Implement new List Diffing algorithm for matching (#49)

* Implement new List Diffing algorithm for matching * Add MatchStrategy interface and classes instead of comparator function * Start updating the tests * Update tests
caxy · May 19, 2016 · 23808d0 · 23808d0
1 parent 755472d
commit 23808d0
Show file tree

Hide file tree

Showing 21 changed files with 651 additions and 16 deletions.
diff --git a/composer.json b/composer.json
@@ -20,7 +20,8 @@
     },
     "require": {
         "php": ">=5.3.3",
-        "ezyang/htmlpurifier": "^4.7"
+        "ezyang/htmlpurifier": "^4.7",
+        "sunra/php-simple-html-dom-parser": "^1.5"
     },
     "require-dev": {
         "phpunit/phpunit": "~4.8",

diff --git a/demo/index.php b/demo/index.php
@@ -41,7 +41,7 @@ function addDebugOutput($value, $key = 'general')
     }
     $diff->setUseTableDiffing($useTableDiffing);
     $diffOutput = $diff->build();
-    $diffOutput = mb_convert_encoding($diffOutput, 'UTF-8');
+    $diffOutput = iconv('UTF-8', 'UTF-8//IGNORE', $diffOutput);
 
     $jsonOutput = json_encode(array('diff' => $diffOutput, 'debug' => $debugOutput));
 

diff --git a/lib/Caxy/HtmlDiff/HtmlDiff.php b/lib/Caxy/HtmlDiff/HtmlDiff.php
@@ -97,11 +97,19 @@ public function build()
             return $this->content;
         }
 
+        // Pre-processing Optimizations
+
+        // 1. Equality
+        if ($this->oldText == $this->newText) {
+            return $this->newText;
+        }
+
         $this->splitInputsToWords();
         $this->replaceIsolatedDiffTags();
         $this->indexNewWords();
 
         $operations = $this->operations();
+
         foreach ($operations as $item) {
             $this->performOperation($item);
         }
@@ -368,7 +376,7 @@ protected function diffElements($oldText, $newText, $stripWrappingTags = true)
      */
     protected function diffList($oldText, $newText)
     {
-        $diff = ListDiff::create($oldText, $newText, $this->config);
+        $diff = ListDiffLines::create($oldText, $newText, $this->config);
 
         return $diff->build();
     }

diff --git a/lib/Caxy/HtmlDiff/HtmlDiffConfig.php b/lib/Caxy/HtmlDiff/HtmlDiffConfig.php
@@ -43,9 +43,9 @@ class HtmlDiffConfig
         'dl' => '[[REPLACE_DEFINITION_LIST]]',
         'table' => '[[REPLACE_TABLE]]',
         'strong' => '[[REPLACE_STRONG]]',
-        'b' => '[[REPLACE_B]]',
+        'b' => '[[REPLACE_STRONG]]',
         'em' => '[[REPLACE_EM]]',
-        'i' => '[[REPLACE_I]]',
+        'i' => '[[REPLACE_EM]]',
         'a' => '[[REPLACE_A]]',
         'img' => '[[REPLACE_IMG]]',
     );

diff --git a/lib/Caxy/HtmlDiff/LcsService.php b/lib/Caxy/HtmlDiff/LcsService.php
@@ -0,0 +1,88 @@
+<?php
+
+namespace Caxy\HtmlDiff;
+
+use Caxy\HtmlDiff\Strategy\EqualMatchStrategy;
+use Caxy\HtmlDiff\Strategy\MatchStrategyInterface;
+
+class LcsService
+{
+    /**
+     * @var MatchStrategyInterface
+     */
+    protected $matchStrategy;
+
+    /**
+     * LcsService constructor.
+     *
+     * @param MatchStrategyInterface $matchStrategy
+     */
+    public function __construct(MatchStrategyInterface $matchStrategy = null)
+    {
+        if (null === $matchStrategy) {
+            $matchStrategy = new EqualMatchStrategy();
+        }
+
+        $this->matchStrategy = $matchStrategy;
+    }
+
+    /**
+     * @param array $a
+     * @param array $b
+     *
+     * @return array
+     */
+    public function longestCommonSubsequence(array $a, array $b)
+    {
+        $c = array();
+
+        $m = count($a);
+        $n = count($b);
+
+        for ($i = 0; $i <= $m; $i++) {
+            $c[$i][0] = 0;
+        }
+
+        for ($j = 0; $j <= $n; $j++) {
+            $c[0][$j] = 0;
+        }
+
+        for ($i = 1; $i <= $m; $i++) {
+            for ($j = 1; $j <= $n; $j++) {
+                if ($this->matchStrategy->isMatch($a[$i - 1], $b[$j - 1])) {
+                    $c[$i][$j] = 1 + (isset($c[$i - 1][$j - 1]) ? $c[$i - 1][$j - 1] : 0);
+                } else {
+                    $c[$i][$j] = max(
+                        isset($c[$i][$j - 1]) ? $c[$i][$j - 1] : 0,
+                        isset($c[$i - 1][$j]) ? $c[$i - 1][$j] : 0
+                    );
+                }
+            }
+        }
+
+        $lcs = array_pad([], $m + 1, 0);
+        $this->compileMatches($c, $a, $b, $m, $n, $lcs);
+
+        return $lcs;
+    }
+
+    /**
+     * @param $c
+     * @param $a
+     * @param $b
+     * @param $i
+     * @param $j
+     * @param $matches
+     */
+    protected function compileMatches($c, $a, $b, $i, $j, &$matches)
+    {
+        if ($i > 0 && $j > 0 && $this->matchStrategy->isMatch($a[$i - 1], $b[$j - 1])) {
+            $this->compileMatches($c, $a, $b, $i - 1, $j - 1, $matches);
+            $matches[$i] = $j;
+        } elseif ($j > 0 && ($i === 0 || $c[$i][$j - 1] >= $c[$i - 1][$j])) {
+            $this->compileMatches($c, $a, $b, $i, $j - 1, $matches);
+        } elseif ($i > 0 && ($j === 0 || $c[$i][$j - 1] < $c[$i - 1][$j])) {
+            $this->compileMatches($c, $a, $b, $i - 1, $j, $matches);
+        }
+    }
+}