I have a blog where I want the user to click a link to be able to invert the colour scheme (from dark on light to light on dark). This link will be available on each page of the blog.
If I wasn't using CodeIgniter, I would create a link with a parameter (e.g. "index.php?inverted=true") and in my controller detect if $_GET['inverted'] was set, set a $_SESSION variable. The link would be changed and the additional CSS file would be added when $_SESSION is set. The link's new href would be something like "index.php?inverted=false" and the $_SESSION variable would be unset. It would work site-wide and regardless of the rest of the URI.
I've tried using the uri_segement functions in CodeIgniter, but because this needs to be across the whole site, my URIs change (sometimes just "index.php/controller/function", other times "index.php/controller/function/para1/para2/para3"). Therefore I can't just make my link: [href="$_SESSION['REQUEST_URI'] . "/inverted"] and then detect if that URI segment exists because I don't know which number the segment "inverted" would be. Also, if a user is on the same page and clicks the link twice, the URI keeps adding "/inverted" again and again.
I made a mock form so I could use the $_POST array instead, but I know this isn't good practice.
Any suggestions?
Make the link submit a form via POST as you described, then set the SESSION variable, after that, issue a redirect to the referring page (HTTP_REFERER). You can do something similar to toggle this option off.
Related
I am trying to get TinyMCE 4's image_list to work with a URL returning JSON data as specified in the example here.
I have setup a GET endpoint http://demo.com/media on my server which gives back a JSON response consisting of a list of objects with their title and value attributes set, for example:
[{"title":"demo.jpg","value":"http://demo.com/demo.jpg"}]
I have also specified the option image_list: "http://demo.com/media" when initializing the plugin.
However, when I click the image icon in the toolbar, nothing pops up. All I can see in the network tab is an OPTIONS request with status 200, but then nothing. The GET request I was expecting never happens.
What is the correct way of using image_list in TinyMCE 4? Also, does anyone have a working demo example? I couldn't find anything.
It is somewhat hard to say what the issue is without seeing the exact data your URL is returning. I have created a TinyMCE Fiddle to show (in general) how this is supposed to work:
http://fiddle.tinymce.com/pwgaab
There is a JavaScript variable at the top (pretendFetchedData) that simulates what you would grab from the server (an array of JavaScript objects) and that is referenced via image_list.
If you enter your URL (http://demo.com/media) into a browser window what is returned? Are you sure its an array of JavaScript objects?
I have the identical problem. No matter what I do with the detail of the format (e.g. putting quotes round title and value), nothing happens.
I guess the only way (for me anyway) is to insert the list into the script with php before sending the web page.
I have searched the net for a solution but can't seem to get anywhere.
My page (php) is loading with one url (let's say www.mysite.com)
in the page several search options on music (albums) can be done and the tracks are shown. (without refreshing the page). the info comes from a database.
So the url stays the same.
In this search process the facebook meta tags (description, url, title) stay the same also because I never reload the page, I only load content into div's.
I would like to be able to 'like' the album, and backlink to it. So I have created the function to load the album by using the url: www.mysite.com?album=12345
I can show a popup with this url to share this.
So, if you go to this url, the content is automatically loaded based on the url parameter.
And on this spot (where you can see the url with the parameter ?album=12345) I would like to show the 'like' button as well. (I generated the url, so I use this in the code:)
echo '<div style="overflow:visable" class="fb-like" data-href="http://mysite.com/?album='.$albumid.'" data-send="false" data-width="300" data-show-faces="false">?</div>';
it works so far... (after I added the parse code to enable the button)
However the like button takes the default meta tags description and title etc.
Not particular on this album or artist - so it's not unique.
Note: if I remove the meta[property=og:url] from the header I can make the button backlink to the right url with the ?album parameter. Otherwise it would go back to the default root of the site mysite.com (this does make the lint tool give an error on the missing meta)
I have tried to add into this same function something like:
$("meta[property=og\\:url]").attr("content", "http://mysite.com/?album=<?php echo $albumid; ?>");
$("meta[property=og\\:title]").attr("content", "<?php echo $artistname; ?>");
$("meta[property=og\\:description]").attr("content", "<?php echo $albumname; ?>");
I did this so the meta tags will be changed, just to let the like button show the right description etc. However this doesn't work.
I understand that facebook scrapes the page (I used the lint tool etc.) but I will never executes javascript, so the meta tags wil stay as default (when first loading the page)
What can I do to make a unique like button, with it's own description (albumname etc) without making a html page for each one of them (millions of albums in the database...)
I hope it makes sense.
I can't seem to figure this one out, help please :-)
Based on the comments below I used the following solution:
you should create the right fb meta tags when the url (with the params ?alb_id=12345) is opened.
That's enough for the like button to do its job.
Your logic is fine, up to the point where you're setting the meta tags using jquery.
They should be set using PHP. As you can imagine the scraper won't execute the jquery, but if it's fed the already PHP-customized meta tags it will use them (as provided).
Just have the og:tags prepared server-side, depending on the albumId requested, and it should work. It might not work right away, I remember there used to be occasional caching issues with the scraper before.
In short, index.php?album=123 will send a different set of og:tags to the scraper than say index.php?album=321. Just set them up server-side.
<meta property="og:title" content="<?php echo $artistTitle; ?>"/>
What can I do to make a unique like button, with it's own description (albumname etc) without making a html page for each one of them (millions of albums in the database...)
