Generators HTML/Markdown: consistent encoding cross-PHP

The default value for the `$flags` parameter of the `htmlspecialchars()` function changed in PHP 8.1.0.
Previously, the default was `ENT_COMPAT`. Now the default is `ENT_QUOTES | ENT_SUBSTITUTE | ENT_HTML401`.

The most notable differences are:
* Single quotes will be encoded.
* Invalid code unit sequences will be replace by a Unicode Replacement Character.

For consistent output cross-version PHP, it is advised to always explicitly pass the `$flags` parameter` and not rely on the default value.

Fixed now and using the _new_ `$flags` default value as the parameter value.

This commit allows for the tests to have the same output expectations cross-version PHP. For end-users, the differences shouldn't have been noticeable.

Author: jrfnl

src/Generators/HTML.php

@@ -235,7 +235,7 @@ public function processSniff(DOMNode $doc)
     protected function printTextBlock(DOMNode $node)
     {
         $content = trim($node->nodeValue);
-        $content = htmlspecialchars($content);
+        $content = htmlspecialchars($content, (ENT_QUOTES | ENT_SUBSTITUTE | ENT_HTML401));
 
         // Use the correct line endings based on the OS.
         $content = str_replace("\n", PHP_EOL, $content);

src/Generators/Markdown.php

@@ -116,7 +116,7 @@ protected function processSniff(DOMNode $doc)
     protected function printTextBlock(DOMNode $node)
     {
         $content = trim($node->nodeValue);
-        $content = htmlspecialchars($content);
+        $content = htmlspecialchars($content, (ENT_QUOTES | ENT_SUBSTITUTE | ENT_HTML401));
 
         // Use the correct line endings based on the OS.
         $content = str_replace("\n", PHP_EOL, $content);