self::_decode($str);
$points[] = array($lat * 1e-5, $lon * 1e-5);
}
return $points;
}
// private
private static function _encode($dec)
{
$dec = round($dec * 1e5, 0);
if ( !(self::$_calls % 2) )
{
//lon
if ( self::$_calls >= 2 )
$ndec    = $dec - self::$_lastLon;
self::$_lastLon = $dec;
}
else
{
//lat
if ( self::$_calls >= 2 )
$ndec    = $dec - self::$_lastLat;
self::$_lastLat = $dec;
}
$dec = isset($ndec) ? $ndec : $dec;
$is_neg  = stristr($dec, '-');
$dec   <<= 1;
// invert bits if negative
if ( $is_neg )
$dec = (~$dec);
//0 pad to 32 bits
$dec     = str_pad(sprintf('%b', $dec), 30, '0', STR_PAD_LEFT);
// chunk into 5 char strings and reverse
$dec     = array_reverse(str_split($dec, 5));
// or with 0x20 except last one ( add 63 to each and convert to ascii )
$c = count($dec);
for ( $i = 0; $i < $c; ++$i )
$dec[$i] = (isset($dec[$i+1]) && $dec[$i+1] & 31) ?  ((bindec($dec[$i]) | 32) + 63) : (((bindec($dec[$i])) > 0) ?  bindec($dec[$i]) + 63 : '');
// set times called
self::$_calls++;
return vsprintf('%c%c%c%c%c%c', $dec);
}
private static function _decode(&$str)
{
$shift  = 0;
$result = 0;
$i      = 0;
do { // while ascii($str[$i++]) > ascii([space] " ")
$b       = ord($str[$i++]) - 63;
$result |= ($b & 0x1f) << $shift;
$shift += 5;
} while ($b >= 0x20);
$str = substr($str, $i);
return (($result & 1) ? ~($result >> 1) : ($result >> 1));
}
private function __construct() {}
private function __clone()     {}
}
?>



not that it matters much, since the rest of the application in which this class resides went to design shit, but; is this bad practice in your opinion?

Hi. I've been learning OOP and watched a video on YouTube about a "secure" OOP PHP login script. It was pretty good, and helped me to understand some OOP approaches, but I couldn't help thinking some of it was wrong. I'm not sure so feedback appreciated!

The class starts of something like this:
Code: [Select]
<?php
class Login
{
  private $_id;
  private $_username;
  private $_password;
  private $_passmd5;

  private $_errors;
  private $_access;
  private $_login;
  private $_token;

  public function __construct()
  {
    $this->_errors = array();
    $this->_login  = isset($_POST['login'])? 1 : 0;
    $this->_access = 0;
    $this->_token  = $_POST['token'];

    $this->_id       = 0;
    $this->_username = ($this->_login)? $this->filter($_POST['username']) : $_SESSION['username'];
    $this->_password = ($this->_login)? $this->filter($_POST['password']) : '';
    $this->_passmd5  = ($this->_login)? md5($this->_password) : $_SESSION['password'];

  }


Isn't "hard" setting variables, like the POST vars in the contruct bad? Shouldn't they be passed through elsewhere?

I've learnt that OOP needs to be reusable and manageable as it's primarily the point of using OOP in the first place. I might be wrong but I noticed straight away that the above class doesn't seem reusable (in the true sense).

Hopefully i'm getting the hang of it...

Thanks

Hi guys, so I tried searching this, but I had some issues being able to find the proper wording and keywords to query. Everything I searched PHP, HTML, Coding Standard, Coding Style, Coding, Style, Standard, etc. I got varying results. With that being said, if anyone knows of a topic that already addresses this, please feel free to link it and I will gladly do the footwork.   ^ TL;DR --> I searched, but couldn't find a topic for this...   Alright, now down to the brass tacks... I am having some trouble building a function in PHP that will display the header. I believe (if I am not mistaken) that I have seen people warning against peppering PHP code with echo <HTML HERE> throughout your code. I can sort of understand why, but I admit I am not completely sure why this is. I am trying to figure out a way to display HTML without doing the following...

