|
sivo posted:Zend_Form. This class seems to "suffer" from the ZF use-at-will design in that it blurs together aspects of the view (decorators) and the model (validation logic) - I just have no idea where to stick my Zend_Forms. Your model should just be the data and thats it. Ideally if you dropped zend framework and went with code ignitor, you should just be able to take the zend model class and maybe the zend pdo class with you and your models should work fine. If not, then stuff isn't separated like it should be. Think of it this way, you should be able to code your model so that you could satisfy all requirements of your app by running it through command line scripts and no GUI. If forms became tied to your model this would break. So stick your forms outside the model. sivo posted:To me, it looks like B is the way to go, but I can't help but feel worried that I am robbing Controllers of what seems to be, fairly often, their only responsibility (that you write yourself, at least) - that A is so popular doesn't help my confidence in B. A is the way to go. You want your views to just display data without actually doing any logic. Your controllers should do the querying and decisions. Overall though, I think I went down the same path as you previously. Just because it is included in Zend, doesn't mean you have to use it. Its good to learn how to use things and learn the theory, but more often than not using all of the framework for everything leads to pattern overuse and unnecessary complexity. Sometimes its better to just hand write forms in HTML instead of using form generators.
|
#
?
May 1, 2009 15:26
|
|
|
|
| # ? May 15, 2024 17:09 |
|
|
sivo posted:Secondly, what exactly are Controllers supposed to do? It seems like everyone agrees that Controllers should do as little as possible, but where you draw the line seems to differ for everyone. It seems to me that it's either: Anyone that is saying controllers should do as little as possible is a moron. MVC is simple: Model; this is an interface to your data storage (db, csv, inifile, punched card reader - it shouldn't matter AT ALL to anything outside the model). You should have a model per conceptual 'entity' collection in your system; Users, Articles, Images, Newsletters, Gallery Albums - these are all candidates to have a model dedicated to abstracting the work of extracting the data from storage, or making changes to it there. Controller; this is where the 'business logic' of your application should live. Any decisions being made, this is where it should happen. Controllers decide both which elements of data are required to service the request (i.e. build a page, handle a post) and which views are required to communicate the results back to the user. Controllers are the brains, they are, really, where the application lies. Views; this is where all data is packeged up and formatted for output. Note that this CAN mean 'formatted to be written to a file' if the file is output not storage (e.g. if your app creates graphs of data, the graph drawing is a view, even if it spits out a jpeg to disk). All presentation logic should be in views, and good frameworks should allow views to delegate presentation to other views. But NO framework should allow views direct access to setup or control models. Also views shouldn't be asking controllers for things; controllers should be invoking views to display/present things. If you are then changing your app, MVC should, if used right, make the job simple. Change the look of your app; change views. (i.e. go from html to pdf output; new view) Change the way your app reads/writes its data; change models. (i.e. go from csv to db storage; new model) Change the behaviour of your app; change controllers. (i.e. want to make adding a blog post save i in the db and punt a notification to twitter? change the controller to also invoke the view that renders the tweet) Note the most fundamental change to your app comes if you modify the controller. If changing from flat files to DB involves a change in the controller; you did it wrong. If changing from flat files to DB involves a change in a view; you hosed up massively. sivo posted:To me, it looks like B is the way to go, but I can't help but feel worried that I am robbing Controllers of what seems to be, fairly often, their only responsibility (that you write yourself, at least) - that A is so popular doesn't help my confidence in B. B is entirely wrong. You would basically be making the view do the job of the controller as well as the view. Also, everything I have heard about Zend Framework is that it is a bloated and puffy framwork that is overly restrictive and clunky. But has a few nice bits hidden inside it. KuruMonkey fucked around with this message at 00:28 on May 2, 2009 |
|
#
?
May 2, 2009 00:19
|
|
|
Are there any CMS systems that are as elegant and simple as Wordpress that are ideal for simple mostly-static non-blog sites? I like alot of the functionality that Wordpress provides (namely the creating pages on the fly and site-wide settings), but am just curious if there's a better option to use for non-blog sites than to just hide the blog portion of Wordpress.
|
|
#
?
