Clean up attributes of image elements

javatests/org/chromium/distiller/ContentExtractorTest.java

Index: javatests/org/chromium/distiller/ContentExtractorTest.java
diff --git a/javatests/org/chromium/distiller/ContentExtractorTest.java b/javatests/org/chromium/distiller/ContentExtractorTest.java
index 0b03d6a9dd95b2177f5233aaa61e93beb6229ef7..c3e6bcdddbe9018588f7641e448375f6081faefc 100644
--- a/javatests/org/chromium/distiller/ContentExtractorTest.java
+++ b/javatests/org/chromium/distiller/ContentExtractorTest.java
@@ -86,14 +86,17 @@ public class ContentExtractorTest extends DomDistillerJsTestCase {
                 MARKUP_PARSER_TITLE, extractor.extractTitle());
     }
 
-    public void testImageWithSrcset() {
+    public void testImage() {
         // Test the absolute and different kinds of relative URLs for image sources,
         // and also add an extra comma (,) as malformed srcset syntax for robustness.
+        // Also test images in WebImage, WebText, and WebTable.

[wychen, 2015/12/09 05:17:40]

TODO(wychen): how do I get images inside a WebText?

[mdjones, 2015/12/09 17:40:16]

This is an unsolved problem. Since tables are currently treated as a singular
element, it can be argued that we simply ignore images contained in them, or at
least leave them unaltered. If we decide to start traversing into tables, we
would have to make a number of changes to our extraction logic and possibly eval
data. A mix of both is complicated; we can discuss this more offline.

         final String html =
             "<h1>" + CONTENT_TEXT + "</h1>" +
-            "<img src=\"image\" srcset=\"image200 200w, //example.org/image400 400w\">" +
+            "<img id=\"a\" style=\"a\" align=\"left\" src=\"image\" srcset=\"image200 200w, //example.org/image400 400w\">" +
+            "<img id=\"a\" style=\"a\" align=\"left\" src=\"image2\">" +
             "<table role=\"grid\"><tbody><tr><td>" +
-                "<img src=\"/image\" srcset=\"https://example.com/image2x 2x, /image4x 4x,\">" +
+                "<img id=\"a\" style=\"a\" align=\"left\" src=\"/image\" srcset=\"https://example.com/image2x 2x, /image4x 4x,\">" +
+                "<img id=\"a\" style=\"a\" align=\"left\" src=\"/image2\">" +
                 "</td></tr></tbody></table>" +
             "<p>" + CONTENT_TEXT + "</p>";
 
@@ -101,9 +104,11 @@ public class ContentExtractorTest extends DomDistillerJsTestCase {
             "<h1>" + CONTENT_TEXT + "</h1>" +
             "<img src=\"http://example.com/path/image\" " +
                  "srcset=\"http://example.com/path/image200 200w, http://example.org/image400 400w\">" +
+            "<img src=\"http://example.com/path/image2\">" +
             "<table role=\"grid\"><tbody><tr><td>" +
                 "<img src=\"http://example.com/image\" " +
                      "srcset=\"https://example.com/image2x 2x, http://example.com/image4x 4x, \">" +
+                "<img src=\"http://example.com/image2\">" +
             "</td></tr></tbody></table>" +
             "<p>" + CONTENT_TEXT + "</p>";
 
@@ -151,10 +156,13 @@ public class ContentExtractorTest extends DomDistillerJsTestCase {
             "<p style=\"\">" +
                 CONTENT_TEXT +
             "</p>" +
+            "<img style=\"align: left\" src=\"/test.png\">" +
             "<table style=\"position: absolute\">" +
                 "<tbody style=\"font-size: 2\">" +
                     "<tr style=\"z-index: 0\">" +
-                        "<th style=\"top: 0px\">" + CONTENT_TEXT + "</th>" +
+                        "<th style=\"top: 0px\">" + CONTENT_TEXT +
+                            "<img style=\"align: left\" src=\"/test.png\">" +
+                        "</th>" +
                         "<th style=\"width: 20px\">" + CONTENT_TEXT + "</th>" +
                     "</tr><tr style=\"left: 0\">" +
                         "<td style=\"display: block\">" + CONTENT_TEXT + "</td>" +
@@ -170,10 +178,13 @@ public class ContentExtractorTest extends DomDistillerJsTestCase {
             "<p>" +
                 CONTENT_TEXT +
             "</p>" +
+            "<img src=\"http://example.com/test.png\">" +
             "<table>" +
                 "<tbody>" +
                     "<tr>" +
-                        "<th>" + CONTENT_TEXT + "</th>" +
+                        "<th>" + CONTENT_TEXT +
+                            "<img src=\"http://example.com/test.png\">" +
+                        "</th>" +
                         "<th>" + CONTENT_TEXT + "</th>" +
                     "</tr><tr>" +
                         "<td>" + CONTENT_TEXT + "</td>" +
@@ -182,6 +193,7 @@ public class ContentExtractorTest extends DomDistillerJsTestCase {
                 "</tbody>" +
             "</table>";
 
+        mHead.setInnerHTML("<base href=\"http://example.com/\">");
         mBody.setInnerHTML(html);
 
         ContentExtractor extractor = new ContentExtractor(mRoot);