Differences

This shows you the differences between two versions of the page.

--- oracle:fixhtml [2010/09/13 17:07] – creado rlunaro
+++ oracle:fixhtml [2022/12/02 22:02] (current) – external edit 127.0.0.1
@@ Line 1: / Line 1: @@
-====== Fix or Sanitize HTML code from Word ======
+====== Fix or Sanitize HTML code from Microsoft Word ======
 Yes: I've found the silver bullet for those of you who are seeking for a function that clean html code or sanitize it, specially if it comes from a cut and paste operation from word.
@@ Line 17: / Line 17: @@
 Wipes out all the garbage who is in the html code, leaving it --more or less-- "clean". And yes, I am using regular expresions to perform the fixing, so it is easy to move it to java or other programming languages. And without more preamble, here is the code:
-<code>
+<code plsql>
-/*
-*/
+ create or replace function strip_html(dirty in clob,
-create or replace function strip_html(dirty in clob,
                                       to_cvs in number default 0)
   return clob is out clob ;
@@ Line 296: / Line 295: @@
   if to_cvs = 2 then
      -- sanitize (not clean) the html
+     -- clean the tag <?xml:whatever>
+     out := regexp_replace(out, '<\?xml:.*?>', '', 1, 0, 'ni');
+     -- clean the tags <img whatever>
+     out := regexp_replace(out, '<img.*?>', '', 1, 0, 'ni');
      -- clean comments
      out := regexp_replace(out,'<!--.*?-->','', 1, 0, 'ni');
@@ Line 309: / Line 312: @@
      -- clean "class inside tags"
      out := regexp_replace(out,'(<.*?)class="?[a-zA-Z0-9-_]*"?(.*?>)', '\1\2', 1, 0, 'ni');
+     -- clean "style" inside the following tags: i b p
+     out := regexp_replace(out,'(<[ibp] .*?)style=".*?"(.*?>)', '\1\2', 1, 0, 'ni');
      -- clean namespaces <o:p> </o:p>
      out := regexp_replace(out, '(<)[a-zA-Z0-9-_]*:(.*?>)', '\1\2', 1, 0, 'ni');
@@ Line 354: / Line 359: @@
   return(out);
 end strip_html;