emulate Word's treatment of xml:space="preserve"

(when converting DOCX content to Unicode strings)

Author: Lowell Stewart

OpenXmlPowerTools.Tests/UnicodeMapperTests.cs

@@ -142,6 +142,40 @@ public void CanStringifySymbols()
             Assert.Equal(symFromChar1.ToString(SaveOptions.None), symFromChar2.ToString(SaveOptions.None));
             Assert.Equal(symFromChar1.ToString(SaveOptions.None), symFromChar3.ToString(SaveOptions.None));
         }
+
+        [Fact]
+        public void HonorsXmlSpace()
+        {
+            XDocument partDocument = XDocument.Parse(PreserveSpacingXmlString);
+            XElement p = partDocument.Descendants(W.p).Last();
+            string innerText = p.Descendants(W.r)
+                .Select(UnicodeMapper.RunToString)
+                .StringConcatenate();
+            Assert.Equal(@"The following space is retained: but this one is not:. Similarly these two lines should have only a space between them: Line 1! Line 2!", innerText);
+        }
+
+        private const string PreserveSpacingXmlString =
+@"<w:document xmlns:w=""http://schemas.openxmlformats.org/wordprocessingml/2006/main"">
+  <w:body>
+    <w:p>
+      <w:r>
+        <w:t xml:space=""preserve"">The following space is retained: </w:t>
+      </w:r>
+      <w:r>
+        <w:t>but this one is not: </w:t>
+      </w:r>
+      <w:r>
+        <w:t xml:space=""preserve"">. Similarly these two lines should have only a space between them: </w:t>
+      </w:r>
+      <w:r>
+        <w:t>
+          Line 1!
+Line 2!
+        </w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>";
     }
 }
 

OpenXmlPowerTools/UnicodeMapper.cs

@@ -69,7 +69,12 @@ public static string RunToString(XElement element)
 
             // For w:t elements, we obviously want the element's value.
             if (element.Name == W.t)
-                return (string) element;
+            {
+                // Emulate Word's handling of the xml:space attribute on text elements
+                XAttribute spaceAttribute = element.Attribute(XNamespace.Xml + "space");
+                string space = spaceAttribute != null ? spaceAttribute.Value : null;
+                return space == "preserve" ? (string)element : IgnoreTextSpacing((string)element);
+            }
 
             // Turn elements representing special characters into their corresponding
             // unicode characters.
@@ -125,6 +130,22 @@ public static string RunToString(XElement element)
             return StartOfHeading.ToString();
         }
 
+        /// <summary>
+        /// Emulate the way Word treats text elements when attribute xml:space="preserve"
+        /// is NOT present.
+        /// </summary>
+        /// <param name="text">The entire content of the w:t element.</param>
+        /// <returns>The corresponding text string Word would display, print, and
+        /// allow to be edited.</returns>
+        private static string IgnoreTextSpacing(string text)
+        {
+            // all whitespace at beginning and end of entire string is ignored
+            // if text contains line breaks, they are ignored/replaced with a single space
+            return string.Join(" ",
+                text.Split(new char[] { '\n', '\r' }, StringSplitOptions.RemoveEmptyEntries)
+            ).Trim();
+        }
+
         /// <summary>
         /// Translate a symbol into a Unicode character, using the specified w:font attribute
         /// value and unicode value (represented by the w:sym element's w:char attribute),