SAS7BDAT html tests

MikeThomsen · May 3, 2018 · 90720ae · 90720ae
1 parent 79f313d
commit 90720ae
Show file tree

Hide file tree

Showing 2 changed files with 21 additions and 2 deletions.
diff --git a/tika-parsers/src/main/java/org/apache/tika/parser/sas/SAS7BDATParser.java b/tika-parsers/src/main/java/org/apache/tika/parser/sas/SAS7BDATParser.java
@@ -119,8 +119,11 @@ public void parse(InputStream stream, ContentHandler handler,
         // Do the column headings
         xhtml.startElement("tr");
         for (Column c : sas.getColumns()) {
+            String label = c.getLabel();
+            if (label == null || label.isEmpty()) label = c.getName();
+
             xhtml.startElement("th", "title", c.getName());
-            xhtml.characters(c.getLabel());
+            xhtml.characters(label);
             xhtml.endElement("th");
         }
         xhtml.endElement("tr");

diff --git a/tika-parsers/src/test/java/org/apache/tika/parser/sas/SAS7BDATParserTest.java b/tika-parsers/src/test/java/org/apache/tika/parser/sas/SAS7BDATParserTest.java
@@ -118,6 +118,22 @@ public void testMultiColumns() throws Exception {
         assertContains("\t08Feb1904\t", content);
     }
 
-    // TODO HTML contents unit test
+    @Test
+    public void testHTML() throws Exception {
+        XMLResult result = getXML("testSAS.sas7bdat");
+        String xml = result.xml;
+
+        // Check the title came through
+        assertContains("<h1>TESTING</h1>", xml);
+        // Check the headings
+        assertContains("<th title=\"recnum\">recnum</th>", xml);
+        assertContains("<th title=\"label\">label</th>", xml);
+        // Check some rows
+        assertContains("<td>3</td>", xml);
+        assertContains("<td>This is row", xml);
+        assertContains("10</td>", xml);
+    }
+
     // TODO Column names vs labels, with a different test file
+    // TODO Columnar consistency test
 }