tools,doc: add Google Analytics tracking.

Adds Google Analytics tracking script to all doc pages when
`DOCS_ANALYTICS` is set when running `make`:

```bash
$ DOCS_ANALYTICS=<GOOGLE ANALYTICS ID> make
```

By default (when `DOCS_ANALYTICS` is not set), no tracking scripts are
included.

It respects "Do Not Track" settings end users might have in their
browser.

Also changes make target `doc-upload` from depending on the
`$(TARBALL)` target, to only depend on `doc` directly.

PR-URL: #6601
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>

Makefile

@@ -300,6 +300,11 @@ test-v8 test-v8-intl test-v8-benchmarks test-v8-all:
 		"$ git clone https://github.com/nodejs/node.git"
 endif
 
+# Google Analytics ID used for tracking API docs page views, empty
+# DOCS_ANALYTICS means no tracking scripts will be included in the
+# generated .html files
+DOCS_ANALYTICS ?=
+
 apidoc_sources = $(wildcard doc/api/*.md)
 apidocs_html = $(apidoc_dirs) $(apiassets) $(addprefix out/,$(apidoc_sources:.md=.html))
 apidocs_json = $(apidoc_dirs) $(apiassets) $(addprefix out/,$(apidoc_sources:.md=.json))
@@ -333,7 +338,8 @@ out/doc/api/%.json: doc/api/%.md
 	[ -x $(NODE) ] && $(NODE) $(gen-json) || node $(gen-json)
 
 # check if ./node is actually set, else use user pre-installed binary
-gen-html = tools/doc/generate.js --node-version=$(FULLVERSION) --format=html --template=doc/template.html $< > $@
+gen-html = tools/doc/generate.js --node-version=$(FULLVERSION) --format=html \
+			--template=doc/template.html --analytics=$(DOCS_ANALYTICS) $< > $@
 out/doc/api/%.html: doc/api/%.md
 	@[ -e tools/doc/node_modules/js-yaml/package.json ] || \
 		[ -e tools/eslint/node_modules/js-yaml/package.json ] || \
@@ -587,7 +593,7 @@ ifeq ($(XZ), 0)
 	ssh $(STAGINGSERVER) "touch nodejs/$(DISTTYPEDIR)/$(FULLVERSION)/node-$(FULLVERSION).tar.xz.done"
 endif
 
-doc-upload: tar
+doc-upload: doc
 	ssh $(STAGINGSERVER) "mkdir -p nodejs/$(DISTTYPEDIR)/$(FULLVERSION)"
 	chmod -R ug=rw-x+X,o=r+X out/doc/
 	scp -pr out/doc/ $(STAGINGSERVER):nodejs/$(DISTTYPEDIR)/$(FULLVERSION)/docs/