May 2, 2009 00:25
|
|
|
supster posted:Are there any CMS systems that are as elegant and simple as Wordpress that are ideal for simple mostly-static non-blog sites? I like alot of the functionality that Wordpress provides (namely the creating pages on the fly and site-wide settings), but am just curious if there's a better option to use for non-blog sites than to just hide the blog portion of Wordpress. Probably CMS Made Simple. It's decent and the only real downside I've noticed is that templates and CSS are stored in the database and there is no separate storage for template images and content images. I've used Website Baker as well, but its backend is pretty crude in comparison.
|
|
#
?
May 2, 2009 01:50
|
|
|
Thanks for the replies. I'm definitely worried about going overboard with patterns/complexity, it's something I think I'm very susceptible to  (and I know it's not a good thing) It's just a bit of a struggle to understand when all the examples on the internet are very simple and descriptions ambiguous/genericSo, it sounds like I was seduced by a couple of oddball websites. But I'm not clear on a couple of things. I think my underlying problem is a lack of understanding what ways programs "change" that are supposed to be absorbed and localized. Say for instance you have a Model Person with two properties: Name, Address. If you pass this Model to the view, and the view has something like code:In either case, It doesn't seem like changing the back-end storage media would need either to change. Begby: Thanks for the warning about it sometimes being simpler to avoid framework implementations. As I mentioned before I think I am prone to doing things more complicated/awkwardly than they need to be. I think I will have to try swapping framework as an exercise to understand. But, I'm still not sure how having the forms in there is a problem. I (think I) understand what you are saying about it being desirable for the model to stand alone, it's definitely something I was aiming for... but the Models only use the forms to validate data, so loading the Zend MVC stack isn't necessary to use them. What piece of the puzzle am I missing here? I was definitely surprised by you guys' thoughts on the Controller. Is "fat model thin controller" a fringe philosophy after all? did I just misunderstand it? or is it just a matter of preference and not a rule?
|
|
#
?
May 2, 2009 05:17
|
|
|
sivo posted:Say for instance you have a Model Person with two properties: Name, Address. If you pass this Model to the view, and the view has something like Yes; because in the example you've given, its reasonable to assume you meant for the controller to have code along the lines of: code:So storage change will only involve the model. Showing an extra field from the person is presentation, and is only going to involve the view. But changing which, how many and whether the person is displayed is behavious, and is a change to the controller. Thats it; anyone who makes MVC seem more complex than that is also a moron. (and more importantly, doesn't understand what they're trying to explain to you) Even if your situation involves handling a collection (of people, for example), this will still apply: code:code:Now, this is straying into in my opinion, but the essence of keeping your MVC separation clean, neat and consistent is to only pass collections that can be handled in full to the view. So I would say that selection, sorting and pagination is the job of the MC section not the V section. That is; the controller decides what to select, how/if to sort and how/if to reduce the selection for pagination. The view then just formats and outputs what its told. sivo posted:I think I will have to try swapping framework as an exercise to understand. But, I'm still not sure how having the forms in there is a problem. I (think I) understand what you are saying about it being desirable for the model to stand alone, it's definitely something I was aiming for... but the Models only use the forms to validate data, so loading the Zend MVC stack isn't necessary to use them. What piece of the puzzle am I missing here? If you try another framework I'd suggest Kohana or CodeIgniter in that order of preference. (CodeIgniter has bette docs though, I just prefer Kohanan now I've used both) DEFINITELY have a test drive of at least one other framework, even just to look at/play with. It will let you evaluate how Zend actually is. (I've not used Zend) I think to clarify the forms issue, I would follow this train of thought. Conceptuall I think of a web based DB driven app (as a reasonable example to work on) as having a central PHP core. This handles data in 'PHP form', arrays, strings, objects etc. Off one end, there's the web pages it generates. Making these involves a translation out of 'PHP form' and into 'html form'. The view handles that job, and as far as possible, ONLY the view. Off the other end, there's the DB, and its not handling 'PHP form' either. The models handle that translation. And only the models. SO; if you're considering something that is handling stuff thats not PHP data, its likely that either a model ora view is the right thing for the job. Conceptually, for me, the decision is a case of 'is it going towards the user, or towards PHP or the DB?' if towards the user, then a view. If towards PHP or the DB, then a model. Forms are a lgitimately awkward case; I think in a perfect world (an in fact in my code) you have to split them up, mentally. You see a controller needs to know which form to use and when. A view has to render it for the user, but then I actually use a model to handle the acceptance of data from post/get, and the validation. (and somewhere in there, you need a record of what fields are in the form and what types they are for validation - this lets you generate and validate a form 'by id' if done right) Handling forms well isn't simple, because they are 2-way, basically.
