Add src_lazy_load_attr in siteconfig

This allow to define the image attributes replacement for lazyloaded image
j0k3r · Mar 4, 2018 · 5ed0fd3 · 5ed0fd3
1 parent 6b4de37
commit 5ed0fd3
Show file tree

Hide file tree

Showing 5 changed files with 17 additions and 1 deletion.
diff --git a/src/Extractor/ContentExtractor.php b/src/Extractor/ContentExtractor.php
@@ -179,6 +179,11 @@ public function process($html, $url, SiteConfig $siteConfig = null, $smartTidy =
             $this->siteConfig = $this->buildSiteConfig($url, $html);
         }
 
+        // add lazyload information from siteconfig
+        if ($this->siteConfig->src_lazy_load_attr && !in_array($this->siteConfig->src_lazy_load_attr, $this->config['src_lazy_load_attributes'], true)) {
+            $this->config['src_lazy_load_attributes'][] = $this->siteConfig->src_lazy_load_attr;
+        }
+
         // do string replacements
         if (!empty($this->siteConfig->find_string)) {
             if (count($this->siteConfig->find_string) === count($this->siteConfig->replace_string)) {

diff --git a/src/SiteConfig/ConfigBuilder.php b/src/SiteConfig/ConfigBuilder.php
@@ -339,7 +339,7 @@ public function parseLines(array $lines)
             } elseif (in_array($command, ['tidy', 'prune', 'autodetect_on_failure', 'requires_login'], true)) {
                 $config->$command = ('yes' === $val || 'true' === $val);
                 // check for single statement commands stored as strings
-            } elseif (in_array($command, ['parser', 'login_username_field', 'login_password_field', 'not_logged_in_xpath', 'login_uri'], true)) {
+            } elseif (in_array($command, ['parser', 'login_username_field', 'login_password_field', 'not_logged_in_xpath', 'login_uri', 'src_lazy_load_attr'], true)) {
                 $config->$command = $val;
                 // check for replace_string(find): replace
             } elseif ((')' === substr($command, -1)) && preg_match('!^([a-z0-9_]+)\((.*?)\)$!i', $command, $match) && 'replace_string' === $match[1]) {

diff --git a/src/SiteConfig/SiteConfig.php b/src/SiteConfig/SiteConfig.php
@@ -32,6 +32,9 @@ class SiteConfig
     // Strip attributes matching these xpath expressions (0 or more)
     public $strip_attr = [];
 
+    // Attribute used to replace lazyload image (like `data-lazy-src`)
+    public $src_lazy_load_attr = null;
+
     // Strip elements which contain these strings (0 or more) in the id or class attribute
     public $strip_id_or_class = [];
 

diff --git a/tests/Extractor/ContentExtractorTest.php b/tests/Extractor/ContentExtractorTest.php
@@ -744,6 +744,11 @@ public function dataForlazyLoad()
                 '<div>' . str_repeat('this is the best part of the show', 10) . '<img src="data:image/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==" data-sources="http://0.0.0.0/big_image.jpg"/></div>',
                 '<img src="http://0.0.0.0/big_image.jpg"',
             ],
+            // test with img attribute from site config
+            [
+                '<div>' . str_repeat('this is the best part of the show', 10) . '<img src="data:image/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==" data-toto-src="http://0.0.0.0/big_image.jpg"/></div>',
+                '<img src="http://0.0.0.0/big_image.jpg"',
+            ],
         ];
     }
 
@@ -758,6 +763,7 @@ public function testConvertLazyLoadImages($html, $htmlExpected)
 
         $config = new SiteConfig();
         $config->body = ['//div'];
+        $config->src_lazy_load_attr = 'data-toto-src';
 
         $res = $contentExtractor->process(
             $html,

diff --git a/tests/SiteConfig/ConfigBuilderTest.php b/tests/SiteConfig/ConfigBuilderTest.php
@@ -35,6 +35,7 @@ public function testBuildFromArray()
             'title: hoho',
             'tidy: yes',
             'parser: bob',
+            'src_lazy_load_attr: data-toto-src',
             'date: foo',
             'replace_string(toto): titi',
             'http_header(user-agent): my-user-agent',
@@ -47,6 +48,7 @@ public function testBuildFromArray()
         $configExpected->title = ['hoho'];
         $configExpected->tidy = true;
         $configExpected->parser = 'bob';
+        $configExpected->src_lazy_load_attr = 'data-toto-src';
         $configExpected->find_string = ['toto'];
         $configExpected->replace_string = ['titi'];
         $configExpected->http_header = [