Include embedded style map when explicit style map is set

6 files changed, 119 insertions, 88 deletions

diff --git a/NEWS b/NEWS
index 1f5db7e..facee0e 100644
--- a/NEWS
+++ b/NEWS
@@ -10,6 +10,11 @@
 
 * Allow a separator to be specified when HTML elements are collapsed.
 
+* Add includeEmbeddedStyleMap option to allow embedded style maps to be
+  disabled.
+
+* Include embedded styles when explicit style map is passed.
+
 # 1.2.5
 
 * Ignore bold, italic, underline and strikethrough elements that have a value of
diff --git a/README.md b/README.md
index fa4cbd6..027c494 100644
--- a/README.md
+++ b/README.md
@@ -345,6 +345,11 @@ Converts the source document to HTML.
      each element is expected to be a string representing a single style mapping.
      See "Writing style maps" for a reference to the syntax for style maps.
 
+  * `includeEmbeddedStyleMap`: by default,
+     if the document contains an embedded style, then it is combined with the default style map.
+     To ignore any embedded style maps,
+     set `options.includeEmbeddedStyleMap` to `false`.
+
   * `includeDefaultStyleMap`: by default,
      the style map passed in `styleMap` is combined with the default style map.
      To stop using the default style map altogether,
@@ -407,7 +412,7 @@ Each paragraph is followed by two newlines.
 Given an existing docx file,
 `embedStyleMap` will generate a new docx file with the passed style map embedded.
 When the new docx file is read by Mammoth,
-it will use the embedded style if no explicit style map is provided.
+it will use the embedded style map.
 
 * `input`: an object describing the source document.
   On node.js, the following inputs are supported:
diff --git a/lib/index.js b/lib/index.js
index ae95371..212b0b7 100644
--- a/lib/index.js
+++ b/lib/index.js
@@ -1,3 +1,5 @@
+var _ = require("underscore");
+
 var docxReader = require("./docx/docx-reader");
 var docxStyleMap = require("./docx/style-map");
 var DocumentConverter = require("./document-to-html").DocumentConverter;
@@ -27,24 +29,21 @@ function convertToMarkdown(input, options) {
 }
 
 function convert(input, options) {
-    options = Object.create(options || {});
+    options = readOptions(options);
     
     return unzip.openZip(input)
         .tap(function(docxFile) {
-            if (!options.styleMap) {
-                return docxStyleMap.readStyleMap(docxFile).then(function(styleMap) {
-                    options.styleMap = styleMap || "";
-                });
-            }
+            return docxStyleMap.readStyleMap(docxFile).then(function(styleMap) {
+                options.embeddedStyleMap = styleMap;
+            });
         })
         .then(function(docxFile) {
-            var fullOptions = readOptions(options);
             return docxReader.read(docxFile, input)
                 .then(function(documentResult) {
-                    return documentResult.map(fullOptions.transformDocument);
+                    return documentResult.map(options.transformDocument);
                 })
                 .then(function(documentResult) {
-                    return convertDocumentToHtml(documentResult, fullOptions);
+                    return convertDocumentToHtml(documentResult, options);
                 });
         });
 }
@@ -55,9 +54,10 @@ function readEmbeddedStyleMap(input) {
 }
 
 function convertDocumentToHtml(documentResult, options) {
-    var parsedOptions = Object.create(options);
-    var styleMapResult = parseStyleMap(options.styleMap);
-    parsedOptions.styleMap = styleMapResult.value;
+    var styleMapResult = parseStyleMap(options.readStyleMap());
+    var parsedOptions = _.extend({}, options, {
+        styleMap: styleMapResult.value
+    });
     var documentConverter = new DocumentConverter(parsedOptions);
     
     return documentResult.flatMapThen(function(document) {
diff --git a/lib/options-reader.js b/lib/options-reader.js
index 4a7db9d..e7e94fd 100644
--- a/lib/options-reader.js
+++ b/lib/options-reader.js
@@ -3,79 +3,81 @@ exports.readOptions = readOptions;
 
 var _ = require("underscore");
 