<?php
// --------------------------------------------------------------------------
// ADMIN BAR HANDLER
// --------------------------------------------------------------------------

/**
 * First we need to determine if the user
 * is logged in or not. After determining
 * the user's logged in status, we will then
 * choose to display the admin bar or
 * redirect the user to the login page.
 */
function admin_bar()
{
	if ( defined( USER_IS_LOGGED_IN ) )
	{
		echo "<div class='admin_bar'>
					Admin Bar Here
				</div>";
	}
	
	else if ( defined( USER_IS_NOT_LOGGED_IN ) )
	{
		header( "Location: " . LOGIN_PAGE ) ;
	}
}

?>

2 PART QUESTION (so I don't have to ask two questions 😀  )

1.)
My script has tons of PHP variables. Is it better to declare them all at the top of my page like this:
$cat = $words = $day = $lunch = $user = $address = $potato = $_SESSION["goodpeople"] = $dumbquestionList = $b = $hundredMoreVariables = "";
-or-
as needed, like this:
$cat = "";  if($day == 'Thursday') { $cat = 'time_to_feed'; }
$words = "";  $school = array('teachers'=>'boring','lunch'=>'free'); $words = $school['teachers'];
$_SESSION["goodpeople"] = ""; if($_SESSION["goodpeople"] == "" { $_SESSION["goodpeople"] = 'Gandhi'; }

2.)
What are the pros and cons of declaring variables as NULL and using " isset " versus declaring variables as "" and using ==
Example:
$cat = $words = $day = $lunch = $user = $address = NULL;
if( !isset($cat)) { $cat = 'black'; }.........................if(isset($day)) { $sleep = 'none'; }..............etc., etc.
-or-
$cat = $words = $day = $lunch = $user = $address = "";
if($cat == "") { $cat = 'black'; }............................ if($day != "") { $sleep = 'none'; }................etc., etc.

On my particular coding, I can use either the NULL/isset style, or, the ""/== style, with no errors or ill effects.... but something tells me the expert PHP coders prefer one over the other...

Thank you!!

Edited July 10 by ChenXiu

I don't really have an issue with my script... I simply want to know if it is a bad practice to create a script that recurses through your root directory and zips up everything into one large zip file... I am creating a backup deal for my site and my site is rather large...

Is this considered bad practice?

Hi Guys,

I have a query regarding best practice for performance on a web application I'm developing.

I am retrieving data via an API that I then need to display on a website.

The data retrieved all relates to a product ID. The issue I am having is that each product has a UniqueID, but also a GroupID. The GroupID is used to display the same products but of different colour/size combinations.

ie.

UniqueID GroupID Size Colour 1 1 n/a n/a 2 2 Small Red 3 2 Medium Blue 4 2 Large Green
I need to display the products in a list, but where there are multiple products in a group, I need to provide a dropdown selector for size and a dropdown selector for colour.

I'm struggling a bit with how to structure it with just the array spat out by the API, so I was considering saving the data to a MySQL database and then querying that to get what I need.

Is this a good way to go or is the double query hit going to be horribly inefficient?

Cheers!

Hey everyone,

 

I am working on a project for fun. This is a LAMP application that is going to run on my intranet server, and hold NO VALUABLE data. I am doing it just to get better at programming, and learn some best practices and techniques.

at this point, I am working on some things, and I don't know what the best practices are. Can someone help me go down the pest route? 

 

1. A log-in system and I want to include a "remember me" button. What is the best practice for this? Obviously leaving user data in a cookie is asking for trouble, so I was thinking of leaving a unique id of some sort?

2. Information in a login SESSION. What information do you put in a log in session? I have seen lots of different techniques on this. I don't really know which is the best. For the moment, I keep an array like this: array('Status'=>True, 'Username'=>'Users name"  ,'email'=>'users Email'). If a hacker can inject session information, this seems like it would be really easy to break, because they only need a user's name and email to gain access. Is there something more I should do?

