From 737ec72b652f0cb1e84d6d39aff451a59f2b92f8 Mon Sep 17 00:00:00 2001
From: Alexej Kossmann <32612134+AlexejKossmann@users.noreply.github.com>
Date: Mon, 12 Apr 2021 15:08:36 +0200
Subject: [PATCH] Adds Configuration to skipp html tags (#5)

* added logic to skipp tags

* added script tag to be skipped default

* adjusted tests

* added ci workflow configuration
---
 .github/workflows/ci.yml                      | 56 +++++++++++++++++++
 README.md                                     | 13 ++++-
 phpunit.xml.dist                              |  9 ++-
 src/DependencyInjection/Configuration.php     | 44 +++++++++++++++
 .../HyphenatorExtension.php                   |  9 ++-
 src/Hyphenator/FrontendHyphenator.php         | 39 ++++++++++++-
 tests/Hyphenator/FrontendHyphenatorTest.php   | 15 ++++-
 7 files changed, 179 insertions(+), 6 deletions(-)
 create mode 100644 .github/workflows/ci.yml
 create mode 100644 src/DependencyInjection/Configuration.php

diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
new file mode 100644
index 0000000..6fc83f7
--- /dev/null
+++ b/.github/workflows/ci.yml
@@ -0,0 +1,56 @@
+name: CI
+
+on: [ push ]
+
+jobs:
+  tests:
+    name: PHP ${{ matrix.php }}
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        php: [ 7.2., 7.3, 7.4 ]
+    steps:
+      - name: Setup PHP
+        uses: shivammathur/setup-php@v2
+        with:
+          php-version: ${{ matrix.php }}
+          extensions: dom, fileinfo, filter, gd, hash, intl, json, mbstring, pcre, pdo, zlib
+          tools: phpunit
+          coverage: none
+
+      - name: Checkout
+        uses: actions/checkout@v2
+
+      - name: Install the dependencies
+        run: composer install --no-interaction
+
+      - name: Run the unit tests
+        run: php vendor/bin/phpunit -c phpunit.xml.dist --colors=always
+
+  coverage:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Setup PHP
+        uses: shivammathur/setup-php@v2
+        with:
+          php-version: 7.4
+          extensions: dom, fileinfo, filter, gd, hash, intl, json, mbstring, pcre, pdo, zlib
+          coverage: xdebug
+          tools: php-cs-fixer, phpunit
+
+      - name: Checkout
+        uses: actions/checkout@v2
+
+      - name: Install the dependencies
+        run: composer install --no-interaction
+
+      - name: Generate the coverage report
+        run: php vendor/bin/phpunit -c phpunit.xml.dist --coverage-clover build/logs/clover.xml
+
+      - name: Coveralls
+        env:
+          COVERALLS_REPO_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          composer global require php-coveralls/php-coveralls
+          php-coveralls --coverage_clover=build/logs/clover.xml -v
\ No newline at end of file
diff --git a/README.md b/README.md
index 6468114..f801a81 100644
--- a/README.md
+++ b/README.md
@@ -29,6 +29,18 @@ hyphenator_locale_language_mapping | array | ['en' => 'en-us', 'cz' => 'cs'] | M
 
 If you want to skip several tags from hyphenation simply add `hyphen-none` as css-class to the appropriate element or use the `tl_page.hyphenation` field. 
 
+You also can add tags to be skipped to your project configuration. See configuration reference below
+
+## Configuration reference
+
+```yaml
+# Default configuration for extension with alias: "huh_hyphenator"
+huh_hyphenator:
+  
+  # Add tags you want to be skipped from hyphenating, to array (string without <>)
+  skip_tags: []
+
+```
 
 ## Line break exceptions
 
@@ -44,4 +56,3 @@ As you can see, if you provide an replace pattern, than an regular expression wi
 
 * [vanderlee/phpSyllable](https://github.com/vanderlee/phpSyllable)
 * [wa72/htmlpagedom](https://github.com/wasinger/htmlpagedom)
-
diff --git a/phpunit.xml.dist b/phpunit.xml.dist
index a27a59c..337c4e6 100644
--- a/phpunit.xml.dist
+++ b/phpunit.xml.dist
@@ -17,8 +17,9 @@
         <ini name="error_reporting" value="-1"/>
         <ini name="display_errors" value="1"/>
         <ini name="display_startup_errors" value="1"/>
-        <ini name="intl.default_locale" value="en" />
-        <ini name="intl.error_level" value="0" />
+        <ini name="intl.default_locale" value="en"/>
+        <ini name="intl.error_level" value="0"/>
+        <env name="SYMFONY_DEPRECATIONS_HELPER" value="weak" />
     </php>
 