|
|
#
?
May 2, 2009 08:02
|
|
|
Thanks a bunch. I think I'm clicking on to it. I realized that in that post I confused passing the model with passing data as a package. I think the original point of confusion that started it all was how to deal with related child-type data, but the idea that the controller handles which and what while the view is passive clears this up. Also thanks for the recommendation. I'll definitely look into Kohana.
|
|
#
?
May 2, 2009 09:53
|
|
|
I have a bear of a form that I can't get to function correctly. I am helping setup a conference, and we have a php form that allows people to upload their powerpoint to us. The form was created using one of those "custom forms" websites, as no one here knows enough php to do it on our own. The main problem seems to reside in the way that the form handles the uploaded powerpoint file; we would like it to enter the name of the uploaded file into our mysql database, just as all the other fields in the form, and to also save the powerpoint file itself in a designated folder in our server. Right now all the fields populate the mysql database, except for the field that has the name of the file. the database table has the column for that field, but the entries are empty. Also, the way the form is setup, the only action performed with the uploaded file is to email it to a designated account, which is pretty impractical for obvious reasons. Is there any way I could get help with this? I'm hesitant to paste the code here because of the copyright disclaimer in the code, but I can pm/email someone the php code. walumachoncha fucked around with this message at 21:08 on May 4, 2009 |
|
#
?
May 4, 2009 20:44
|
|
|
Still getting used to Prepared Statements with MySQLi and have another question... I can't figure out how to make another database call in the middle of this code to fetch more results: php:<?
if (mysqli_stmt_prepare($stmt, $sql)) {
mysqli_stmt_bind_param($stmt, 'i', $topid);
mysqli_stmt_execute($stmt);
mysqli_stmt_bind_result($stmt, $fieldid, $fieldtype, $parentid, $fieldvalue);
while(mysqli_stmt_fetch($stmt)) {
echo '<p>';
echo '<label for="'.$fieldid.'">'.$fieldvalue.'</label><br />';
if($fieldtype == '1') {
echo '<input type="text" />';
} elseif($fieldtype == '2') {
echo '<select>';
//there needs to be another database call here to pull in the option values for the select
echo '</select>';
} elseif($fieldtype == '3') {
echo '<textarea></textarea>';
}
echo '</p>';
}
}?>
|
|
#
?
May 4, 2009 20:49
|
|
|
VerySolidSnake posted:Still getting used to Prepared Statements with MySQLi and have another question... I'm a mysqli newbie myself, but wouldn't it work to just do the same code, but with a different stmt variable? php:<?
if (mysqli_stmt_prepare($stmt, $sql)) {
mysqli_stmt_bind_param($stmt, 'i', $topid);
mysqli_stmt_execute($stmt);
mysqli_stmt_bind_result($stmt, $fieldid, $fieldtype, $parentid, $fieldvalue);
while(mysqli_stmt_fetch($stmt)) {
echo '<p>';
echo '<label for="'.$fieldid.'">'.$fieldvalue.'</label><br />';
if($fieldtype == '1') {
echo '<input type="text" />';
} elseif($fieldtype == '2') {
echo '<select>';
$stmt2 = 'select * from whatever';
if (mysqli_stmt_prepare($stmt2, $sql)) {
mysqli_stmt_bind_param($stmt2, $whatever);
mysqli_stmt_execute($stmt2);
mysqli_stmt_bind_result($stmt2, $stuff2);
while (mysqli_stmt_fetch($stmt2)) echo $stuff2;
}
echo '</select>';
} elseif($fieldtype == '3') {
echo '<textarea></textarea>';
}
echo '</p>';
}
}?>
|
|
#
?
