EPUB: forward dir and lang attributes from the html element

In EPUB, the <html> node of each embedded HTML file is not included in the generated single DOM. We now parse its attributes and forward them to be included as attribute of the followup <docFragment> element, so they are part of the DOM.
koreader · Sep 14, 2019 · 12e542e · 12e542e
1 parent ed2c999
commit 12e542e
Show file tree

Hide file tree

Showing 2 changed files with 32 additions and 3 deletions.
diff --git a/crengine/include/lvtinydom.h b/crengine/include/lvtinydom.h
@@ -2456,6 +2456,10 @@ class ldomDocumentFragmentWriter : public LVXMLParserCallback
     lString8 headStyleText;
     int headStyleState;
 
+    lString16 htmlDir;
+    lString16 htmlLang;
+    bool insideHtmlTag;
+
 public:
 
     /// return content of html/head/style element
@@ -2486,6 +2490,9 @@ class ldomDocumentFragmentWriter : public LVXMLParserCallback
         insideTag = false;
         headStyleText.clear();
         headStyleState = 0;
+        insideHtmlTag = false;
+        htmlDir.clear();
+        htmlLang.clear();
     }
     /// called on parsing end
     virtual void OnStop()
@@ -2525,7 +2532,8 @@ class ldomDocumentFragmentWriter : public LVXMLParserCallback
     /// constructor
     ldomDocumentFragmentWriter( LVXMLParserCallback * parentWriter, lString16 baseTagName, lString16 baseTagReplacementName, lString16 fragmentFilePath )
     : parent(parentWriter), baseTag(baseTagName), baseTagReplacement(baseTagReplacementName),
-    insideTag(false), styleDetectionState(0), pathSubstitutions(100), baseElement(NULL), lastBaseElement(NULL), headStyleState(0)
+    insideTag(false), styleDetectionState(0), pathSubstitutions(100), baseElement(NULL), lastBaseElement(NULL),
+    headStyleState(0), insideHtmlTag(false)
     {
         setCodeBase( fragmentFilePath );
     }

diff --git a/crengine/src/lvtinydom.cpp b/crengine/src/lvtinydom.cpp
@@ -10043,7 +10043,15 @@ void ldomDocumentFragmentWriter::OnAttribute( const lChar16 * nsname, const lCha
             parent->OnAttribute(nsname, attrname, attrvalue);
         }
     } else {
-        if ( styleDetectionState ) {
+        if (insideHtmlTag) {
+            // Grab attributes from <html dir="rtl" lang="he"> (not included in the DOM)
+            // to reinject them in <DocFragment>
+            if ( !lStr_cmp(attrname, "dir") )
+                htmlDir = attrvalue;
+            else if ( !lStr_cmp(attrname, "lang") )
+                htmlLang = attrvalue;
+        }
+        else if ( styleDetectionState ) {
             if ( !lStr_cmp(attrname, "rel") && lString16(attrvalue).lowercase() == L"stylesheet" )
                 styleDetectionState |= 2;
             else if ( !lStr_cmp(attrname, "type") ) {
@@ -10079,8 +10087,13 @@ ldomNode * ldomDocumentFragmentWriter::OnTagOpen( const lChar16 * nsname, const
     } else {
         if ( !lStr_cmp(tagname, "link") )
             styleDetectionState = 1;
-        if ( !lStr_cmp(tagname, "style") )
+        else if ( !lStr_cmp(tagname, "style") )
             headStyleState = 1;
+        else if ( !lStr_cmp(tagname, "html") ) {
+            insideHtmlTag = true;
+            htmlDir.clear();
+            htmlLang.clear();
+        }
     }
 
     // When meeting the <body> of each of an EPUB's embedded HTML files,
@@ -10115,6 +10128,11 @@ ldomNode * ldomDocumentFragmentWriter::OnTagOpen( const lChar16 * nsname, const
             }
             if ( !codeBasePrefix.empty() ) // add attribute <DocFragment id="..html_file_name"
                 parent->OnAttribute(L"", L"id", codeBasePrefix.c_str() );
+            if ( !htmlDir.empty() ) // add attribute <DocFragment dir="rtl" from <html dir="rtl"> tag
+                parent->OnAttribute(L"", L"dir", htmlDir.c_str() );
+            if ( !htmlLang.empty() ) // add attribute <DocFragment lang="ar" from <html lang="ar"> tag
+                parent->OnAttribute(L"", L"lang", htmlLang.c_str() );
+
             parent->OnTagBody(); // inside <DocFragment>
             if ( !headStyleText.empty() || stylesheetLinks.length() > 0 ) {
                 // add stylesheet element as child of <DocFragment>: <stylesheet href="...">
@@ -10178,6 +10196,9 @@ void ldomDocumentFragmentWriter::OnTagBody()
     if ( insideTag ) {
         parent->OnTagBody();
     }
+    else if ( insideHtmlTag ) {
+        insideHtmlTag = false;
+    }
     if ( styleDetectionState == 11 ) {
         // incomplete <link rel="stylesheet", href="..." />; assuming type="text/css"
         if ( !stylesheetFile.empty() )