3. Config file that holds Mysql Information. I made a file that contains all the values that might change over time. That way I only need to change it in one spot. In this file I have things like the Mysql Database information. Should these files be encrypted? Or can I use a .htaccess file to make sure it isn't accessible to a hacker (as I mentioned before, this isn't a project that's going live, its more of an exercise to help me learn)

 

Thanks everyone

Hi,

I have a working PHP app with limited functionality that I use daily. Some months ago I wanted to add functionality, but soon came to a conclusion I should convert my procedural code to OO PHP. So I read some books by David Powers and Matt Zandstra (the latter was sometimes a little bit too complex for me) and started converting.

I now have converted some parts of my app and created new functionality and everything was working perfectly, until I started to use methods from class A inside methods of class B. I started getting irrational errors: sometimes it worked, sometimes it returned an error.

I think I have narrowed it down to the use of __construct() and __destruct() in all of my classes, and I'm wondering what is the (better/best) solution. My classes at the moment are all built the same way:

Code: [Select]
class A {
function __construct() {
$this->_conn = @ pg_connect( "host=localhost dbname=* user=* password=*" );
}

function get_detail() {
$this->_res = @ pg_query($this->_conn, "SELECT detail FROM table WHERE ph_id='$id'" );
}

function __destruct() {
@ pg_free_result( $this->_res );
@ pg_close( $this->_conn );}
}
I thought I was doing the correct thing by instead of repeating the setting of $this->_conn inside every method, setting that variable upon instantiating the object.

But now, I have to make a $this->_conn2 and so on to circumvent the irrational errors (always the same: '5 is not a valid result resource', where '5' should be the '$this->_conn'.

Is this a bad practice, or am I doing something else wrong? I also tried 'unset( $this->_conn )' after calling the method, but that didn't seem to work.

I'm thinking of looking for a generic PostgreSQL-class so that I can establish the connection inside every method itself, rather than in the __construct() - do you think that would solve my problem?

This might be an easy question for someone who is used to OO PHP, so I'm hoping to get an answer which helps me understand the problem. thanks for all the help I can get!

Hey all.  I was curious what is the best practice when creating a user login system? I've seen them done in the following 2 ways.

First I've seen tutorials on logins where after the post data is verified against the database a username session is created and member pages are accessed if the user session is set.

Second I've seen tutorials on logins where the username session is verified against the database on every single page.

What is the best practice along these lines?

Cheers!

Hello. I'm working on a website with an editor that allows image uploading. Ideally I want to be able to develop a framework for this and use it in later projects. What I have in mind is this:
  Database has an images table Entries contain these fields: id, filename, and a short description Images are referred to by their ids in other parts of the application. That part seems simple to me, but now there are two details I need to determine: the file name, and storing the images.
Big sites like Facebook - as far as I know - parse uploads and store them all in the same format. I can see the huge security benefit there. What are the best ways of doing that? As for the file name, I know PHP has a function to generate a file with a unique name. Is there any benefit to doing that over using the id? (i.e. 1.jpg, 2.png, 3.jpg, etc)

Hi   I have a self calling script which does the following:   Stage 1: First run, it detects that $_POST is empty, and so displays a chunk of html which prompts for a password. Stage 2: When resubmitted, it detects $_POST and displays another chunk of html which prompts the user for some values Stage 3; When resubmitted the second time, $_POST is detected, along with the user values and some php is executed and a menu displayed.   So, question is this, what is the best method to output the html at stage 1 and stage 2? I have tried using echo statements and wrapping each chunk in a function = messy.  I have tried using HEREDOCS (<<<VARNAME), better but ties my HTML to my script which is a pain I am thinking to use file_get_contents("../html/chunk1.htm") this seems quite elegant and allows me to get someone else on our team to design the HTML keeping it out of my script   Thoughts and suggestions?   Thanks as ever   p.s. who pays for this site, are donation accepted?