     <testsuites>
@@ -35,4 +36,8 @@
             </exclude>
         </whitelist>
     </filter>
+
+    <listeners>
+        <listener class="Symfony\Bridge\PhpUnit\SymfonyTestsListener"/>
+    </listeners>
 </phpunit>
\ No newline at end of file
diff --git a/src/DependencyInjection/Configuration.php b/src/DependencyInjection/Configuration.php
new file mode 100644
index 0000000..926c844
--- /dev/null
+++ b/src/DependencyInjection/Configuration.php
@@ -0,0 +1,44 @@
+<?php
+
+/*
+ * Copyright (c) 2021 Heimrich & Hannot GmbH
+ *
+ * @license LGPL-3.0-or-later
+ */
+
+namespace HeimrichHannot\HyphenatorBundle\DependencyInjection;
+
+use Symfony\Component\Config\Definition\Builder\TreeBuilder;
+use Symfony\Component\Config\Definition\ConfigurationInterface;
+
+class Configuration implements ConfigurationInterface
+{
+    const ROOT_ID = 'huh_hyphenator';
+
+    /**
+     * {@inheritdoc}
+     */
+    public function getConfigTreeBuilder()
+    {
+        $treeBuilder = new TreeBuilder(static::ROOT_ID);
+
+        // Keep compatibility with symfony/config < 4.2
+        if (!method_exists($treeBuilder, 'getRootNode')) {
+            $rootNode = $treeBuilder->root(static::ROOT_ID);
+        } else {
+            $rootNode = $treeBuilder->getRootNode();
+        }
+
+//        $rootNode = $treeBuilder->root(static::ROOT_ID);
+        $rootNode
+            ->children()
+                ->arrayNode('skip_tags')
+                    ->scalarPrototype()
+                        ->defaultValue(['script'])
+                    ->end()
+                ->end()
+            ->end();
+
+        return $treeBuilder;
+    }
+}
diff --git a/src/DependencyInjection/HyphenatorExtension.php b/src/DependencyInjection/HyphenatorExtension.php
index d65ac6c..386c145 100644
--- a/src/DependencyInjection/HyphenatorExtension.php
+++ b/src/DependencyInjection/HyphenatorExtension.php
@@ -20,8 +20,15 @@ class HyphenatorExtension extends Extension
      */
     public function load(array $configs, ContainerBuilder $container)
     {
-        $loader = new YamlFileLoader($container, new FileLocator(__DIR__.'/../Resources/config'));
+        $configuration = new Configuration();
+        $container->setParameter(Configuration::ROOT_ID, $this->processConfiguration($configuration, $configs));
 
+        $loader = new YamlFileLoader($container, new FileLocator(__DIR__.'/../Resources/config'));
         $loader->load('services.yml');
     }
+
+    public function getAlias()
+    {
+        return Configuration::ROOT_ID;
+    }
 }
diff --git a/src/Hyphenator/FrontendHyphenator.php b/src/Hyphenator/FrontendHyphenator.php
index 063b220..437022e 100644
--- a/src/Hyphenator/FrontendHyphenator.php
+++ b/src/Hyphenator/FrontendHyphenator.php
@@ -1,7 +1,7 @@
 <?php
 
 /*
- * Copyright (c) 2020 Heimrich & Hannot GmbH
+ * Copyright (c) 2021 Heimrich & Hannot GmbH
  *
  * @license LGPL-3.0-or-later
  */
@@ -108,10 +108,47 @@ function (HtmlPageCrawler $node, $i) use ($h, &$cache, $cacheEnabled, $objPage)
                 $html = $this->handleLineBreakExceptions($html, $objPage);
 
