Merge pull request #105 from tawk/fix-malformed-td

Treat non-numeric colspans as zero and handle them gracefully
html-to-text · Aug 17, 2016 · b73cc33 · b73cc33
2 parents 15ed4c9 + 64ae3e7
commit b73cc33
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 1 deletion.
diff --git a/lib/formatter.js b/lib/formatter.js
@@ -196,7 +196,7 @@ function formatTable(elem, fn, options) {
 								rows.push(_.compact(tokens));
 								// Fill colspans with empty values
 								if (elem.attribs && elem.attribs.colspan) {
-									times = elem.attribs.colspan - 1;
+									times = elem.attribs.colspan - 1 || 0;
 									_.times(times, function() {
 										rows.push(['']);
 									});

diff --git a/test/html-to-text.js b/test/html-to-text.js
@@ -168,6 +168,23 @@ describe('html-to-text', function() {
       var result = htmlToText.fromString(html, { tables: true });
       expect(result).to.equal(resultExpected);
     });
+
+    it('does handle non-integer colspan on td element gracefully', function () {
+      var html = 'Good morning Jacob, \
+        <TABLE> \
+        <CENTER> \
+        <TBODY> \
+        <TR> \
+        <TD colspan="abc">Lorem ipsum dolor sit amet.</TD> \
+        </TR> \
+        </CENTER> \
+        </TBODY> \
+        </TABLE> \
+      ';
+      var resultExpected = 'Good morning Jacob, Lorem ipsum dolor sit amet.';
+      var result = htmlToText.fromString(html, { tables: true });
+      expect(result).to.equal(resultExpected);
+    });
   });
 
   describe('lists', function() {