Bug #76285 [Opn->Csd]: DOMDocument::formatOutput attribute sometimes ignored

Edit report at https://bugs.php.net/bug.php?id=76285&edit=1 ID: 76285 Updated by: ab@php.net Reported by: daniel dot hemberger at gmail dot com Summary: DOMDocument::formatOutput attribute sometimes ignored -Status: Open +Status: Closed Type: Bug Package: DOM XML related PHP Version: Irrelevant -Assigned To: +Assigned To: ab Block user comment: N Private report: N New Comment: The linked PR has been merged. Thanks. Previous Comments: ------------------------------------------------------------------------ [2018-05-01 12:08:01] andrew dot nester dot dev at gmail dot com Thanks for reporting the issue! I added PR fixing this: https://github.com/php/php-src/pull/3229 ------------------------------------------------------------------------ [2018-04-29 06:23:16] daniel dot hemberger at gmail dot com Description: ------------ When using DOMDocument::saveHTML ([ DOMNode $node = NULL ] ), whether or not the formatOutput attribute of the class instance is respected depends on whether or not the $node argument is specified when calling the method. We can see where the problem arises in the source code (here: https://github.com/php/php-src/blob/701437a94872853026c97b225f9882ed7fa3164c/ext/dom/document.c). If nodep == NULL (i.e. $node is NOT passed to saveHTML) then the following function call is made in dom_document_save_html: htmlDocDumpMemoryFormat(docp, &mem, &size, format); But if nodep != NULL (i.e. $node IS passed to saveHTML), it calls: htmlNodeDump(buf, docp, node); We can see already that the format variable is not passed into htmlNodeDump, and thus the output cannot respect the DOMDocument::formatOutput attribute. This should be fairly simple to fix, because there is a similar function htmlNodeDumpFormatOutput which does take an argument specifying the formatting. I think that the problem is solved by replacing the instances of htmlNodeDump with htmlNodeDumpFormatOutput and passing the appropriate formatting arguments (see: http://xmlsoft.org/html/libxml-HTMLtree.html#htmlNodeDumpFormatOutput). Thank you for your time! Test script: --------------- <?php $dom = new DOMDocument(); $dom->formatOutput = false; $html = '<div><div><a>test</a></div><div><a>test2</a></div></div>'; $dom->loadHTML($html); $rootNode = $dom->documentElement; foreach ($rootNode->firstChild->childNodes as $child) { var_dump($dom->saveHTML($child)); $dom2 = new DOMDocument(); $dom2->formatOutput = false; $child2 = $dom2->importNode($child, true); $dom2->appendChild($child2); echo "\n"; var_dump($dom2->saveHTML()); } Expected result: ---------------- string(57) "<div><div><a>test</a></div><div><a>test2</a></div></div> " string(57) "<div><div><a>test</a></div><div><a>test2</a></div></div> " Actual result: -------------- string(59) "<div> <div><a>test</a></div> <div><a>test2</a></div> </div>" string(57) "<div><div><a>test</a></div><div><a>test2</a></div></div> " ------------------------------------------------------------------------ -- Edit this bug report at https://bugs.php.net/bug.php?id=76285&edit=1