May 4, 2009 23:40
|
|
|
How do you get cURL to handle sessions? Using cURL to log into a website, the site seems to return some session data, how do I store that session data and send it again with all subsequent cURLs? Basically I use cURL to POST my username and password to a website, the website sends back a cookie and cURL stores it in the cookiejar. But the website sends back some session data, too. How do I store that? My code below shows first I log into a website, storing the cookies in COOKIEFILE, then it tries to visit a page that only logged in users can see, using COOKIEJAR. But I get rejected, most likely due to not preserving the session data. code:
|
|
#
?
May 5, 2009 01:43
|
|
|
Golbez posted:I'm a mysqli newbie myself, but wouldn't it work to just do the same code, but with a different stmt variable? I believe that will not work, because from what I understand, you cannot start another prepared statement while another is running
|
|
#
?
May 5, 2009 15:07
|
|
|
I've having a hell of a time trying to get my head around working with basic hierarchical data. I am storing a list of "pages" in a mySQL database and as you'd expect each entry has a parent_id. I know there are better methods for doing this (pre-ordered traversal trees?) but since there is unlikely to be massive amounts of entries in this table, I'd like to keep it as simple as possible. I'm trying to limit myself to using only one SQL query. I can't for the life of me figure out how to structure my array after I've retrieved the pages from the database. Any suggestions / code would be really appreciated.
|
|
#
?
May 5, 2009 23:24
|
|
|
I think I have something to fix my prepared statements. Lets say I have a query like this. php:<?
$sql = 'SELECT papers.papid AS papid, keywords.keyid AS keyid, keywords.keyword AS keyword
FROM papers,keywords
WHERE papers.topid = ?
AND keywords.topid = papers.topid';
$stmt = mysqli_stmt_init($connect);
if (mysqli_stmt_prepare($stmt, $sql)) {
mysqli_stmt_bind_param($stmt, 'i', $topid);
mysqli_stmt_execute($stmt);
mysqli_stmt_bind_result($stmt, $papid, $keyid, $keyword);
mysqli_stmt_fetch($stmt);
//echo stuff here?
}?>indulgenthipster fucked around with this message at 22:13 on May 6, 2009 |
|
#
?
May 6, 2009 22:10
|
|
|
I apologise in advance for the 'n00b'-ness of this question. I've googled but I'm still getting nowhere so you guys are kinda my last hope. php:<?php $dirPath = '.'; if ($handle = opendir($dirPath)) { while (false !== ($file = readdir($handle))) { if ($file != "." && $file != "..") { if (is_dir("$dirPath/$file")) { echo "$file <br>"; } } } closedir($handle); } ?> EDIT - ooh. I think I did it. I added if (("$dirPath/$file") != "./img") and it seems to be working. roxanne fucked around with this message at 00:04 on May 7, 2009 |
|
#
?
May 6, 2009 23:47
|
|
|
Whats a good way to encrypt and decrypt a variable with a specific salt?
|
|
#
?
May 7, 2009 04:13
|
|
|
roxanne posted:EDIT - ooh. I think I did it. I added if (("$dirPath/$file") != "./img") and it seems to be working. Thats only working because you've set $dirPath to '.' above; change that and your test is broken. You should be able to use if($file != 'img') and it should work You could restructure your code a bit to make this simpler to modify/extend: php:<?
