Hanna Code

[teppo]

Agreed, this is a problem for anyone using CKEditor together with Hanna Code. Entities config setting seems like a good solution, at least I'm not aware of any serious drawbacks to that.

There's another option (config.basicEntities) for things like <, > etc. which could be potentially harmful if left unescaped. We're talking about HTML markup here, after all. If I ever needed to use comparison operators as parts of Hanna Code variables, I'd probably let CKEditor escape them and then handle decoding in the code snippet with str_replace() or html_entity_decode().

If you're wondering what these config settings can cause, I'd suggest taking a look at CKEditor source. There you can see a list of all entities separated to couple of different variables; htmlbase is only affected by config.basicEntities while encoding all the others (entities, latin, and greek) is going to be disabled if you set config.entities to false. 

[Macrura]

@Teppo - thanks again for your help and comments  - yeah i should definitely go with the basicEntities and decode in Hanna.. will try asap and report back...

[ryan]

I recently debugged similar issue and eventually found out that it was caused by CKEditor encoding " to " (default behavior) before Hanna Code had a chance to grab it. Similar issues could be caused by other editors, textformatters placed before Hanna Code etc.

It might make sense for Hanna code's parser to just allow " as an allowed attribute container, in the same way it does unencoded quotes? I'm happy to implement this change if you guys think it would solve the issue without creating other problems (I can't think of any at the moment). 

[Macrura]

Yes! that would be great! - seems the only reliable way to use HC with an editor? otherwise you'd have to adjust the config settings per the above conversation...

[ryan]

Yes! that would be great! - seems the only reliable way to use HC with an editor? otherwise you'd have to adjust the config settings per the above conversation...

Strangely I've not ever run into this issue, and I guess it just depends on whether the editor is entity encoding quotes or not (it looks like CKEditor does not entity encode quotes, so no problems with Hanna code). I've gone ahead and added support for " entity encoded quotes to the latest version. 

[Macrura]

for some reason both times i was using ckeditor and i had to either decode entities in the hanna code or change the ck editor config...

[cmscritic]

I've just tried using it to input something from rafflecopter into my post but it does not seem to work correctly. The code from rafflecopter is:

<a id="rc-603abd0" class="rafl" href="http://www.rafflecopter.com/rafl/display/603abd0/" rel="nofollow">a Rafflecopter giveaway</a>
<script src="//d12vno17mo87cx.cloudfront.net/embed/rafl/cptr.js"></script>

but when I add to hanna code and insert the tag into my post, nothing shows up. I've selected javascript as the options as well.

[WillyC]

i see htmls in there

sholdunt u.use html as mode ?

their is no javascripts here

just html script link to javascript

tl;dr use.html mode

[ryan]

Just to follow up here in case anyone else ever runs into something similar: WillyC was correct about the issue here. It just needed to be an HTML HannaCode rather than a javascript one. 

Also wanted to mention I've posted an update to HannaCode, version 0.1.4 which fixes a couple of bugs having to do with the Ace editor. It fixes the issue where extra whitespace was getting inserted at the end of the code. It also fixes the issue where Ace sometimes wasn't active after doing a 'save & test'. 

[Neeks]

This I almost feel like i'm building my own CMS when I use this module it is so easy! + awesome. 

Idea: it could be cool but not totally necessary of Hanna supported shorthand single arguments for easy to remember usage. ie. [[block=myblock]]  rather than [[block title=myblock]. Just idea to shorten hanna codes by a few character and make them easier to remember.  if you have an attribute that matches the name of your Hanna code it could work that way...

question: is it possible to render hannacode inside of hanna? so far I have not been able to do it, even by calling a new instance of Hanna: 

<?php

//my Hanna block
$mypage = $pages->get("/settings/blocks/".$title);
$block_hanna = $modules->get('TextformatterHannaCode'); 
echo $block_hanna->render($page->body); 

/* 
This does no seem to work as planned. 
It currently returns nothing: and breaks my existing Hanna codes.  Also does not work if I use an existing 
instance of hanna i call from my template which would look like this: echo $hanna->render($page->body); 
*/
?>

I might just be a Hanna obsessed and should chill-out with with that I want it to do, and find other ways. The main reason i'm running into this is that I am using hanna to build my css grids and call the the [[blocks]] of hanna formatted feilds from other hanna formatted fields. 

[[grid_start col=8 ]]
My left column body content
[[grid col=4 col_class=print]]
my right column content
[[block title="load my hanna formatted field"]]
[[grid_end]]

I might have been able to accomplish this all with repeaters rather than using hanna, but I liked the freedom Hanna offered.

[teppo]

@Neeks,  your suggested idea sounds kind of fun, but also not very useful. [[block=myblock]] would be equal to [[myblock]], except that you'd have a lot of your "blocks" within one single Hanna Code.. which is exactly what individual Hanna Codes are meant for.

