You are using method="POST"
in your form tag. So you're trying to have a page with both a query string (URL scope) and a form body (FORM scope), correct?
I'm not sure that's best practice or even allowed by some browsers (I read elsewhere they'll strip query strings on POST actions).
The best solution might be to make the action either GET or POST, and loop through the query string making each item a hidden input?
<cfloop list="#CGI.query_string#" delimiters="&" index="i">
<input
type='hidden'
name='#listFirst(i, "=")#'
value='#listLast(i, "=")#'
/>
</cfloop>
As you say, you can't do this. Your specific question was whether you should do any escaping. The answer to that is "yes" and the location is going to be on the backend, parsing the query string.
<cfoutput>
<form action='?#CGI.query_string#' method='POST' class='form-horizontal bordered-group' role='form' id='test'>
<input
class='form-control'
type='text'
name='formvar'
/>
<input
class="btn btn-primary btn-lg btn-block"
type="submit"
value="Submit"
/>
</form>
</cfoutput>
Will submit a form to the same page, with the FORM scope present, the URL scope present, and the CGI.query_string defined. The CGI.query_string will have url formatting (%20 for space, etc). The FORM and URL scopes will already be decoded (%20 converted to space, etc).
It seems the crux of your question is really about security and sanitization. In which case you'll want to examine encodeForHTML()
(Adobe Docs for encodeForHTML()).
Obviously, this isn't 100% foolproof, since I don't know the details of your code and what you do with the input. But those sanitization functions should be a good start.
So very generally, if you use the URL scope, use encodeForHTML()
, and if you use #CGI.query_string#
, it will be URL-encoded.