feat: check the epub:textref attribute on Media Overlays body and seq elements

mattgarrish · rdeltour · commit eea1574c0fd4 · 2020-12-28T12:19:21.000+01:00
- Check the value of the `epub:textref` attribute when found on `body`
  or `seq` elements
- New error MED-014 is reported when a fragment has been specified in
  `epub:textref` and `src` attributes
- RSC-012 is reported when the resource referenced in `epub.textref`
  could not be found
diff --git a/src/main/java/com/adobe/epubcheck/messages/DefaultSeverities.java b/src/main/java/com/adobe/epubcheck/messages/DefaultSeverities.java
@@ -150,6 +150,7 @@ private void initialize()
     severities.put(MessageId.MED_011, Severity.ERROR);
     severities.put(MessageId.MED_012, Severity.ERROR);
     severities.put(MessageId.MED_013, Severity.ERROR);
+    severities.put(MessageId.MED_014, Severity.ERROR);
 
     // NAV
     severities.put(MessageId.NAV_001, Severity.ERROR);
diff --git a/src/main/java/com/adobe/epubcheck/messages/MessageId.java b/src/main/java/com/adobe/epubcheck/messages/MessageId.java
@@ -144,6 +144,7 @@ public enum MessageId implements Comparable<MessageId>
   MED_011("MED_011"),
   MED_012("MED_012"),
   MED_013("MED_013"),
+  MED_014("MED_014"),
 
   // Epub3 based table of content errors
   NAV_001("NAV-001"),
diff --git a/src/main/java/com/adobe/epubcheck/overlay/OverlayHandler.java b/src/main/java/com/adobe/epubcheck/overlay/OverlayHandler.java
@@ -64,29 +64,28 @@ public void startElement()
     XMLElement e = parser.getCurrentElement();
     String name = e.getName();
 
-    if (name.equals("smil"))
-    {
-      vocabs = VocabUtil.parsePrefixDeclaration(
-          e.getAttributeNS(EpubConstants.EpubTypeNamespaceUri, "prefix"), RESERVED_VOCABS,
-          KNOWN_VOCAB_URIS, DEFAULT_VOCAB_URIS, report,
-          EPUBLocation.create(path, parser.getLineNumber(), parser.getColumnNumber()));
-    }
-    else if (name.equals("seq"))
-    {
-      processSeq(e);
-    }
-    else if (name.equals("text"))
-    {
-      processSrc(e);
-    }
-    else if (name.equals("audio"))
-    {
-      processRef(e.getAttribute("src"), XRefChecker.Type.AUDIO);
-      checkTime(e.getAttribute("clipBegin"), e.getAttribute("clipEnd"));
-    }
-    else if (name.equals("body") || name.equals("par"))
-    {
-      checkType(e.getAttributeNS(EpubConstants.EpubTypeNamespaceUri, "type"));
+    switch (name) {
+      case "smil":
+        vocabs = VocabUtil.parsePrefixDeclaration(
+            e.getAttributeNS(EpubConstants.EpubTypeNamespaceUri, "prefix"), RESERVED_VOCABS,
+            KNOWN_VOCAB_URIS, DEFAULT_VOCAB_URIS, report,
+            EPUBLocation.create(path, parser.getLineNumber(), parser.getColumnNumber()));
+        break;
+        
+      case "body":
+      case "seq":
+      case "par":
+        processGlobalAttrs(e);
+        break;
+       
+      case "text":
+        processSrc(e);
+        break;
+      
+      case "audio":
+        processRef(e.getAttribute("src"), XRefChecker.Type.AUDIO);
+        checkTime(e.getAttribute("clipBegin"), e.getAttribute("clipEnd"));
+        break;
     }
   }
   
@@ -163,6 +162,7 @@ private void processRef(String ref, XRefChecker.Type type)
         }
       }
       else {
+        checkFragment(ref);
         String uniqueResource = PathUtil.removeFragment(ref);
         if (!Strings.isNullOrEmpty(uniqueResource)) {
           if (!context.overlayTextChecker.get().add(uniqueResource, context.opfItem.get().getId())) {
@@ -175,10 +175,12 @@ private void processRef(String ref, XRefChecker.Type type)
     }
   }
 
-  private void processSeq(XMLElement e)
+  private void processGlobalAttrs(XMLElement e)
   {
-    processRef(e.getAttributeNS(EpubConstants.EpubTypeNamespaceUri, "textref"),
-        XRefChecker.Type.HYPERLINK);
+    if (!e.getName().equals("audio")) {
+      processRef(e.getAttributeNS(EpubConstants.EpubTypeNamespaceUri, "textref"),
+          XRefChecker.Type.HYPERLINK);
+    }
     checkType(e.getAttributeNS(EpubConstants.EpubTypeNamespaceUri, "type"));
   }
 
