Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(486)

Issues Search
    My Issues | Starred     Open | Closed | All

Unified Diff: javatests/org/chromium/distiller/ContentExtractorTest.java

Issue 1230583006: Fix for keeping lists structure (Closed) Base URL: https://github.com/chromium/dom-distiller.git@master
Patch Set: Small code refactor for more appropriate names. Created 5 years, 5 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« java/org/chromium/distiller/webdocument/filters/RelevantElements.java ('K') | « java/org/chromium/distiller/webdocument/filters/RelevantElements.java ('k') | javatests/org/chromium/distiller/TestUtil.java » ('j') | javatests/org/chromium/distiller/webdocument/filters/RelevantElementsTest.java » ('J')
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: javatests/org/chromium/distiller/ContentExtractorTest.java
diff --git a/javatests/org/chromium/distiller/ContentExtractorTest.java b/javatests/org/chromium/distiller/ContentExtractorTest.java
index 63d349c923a4c341670a5bd85a4632cab8903c8f..5b5154da9071ba818d2dd70e0526f0f0c53a0ed2 100644
--- a/javatests/org/chromium/distiller/ContentExtractorTest.java
+++ b/javatests/org/chromium/distiller/ContentExtractorTest.java
@@ -143,6 +143,206 @@ public class ContentExtractorTest extends DomDistillerJsTestCase {
TestUtil.removeAllDirAttributes(extractedContent));
}
+ public void testPreserveOrderedList() {
+ Element outerListTag = Document.get().createElement("OL");
+ mBody.appendChild(outerListTag);
+
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ol>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ol>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
+ public void testPreserveNestedOrderedList() {
+ Element outerListTag = Document.get().createElement("OL");
+ Element outerListItem = Document.get().createElement("LI");
+
+ Element innerListTag = Document.get().createElement("OL");
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+
+ outerListItem.appendChild(innerListTag);
+ outerListTag.appendChild(outerListItem);
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+
+ mBody.appendChild(outerListTag);
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ol>" +
+ "<li>" +
+ "<ol>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
wychen 2015/08/01 01:00:20 nitpick: nested html would be more readable if ind
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ol>" +
+ "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ol>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
+ public void testPreserveNestedOrderedListWithOtherElementsInside() {
+ Element outerListTag = Document.get().createElement("OL");
+ Element outerListItem = Document.get().createElement("LI");
+ outerListItem.appendChild(TestUtil.createText(CONTENT_TEXT));
+ outerListItem.appendChild(TestUtil.createParagraph(CONTENT_TEXT));
+
+ Element innerListTag = Document.get().createElement("OL");
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createParagraph(""));
+
+ outerListItem.appendChild(innerListTag);
+ outerListTag.appendChild(outerListItem);
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createParagraph(CONTENT_TEXT));
+
+ mBody.appendChild(outerListTag);
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ol>" +
+ "<li>" + CONTENT_TEXT +
+ "<p>" + CONTENT_TEXT + "</p>" +
+ "<ol>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ol>" +
+ "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<p>" + CONTENT_TEXT + "</p>" +
+ "</ol>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
+ public void testPreserveUnorderedList() {
+ Element outerListTag = Document.get().createElement("UL");
+ mBody.appendChild(outerListTag);
+
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ul>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ul>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
+ public void testPreserveNestedUnorderedList() {
wychen 2015/08/01 01:00:20 We can also create a test with malformed html, by
+ Element outerListTag = Document.get().createElement("UL");
+ Element outerListItem = Document.get().createElement("LI");
+
+ Element innerListTag = Document.get().createElement("UL");
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+
+ outerListItem.appendChild(innerListTag);
+ outerListTag.appendChild(outerListItem);
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+
+ mBody.appendChild(outerListTag);
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ul>" +
+ "<li>" +
+ "<ul>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ul>" +
+ "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ul>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
+ public void testPreserveNestedUnorderedListWithOtherElementsInside() {
+ Element outerListTag = Document.get().createElement("UL");
+ Element outerListItem = Document.get().createElement("LI");
+ outerListItem.appendChild(TestUtil.createText(CONTENT_TEXT));
+ outerListItem.appendChild(TestUtil.createParagraph(CONTENT_TEXT));
+
+ Element innerListTag = Document.get().createElement("UL");
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ innerListTag.appendChild(TestUtil.createParagraph(""));
+
+ outerListItem.appendChild(innerListTag);
+ outerListTag.appendChild(outerListItem);
+ outerListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ outerListTag.appendChild(TestUtil.createParagraph(CONTENT_TEXT));
+
+ mBody.appendChild(outerListTag);
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ul>" +
+ "<li>" + CONTENT_TEXT +
+ "<p>" + CONTENT_TEXT + "</p>" +
+ "<ul>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ul>" +
+ "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<p>" + CONTENT_TEXT + "</p>" +
+ "</ul>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
+ public void testPreserveUnorderedListWithNestedOrderedList() {
+ Element unorderedListTag = Document.get().createElement("UL");
+ Element li = Document.get().createElement("LI");
+ Element orderedList = Document.get().createElement("OL");
+ orderedList.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ orderedList.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ li.appendChild(orderedList);
+ unorderedListTag.appendChild(li);
+ unorderedListTag.appendChild(TestUtil.createListItem(CONTENT_TEXT));
+ mBody.appendChild(unorderedListTag);
+ ContentExtractor extractor = new ContentExtractor(mRoot);
+ String extractedContent = extractor.extractContent();
+ assertEquals("<ul>" +
+ "<li>" +
+ "<ol>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ol>" +
+ "</li>" +
+ "<li>" + CONTENT_TEXT + "</li>" +
+ "</ul>",
+ TestUtil.removeAllDirAttributes(extractedContent));
+ }
+
private void assertExtractor(String expected, String html) {
mBody.setInnerHTML("");
Element div = TestUtil.createDiv(0);
« java/org/chromium/distiller/webdocument/filters/RelevantElements.java ('K') | « java/org/chromium/distiller/webdocument/filters/RelevantElements.java ('k') | javatests/org/chromium/distiller/TestUtil.java » ('j') | javatests/org/chromium/distiller/webdocument/filters/RelevantElementsTest.java » ('J')

Powered by Google App Engine
This is Rietveld 408576698