ID: 30887 Updated by: [EMAIL PROTECTED] Reported By: maddam at volny dot cz -Status: Open +Status: Bogus Bug Type: *XML functions Operating System: Win XP PHP Version: 5.0.2 New Comment:
You had to concatenate character_data as well in PHP 4, if you wanted to write a proper XML parser based on ext/xml. In PHP 5, it just happens that the "breakup" happens on different places. Read the usercomments at http://www.php.net/manual/en/function.xml-set-character-data-handler.php and the dozens of tutorials available on the net please do not reopen this "bug". It ain't no bug. This behaviour exists since ages. Previous Comments: ------------------------------------------------------------------------ [2004-11-29 22:26:30] maddam at volny dot cz Yes I will not rely on getting one event for each CDATA section now for PHP5 and above. But explain me please this behavior of XML parser for this four CDATA sections: function characterData($parser, $data) { echo $data . '<br />'; } 1. ------------------------------------------------------------------------------------------------------------ <data>This is expected, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we have in PHP 5 does it like this).</data> echo is same, the all sentence in one $data: This is expected, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we have in PHP 5 does it like this). 2. ------------------------------------------------------------------------------------------------------------ <data>This is expácted, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we have in PHP 5 does it like this).</data> here I add character 'á' at the begening of CDATA (expácted) echo is dividet into $data first This is exp $data second ácted, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we have in PHP 5 does it like this). 3. ------------------------------------------------------------------------------------------------------------ <data>This is expected, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we háve in PHP 5 does it like this).</data> here I add character 'á' at the end of CDATA (háve) echo is dividet into $data first This is expected, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we h $data second áve in PHP 5 does it like this). 4. ------------------------------------------------------------------------------------------------------------ <data>This is expácted, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we háve in PHP 5 does it like this).</data> here I add character 'á' at the begening and at the end of CDATA (expácted and háve) echo is divided into $data first This is exp $data second ácted, and definitely not wrong. There is never said that XML parsers can't break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we háve in PHP 5 does it like this). ------------------------------------------------------------------------------------------------------------ Does not mather what long CDATA section is, the divided CDATA section strongly depend on first occurence of character 'á' other occurencies of 'á' parser do not touch. I think you are wrong and THIS IS BIG BUG ! PHP 4 is OK for this. I think many peoples who was used PHP4 and goes to PHP5 have big problems if they worked at other languages than English as me. My internet provider always goes to last software versions of PHP and now he has 5.0.2 and cannot for one person as me who is using XML with PHP go back to working PHP4. Instead off I must use correction code for PHP5, what is watching division of $data and add them back to one $data. THE SECOND BUG FOR THIS IS when parser give back first $data when do division, he add CRLF at the end of first $data - for point 4: (This is expCRLF). IS NOT THAT BUG ? when parser add characters who is not present in CDATA section ? I THING THIS IS BIG BIG BUG ! When parser give back $data 'I am last' than strlen($data) = 9 but when $data is 'I am lást' with 'á' and you use say $getdata = $getdata . $data (for connection divided $data) you get strlen($getdata) = 11 andy put the question WHERE ARE 11 characters in <data>I am lást</data> ? THIS BUG is BIG problem when you want to search occurence of CDATA string. You are looking for text 'I am lást' in XML file with function strcmp if (strcmp($currentTag, "data") == 0) { $sentence = $data; if (strcmp($sentence, "I am lást") == 0) { // here you never find it because when you get two $data from parser (I am l) and (ást) you get string $sentence with strlen 11 not 9 } } but you never find it!!! I know what I am talking about. My own search problem. My code definitely stop work with PHP5! That it is my correction code must think about connection CDATA dividet into two sections and even must trim second BUG - trim all not printable characters from $data. Why to spend time for develop correction code when PHP4 works correct ? When you goes from PHP4 to PHP5 the all sites with PHP4 and XML definitely stop works (with special other Language characters not with real English). The text will bu cutted with PHP5 to first occurence of special Language character and all the search functions in XML with parser will not be capable to catch any looked string. And must read and see this text above not only say your conventional ' Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://www.php.net/manual/ and the instructions on how to report a bug at http://bugs.php.net/how-to-report.php'. Right ? 