// use a portable separator, shorten it for convenience
define('DS', 'DIRECTORY_SEPARATOR');
$dirPath = '.';
$rejects = array('.', '..', 'img'); // and any others...
if($handle = opendir($dirPath))
{
while(false !== ($file = readdir($handle)))
{
// put the test most likely to fail first
if(is_dir($dirPath.DS.$file) && !in_array($file, $rejects))
{
echo $file.'<br>';
}
}
closedir($handle);
}
?>
|
|
#
?
May 7, 2009 08:51
|
|
|
drcru posted:Whats a good way to encrypt and decrypt a variable with a specific salt? What exactly are you trying to accomplish? Do you want to encrypt some text and store it and decode it with a password or keyphrase later? Or are you trying to encrypt a password for storing in a db? something else?
|
|
#
?
May 7, 2009 17:49
|
|
|
Begby posted:Do you want to encrypt some text and store it and decode it with a password or keyphrase later? Or are you trying to encrypt a password for storing in a db? This. I want to be able to encrypt and store in the DB and later decrypt it. Any easy ways of doing this?
|
|
#
?
May 7, 2009 23:47
|
|
|
drcru posted:This. I want to be able to encrypt and store in the DB and later decrypt it. Any easy ways of doing this? You want to look into mcrypt.
|
|
#
?
May 8, 2009 02:20
|
|
|
I am having a hard time figuring out character encodings. I have a function that crawls a webpage that happens to be encoded in ISO-8859-1. Some text is lifted off that site and then ran through utf_encode() and then stored in a database. When I view the contents of that field in phpMyAdmin, it looks like this: "México", with a "A" with a squiggly line above it and a copyright logo instead of an "e" with an accent mark as how it looks on the original ISO-8859-1 page. The phpMyAdmin page is encoded in UTF-8, so I don't know what is happening. Is this caused by utf_encode() not working? What could be causing this? The "collation" of the databse is "latin1_swedish_ci", could that be the problem? I think I remember messing with encoding a few months ago, and changing the collation didn't change anything. When I drawl the info from the database it works fine, with the accent mark displaying correctly, which gets me because the page I use to display the data is set to UTF-8 as well...
|
|
#
?
May 9, 2009 01:04
|
|
|
nbv4 posted:it looks like this: "México", with a "A" with a squiggly line above it and a copyright logo Looks like utf8_encode is working fine, check your page charset.
|
|
#
?
May 9, 2009 09:25
|
|
|
I have a similar problem... One of my PHP scripts receives a tab separated file from FileMaker on a Mac. I read the file with file_get_content(), then explode() it on "\r" to get records, and then I explode() these on "\t" to get the fields. That part is working fine. The problem is that I have no idea what the encoding is on that file. When I look at the data, all accented characters are wrong. Where I expect to see an "ê" (that's E-circ) I see a blank space. I checked the ascii value for that space and it's 143. mb_convert_encoding() seems to say the encoding is "ISO-8859-1", so I try to convert it to UTF-8 using: code:What am I doing wrong ?
|
|
#
?
May 12, 2009 02:40
|
|
|
Im working on a site hosted at a server running mysql 4, my dev machine's php has mysql 5 and when I attempt to connect to the remote server I get Unknown MySQL server host '<the server>' (11004) the same when I try to shell in. is there an easy way to tell my dev php to use the mysql4 driver? and also is there a way to do the same when I use the shell?
|
|
#
?
May 13, 2009 02:04
|
|
|
Does anyone have any experience with reliable ways of stripping out Javascript and any malicious HTML (e.g., iframes) out of user-submitted HTML?
|
|
#
?
May 13, 2009 02:16
|
|
|
a lot of existing php driven sites like phpbb, wordpress etc have tried and true stuff. preg_replace() is one of your best options, write a regex for stuff like <script>.*?</script> etc worst case scenario, pick apart wordpress's comment filtering algorithms, or look into akismet?
|
|
#
?
May 13, 2009 02:19
|
|
|
supster posted:Does anyone have any experience with reliable ways of stripping out Javascript and any malicious HTML (e.g., iframes) out of user-submitted HTML? To build on LB4's answer: php:<?