About your question: in your example code you're assigning /settings/blocks/".$title to $mypage and later asking Hanna Code to render $page->body -- while you probably meant to render $mypage->body. Is this just a typo in your example code here or could this be the issue with your original Hanna Code?

Anyway, even if this can be fixed, wouldn't it make more sense to have a [[block type=something' title='something else]] Hanna Code that finds a page (let's call it $block_page) and then renders it with $block_page->render(), i.e. by using ProcessWire's full template system? Is there a particular reason why you're trying to move your template level stuff into Hanna Codes?

I'm afraid that's not really what Hanna Code was intended for, at least as far as I'm aware. Hanna Codes are snippets you can drop in your code -- once you find yourself building complex logic with them, you'd probably be better off by using other PW features instead 

[teppo]

Strangely I've not ever run into this issue, and I guess it just depends on whether the editor is entity encoding quotes or not (it looks like CKEditor does not entity encode quotes, so no problems with Hanna code). I've gone ahead and added support for " entity encoded quotes to the latest version. 

@ryan: that definitely is strange, considering that with default config CKEditor encodes everything it possibly can; even single quotes get encoded per the entities_additional default value. Perhaps you've modified the config already?

Anyway, my solution for this (and most CKEditor issues I've been dealing with) was simply adding config.entities = false to CKEditor's config.js. As you can see from CKEditor source, so-called base HTML entities (nbsp, gt, lt and amp) still get encoded, making this a very good default. As long as we're dealing with UTF-8 data, there shouldn't be any reason to encode everything.

This also prevents issues such as search not working properly with Scandinavian characters ä/Ä, ö/Ö etc. Sure, searching from CKEditor fields is possible even with default config, but how many users would really try something like "Väinö" when they want to search for "Väinö"? Of course encoding search queries would be doable, but that would mean searching with both encoded and non-encoded versions..

As a matter of fact, I would like to suggest this as the default value for upcoming Inputfield CKEditor releases, unless someone knows of some terrible security issue this introduces.

Edit: by the way, I do realize that the default config is much less of a problem for English speakers than us who have to deal with Scandinavian characters, umlauts etc. regularly. Still, it becomes a problem for just about everyone once they try to create multilanguage sites and/or use modules like Hanna Code, which is essentially why I believe that this would make sense as a default.

Edited December 27, 2013 by teppo

[ryan]

Teppo, the way you describe the entities config option, that makes sense to me. Given these issues, I wonder why they have it true by default. I have updated the InputfieldCKEditor module to have the entities option set to false by default. I really can't think of any security issues with doing that, especially given that we're requiring HTMLPurifier for the inline editor. As for why I seem to be getting the entities=false behavior for quotes already, I have no idea. Maybe something to do with HTMLPurifier and inline mode vs. regular mode. 

[Soma]

Just a little tip when working with hanna code.

I for some project prefer to have the code on file system and not db.

This can be achieved simply by including the php file in the hanna code.

For example a Hanna code:

include($config->paths->templates . "hannas/archive.php"); 

And have your Hanna code in archive.php.

[ryan]

Hanna Code doesn't actually execute from the DB. It writes each Hanna code to a file in /site/assets/cache/HannaCode/ and then uses the DB copy of the code to compare against the one on the file system, to make sure something hasn't changed the one on the file system. It executes the PHP code from a file in order to ensure it is as fast as possible (avoiding eval) and in the same context as a template file. That's what it does behind the scenes, but Soma's tip is just as useful either way, as Hanna Code wouldn't let you edit its own PHP files directly. 

[Soma]

I just noticed that the complete ACE editor is bundled with TextformatterHannaCode. Wouldn't it be nice to make it a separate js module and use it if it's installed? Also there's already a ACE editor fieldtype from Adam. Since ACE editor source comes with 1million! files I think it would be the way to go. Much like jQueryDataTables or Magnific etc.

[SiNNuT]

I've just been playing around wit Hanna code for the first time, pretty cool. I think i might have discovered some bugs, or at least an error/omission in the instructions.

If you make a Hanna code with a hyphen in it and you then use parameters the Hanna code doesn't get executed. Given the example Hanna on https://github.com/ryancramerdesign/ProcessHannaCode:

if(isset($parent)) {
  // If $parent is an ID or path, lets convert it to a Page
  $parent = $pages->get($parent);
} else {
  // otherwise lets assume the current page is the parent
  $parent = $page; 
}

foreach($parent->children as $child) {
  echo "<p><a href='$child->url'>$child->title</a>";
  // echo "<p><a href='$child->url'>$child->body</a>"; does not work as i expected, it only outputs the last child->body instead of foreaching over them
}

// below is used in the body field of a fresh install from dev  2.3.15

[[menu_items]] //this works
[[menu_items, parent=1006]] //this works
[[menu-items]] //this works
[[menu-items, parent=1006]] //this does not work, it doesn't execute and just displays the literal text

Also, in the example above, if i echo $child->body the Hanna code stops working. It's looks like echoing $child->body from the body field of a parent does not work, or at least not in the example above.