+var defaultStyleMap = exports._defaultStyleMap = [
+    "p.Heading1 => h1:fresh",
+    "p.Heading2 => h2:fresh",
+    "p.Heading3 => h3:fresh",
+    "p.Heading4 => h4:fresh",
+    "p.Heading5 => h5:fresh",
+    "p.Heading6 => h6:fresh",
+    "p[style-name='Heading 1'] => h1:fresh",
+    "p[style-name='Heading 2'] => h2:fresh",
+    "p[style-name='Heading 3'] => h3:fresh",
+    "p[style-name='Heading 4'] => h4:fresh",
+    "p[style-name='Heading 5'] => h5:fresh",
+    "p[style-name='Heading 6'] => h6:fresh",
+    "p[style-name='heading 1'] => h1:fresh",
+    "p[style-name='heading 2'] => h2:fresh",
+    "p[style-name='heading 3'] => h3:fresh",
+    "p[style-name='heading 4'] => h4:fresh",
+    "p[style-name='heading 5'] => h5:fresh",
+    "p[style-name='heading 6'] => h6:fresh",
+    
+    "r[style-name='Strong'] => strong",
+    
+    "p[style-name='footnote text'] => p",
+    "r[style-name='footnote reference'] =>",
+    "p[style-name='endnote text'] => p",
+    "r[style-name='endnote reference'] =>",
+    "p[style-name='annotation text'] => p",
+    "r[style-name='annotation reference'] =>",
+    
+    // LibreOffice
+    "p[style-name='Footnote'] => p",
+    "r[style-name='Footnote anchor'] =>",
+    "p[style-name='Endnote'] => p",
+    "r[style-name='Endnote anchor'] =>",
+    
+    "p:unordered-list(1) => ul > li:fresh",
+    "p:unordered-list(2) => ul|ol > li > ul > li:fresh",
+    "p:unordered-list(3) => ul|ol > li > ul|ol > li > ul > li:fresh",
+    "p:unordered-list(4) => ul|ol > li > ul|ol > li > ul|ol > li > ul > li:fresh",
+    "p:unordered-list(5) => ul|ol > li > ul|ol > li > ul|ol > li > ul|ol > li > ul > li:fresh",
+    "p:ordered-list(1) => ol > li:fresh",
+    "p:ordered-list(2) => ul|ol > li > ol > li:fresh",
+    "p:ordered-list(3) => ul|ol > li > ul|ol > li > ol > li:fresh",
+    "p:ordered-list(4) => ul|ol > li > ul|ol > li > ul|ol > li > ol > li:fresh",
+    "p:ordered-list(5) => ul|ol > li > ul|ol > li > ul|ol > li > ul|ol > li > ol > li:fresh",
+    
+    "r[style-name='Hyperlink'] =>",
+    
+    "p[style-name='Normal'] => p:fresh"
+];
 