$input = '<script type="text/javascript">alert("hacked!");</script>';
$magic = '#<(script|iframe).*?>(.*?)</\\1>#i';
$output = preg_replace($magic, '\\2', $input);
echo $output; // safe, plain text: alert("hacked");
?>Edit: vvv This is not a catchall, nor was it supposed to be. As mentioned, if you want to catch everything, catch EVERYTHING and write up a bbcode-style parser. No blacklist, no whitelist, there will be ways around both. filter_var($input, FILTER_SANITIZE_STRING) is my favorite. Supervillin fucked around with this message at 16:14 on May 13, 2009 |
|
#
?
May 13, 2009 02:36
|
|
|
There's now way you can trust a simple regular expression like that to catch everything. What about <a href="java script:_____">, what about DOM element event handlers (e.g., onclick, onmouseover, etc). There's a ton of ways to obfuscate malicious javascript/html so that naive parsers (such as a simple regex) don't catch them. Wordpress does not do any sort of filtering, you are free to post whatever you want on your blog. Comments are plain text only (right? I think so at least, I will look into it). phpBB and other forums do not allow direct HTML input and resort to BBCode or other forms of input such as textile or markup. Their filtering will first parse ALL html out of the post and then parse the BBCode/textile/markup/etc into safe predefined html. It is trivial to filter out ALL html/javascript, but it is not so simple to filter out only potentially malicious html/javascript. edit: I found http://htmlpurifier.org/ which looks promising. edit2: and http://www.bioinformatics.org/phplabware/internal_utilities/htmLawed/index.php which seems like a more light weight alternative. supster fucked around with this message at 04:00 on May 13, 2009 |
|
#
?
May 13, 2009 03:43
|
|
|
ours were just examples of using regex, obviously you're gonna want to use as many patterns as you want or need. Maybe you're best off doing like phpbb and removing all but the basic styling tags and further removing any handlers? As well as scrubbing out script tags completely? edit: that DOES look promising! *bookmarks*
|
|
#
?
May 13, 2009 03:58
|
|
|
A (very restrictive) whitelist is the only remotely sane and safe way to do it. Pick a set of safe tags, safe attributes for those tags, and safe protocols for urls and strip everything else. If you have any doubt about if something is safe, don't include it in the whitelist.
|
|
#
?
May 13, 2009 04:01
|
|
|
Lamb-Blaster 4000 posted:Im working on a site hosted at a server running mysql 4, my dev machine's php has mysql 5 and when I attempt to connect to the remote server I get This error does not have to do with mysql versions, but rather there is a problem connecting to the remote mysql machine. Either you are using an invalid IP address, or there is a DNS issue where the name is not resolving correctly, or its trying to connect over a proxy or something.
|
|
#
?
May 13, 2009 14:37
|
|
|
This is interesting http://bugs.php.net/bug.php?id=48139
|
|
#
?
May 13, 2009 14:38
|
|
|
Begby posted:This is interesting How much of a moron do you have to be not to be able to use PHP with MySQL on windows? Install any WAMP stack and you're good to go. The guy is on windows and there are plenty of stacks available; I can't even comprehend how he can't get it to load dlls when it should be done without touching the config.
|
|
#
?
May 13, 2009 15:17
|
|
|
I want a script that will return a random picture from Flickr by tag, above a certain resolution, which I can then resize to 1024*768 using image magic. ex: Tag specified is "stairs," script returns a 2048*1024 picture of a staircase. On refresh, it returns a different one. Is this easily accomplished by someone who is a php retard? (i've basically used it to make forms, nothing more complicated than that.)
|
|
#
?
May 13, 2009 19:09
|
|
|
not only do you need to know the gd and exif libraries, but you'll need some knowledge of the Flickr API (and an API key- both of which are easily attained at flickr) There may be a flickr class floating around on the net already http://phpflickr.com/ (first result googling for "php flickr class") hey look at that! Anyways, now you just need to do the gd, which is pretty basic, php.net will tell you all about the functions you need look around for imagecreatefromjpeg, imagejpeg, imagecopyresampled and imagecreatetruecolor or google around for an image class (http://code.google.com/p/php-image/wiki/PHPImage) was my first result
|
|
#
?
