- From: Leif Halvard Silli <lhs@malform.no>
- Date: Wed, 29 Jul 2009 04:04:26 +0200
- To: Thomas Broyer <t.broyer@ltgt.net>
- CC: HTMLWG <public-html@w3.org>
Thomas Broyer On 09-07-23 10.23: > On Wed, Jul 22, 2009 at 7:50 PM, Leif Halvard Silli wrote: >> Thomas Broyer On 09-07-22 15.56: >> Making PHP pages that are valid before execution is a choice of the author >> or the authoring tool. > > If the author has to know that: > - he must not use ">" (i.e. use \x3E instead) to be valid HTML 4 (and > to comply with how Firefox and Opera will parse the doc [1]) > - he must use paired quotes to comply with how IE will parse the doc [1] > - he should not output HTML that would change the way the HTML is > parsed and alter its validity (i.e. <?php if ($foo) { echo > "<select\x3E"; } else { echo "<select multiple\x3E"; } > ?><option>A<option>B</select>) > - he must pay attention to the quote characters he uses in his PHP > code when placed within an HTML attribute value (e.g. value="<?php > echo "$foo $bar"; ?>" is invalid HTML) It is not so uncommon to have to manage quote characters ... > - he must not use "--" in his PHP code when placed within an HTML > comment (could be "--" in a string, or the decrement operator) What was the link between '--' and PIs, here? I don't think '--' will hamper the processing of the PIs. In fact, "<?php <!-- -- -->" validates, whereas "<!-- -- -->" doesn't (in HTML 4). > - and if he also wants his document to be valid HTML *and* valid > XHTML (or at least well-formed XML), he must not use PHP within start > tags (neither in the tag itself or within attribute values) > > If he does know all these things, cannot it just ignore any > warning/error regarding "<?" being invalid HTML5 ? Would it not be better avoid the validator then? Isn't the point of the validating to squish bugs? Are authors supposed to think that "but I know better than the validator"? > (I said "author" above, and not "authoring tool", because I can hardly > imagine a tool that would do any/all of this without the author being > involved) > > [1] You could tell me that he won't open the page in a browser without > first processing the PHP code; but then what would be the point of > arguing that UAs do support the <? > syntax? Absolutely. The point was to create a page that is valid before it is processed. -- leif halvard silli
Received on Wednesday, 29 July 2009 02:05:07 UTC