 var standardOptions = exports._standardOptions = {
-    styleMap: [
-        "p.Heading1 => h1:fresh",
-        "p.Heading2 => h2:fresh",
-        "p.Heading3 => h3:fresh",
-        "p.Heading4 => h4:fresh",
-        "p.Heading5 => h5:fresh",
-        "p.Heading6 => h6:fresh",
-        "p[style-name='Heading 1'] => h1:fresh",
-        "p[style-name='Heading 2'] => h2:fresh",
-        "p[style-name='Heading 3'] => h3:fresh",
-        "p[style-name='Heading 4'] => h4:fresh",
-        "p[style-name='Heading 5'] => h5:fresh",
-        "p[style-name='Heading 6'] => h6:fresh",
-        "p[style-name='heading 1'] => h1:fresh",
-        "p[style-name='heading 2'] => h2:fresh",
-        "p[style-name='heading 3'] => h3:fresh",
-        "p[style-name='heading 4'] => h4:fresh",
-        "p[style-name='heading 5'] => h5:fresh",
-        "p[style-name='heading 6'] => h6:fresh",
-        
-        "r[style-name='Strong'] => strong",
-        
-        "p[style-name='footnote text'] => p",
-        "r[style-name='footnote reference'] =>",
-        "p[style-name='endnote text'] => p",
-        "r[style-name='endnote reference'] =>",
-        "p[style-name='annotation text'] => p",
-        "r[style-name='annotation reference'] =>",
-        
-        // LibreOffice
-        "p[style-name='Footnote'] => p",
-        "r[style-name='Footnote anchor'] =>",
-        "p[style-name='Endnote'] => p",
-        "r[style-name='Endnote anchor'] =>",
-        
-        "p:unordered-list(1) => ul > li:fresh",
-        "p:unordered-list(2) => ul|ol > li > ul > li:fresh",
-        "p:unordered-list(3) => ul|ol > li > ul|ol > li > ul > li:fresh",
-        "p:unordered-list(4) => ul|ol > li > ul|ol > li > ul|ol > li > ul > li:fresh",
-        "p:unordered-list(5) => ul|ol > li > ul|ol > li > ul|ol > li > ul|ol > li > ul > li:fresh",
-        "p:ordered-list(1) => ol > li:fresh",
-        "p:ordered-list(2) => ul|ol > li > ol > li:fresh",
-        "p:ordered-list(3) => ul|ol > li > ul|ol > li > ol > li:fresh",
-        "p:ordered-list(4) => ul|ol > li > ul|ol > li > ul|ol > li > ol > li:fresh",
-        "p:ordered-list(5) => ul|ol > li > ul|ol > li > ul|ol > li > ul|ol > li > ol > li:fresh",
-        
-        "r[style-name='Hyperlink'] =>",
-        
-        "p[style-name='Normal'] => p:fresh"
-    
-    ],
     transformDocument: identity,
-    includeDefaultStyleMap: true
+    includeDefaultStyleMap: true,
+    includeEmbeddedStyleMap: true
 };
 
 function readOptions(options) {
     options = options || {};
-    var fullOptions = {};
-    _.extend(fullOptions, standardOptions, options);
-    
-    fullOptions.styleMap = readStyleMap(options.styleMap, fullOptions);
-    
-    return fullOptions;
+    return _.extend({}, standardOptions, options, {
+        customStyleMap: readStyleMap(options.styleMap),
+        readStyleMap: function() {
+            var styleMap = this.customStyleMap;
+            if (this.includeEmbeddedStyleMap) {
+                styleMap = styleMap.concat(readStyleMap(this.embeddedStyleMap));
+            }
+            if (this.includeDefaultStyleMap) {
+                styleMap = styleMap.concat(defaultStyleMap);
+            }
+            return styleMap;
+        }
+    });
 }
 
-function readStyleMap(styleMap, options) {
-    var customStyleMap = readCustomStyleMap(styleMap);
-    return customStyleMap.concat(options.includeDefaultStyleMap ? standardOptions.styleMap : []);
-}
-
-function readCustomStyleMap(styleMap) {
+function readStyleMap(styleMap) {
     if (!styleMap) {
         return [];
     } else if (_.isString(styleMap)) {
@@ -91,7 +93,6 @@ function readCustomStyleMap(styleMap) {
     }
 }
 
-
 function identity(value) {
     return value;
 }
diff --git a/test/mammoth.tests.js b/test/mammoth.tests.js
index db2b338..19e0c47 100644
--- a/test/mammoth.tests.js
+++ b/test/mammoth.tests.js
@@ -100,6 +100,17 @@ test('explicit style map takes precedence over embedded style map', function() {
     });
 });
 
