Regular Expressions for converting HTML to Structured Plain Text

Question

I'm writing a PL/SQL function that will convert HTML to plain text, but still preserve some of the formatting/line breaks. One of my challenges is in writing a regular expression to capture the text blocks while ignoring the markup. I'm trying to write an expression that will grab all of the text between start/end tags, but discard the tags. For example, to find all of the text between a start/end paragraph, I want to do something like:
REGEXP_REPLACE('<p style="text-align:center&#59;">This is the body of the paragraph</p>', '<p.*>(.*)</p>', '\1||v_crlf' )
where \1 returns the contents of the paragraph and v_crlf (declared earlier in the function) inserts a line break. I know there are more general expressions that will remove all tags, but I want to specifically identify the tags so I can process them appropriately. This way I can easily convert HTML to plain text for email and reporting without having to keep two versions around. Any help would be greatly appreciated. Once I get this worked out, I will repost with the function code for others to use. Thanks.
Edited by: jritschel on Oct 26, 2010 9:58 AM

Answer 1

Here's a function I wrote for an app. I'm not making in promises on it's accuracy as the app was just a proof of concept and never made it to production.
function strip_html( p_clob in clob )
return clob
is
    l_out clob;
    l_test  number := 0;
    l_max_loops constant number := 20;
    i   pls_integer := 0;
begin
    l_out := regexp_replace(p_clob,'<br>|<br />',chr(13)||chr(10),1,0,'imn');
    l_out := regexp_replace(l_out,'<p>',chr(13)||chr(10),1,0,'imn');
    l_out := replace(l_out,'<li>',chr(13)||chr(10)||'*<li>');
    l_out := regexp_replace(l_out,'<b>(.+?)</b>','*\1*',1,0,'imn');
    l_out := regexp_replace(l_out,'<u>(.+?)</u>','_\1_',1,0,'imn');
    loop
        l_test := regexp_instr(l_out,'<([A-Z][A-Z0-9]*)[^>]*>.*?</\1>',1,1,0,'imn');
        exit when l_test = 0 or i > l_max_loops;
        l_out := regexp_replace(l_out,'<([A-Z][A-Z0-9]*)[^>]*>(.*?)</\1>','\2',1,0,'imn');
        i := i + 1;
    end loop;
    return l_out;
end strip_html;{code}
The loop is there to handle nested HTML.
Tyler Muth
http://tylermuth.wordpress.com
"Applied Oracle Security: Developing Secure Database and Middleware Environments": http://sn.im/aos.book
Edited by: Tyler on Oct 26, 2010 10:03 AM