Merge pull request #146 from tgalopin/remove-input-streams

Improve performance by relying on a native string instead of InputStream

Author: goetas

README.md

@@ -50,8 +50,8 @@ Here is how you use the high-level `HTML5` library API:
 <?php
 // Assuming you installed from Composer:
 require "vendor/autoload.php";
-use Masterminds\HTML5;
 
+use Masterminds\HTML5;
 
 // An example HTML document:
 $html = <<< 'HERE'
@@ -115,8 +115,6 @@ The following options are supported:
 This library provides the following low-level APIs that you can use to
 create more customized HTML5 tools:
 
-- An `InputStream` abstraction that can work with different kinds of
-input source (not just files and strings).
 - A SAX-like event-based parser that you can hook into for special kinds
 of parsing.
 - A flexible error-reporting mechanism that can be tuned to document
@@ -130,7 +128,6 @@ is well-documented.
 
 The parser is designed as follows:
 
-- The `InputStream` portion handles direct I/O.
 - The `Scanner` handles scanning on behalf of the parser.
 - The `Tokenizer` requests data off of the scanner, parses it, clasifies
 it, and sends it to an `EventHandler`. It is a *recursive descent parser.*

src/HTML5.php

@@ -55,7 +55,7 @@ public function getOptions()
      *
      * The rules governing parsing are set out in the HTML 5 spec.
      *
-     * @param string $file
+     * @param string|resource $file
      *            The path to the file to parse. If this is a resource, it is
      *            assumed to be an open stream whose pointer is set to the first
      *            byte of input.
@@ -68,13 +68,10 @@ public function load($file, array $options = array())
     {
         // Handle the case where file is a resource.
         if (is_resource($file)) {
-            // FIXME: We need a StreamInputStream class.
-            return $this->loadHTML(stream_get_contents($file), $options);
+            return $this->parse(stream_get_contents($file), $options);
         }
 
-        $input = new FileInputStream($file);
-
-        return $this->parse($input, $options);
+        return $this->parse(file_get_contents($file), $options);
     }
 
     /**
@@ -92,9 +89,7 @@ public function load($file, array $options = array())
      */
     public function loadHTML($string, array $options = array())
     {
-        $input = new StringInputStream($string);
-
-        return $this->parse($input, $options);
+        return $this->parse($string, $options);
     }
 
     /**
@@ -121,19 +116,15 @@ public function loadHTMLFile($file, array $options = array())
     /**
      * Parse a HTML fragment from a string.
      *
-     * @param string $string
-     *            The html5 fragment as a string.
-     * @param array $options
-     *            Configuration options when parsing the HTML
+     * @param string $string The HTML5 fragment as a string.
+     * @param array $options Configuration options when parsing the HTML
      *
      * @return \DOMDocumentFragment A DOM fragment. The DOM is part of libxml, which is included with
      *         almost all distributions of PHP.
      */
     public function loadHTMLFragment($string, array $options = array())
     {
-        $input = new StringInputStream($string);
-
-        return $this->parseFragment($input, $options);
+        return $this->parseFragment($string, $options);
     }
 
     /**
@@ -162,12 +153,12 @@ public function hasErrors()
      * Lower-level loading function. This requires an input stream instead
      * of a string, file, or resource.
      *
-     * @param InputStream $input
+     * @param string $input
      * @param array $options
      *
      * @return \DOMDocument
      */
-    public function parse(InputStream $input, array $options = array())
+    public function parse($input, array $options = array())
     {
         $this->errors = array();
         $options = array_merge($this->getOptions(), $options);
@@ -187,14 +178,12 @@ public function parse(InputStream $input, array $options = array())
      * Lower-level loading function. This requires an input stream instead
      * of a string, file, or resource.
      *
-     * @param InputStream $input
-     *            The input data to parse in the form of a InputStream instance.
-     * @param array $options
-     *            An array of options
+     * @param string $input The input data to parse in the form of a string.
+     * @param array $options An array of options
      *
      * @return \DOMDocumentFragment
      */
-    public function parseFragment(InputStream $input, array $options = array())
+    public function parseFragment($input, array $options = array())
     {
         $options = array_merge($this->getOptions(), $options);
         $events = new DOMTreeBuilder(true, $options);

src/HTML5/Parser/FileInputStream.php

@@ -11,11 +11,10 @@
  * really like to rewrite this class to efficiently handle lower level
  * stream reads (and thus efficiently handle large documents).
  *
- * @todo A buffered input stream would be useful.
+ * @deprecated since 2.4, to remove in 3.0. Use a string in the scanner instead.
  */
 class FileInputStream extends StringInputStream implements InputStream
 {
-
     /**
      * Load a file input stream.
      *

src/HTML5/Parser/InputStream.php

@@ -1,4 +1,5 @@
 <?php
+
 namespace Masterminds\HTML5\Parser;
 
 /**
@@ -9,10 +10,11 @@
  *
  * Currently provided InputStream implementations include
  * FileInputStream and StringInputStream.
+ *
+ * @deprecated since 2.4, to remove in 3.0. Use a string in the scanner instead.
  */
 interface InputStream extends \Iterator
 {
-
     /**
      * Returns the current line that is being consumed.
      *