+test('embedded style maps can be disabled', function() {
+    var docxPath = path.join(__dirname, "test-data/embedded-style-map.docx");
+    var options = {
+        includeEmbeddedStyleMap: false
+    };
+    return mammoth.convertToHtml({path: docxPath}, options).then(function(result) {
+        assert.equal(result.value, "<p>Walking on imported air</p>");
+        assert.deepEqual(result.messages, []);
+    });
+});
+
 test('embedded style map can be written and then read', function() {
     var docxPath = path.join(__dirname, "test-data/single-paragraph.docx");
     return promises.nfcall(fs.readFile, docxPath)
diff --git a/test/options-reader.tests.js b/test/options-reader.tests.js
index b0f202b..bad4c56 100644
--- a/test/options-reader.tests.js
+++ b/test/options-reader.tests.js
@@ -1,5 +1,7 @@
 var assert = require("assert");
 
+var _ = require("underscore");
+
 var optionsReader = require("../lib/options-reader");
 var standardOptions = optionsReader._standardOptions;
 var readOptions = optionsReader.readOptions;
@@ -7,42 +9,49 @@ var test = require("./test")(module);
 
 
 test('standard options are used if options is undefined', function() {
-    assert.deepEqual(standardOptions, readOptions(undefined));
+    var options = readOptions(undefined);
+    assert.deepEqual(standardOptions, _.omit(options, "customStyleMap", "readStyleMap"));
+    assert.deepEqual(options.customStyleMap, []);
 });
 
 test('standard options are used if options is empty', function() {
-    assert.deepEqual(standardOptions, readOptions({}));
+    var options = readOptions({});
+    assert.deepEqual(standardOptions, _.omit(options, "customStyleMap", "readStyleMap"));
+    assert.deepEqual(options.customStyleMap, []);
 });
 
 test('custom style map as string is prepended to standard style map', function() {
     var options = readOptions({
         styleMap: "p.SectionTitle => h2"
     });
-    assert.deepEqual("p.SectionTitle => h2", options.styleMap[0]);
-    assert.deepEqual(standardOptions.styleMap, options.styleMap.slice(1));
+    var styleMap = options.readStyleMap();
+    assert.deepEqual("p.SectionTitle => h2", styleMap[0]);
+    assert.deepEqual(optionsReader._defaultStyleMap, styleMap.slice(1));
 });
 
 test('custom style map as array is prepended to standard style map', function() {
     var options = readOptions({
         styleMap: ["p.SectionTitle => h2"]
     });
-    assert.deepEqual("p.SectionTitle => h2", options.styleMap[0]);
-    assert.deepEqual(standardOptions.styleMap, options.styleMap.slice(1));
+    var styleMap = options.readStyleMap();
+    assert.deepEqual("p.SectionTitle => h2", styleMap[0]);
+    assert.deepEqual(optionsReader._defaultStyleMap, styleMap.slice(1));
 });
 
 test('lines starting with # in custom style map are ignored', function() {
     var options = readOptions({
         styleMap: "# p.SectionTitle => h3\np.SectionTitle => h2"
     });
-    assert.deepEqual("p.SectionTitle => h2", options.styleMap[0]);
-    assert.deepEqual(standardOptions.styleMap, options.styleMap.slice(1));
+    var styleMap = options.readStyleMap();
+    assert.deepEqual("p.SectionTitle => h2", styleMap[0]);
+    assert.deepEqual(optionsReader._defaultStyleMap, styleMap.slice(1));
 });
 
 test('blank lines in custom style map are ignored', function() {
     var options = readOptions({
         styleMap: "\n\n"
     });
-    assert.deepEqual(standardOptions.styleMap, options.styleMap);
+    assert.deepEqual(optionsReader._defaultStyleMap, options.readStyleMap());
 });
 
 test('default style mappings are ignored if includeDefaultStyleMap is false', function() {
@@ -50,5 +59,5 @@ test('default style mappings are ignored if includeDefaultStyleMap is false', fu
         styleMap: "p.SectionTitle => h2",
         includeDefaultStyleMap: false
     });
-    assert.deepEqual(["p.SectionTitle => h2"], options.styleMap);
+    assert.deepEqual(["p.SectionTitle => h2"], options.readStyleMap());
 });