May 13, 2009 19:42
|
|
|
sorry to double post, it's been 10 hours since the above post and this is a different issue: what's the deal with the difference between include paths in php4 and php5 and how can I make include work in both versions? As noted before, my dev machine is php5 and the production machine is php4
|
|
#
?
May 14, 2009 05:58
|
|
|
Describe in more detail the setup on you dev machine; is it a wamp variant? Windows? Unix? Running on your machine / under your control? Shared with others / for you own personal dev work? (knowingly having the dev machine run a newer version of php than you will deploy on is STUPID; I deem to recall its your boss who made this clever decision?) If your include paths are troublesome the problem is most likely not the difference between php4 and php5 (as so far as I know there is no difference between them in this regard). It is more likely a difference in the setup of php on the 2 machines. (e.g. is one unix and the other windows?, does one have the include path setup and the other not etc) write yourself the usual phpinfo.php file, and run that on both machines; compare the settings, looking for path differences. (and I'm talking about one machine having you tied down to only certain include paths, although excluding '.' from that would be a case of misconfiguration) And be sure you are specifying paths cross-platformly, and in a manner that allows you to move the files without re-editing e.g. a start is to do: php:<?
define('MYPATH', dirname(realpath(__FILE__)));
include(MYPATH.DIRECTORY_SEPARATOR.'does_this_work_on_both.php');
// assumes the included file is in the same dir
?>KuruMonkey fucked around with this message at 08:23 on May 14, 2009 |
|
#
?
May 14, 2009 08:17
|
|
|
Got this email the other day:quote:Dear Philip, For obvious reasons I have no intention of giving this guy the PHP scripts that make up my website. But I am not averse to adding the functionality to my site that would allow it to be displayed in different languages (translated by users). I'm just not sure how I'd go about it. One idea: replace all "echo" statements with calls to "multiecho" where multiecho is defined in a central configuration file as follows: php:<?
function multiecho ($english, $turkish = '') {
global $user_language;
switch ( $user_language ) {
case 'turkish': echo $turkish; break;
// where 'turkish' would more likely
// be a numerical code I guess
default: echo $english; break;
}
}?>php:<?
echo 'Please enter your password.';
// becomes
multiecho('Please enter your password.',
'Lütfen şifrenizi giriniz.'
);?>Second option: Associate to each public-accessible script a file containing the text that might be required for that file. So for login.php, we might have a file either inside or outside of webroot called "login_text.php" (say) that looks something like php:<?
$text=array(
'enter-name'=>array('Please enter your login name.',
'Lütfen giriş adı girin.'),
'enter-password'=>array('Please enter your password.',
'Lütfen şifrenizi giriniz.'),
'click-submit'=>array('When you are done, click the "submit" button.',
'the turkish is long and breaks tables')
);?>php:<? echo $text['enter-password'][$user_language];?> Third option: As with the second option, but with a file for each language as well as for each script, and include statements that look something like php:<? include 'path/login_text_'.$user_language.'.php';?> Additional question: the email guy says he and his friends "experience some problems when connecting to" my site. What kind of problems is he likely to be talking about? Do they just have unreliable internet access over there?
|
|
#
?
May 15, 2009 18:13
|
|
|
They probably have terrible pings since you're likely somewhere in North America and they're XX hours away by plane.
|
|
#
?
May 15, 2009 18:45
|
|
|
|
| # ? May 15, 2024 17:09 |
|
|
Hammerite posted:i10n & i18n stuff Turkey is one of the hardest countries to localize, if there is a standard way of doing something they do the opposite (I believe this might be just out of spite.. but I'm not sure) http://www.moserware.com/2008/02/does-your-code-pass-turkey-test.html Don't reinvent the wheel.. Use gettext http://php.net/gettext
|
|
#
?
May 15, 2009 18:49
|
|
