fix: do not treat escaped <a> elements as hyperlinks in HTM-053

Fix the regex used to report "file:" hyperlinks as `HTM-053` (informative) to only consider HTML elements and not plain text. This regex-based parsing is still brittle, but we'll refactor this whole package later. For now this simple fix will do. Fixes #1182
w3c · Feb 26, 2021 · 5949b6c · 5949b6c
1 parent 5ee72e7
commit 5949b6c
Show file tree

Hide file tree

Showing 7 changed files with 56 additions and 1 deletion.
diff --git a/src/main/java/com/adobe/epubcheck/ctc/FileLinkSearch.java b/src/main/java/com/adobe/epubcheck/ctc/FileLinkSearch.java
@@ -20,7 +20,7 @@
  *  ========================================================<br/>
  */
 public class FileLinkSearch extends TextSearch {
-    private static final Pattern fileLinkPattern = Pattern.compile("href=[\"']file://");
+    private static final Pattern fileLinkPattern = Pattern.compile("<a\\s([^<>]*\\s)?href=[\"']file://");
 
     public FileLinkSearch(EPUBVersion version, ZipFile zip, Report report)
     {

diff --git a/src/test/resources/epub3/content-publication.feature b/src/test/resources/epub3/content-publication.feature
@@ -66,6 +66,12 @@ Feature: EPUB 3 ▸ Content Documents ▸ Full Publication Checks
     When checking EPUB 'content-xhtml-link-to-local-file-valid'
     Then info HTM-053 is reported
     And no errors or warnings are reported
+
+  Scenario: Do not report escaped hyperlinks to resources in the local file system
+    See issue #1182
+    When checking EPUB 'content-xhtml-link-to-local-file-escaped-valid'
+    Then info HTM-053 is reported 0 times
+    And no errors or warnings are reported
 
   Scenario: Report a hyperlink to a resource missing from the publication
     When checking EPUB 'content-xhtml-link-to-missing-doc-error'

diff --git a/...es/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/EPUB/content_001.xhtml b/...es/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/EPUB/content_001.xhtml
@@ -0,0 +1,12 @@
+<!DOCTYPE html>
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+	<head>
+		<meta charset="utf-8"/>
+		<title>Minimal EPUB</title>
+	</head>
+	<body>
+		<h1>Loomings</h1>
+		<p>Call me Ishmael.</p>
+		&lt;a class="external" href="file:///C:/path/file.pdf"&gt;link to local file&lt;/a&lt;
+	</body>
+</html>
diff --git a/.../resources/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/EPUB/nav.xhtml b/.../resources/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/EPUB/nav.xhtml
@@ -0,0 +1,14 @@
+<!DOCTYPE html>
+<html xmlns="http://www.w3.org/1999/xhtml" xmlns:epub="http://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/...esources/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/EPUB/package.opf b/...esources/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/EPUB/package.opf
@@ -0,0 +1,16 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<package xmlns="http://www.idpf.org/2007/opf" version="3.0" xml:lang="en" unique-identifier="q">
+<metadata xmlns:dc="http://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>
+</metadata>
+<manifest>
+  <item id="content_001"  href="content_001.xhtml" media-type="application/xhtml+xml"/>
+  <item id="nav"  href="nav.xhtml" media-type="application/xhtml+xml" properties="nav"/>
+</manifest>
+<spine>
+  <itemref idref="content_001" />
+</spine>
+</package>
diff --git a/...es/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/META-INF/container.xml b/...es/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/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/content-xhtml-link-to-local-file-escaped-valid/mimetype b/src/test/resources/epub3/files/epub/content-xhtml-link-to-local-file-escaped-valid/mimetype
@@ -0,0 +1 @@
+application/epub+zip