Custom markup to HTML parser

Anon

I'm building an interface which will be used to update the web pages at a website.
The user will not only be able to input the text to be shown on the website webpages but he/she should also be able to format the text manually by using markup tags. The parser will then convert the markup tags to real HTML (to be viewed in the browser).
Examples: for bold text, [link="...url..."]name-of-link[/link] for links and so on. The parser also allow input of raw HTML placed within

tags. All text between these two tags (

) should be ignored by the parser (except converting special characters to their character entities).

I've got parts of the parser completed but I haven't found a way to support this

-tag which will enable users to input HTML themselves.
If you have any idea how to implement such a feature I'd be happy to hear it.

Thanks in advance.

Jakob Persson

stevo

Hi,

it strikes me that this <code></code> and <link=""></link> system is actually not particularly easier or clearer to use than normal HTML (? maybe i have misunderstood what you are trying to do?)

Why not just let the user enter plain HTML encoded text,- as it seems they will need to use similar but incompatible tags for your system anyway?

If it is necessary, then i would suspect that the best way to do this would be to run through line by line replacing your tags with HTML until you reach a <code> tag, then continue streaming the real HTML until you reach a </code> tag etc etc. This would be possible in a basic form using while loops.... it would be easier if you ensured that the tags were always at the start of the line first.
(ie replace <code> with \n<code> and </code> with \n</code> and the same with all of your custom tags)

If you are feeling clever (read insane🙂 then you could try using a java applet as a basic wysiwyg editor on a webpage - there may even be some prewritten applets to do this ?- , but that might be overkill.

cheers,
Stevo

Anon

I had a similar need and the easiest way of doing this is in JavaScript. WebFX has done all the work for you and you can find the code at:

http://webfx.eae.net/dhtml/richedit/richedit.html

Dan

Anon

The easiest way to deal with this is regular expressions:

preg_match_all('|[([^]]*)]|', $test_string, $matches);

that will fill $matches with all the occurences of [whatever] and [/whatever]s in the $test_string variable.

You can then just search/replace.

If that stumps you, you can just iterate through the code, until you get a '[', find the next ']' and that'll give you the positions of the tags, and you can easily read the contents.

hope that helps!

Anon

The primary reason why I want to do this is to force the user not to break the graphical design.
A worst case scenario: a newbie who has just learn the <FONT>-tag decides to apply his new skills and replaces the font on the page with Comic Sans, 24 points, pink.
That'll look nice, huh?

I've been thinking about using an applet instead. This is a rich-text editor applet which generates HTML-code. If I had the Java source I could limit it so that it only supported certain tags that don't pose a threat to the graphical design of the web site.

When it comes to the parser, replacing tags isn't the problem...the problem is to rip all HTML-tags outside the

-tags. 

Thanks for the advice and tips...I think I've got a fairly detailed idea how to write this parser now. I'll give that a try.
Thanks!!!

Jakob Persson

recompiler

I am doing just that with a modified open tables code from the Code Library.