Add safer method for text extraction from elements

2 files changed, 10 insertions, 1 deletions

diff --git a/lib/docx/document-xml-reader.js b/lib/docx/document-xml-reader.js
index a23aa5f..15a2dae 100644
--- a/lib/docx/document-xml-reader.js
+++ b/lib/docx/document-xml-reader.js
@@ -104,7 +104,7 @@ function DocumentXmlReader(relationships, contentTypes, docxFile, numbering) {
             return new Result(properties);
         },
         "w:t": function(element) {
-            return new Result(new documents.Text(element.children[0].value));
+            return new Result(new documents.Text(element.text()));
         },
         "w:tab": function(element) {
             return new Result(new documents.Tab());
diff --git a/lib/xmlreader.js b/lib/xmlreader.js
index 51f24a7..f219d2b 100644
--- a/lib/xmlreader.js
+++ b/lib/xmlreader.js
@@ -105,6 +105,15 @@ Element.prototype.getElementsByTagName = function(name) {
     return toElementList(elements);
 };
 
+Element.prototype.text = function() {
+    if (this.children.length === 0) {
+        return "";
+    } else if (this.children.length !== 1 || this.children[0].type !== "text") {
+        throw new Error("Not implemented");
+    }
+    return this.children[0].value;
+};
+
 var elementListPrototype = {
     getElementsByTagName: function(name) {
         return toElementList(_.flatten(this.map(function(element) {