@@ -212,4 +214,13 @@ private void checkItemReferences() {
     
   }
 
+  private void checkFragment(String ref) {
+  
+    String frag = PathUtil.getFragment(ref.trim());
+    
+    if (ref.indexOf("#") == -1 || Strings.isNullOrEmpty(frag)) {
+      // must include a non-empty fragid
+      report.message(MessageId.MED_014, EPUBLocation.create(path, parser.getLineNumber(), parser.getColumnNumber()));
+    }
+  }
 }
diff --git a/src/main/resources/com/adobe/epubcheck/messages/MessageBundle.properties b/src/main/resources/com/adobe/epubcheck/messages/MessageBundle.properties
@@ -137,6 +137,7 @@ MED_010=EPUB Content Documents referenced from a Media Overlay must specify the
 MED_011=EPUB Content Document referenced from multiple Media Overlay Documents.
 MED_012=The "media-overlay" attribute does not match the ID of the Media Overlay that refers to this document.
 MED_013=Media Overlay Document referenced from the "media-overlay" attribute does not contain a reference to this Content Document.
+MED_014=A non-empty fragment identifier is required.
 
 #NAV EPUB v3 Table of contents
 NAV_001=The nav file is not supported for EPUB v2.
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-audio-non-cmt-error/EPUB/content_001.smil b/src/test/resources/epub3/files/epub/mediaoverlays-audio-non-cmt-error/EPUB/content_001.smil
@@ -2,7 +2,7 @@
 <smil xmlns="https://www.w3.org/ns/SMIL" xmlns:epub="https://www.idpf.org/2007/ops" version="3.0">
     <body>
         <par id="par1">
-            <text src="content_001.xhtml"/>
+            <text src="content_001.xhtml#c01"/>
             <audio src="content_001.m4a"/>
         </par>
     </body>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-audio-non-cmt-error/EPUB/content_001.xhtml b/src/test/resources/epub3/files/epub/mediaoverlays-audio-non-cmt-error/EPUB/content_001.xhtml
@@ -5,7 +5,7 @@
 		<title>Minimal EPUB</title>
 	</head>
 	<body>
-		<h1>Loomings</h1>
+		<h1 id="c01">Loomings</h1>
 		<p>Call me Ishmael.</p>
 	</body>
 </html>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/content_001.mp3 b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/content_001.mp3
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/content_001.smil b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/content_001.smil
@@ -0,0 +1,9 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<smil xmlns="https://www.w3.org/ns/SMIL" xmlns:epub="https://www.idpf.org/2007/ops" version="3.0">
+    <body epub:textref="content_001.xhtml#">
+    	<par id="par1">
+    		<text src="content_001.xhtml"/>
+    		<audio src="content_001.mp3"/>
+    	</par>
+    </body>
+</smil>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/content_001.xhtml b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/content_001.xhtml
@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html xmlns="https://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+	<head>
+		<meta charset="utf-8" />
+		<title>Minimal EPUB</title>
+	</head>
+	<body>
+		<h1 id="c01">Loomings</h1>
+		<p>Call me Ishmael.</p>
+	</body>
+</html>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/nav.xhtml b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/nav.xhtml
@@ -0,0 +1,14 @@
+<!DOCTYPE html>
+<html xmlns="https://www.w3.org/1999/xhtml" xmlns:epub="https://www.idpf.org/2007/ops" xml:lang="en" lang="en">
+	<head>
+		<meta charset="utf-8"/>
+		<title>Minimal Nav</title>
+	</head>
+	<body>
+		<nav epub:type="toc">
+			<ol>
+				<li><a href="content_001.xhtml">content 001</a></li>
+			</ol>
+		</nav>
+	</body>
+</html>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/package.opf b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/EPUB/package.opf
@@ -0,0 +1,20 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<package xmlns="https://www.idpf.org/2007/opf" version="3.0" xml:lang="en" unique-identifier="q">
+<metadata xmlns:dc="https://purl.org/dc/elements/1.1/">
+  <dc:title id="title">Minimal EPUB 3.0</dc:title>
+  <dc:language>en</dc:language>
+  <dc:identifier id="q">NOID</dc:identifier>
+  <meta property="dcterms:modified">2017-06-14T00:00:01Z</meta>
+  <meta property="media:duration">2.5s</meta>
+  <meta property="media:duration" refines="#mo_001">2.5s</meta>
+</metadata>
+<manifest>
+  <item id="content_001"  href="content_001.xhtml" media-type="application/xhtml+xml" media-overlay="mo_001"/>
+  <item id="mo_001"  href="content_001.smil" media-type="application/smil+xml"/>
+  <item id="audio_001"  href="content_001.mp3" media-type="audio/mpeg"/>
+  <item id="nav"  href="nav.xhtml" media-type="application/xhtml+xml" properties="nav"/>
+</manifest>
+<spine>
+  <itemref idref="content_001" />
+</spine>
+</package>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/META-INF/container.xml b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/META-INF/container.xml
@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8" ?>
+<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
+	<rootfiles>
+		<rootfile full-path="EPUB/package.opf" media-type="application/oebps-package+xml"/>
+	</rootfiles>
+</container>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/mimetype b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-invalid-error/mimetype
