Add Page.content() method. Fixes puppeteer#406. (puppeteer#419)

Go-Labsss · Aug 21, 2017 · 598f439 · 598f439
1 parent 6fcf3d9
commit 598f439
Show file tree

Hide file tree

Showing 3 changed files with 36 additions and 2 deletions.
diff --git a/docs/api.md b/docs/api.md
@@ -30,6 +30,7 @@
     + [page.addScriptTag(url)](#pageaddscripttagurl)
     + [page.click(selector[, options])](#pageclickselector-options)
     + [page.close()](#pageclose)
+    + [page.content()](#pagecontent)
     + [page.emulate(options)](#pageemulateoptions)
     + [page.emulateMedia(mediaType)](#pageemulatemediamediatype)
     + [page.evaluate(pageFunction, ...args)](#pageevaluatepagefunction-args)
@@ -317,6 +318,11 @@ If there's no element matching `selector`, the method throws an error.
 #### page.close()
 - returns: <[Promise]>
 
+#### page.content()
+- returns: <[Promise]<[String]>>
+
+Gets the full HTML contents of the page, including the doctype.
+
 #### page.emulate(options)
 - `options` <[Object]>
   - `viewport` <[Object]>

diff --git a/lib/Page.js b/lib/Page.js
@@ -260,6 +260,20 @@ class Page extends EventEmitter {
     return this.mainFrame().url();
   }
 
+  /**
+   * @return {!Promise<String>}
+   */
+  async content() {
+    return await this.evaluate(() => {
+      let retVal = '';
+      if (document.doctype)
+        retVal = new XMLSerializer().serializeToString(document.doctype);
+      if (document.documentElement)
+        retVal += document.documentElement.outerHTML;
+      return retVal;
+    });
+  }
+
   /**
    * @param {string} html
    * @return {!Promise}

diff --git a/test/test.js b/test/test.js
@@ -1437,10 +1437,24 @@ describe('Page', function() {
     }));
   });
   describe('Page.setContent', function() {
+    const expectedOutput = '<html><head></head><body><div>hello</div></body></html>';
     it('should work', SX(async function() {
       await page.setContent('<div>hello</div>');
-      let result = await page.evaluate(() => document.body.innerHTML);
-      expect(result).toBe('<div>hello</div>');
+      let result = await page.content();
+      expect(result).toBe(expectedOutput);
+    }));
+    it('should work with doctype', SX(async function() {
+      const doctype = '<!DOCTYPE html>';
+      await page.setContent(`${doctype}<div>hello</div>`);
+      let result = await page.content();
+      expect(result).toBe(`${doctype}${expectedOutput}`);
+    }));
+    it('should work with HTML 4 doctype', SX(async function() {
+      const doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN" ' +
+        '"http://www.w3.org/TR/html4/strict.dtd">';
+      await page.setContent(`${doctype}<div>hello</div>`);
+      let result = await page.content();
+      expect(result).toBe(`${doctype}${expectedOutput}`);
     }));
   });
   describe('Network Events', function() {