Issue #1 modernize library

.gitignore

@@ -0,0 +1 @@
+/vendor/

README.md

@@ -14,32 +14,54 @@ than simply calling `strpos` many times, and it's much faster than calling
 I originally wrote this to use with [F5Bot](https://f5bot.com), since it's
 searching for the same set of a few thousand keywords over and over again.
 
+# Install via Composer
+
+Add the following to your project's `composer.json`:
+
+```json
+    "repositories": [
+        {
+            "type": "vcs",
+            "url": "https://github.com/codeplea/ahocorasickphp"
+        }
+    ],
+    "require": {
+        "codeplea/ahocorasickphp": "dev-master"
+    }
+```
+
+Then, install the package itself:
+
+```bash
+$ composer update
+```
+
 # Usage
 
-It's designed to be really easy to use. You create the `ahocorasick` object,
+It's designed to be really easy to use. You create the `Search` object,
 add your keywords, call `finalize()` to finish setup, and then search your
 text. It'll return an array of the keywords found and their position in the
 search text.
 
 Create, add keywords, and `finalize()`:
 
 ```php
-require('ahocorasick.php');
+use codeplea\AhoCorasick\Search;
 
-$ac = new ahocorasick();
+$ac = new Search();
 
-$ac->add_needle('art');
-$ac->add_needle('cart');
-$ac->add_needle('ted');
+$ac->addNeedle('art');
+$ac->addNeedle('cart');
+$ac->addNeedle('ted');
 
 $ac->finalize();
 
 ```
 
-Call `search()` to preform the actual search. It'll return an array of matches.
+Call `execute()` to preform the actual search. It'll return an array of matches.
 
 ```php
-$found = $ac->search('a carted mart lot one blue ted');
+$found = $ac->execute('a carted mart lot one blue ted');
 print_r($found);
 ```
 