@@ -0,0 +1 @@
+application/epub+zip
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/content_001.mp3 b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/content_001.mp3
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/content_001.smil b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/content_001.smil
@@ -0,0 +1,11 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<smil xmlns="https://www.w3.org/ns/SMIL" xmlns:epub="https://www.idpf.org/2007/ops" version="3.0">
+    <body>
+    	<seq epub:textref="content_001.xhtml#c1111">
+    		<par id="par1">
+    			<text src="content_001.xhtml#c01"/>
+    			<audio src="content_001.mp3"/>
+    		</par>
+    	</seq>
+    </body>
+</smil>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/content_001.xhtml b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/content_001.xhtml
@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html xmlns="https://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+	<head>
+		<meta charset="utf-8" />
+		<title>Minimal EPUB</title>
+	</head>
+	<body>
+		<h1 id="c01">Loomings</h1>
+		<p>Call me Ishmael.</p>
+	</body>
+</html>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/nav.xhtml b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/nav.xhtml
@@ -0,0 +1,14 @@
+<!DOCTYPE html>
+<html xmlns="https://www.w3.org/1999/xhtml" xmlns:epub="https://www.idpf.org/2007/ops" xml:lang="en" lang="en">
+	<head>
+		<meta charset="utf-8"/>
+		<title>Minimal Nav</title>
+	</head>
+	<body>
+		<nav epub:type="toc">
+			<ol>
+				<li><a href="content_001.xhtml">content 001</a></li>
+			</ol>
+		</nav>
+	</body>
+</html>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/package.opf b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/EPUB/package.opf
@@ -0,0 +1,20 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<package xmlns="https://www.idpf.org/2007/opf" version="3.0" xml:lang="en" unique-identifier="q">
+<metadata xmlns:dc="https://purl.org/dc/elements/1.1/">
+  <dc:title id="title">Minimal EPUB 3.0</dc:title>
+  <dc:language>en</dc:language>
+  <dc:identifier id="q">NOID</dc:identifier>
+  <meta property="dcterms:modified">2017-06-14T00:00:01Z</meta>
+  <meta property="media:duration">2.5s</meta>
+  <meta property="media:duration" refines="#mo_001">2.5s</meta>
+</metadata>
+<manifest>
+  <item id="content_001"  href="content_001.xhtml" media-type="application/xhtml+xml" media-overlay="mo_001"/>
+  <item id="mo_001"  href="content_001.smil" media-type="application/smil+xml"/>
+  <item id="audio_001"  href="content_001.mp3" media-type="audio/mpeg"/>
+  <item id="nav"  href="nav.xhtml" media-type="application/xhtml+xml" properties="nav"/>
+</manifest>
+<spine>
+  <itemref idref="content_001" />
+</spine>
+</package>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/META-INF/container.xml b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/META-INF/container.xml
@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8" ?>
+<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
+	<rootfiles>
+		<rootfile full-path="EPUB/package.opf" media-type="application/oebps-package+xml"/>
+	</rootfiles>
+</container>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/mimetype b/src/test/resources/epub3/files/epub/mediaoverlays-fragid-resolve-error/mimetype
@@ -0,0 +1 @@
+application/epub+zip
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-minimal-valid/EPUB/content_001.smil b/src/test/resources/epub3/files/epub/mediaoverlays-minimal-valid/EPUB/content_001.smil
@@ -2,7 +2,7 @@
 <smil xmlns="https://www.w3.org/ns/SMIL" xmlns:epub="https://www.idpf.org/2007/ops" version="3.0">
     <body>
         <par id="par1">
-            <text src="content_001.xhtml"/>
+            <text src="content_001.xhtml#c01"/>
             <audio src="content_001.mp3"/>
         </par>
     </body>
diff --git a/src/test/resources/epub3/files/epub/mediaoverlays-minimal-valid/EPUB/content_001.xhtml b/src/test/resources/epub3/files/epub/mediaoverlays-minimal-valid/EPUB/content_001.xhtml
@@ -5,7 +5,7 @@
 		<title>Minimal EPUB</title>
 	</head>
 	<body>
-		<h1>Loomings</h1>
+		<h1 id="c01">Loomings</h1>
 		<p>Call me Ishmael.</p>
 	</body>
 </html>
diff --git a/src/test/resources/epub3/mediaoverlays-publication.feature b/src/test/resources/epub3/mediaoverlays-publication.feature
@@ -28,7 +28,17 @@ Feature: EPUB 3 ▸ Media Overlays ▸ Full Publication Checks
     Then error MED-011 is reported
     And no other errors or warnings are reported
 
-  ### The audio Element
+  Scenario: Report empty fragment identifiers
+    When checking EPUB 'mediaoverlays-fragid-invalid-error'
+    Then error MED-014 is reported 2 times
+    And no other errors or warnings are reported
+
+  Scenario: Report a fragment identifier that does not resolve to an element
+    When checking EPUB 'mediaoverlays-fragid-resolve-error'
+    Then error RSC-012 is reported
+    And no other errors or warnings are reported
+
+  ### 2.4.8 The audio Element
 
   Scenario: Report an audio clip that is not a Core Media Type
     When checking EPUB 'mediaoverlays-audio-non-cmt-error'
