Merge pull request #7475 from Snuffleupagus/api-getTextContent-combineTextItems

[api-minor] Add a parameter to `PDFPageProxy_getTextContent` that controls whether `PartialEvaluator_getTextContent` will attempt to combine same line text items

src/core/document.js

@@ -265,7 +265,8 @@ var Page = (function PageClosure() {
     },
 
     extractTextContent: function Page_extractTextContent(task,
-                                                         normalizeWhitespace) {
+                                                         normalizeWhitespace,
+                                                         combineTextItems) {
       var handler = {
         on: function nullHandlerOn() {},
         send: function nullHandlerSend() {}
@@ -298,7 +299,8 @@ var Page = (function PageClosure() {
                                                task,
                                                self.resources,
                                                /* stateManager = */ null,
-                                               normalizeWhitespace);
+                                               normalizeWhitespace,
+                                               combineTextItems);
       });
     },
 

src/core/evaluator.js

@@ -1132,7 +1132,8 @@ var PartialEvaluator = (function PartialEvaluatorClosure() {
     getTextContent:
         function PartialEvaluator_getTextContent(stream, task, resources,
                                                  stateManager,
-                                                 normalizeWhitespace) {
+                                                 normalizeWhitespace,
+                                                 combineTextItems) {
 
       stateManager = (stateManager || new StateManager(new TextState()));
 
@@ -1443,7 +1444,8 @@ var PartialEvaluator = (function PartialEvaluatorClosure() {
               var isSameTextLine = !textState.font ? false :
                 ((textState.font.vertical ? args[0] : args[1]) === 0);
               advance = args[0] - args[1];
-              if (isSameTextLine && textContentItem.initialized &&
+              if (combineTextItems &&
+                  isSameTextLine && textContentItem.initialized &&
                   advance > 0 &&
                   advance <= textContentItem.fakeMultiSpaceMax) {
                 textState.translateTextLineMatrix(args[0], args[1]);
@@ -1475,7 +1477,8 @@ var PartialEvaluator = (function PartialEvaluatorClosure() {
               // Optimization to treat same line movement as advance.
               advance = textState.calcTextLineMatrixAdvance(
                 args[0], args[1], args[2], args[3], args[4], args[5]);
-              if (advance !== null && textContentItem.initialized &&
+              if (combineTextItems &&
+                  advance !== null && textContentItem.initialized &&
                   advance.value > 0 &&
                   advance.value <= textContentItem.fakeMultiSpaceMax) {
                 textState.translateTextLineMatrix(advance.width,
@@ -1616,7 +1619,8 @@ var PartialEvaluator = (function PartialEvaluatorClosure() {
 
               next(self.getTextContent(xobj, task,
                    xobj.dict.get('Resources') || resources, stateManager,
-                   normalizeWhitespace).then(function (formTextContent) {
+                   normalizeWhitespace, combineTextItems).then(
+                function (formTextContent) {
                   Util.appendToArray(textContent.items, formTextContent.items);
                   Util.extendObj(textContent.styles, formTextContent.styles);
                   stateManager.restore();

src/core/worker.js

@@ -891,12 +891,14 @@ var WorkerMessageHandler = {
     handler.on('GetTextContent', function wphExtractText(data) {
       var pageIndex = data.pageIndex;
       var normalizeWhitespace = data.normalizeWhitespace;
+      var combineTextItems = data.combineTextItems;
       return pdfManager.getPage(pageIndex).then(function(page) {
         var task = new WorkerTask('GetTextContent: page ' + pageIndex);
         startWorkerTask(task);
         var pageNum = pageIndex + 1;
         var start = Date.now();
-        return page.extractTextContent(task, normalizeWhitespace).then(
+        return page.extractTextContent(task, normalizeWhitespace,
+                                       combineTextItems).then(
             function(textContent) {
           finishWorkerTask(task);
           info('text indexing: page=' + pageNum + ' - time=' +

src/display/api.js

@@ -600,6 +600,8 @@ var PDFDocumentProxy = (function PDFDocumentProxyClosure() {
  * @typedef {Object} getTextContentParameters
  * @param {boolean} normalizeWhitespace - replaces all occurrences of
  *   whitespace with standard spaces (0x20). The default value is `false`.
+ * @param {boolean} disableCombineTextItems - do not attempt to combine
+ *   same line {@link TextItem}'s. The default value is `false`.
  */
 
 /**
@@ -891,11 +893,12 @@ var PDFPageProxy = (function PDFPageProxyClosure() {
      * object that represent the page text content.
      */
     getTextContent: function PDFPageProxy_getTextContent(params) {
-      var normalizeWhitespace = (params && params.normalizeWhitespace) || false;
-
       return this.transport.messageHandler.sendWithPromise('GetTextContent', {
         pageIndex: this.pageNumber - 1,
-        normalizeWhitespace: normalizeWhitespace,
+        normalizeWhitespace: (params && params.normalizeWhitespace === true ?
+                              true : /* Default */ false),
+        combineTextItems: (params && params.disableCombineTextItems === true ?
+                           false : /* Default */ true),
       });
     },
 

test/driver.js

@@ -469,12 +469,12 @@ var Driver = (function DriverClosure() {
               textLayerContext.clearRect(0, 0,
                 textLayerCanvas.width, textLayerCanvas.height);
               // The text builder will draw its content on the test canvas
-              initPromise =
-                page.getTextContent({ normalizeWhitespace: true }).then(
-                  function(textContent) {
-                    return rasterizeTextLayer(textLayerContext, viewport,
-                                              textContent);
-                });
+              initPromise = page.getTextContent({
+                normalizeWhitespace: true,
+              }).then(function(textContent) {
+                return rasterizeTextLayer(textLayerContext, viewport,
+                                          textContent);
+              });
             } else {
               textLayerCanvas = null;
 

test/unit/api_spec.js

@@ -771,12 +771,14 @@ describe('api', function() {
     });
     it('gets text content', function (done) {
       var defaultPromise = page.getTextContent();
-      var normalizeWhitespacePromise = page.getTextContent({
-        normalizeWhitespace: true });
+      var parametersPromise = page.getTextContent({
+        normalizeWhitespace: true,
+        disableCombineTextItems: true,
+      });
 
       var promises = [
         defaultPromise,
-        normalizeWhitespacePromise
+        parametersPromise,
       ];
       Promise.all(promises).then(function (data) {
         expect(!!data[0].items).toEqual(true);

web/pdf_page_view.js

@@ -503,12 +503,12 @@ var PDFPageView = (function PDFPageViewClosure() {
         function pdfPageRenderCallback() {
           pageViewDrawCallback(null);
           if (textLayer) {
-            self.pdfPage.getTextContent({ normalizeWhitespace: true }).then(
-              function textContentResolved(textContent) {
-                textLayer.setTextContent(textContent);
-                textLayer.render(TEXT_LAYER_RENDER_DELAY);
-              }
-            );
+            self.pdfPage.getTextContent({
+              normalizeWhitespace: true,
+            }).then(function textContentResolved(textContent) {
+              textLayer.setTextContent(textContent);
+              textLayer.render(TEXT_LAYER_RENDER_DELAY);
+            });
           }
         },
         function pdfPageRenderError(error) {

web/pdf_viewer.js

@@ -784,7 +784,9 @@ var PDFViewer = (function pdfViewer() {
 
     getPageTextContent: function (pageIndex) {
       return this.pdfDocument.getPage(pageIndex + 1).then(function (page) {
-        return page.getTextContent({ normalizeWhitespace: true });
+        return page.getTextContent({
+          normalizeWhitespace: true,
+        });
       });
     },