You can’t, because Open Graph objects are URLs (resp. are represented/identified by their URL).
One URL == one Open Graph object.
But where’s the problem in having one URL for each album? Since it all works using parameters, it’s not like you have to create a page for each album URL manually …
I have two links on a page.
When the user clicks on one link, I'd like to set a ColdFusion variable, and when they click on another link, I'd like to set the variable to something else.
Is this possible to do? I'd like to avoid a JavaScript solution because the contents of the variable needs to be (semi) secure...
Have you tried adding a URL parameter to the link? (index.cfm?link=foo vs index.cfm?link=bar)
Then when the target page reloads, you can use a cfparam variable:
<cfparam name="var" default="">
An Adobe example follows:
http://help.adobe.com/en_US/ColdFusion/9.0/CFMLRef/WSc3ff6d0ea77859461172e0811cbec22c24-7faf.html
I'll explain:
I have a picture gallery, the first page is display.php.
Users can flip through pictures using arrows, when you click an arrow it sends an Ajax request to retrieve the next picture from the db. Now I want the URL to change according to the picture displayed.
So if the first picture is:
www.mydomain.com/display.php?picture=Paris at night
I'll flip to the next one and the URL would be
www.mydomain.com/display.php?picture=The Big Ben
How do I do this?
The trick here are uri's with an anchor fragment.
The part before '#' points to a resource on the internet, and after normally designates to a anchor on the page.
The browser does not refresh if the resource is the same but moves to the anchors position when present.
This way you can keep the convenience of browser history from a usability point of view while replacing certain parts on the page with ajax for a fast and responsive user interface.
Using a plugin like jQuery history (as suggested by others) is really easy: you decorate certain elements with a rel attribute by which the plugin takes care of the rest.
Also kinda related to this topic is something called 'hijax', and it's something I really like.
This means generating html just like you would in the old days before ajax. Then you hijack certain behavior like links and request the content with ajax, only replacing the necessary parts. This in combination with the above technique allows really SEO friendly and accessible webpages.
You can use the jQuery history plugin for example.
changing the search of the url will load the changed url.
See also: stackoverflow, javascript changing the get parameter without redirecting
Do you really want to use AJAX here?
A traditional web request would work like this...
User navigates to display.php
User clicks "next" and location is updated to "display.php?picture=Big-Ben"
Big Ben is shown to user, along with a link to "display.php?picture=Parliment"
User clicks "next" and location is updated to "display.php?picture=Parliment"
And so on.
With AJAX, you essentially replace the GET with a "behind the scenes" GET, that just replaces a portion of your page. You would do this to make things faster... for example...
User navigates to display.php
User clicks "next" and the next image location is obtained using an AJAX request
The image (and image description) is changed to the next image
What you are suggesting is that you retrieve the "next url" using AJAX and then also perform a GET on the whole page. You would be much better off sending the "next" image when you send each page and not using AJAX at all.
this best describes everything i think: http://ajaxpatterns.org/Unique_URLs
I am new to ajax and i wanted to know if we can load a complete new page and not just a part of it using ajax. Please give a small example script for understanding if this is possible. Here i am trying to display only one url to user while i change from one page to another when he clicks on any of the links in the page.
You can of course request for a new page and load it via body.innerHTML = ajax.responseText;
I would strongly recommend against this though for reasons outlined in this post: Why not just using ajax for Page Requests to load the page content?
The whole premise really is that with
AJAX you don't need to reload the
whole page to update a small
percentage of that webpage. This saves
bandwidth and is usually much quicker
than reloading the whole page.
But if you are using AJAX to load the
whole page this is in fact
counterproductive. You have to write
customised routines to deal with the
callback of the AJAX data. Its a whole
lot of extra work for little to no
increase in performance.
General rule for where to use AJAX: If
your updating >50% of your page, just
reload, else use AJAX.
You will not only need to request for the new page, but then also take care of ensuring the old styles on the current page are removed and don't interfere with the new page. Theres all sorts of problems associated with what your trying to do. It's possible, but I recommend not to do it.
edit: actually you might be able to just do document.write(ajax.responseText) which should take care of overwriting everything in the document, including css styles etc. Though still don't recommend it.
When you're finished with processing the AJAX request simply use this JS line:
window.location.replace('<URL of the new page>');
This has exactly the effect of loading the new page via
....
When you make an AJAX request, a request goes off and brings you the contents of the URL that you have requested. Now technically you can do whatever you like with the contents (which could be HTML), you can replace any element within the DOM with it. Be careful however of replacing EVERYTHING on the page, you are more likely just going to want to replace what is within the tags.
If what you want to do is show one URL for multiple pages, AJAX is overkill. Why not just use an IFRAME?
This could be useful if your page was unsure if it was expecting back errors to be inserted onto the page or a "new" submission confirmation page. This can be used when you want to put a validation servlet (or whatever) in front of the submission servlet (or whatever). If the page always hits the validation servlet, you hide the submission servlet which actually performs the data update. In the case where the validation passes, forward to the submission servlet. The user never knows what happened in the background.
When the page gets a response back you could just look at the first portion of the response text and determine if it had a keyword set by the server, which means this is a new page. Remove the keyword from the text, and do document.write(ajax.responseText); as described previously. Otherwise insert the response text into your errorBox div and let the user retry submission.