Add a --disable-astro-inline-modules to prevent inlining external module scripts

src/Crawler/ContentProcessor/AstroProcessor.php

@@ -23,6 +23,8 @@ class AstroProcessor extends BaseProcessor implements ContentProcessor
         Crawler::CONTENT_TYPE_ID_SCRIPT,
     ];
 
+    private bool $disableInlineModules = false;
+
     /**
      * @inheritDoc
      */
@@ -47,6 +49,15 @@ public function findUrls(string $content, ParsedUrl $sourceUrl): ?FoundUrls
         return null;
     }
 
+    /**
+     * Set whether to disable inlining of Astro modules for offline export
+     * @param bool $disable
+     */
+    public function setDisableInlineModules(bool $disable): void
+    {
+        $this->disableInlineModules = $disable;
+    }
+
     /**
      * Astro will need to replace all modules with inline content due to CORS is blocking modules with file:// protocol
      *
@@ -59,6 +70,11 @@ public function applyContentChangesForOfflineVersion(string &$content, int $cont
             return;
         }
 
+        // Skip inlining if disabled via option
+        if ($this->disableInlineModules) {
+            return;
+        }
+
         // stack of already included modules to prevent duplicates
         $alreadyIncludedModules = [];
 

src/Crawler/Export/OfflineWebsiteExporter.php

@@ -59,6 +59,12 @@ class OfflineWebsiteExporter extends BaseExporter implements Exporter
 
     protected bool $ignoreStoreFileError = false;
 
+    /**
+     * Disable inlining of Astro module scripts for offline export
+     * @var bool
+     */
+    protected bool $disableAstroInlineModules = false;
+
     /**
      * Replace HTML/JS/CSS content with `xxx -> bbb` or regexp in PREG format: `/card[0-9]/ -> card`
      *
@@ -103,6 +109,16 @@ public function export(): void
         // user-defined replaceQueryString will deactivate replacing query string with hash and use custom replacement
         OfflineUrlConverter::setReplaceQueryString($this->replaceQueryString);
 
+        // configure AstroProcessor to disable inline modules if requested
+        if ($this->disableAstroInlineModules) {
+            foreach ($this->crawler->getContentProcessorManager()->getProcessors() as $processor) {
+                if ($processor instanceof \Crawler\ContentProcessor\AstroProcessor) {
+                    $processor->setDisableInlineModules(true);
+                    break;
+                }
+            }
+        }
+
         // filter only relevant URLs with OK status codes
         $exportedUrls = array_filter($visitedUrls, function (VisitedUrl $visitedUrl) {
             return in_array($visitedUrl->statusCode, [200, 201, 301, 302, 303, 308]);
@@ -336,6 +352,7 @@ public static function getOptions(): Options
             new Option('--replace-content', null, 'replaceContent', Type::REPLACE_CONTENT, true, "Replace HTML/JS/CSS content with `foo -> bar` or regexp in PREG format: `/card[0-9]/i -> card`", null, true, true),
             new Option('--replace-query-string', null, 'replaceQueryString', Type::REPLACE_CONTENT, true, "Instead of using a short hash instead of a query string in the filename, just replace some characters. You can use simple format 'foo -> bar' or regexp in PREG format, e.g. '/([a-z]+)=([^&]*)(&|$)/i -> $1__$2'", null, true, true),
             new Option('--ignore-store-file-error', null, 'ignoreStoreFileError', Type::BOOL, false, 'Ignores any file storing errors. The export process will continue.', false, false),
+            new Option('--disable-astro-inline-modules', null, 'disableAstroInlineModules', Type::BOOL, false, 'Disables inlining of Astro module scripts for offline export. Scripts will remain as external files with corrected relative paths.', false, false),
         ]));
         return $options;
     }