Maybe this is all for obvious reasons but it took me a good while with trial and error to figure out. All tested with PW dev 2.3.15, Hanna 0.1.5 on a Win7 machine.

[jean-luc]

Probably I'm missing something simple, but the $page variable doesn't seem to be referring to the page where the field including the HannaCode is, as the doc says:

 The $page API variable available to your Hanna code represents the page where the Hanna code exists. It is possible for this to be different from wire('page'), which represents the page that originated the request.

For example, I have a very simple code [[files_url]] to get the files directory url from a text fiel in a page:

echo $config->urls->files . $page->id . "/";

It's used for referencing page images or file from a "body" field:

<img class="video-image" onclick="play_video('NxgHAktpdF8')" src="[[files_url]]an_image.jpg" /> 

When using this code in a "body" field in page $article_page, referenced from the currently displayed page wire("page") like this

function renderArticle($article_page) {
   //...
   $out .= $article_page->body;
   //...
   return $out;
}

echo  renderArticle($article_page); 

the page id in the html page is the wrong one, i.e. the one from  wire("page") and not the article_page's id.

To clarify, here is what I get:

echo $article_page->id;
=> 1048

echo wire("page")->id;
=> 1041

echo $article_page->body
=> <img class="video-image" onclick="play_video('NxgHAktpdF8')" src="/site/assets/files/1041/an_image.jpg" />

Which causes an image-not-found error, as the expected result should have been:

echo $article_page->body
=> <img class="video-image" onclick="play_video('NxgHAktpdF8')" src="/site/assets/files/1048/an_image.jpg" /> 

I'm running version 2.3.15 from dev branch.

So bug or expected behavior? And any idea of how to make it work?

(For such a simple use case I could of course do some kind of string replacement in my template, but I would prefer to stay with HannaCodes for regularity if possible   ) 

Edited February 5, 2014 by jean-luc

[Martijn Geerts]

I read your Post atleast 5 times, But I really don't understand what you're asking.

[jean-luc]

I read your Post atleast 5 times, But I really don't understand what you're asking.

Ok. I've tried to clarify it in the original post.

[Soma]

The page that originated the request is meant the one that is rendered, and that's not your article page.

wire("page") or $page is the requested page, so doing echo $someotherpage->body won't give the hanna code on this body the $someotherpage but $page.

You would need to set the page in this case to the one you get the body from and set it back after to make it work in this case.

Something like this should work:

function renderArticle($article_page) {
    // ...
    $current_page = wire("page");
    wire()->set("page", $article_page);
    $out .= $article_page->body;
    wire()->set("page", $current_page);
    return $out;
}

[jean-luc]

The page that originated the request is meant the one that is rendered, and that's not your article page.

Agreed: my article page is the page containing the field where the Hanna code exists.

That's why the behavior I observe don't look consistent with HannaCode module's documentation saying:

 The $page API variable available to your Hanna code represents the page where the Hanna code exists. It is possible for this to be different from wire('page'), which represents the page that originated the request.

And this behavior  would be great by the way, because this is probably what you expect when putting a Hanna code in a page field.

Thanks anyway for the suggestion. I will try it.

[Soma]

Ah sorry I haven't really ever read the docu that far, seems like a bug cause I see code in there that sets $page to context specific. But I can't get it to work and it's always the same wire("page") and $page. 

I also noticed some strange behavior when using wire()->set("page", $somepage); and then use $page. So Ryan might have a look into this.

For example this code in a template

$p = wire("page");
wire()->set("page",$pages->get("/"));
$content .= "p1: " . $page->id;
wire()->set("page",$p);
$content .= "p2: " . $page->id;

is different as this:

$p = wire("page");
wire()->set("page",$pages->get("/"));
$content .= "p1: " . wire("page")->id;
wire()->set("page",$p);
$content .= "p2: " . wire("page")->id;

when used in a template for example basic-page.php.

[ryan]

This turned out to be an issue with the order of variables returned by core TemplateFile::getArray. It was doing an array_merge which gave preference to API variables, preventing them from being overwritten. Maybe that's a good thing in general, but I don't think we want that limitation. I went ahead and changed it in dev. Now it gives preference to local variables, which enables Hanna Code to provide the behavior it was supposed to. 

[jean-luc]

Thanks Ryan. I tested my $article_page case with the last dev version and it works just as expected now.

This really opens a huge potential for Hanna code regarding cross-fields referencing in a page independently of the rendering context.