Support merged paragraphs when revisions are tracked

4 files changed, 96 insertions, 12 deletions

diff --git a/NEWS b/NEWS
index 5fb2f53..c4309ef 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,7 @@
+# 1.6.0
+
+* Support merged paragraphs when revisions are tracked.
+
 # 1.5.2
 
 * Add transformDocument to the TypeScript declarations.
diff --git a/lib/documents.js b/lib/documents.js
index edc9b7c..d29ad1b 100644
--- a/lib/documents.js
+++ b/lib/documents.js
@@ -203,7 +203,7 @@ function BookmarkStart(options) {
 exports.document = exports.Document = Document;
 exports.paragraph = exports.Paragraph = Paragraph;
 exports.run = exports.Run = Run;
-exports.Text = Text;
+exports.text = exports.Text = Text;
 exports.tab = exports.Tab = Tab;
 exports.Hyperlink = Hyperlink;
 exports.noteReference = exports.NoteReference = NoteReference;
diff --git a/lib/docx/body-reader.js b/lib/docx/body-reader.js
index 61b198b..ca55bb7 100644
--- a/lib/docx/body-reader.js
+++ b/lib/docx/body-reader.js
@@ -23,6 +23,12 @@ function createBodyReader(options) {
 function BodyReader(options) {
     var complexFieldStack = [];
     var currentInstrText = [];
+
+    // When a paragraph is marked as deleted, its contents should be combined
+    // with the following paragraph. See 17.13.5.15 del (Deleted Paragraph) of
+    // ECMA-376 4th edition Part 1.
+    var deletedParagraphContents = [];
+
     var relationships = options.relationships;
     var contentTypes = options.contentTypes;
     var docxFile = options.docxFile;
@@ -213,15 +219,34 @@ function BodyReader(options) {
 
     var xmlElementReaders = {
         "w:p": function(element) {
-            return readXmlElements(element.children)
-                .map(function(children) {
-                    var properties = _.find(children, isParagraphProperties);
-                    return new documents.Paragraph(
-                        children.filter(negate(isParagraphProperties)),
-                        properties
-                    );
-                })
-                .insertExtra();
+            var isDeleted = !!element
+                .firstOrEmpty("w:pPr")
+                .firstOrEmpty("w:rPr")
+                .first("w:del");
+
+            if (isDeleted) {
+                element.children.filter(function(child) {
+                    return child.name !== "w:pPr";
+                }).forEach(function(child) {
+                    deletedParagraphContents.push(child);
+                });
+                return emptyResult();
+            } else {
+                var childrenXml = element.children;
+                if (deletedParagraphContents.length > 0) {
+                    childrenXml = deletedParagraphContents.concat(childrenXml);
+                    deletedParagraphContents = [];
+                }
+                return readXmlElements(childrenXml)
+                    .map(function(children) {
+                        var properties = _.find(children, isParagraphProperties);
+                        return new documents.Paragraph(
+                            children.filter(negate(isParagraphProperties)),
+                            properties
+                        );
+                    })
+                    .insertExtra();
+            }
         },
         "w:pPr": function(element) {
             return readParagraphStyle(element).map(function(style) {
diff --git a/test/docx/body-reader.tests.js b/test/docx/body-reader.tests.js
index f5c8b87..bbc3499 100644
--- a/test/docx/body-reader.tests.js
+++ b/test/docx/body-reader.tests.js
@@ -33,11 +33,19 @@ var testing = require("../testing");
 var test = require("../test")(module);
 var createFakeDocxFile = testing.createFakeDocxFile;
 
-function readXmlElement(element, options) {
+function createBodyReaderForTests(options) {
     options = Object.create(options || {});
     options.styles = options.styles || new Styles({}, {});
     options.numbering = options.numbering || defaultNumbering;
-    return createBodyReader(options).readXmlElement(element);
+    return createBodyReader(options);
+}
+
+function readXmlElement(element, options) {
+    return createBodyReaderForTests(options).readXmlElement(element);
+}
+
+function readXmlElements(element, options) {
+    return createBodyReaderForTests(options).readXmlElements(element);
 }
 
 function readXmlElementValue(element, options) {
@@ -46,6 +54,12 @@ function readXmlElementValue(element, options) {
     return result.value;
 }
 
+function readXmlElementsValue(elements, options) {
+    var result = readXmlElements(elements, options);
+    assert.deepEqual(result.messages, []);
+    return result.value;
+}
+
 var fakeContentTypes = {
     findContentType: function(filePath) {
         var extensionTypes = {
@@ -225,6 +239,47 @@ test("numbering properties are ignored if w:numId is missing", function() {
     assert.equal(numberingLevel, null);
 });
 
+test("content of deleted paragraph is prepended to next paragraph", function() {
+    var styles = new Styles(
+        {
+            "Heading1": {name: "Heading 1"},
+            "Heading2": {name: "Heading 2"}
+        },
+        {}
+    );
+    var bodyXml = [
+        new XmlElement("w:p", {}, [
+            new XmlElement("w:pPr", {}, [
+                new XmlElement("w:pStyle", {"w:val": "Heading1"}, []),
+                new XmlElement("w:rPr", {}, [
+                    new XmlElement("w:del")
+                ])
+            ]),
+            runOfText("One")
+        ]),
+        new XmlElement("w:p", {}, [
+            new XmlElement("w:pPr", {}, [
+                new XmlElement("w:pStyle", {"w:val": "Heading2"}, [])
+            ]),
+            runOfText("Two")
+        ])
+    ];
+
+    var result = readXmlElementsValue(bodyXml, {styles: styles});
+
+    assertThat(result, contains(
+        hasProperties({
+            type: documents.types.paragraph,
+            styleId: "Heading2",
+            children: contains(
+                documents.run([documents.text("One")]),
+                documents.run([documents.text("Two")])
+            )
+        })
+    ));
+
+});
+
 test("complex fields", (function() {
     var uri = "http://example.com";
     var beginXml = new XmlElement("w:r", {}, [