@@ -97,10 +119,17 @@ time: 0.054709911346436
 
 ```
 
-Note: the regex solutions are actually slightly broken. They won't work if you
+**Note:** the regex solutions are actually slightly broken. They won't work if you
 have a keyword that is a prefix or suffix of another. But hey, who really uses
 regex when it's not slightly broken?
 
-Also keep in mind that building the search tree (the `add_needle()` and
+Also keep in mind that building the search tree (the `addNeedle()` and
 `finalize()` calls) takes time. So you'll get the best speed-up if you're
-reusing the same keywords and calling `search()` many times.
+reusing the same keywords and calling `execute()` many times.
+
+# Running tests
+
+```$php
+$ composer install
+$ ./vendor/bin/phpunit
+```

benchmark.php

@@ -1,120 +1,11 @@
 <?php
+use codeplea\AhoCorasick\Benchmark;
 
-/* This program will benchmark searching for 1,000 keywords in a 5,000 word text all at once. */
-/* It compares our ahocorasick method with regex and strpos. */
-
-
-require('ahocorasick.php');
-require('benchmark_setup.php'); /* keywords and text */
-
-$loops = 10;
-
-print("Loaded " . count($needles) . " keywords to search on a text of " .
-  strlen($haystack) . " characters.\n");
-
-print("\nSearching with strpos...\n");
-
-$st = microtime(1);
-for ($loop = 0; $loop < $loops; ++$loop) {
-  $found = array();
-  foreach($needles as $n) {
-    $k = 0;
-    while(($k = strpos($haystack, $n, $k)) !== FALSE) {
-      $found[] = array($n, $k);
-      ++$k;
-    }
-  }
-}
-$et = microtime(1);
-print("time: " . ($et - $st) . "\n");
-$found_strpos = $found;
-
-
-
-
-
-
-print("\nSearching with preg_match...\n");
-//Note, this actually sucks and misses cases where one needle is a prefix or
-//suffix of another.
-$regex = '/' . implode('|', $needles) . '/';
-
-$st = microtime(1);
-for ($loop = 0; $loop < $loops; ++$loop) {
-  $found = array();
-  $k = 0;
-  while(preg_match($regex, $haystack, $m, PREG_OFFSET_CAPTURE, $k)) {
-    $found[] = $m[0];
-    $k = $m[0][1] + 1;
-  }
-}
-$et = microtime(1);
-print("time: " . ($et - $st) . "\n");
-//print_r($found);
-
-
-
-
-
-
-print("\nSearching with preg_match_all...\n");
-//Note, this actually sucks and misses cases where one needle is a prefix or
-//suffix of another.
-$regex = '/' . implode('|', $needles) . '/';
-
-$st = microtime(1);
-for ($loop = 0; $loop < $loops; ++$loop) {
-  $found = array();
-  $k = 0;
-  preg_match_all($regex, $haystack, $found, PREG_OFFSET_CAPTURE);
-  $found = $found[0];
-}
-$et = microtime(1);
-print("time: " . ($et - $st) . "\n");
-
-
-
-
-
-print("\nSearching with aho corasick...\n");
-$ac = new ahocorasick();
-foreach ($needles as $n) $ac->add_needle($n);
-$ac->finalize();
-
-$st = microtime(1);
-for ($loop = 0; $loop < $loops; ++$loop) {
-  $found = array();
-  $found = $ac->search($haystack);
-}
-$et = microtime(1);
-print("time: " . ($et - $st) . "\n");
-
-
-
-
-
-
-//Check that the answers match.
-//First sort the arrays.
-$comp = function($a, $b) {return ($a[1] === $b[1]) ? ($a[0] > $b[0]) : ($a[1] > $b[1]);};
-usort($found, $comp);
-usort($found_strpos, $comp);
-
-if ($found_strpos !== $found) {
-  print("ERROR - Aho Corasick got the wrong result.\n");
-
-  print("strpos size: " . count($found_strpos) . "\n");
-  print("aho corasick size: " . count($found) . "\n");
-
-  for ($i = 0; $i < count($found); ++$i) {
-    if ($found_strpos[$i] !== $found[$i]) {
-      print("Mismatch $i\n");
-      print_r($found_strpos[$i]);
-      print_r($found[$i]);
-    }
-  }
-}
-
-
+require 'vendor/autoload.php';
 
+/* keywords and text */
+require 'benchmark_setup.php';
 
+// Benchmark searching for 1,000 keywords in a 5,000 word text all at once.
+$benchmark = new Benchmark();
+$benchmark->run($needles, $haystack);

benchmark_setup.php

@@ -1,6 +1,6 @@
 <?php
 
-$needles = array('abandonment', 'abashed', 'abashments', 'abduction',
+$needles = ['abandonment', 'abashed', 'abashments', 'abduction',
 'aberrant', 'abiding', 'abidingly', 'abjures', 'ablution', 'abolishes',
 'abominably', 'aborted', 'abrasion', 'abridgment', 'abscesses', 'absconds',
 'absences', 'absinthe', 'absolves', 'absorbingly', 'abundant', 'abused',
@@ -532,7 +532,7 @@
 'remonstrates', 'remorse', 'removes', 'remunerated', 'rendering',
 'renditions', 'reneged', 'renominate', 'renovators', 'reorders',
 'repatriates', 'repave', 'repaying', 'repeatedly', 'repertoires', 'replied',
-'reprehend', 'reprieves', 'reprimanded');
+'reprehend', 'reprieves', 'reprimanded'];
 
 $haystack = 'unscathed grampus antinuclear avenged waste oversee doggies spumes
 senators balk gooseberries grilles respelled ceramists outlaid maladroitly

composer.json

@@ -0,0 +1,25 @@
+{
+    "name": "codeplea/ahocorasickphp",
+    "description": "Aho-Corasick multi-keyword string searching library in PHP.",
+    "authors": [
+        {
+            "name": "Lewis Van Winkle"
+        }
+    ],
+    "type": "library",
+    "license": "zlib",
+    "config": {
+        "sort-packages": true
+    },
+    "require": {
+        "php": ">=7.0"
+    },
+    "require-dev": {
+        "phpunit/phpunit": "^7.3"
+    },
+    "autoload": {
+        "psr-4": {
+            "codeplea\\AhoCorasick\\": "src"
+        }
+    }
+}