5. ------------------------------------------------------------------------------------------------------------ Last question why I cant parse character & and get error as I see in other bug references ? Thanks for me and other users of PHP5. Give you the question, why PHP5 parser stop at the first occurence of 'á' and the all other occurences of 'á' in CDATA section (does not matter what long CDATA it is) do not touch. Give you the question why the PHP4 parser does not have (FOR ME BUG) this feature and always give back all CDATA in one string $data - I think this behaviour of PHP4 is correct. Tell mi why the programmer will program any code, who will wait for adding CDATA section to one string ? I think this must do parser and for this the parser is made - this is main feature of parser - simply get data from CDATA to one string (not divided betwen two strings). Thanks for reply. -----Original Message----- From: Marek [mailto:[EMAIL PROTECTED] Sent: Wednesday, November 24, 2004 11:45 PM To: [EMAIL PROTECTED] Subject: rd [24 Nov 9:28pm CET] [EMAIL PROTECTED] Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://www.php.net/manual/ and the instructions on how to report a bug at http://bugs.php.net/how-to-report.php This is expected, and definitely not wrong. There is never said that XML parsers can\'t break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we have in PHP 5 does it like this). ------------------------------------------------------------------------ [2004-11-24 21:28:10] [EMAIL PROTECTED] Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://www.php.net/manual/ and the instructions on how to report a bug at http://bugs.php.net/how-to-report.php This is expected, and definitely not wrong. There is never said that XML parsers can\'t break up CDATA sections, and you should never rely on getting only one event for each CDATA section. This is just how an XML parser might handle it (and the new libxml2 we have in PHP 5 does it like this). ------------------------------------------------------------------------ [2004-11-24 21:08:48] maddam at volny dot cz Description: ------------ xml file element: <data>Jak se máš holoubátko ?</data> This is Czech language with special characters. <?php function characterData($parser, $data) $getdata = $data; echo $getdata must show 'Jak se máš holoubátko ?' But the parser stop and $getdata will consist of 'Jak se m' and at the next step on same element parser will get all last text and $getdata will consist of 'áš holoubátko ?' This is bug for 5.0.2. In 4.3.9 and sooner is all OK. 5.0.0 and 5.0.1 i was not tested. Description: The parser when get data with language characters as (ìšèøžýáíé) will cut this data to two parts. First part consist of characters to first occurence of some character (ìšèøžýáíé) and second part consist of spare element. THIS BUG WILL NOT SHOW FOR ENGLISH LANGUAGE WHICH NOT USE CHARACTERS AS ìšèøžýáíé Sorry for my english i hope you understand. Contact me at [EMAIL PROTECTED] or ICQ 25684007 Reproduce code: --------------- <?xml version="1.0" encoding="ISO-8859-1"?> <!DOCTYPE rokam [ <!ELEMENT data (#PCDATA)> ]> <rokam> <data>Jak se máš</data> <data>Zde doma je dobrý ocet</data> </rokam> <?php function characterData($parser, $data) $getdata = $data; echo $getdata . <br />; Expected result: ---------------- Echo on screen, need two steps through function characterData: Jak se máš Zde doma je dobrý ocet Actual result: -------------- This output of parser 5.0.2 need four steps through function characterData and will output: Jak se m áš Zde doma je dobr ý ocet ------------------------------------------------- This BUG can be repaired with this code, who connect two parts from parser to one variable say $data. This code connect 'Jak se m' with 'áš' function characterData($parser, $data) { global $currentTag; // <code for repair start> global $lastdata, $lastTag; if (strcmp($lastTag, $currentTag) == 0) { $data = $lastdata . trim($data); $lastdata = $lastTag = ''; }else{ $lastdata = $data; $lastTag = $currentTag; return; } // <code for repair end> here can continue normal code for function characterData // see trim($data) must be there - the parser add to end of the string $data of first part CR(x0D) LF(x0E) (I think) and must be trimed for code to properly work. ------------------------------------------------------------------------ -- Edit this bug report at http://bugs.php.net/?id=30887&edit=1