                 if (!$this->isHyphenationDisabled($objPage, Config::get('hyphenator_skipPages'))) {
+                    // mask tags configured to be skipped
+                    $skipTagCache = [];
+                    $skipTagCacheIndex = 0;
+
+                    foreach ($this->container->getParameter('huh_hyphenator')['skip_tags'] as $tag) {
+                        if (\in_array($tag, $this->voidElements)) {
+                            $html = preg_replace_callback(
+                                '#<\s*?'.$tag.'\b[^>]*>#s',
+                                function ($matches) use (&$skipTagCache, $skipTagCacheIndex) {
+                                    $skipTagCache[$skipTagCacheIndex] = $matches[0];
+                                    ++$skipTagCacheIndex;
+
+                                    return '####skip:open####'.($skipTagCacheIndex - 1).'####skip:close####';
+                                }, $html
+                            );
+                        } else {
+                            $html = preg_replace_callback(
+                                '#<\s*?'.$tag.'\b[^>]*>(.*?)</'.$tag.'\b[^>]*>#s',
+                                function ($matches) use (&$skipTagCache, &$skipTagCacheIndex) {
+                                    $skipTagCache[$skipTagCacheIndex] = $matches[0];
+                                    ++$skipTagCacheIndex;
+
+                                    return '####skip:open####'.($skipTagCacheIndex - 1).'####skip:close####';
+                                }, $html
+                            );
+                        }
+                    }
+
                     // if html contains nested tags, use the hyphenateHtml that excludes HTML tags and attributes
                     libxml_use_internal_errors(true); // disable error reporting when potential using HTML5 tags
                     $html = $h->hyphenateHtml($html);
                     libxml_clear_errors();
+
+                    // replace skipped tags
+                    $html = preg_replace_callback(
+                        '/####skip:open####(.*)####skip:close####/',
+                        function ($matches) use ($skipTagCache) {
+                            return $skipTagCache[$matches[1]];
+                        },
+                        $html
+                    );
                 }
 
                 if (false === preg_match('#<body>(<p>)?(?<content>.+?)(<\/p>)?<\/body>#is', $html, $matches) || !isset($matches['content'])) {
diff --git a/tests/Hyphenator/FrontendHyphenatorTest.php b/tests/Hyphenator/FrontendHyphenatorTest.php
index e46bdda..ed3cb96 100644
--- a/tests/Hyphenator/FrontendHyphenatorTest.php
+++ b/tests/Hyphenator/FrontendHyphenatorTest.php
@@ -1,7 +1,7 @@
 <?php
 
 /*
- * Copyright (c) 2020 Heimrich & Hannot GmbH
+ * Copyright (c) 2021 Heimrich & Hannot GmbH
  *
  * @license LGPL-3.0-or-later
  */
@@ -68,6 +68,7 @@ public function testHyphenate($buffer, array $pageData, array $config, $expected
         $modelUtil = $this->createMock(ModelUtil::class);
 
         $this->container->set('huh.utils.model', $modelUtil);
+        $this->container->setParameter('huh_hyphenator', ['skip_tags' => ['script', 'img']]);
 
         $listener = new FrontendHyphenator($this->container);
 
@@ -171,6 +172,18 @@ public function hyphenationProvider()
                 $this->getConfig(),
                 '<picture><!--[if IE 9]><video style="display: none;"><![endif]--><source srcset="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNkqAcAAIUAgUW0RjgAAAAASUVORK5CYII=" media="(min-width: 853px)"><!--[if IE 9]></video><![endif]--><img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNkqAcAAIUAgUW0RjgAAAAASUVORK5CYII=" data-wrapper="#image-wrapper-1368074709" class="image" alt=""></picture>',
             ],
+            [
+                '<a><script>document.querySelectorAll("#id").addEventListener("click", e => {console.log(e.target})</script></a>',
+                $this->getPage(),
+                $this->getConfig(),
+                '<a><script>document.querySelectorAll("#id").addEventListener("click", e => {console.log(e.target})</script></a>',
+            ],
+            [
+                '<a><picture><!--[if IE 9]><video style="display: none;"><![endif]--><source srcset="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNkqAcAAIUAgUW0RjgAAAAASUVORK5CYII=" media="(min-width: 853px)"><!--[if IE 9]></video><![endif]--><img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNkqAcAAIUAgUW0RjgAAAAASUVORK5CYII=" data-wrapper="#image-wrapper-1368074709" class="image" alt=""></picture></a>',
+                $this->getPage(),
+                $this->getConfig(),
+                '<a><picture><!--[if IE 9]><video style="display: none;"><![endif]--><source srcset="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNkqAcAAIUAgUW0RjgAAAAASUVORK5CYII=" media="(min-width: 853px)"><!--[if IE 9]></video><![endif]--><img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNkqAcAAIUAgUW0RjgAAAAASUVORK5CYII=" data-wrapper="#image-wrapper-1368074709" class="image" alt=""></picture></a>',
+            ],
         ];
     }