doc/api_assets/dnt_helper.js

@@ -0,0 +1,49 @@
+/**
+ * http://schalkneethling.github.io/blog/2015/11/06/respect-user-choice-do-not-track/
+ * https://github.com/schalkneethling/dnt-helper/blob/master/js/dnt-helper.js
+ *
+ * Returns true or false based on whether doNotTack is enabled. It also takes into account the
+ * anomalies, such as !bugzilla 887703, which effect versions of Fx 31 and lower. It also handles
+ * IE versions on Windows 7, 8 and 8.1, where the DNT implementation does not honor the spec.
+ * @see https://bugzilla.mozilla.org/show_bug.cgi?id=1217896 for more details
+ * @params {string} [dnt] - An optional mock doNotTrack string to ease unit testing.
+ * @params {string} [userAgent] - An optional mock userAgent string to ease unit testing.
+ * @returns {boolean} true if enabled else false
+ */
+function _dntEnabled(dnt, userAgent) {
+
+    'use strict';
+
+    // for old version of IE we need to use the msDoNotTrack property of navigator
+    // on newer versions, and newer platforms, this is doNotTrack but, on the window object
+    // Safari also exposes the property on the window object.
+    var dntStatus = dnt || navigator.doNotTrack || window.doNotTrack || navigator.msDoNotTrack;
+    var ua = userAgent || navigator.userAgent;
+
+    // List of Windows versions known to not implement DNT according to the standard.
+    var anomalousWinVersions = ['Windows NT 6.1', 'Windows NT 6.2', 'Windows NT 6.3'];
+
+    var fxMatch = ua.match(/Firefox\/(\d+)/);
+    var ieRegEx = /MSIE|Trident/i;
+    var isIE = ieRegEx.test(ua);
+    // Matches from Windows up to the first occurance of ; un-greedily
+    // http://www.regexr.com/3c2el
+    var platform = ua.match(/Windows.+?(?=;)/g);
+
+    // With old versions of IE, DNT did not exist so we simply return false;
+    if (isIE && typeof Array.prototype.indexOf !== 'function') {
+        return false;
+    } else if (fxMatch && parseInt(fxMatch[1], 10) < 32) {
+        // Can't say for sure if it is 1 or 0, due to Fx bug 887703
+        dntStatus = 'Unspecified';
+    } else if (isIE && platform && anomalousWinVersions.indexOf(platform.toString()) !== -1) {
+        // default is on, which does not honor the specification
+        dntStatus = 'Unspecified';
+    } else {
+        // sets dntStatus to Disabled or Enabled based on the value returned by the browser.
+        // If dntStatus is undefined, it will be set to Unspecified
+        dntStatus = { '0': 'Disabled', '1': 'Enabled' }[dntStatus] || 'Unspecified';
+    }
+
+    return dntStatus === 'Enabled' ? true : false;
+}

doc/template.html

@@ -45,5 +45,6 @@ <h2>Table of Contents</h2>
   <script src="assets/sh_main.js"></script>
   <script src="assets/sh_javascript.min.js"></script>
   <script>highlight(undefined, undefined, 'pre');</script>
+  <!-- __TRACKING__ -->
 </body>
 </html>

test/doctool/test-doctool-html.js

@@ -72,11 +72,18 @@ const testData = [
     '<p>I exist and am being linked to.</p>' +
     '<!-- [end-include:doc_inc_2.md] -->'
   },
+  {
+    file: path.join(common.fixturesDir, 'sample_document.md'),
+    html: '<ol><li>fish</li><li><p>fish</p></li><li><p>Redfish</p></li>' +
+      '<li>Bluefish</li></ol>',
+    analyticsId: 'UA-67020396-1'
+  },
 ];
 
 testData.forEach((item) => {
   // Normalize expected data by stripping whitespace
   const expected = item.html.replace(/\s/g, '');
+  const includeAnalytics = typeof item.analyticsId !== 'undefined';
 
   fs.readFile(item.file, 'utf8', common.mustCall((err, input) => {
     assert.ifError(err);
@@ -89,6 +96,7 @@ testData.forEach((item) => {
           filename: 'foo',
           template: 'doc/template.html',
           nodeVersion: process.version,
+          analytics: item.analyticsId,
         },
         common.mustCall((err, output) => {
           assert.ifError(err);
@@ -97,6 +105,16 @@ testData.forEach((item) => {
           // Assert that the input stripped of all whitespace contains the
           // expected list
           assert.notStrictEqual(actual.indexOf(expected), -1);
+
+          // Testing the insertion of Google Analytics script when
+          // an analytics id is provided. Should not be present by default
+          if (includeAnalytics) {
+            assert.notStrictEqual(actual.indexOf('google-analytics.com'), -1,
+                                  'Google Analytics script was not present');
+          } else {
+            assert.strictEqual(actual.indexOf('google-analytics.com'), -1,
+                               'Google Analytics script was present');
+          }
         }));
     }));
   }));

tools/doc/generate.js

@@ -11,16 +11,19 @@ let format = 'json';
 let template = null;
 let inputFile = null;
 let nodeVersion = null;
+let analytics = null;
 
 args.forEach(function(arg) {
-  if (!arg.match(/^--/)) {
+  if (!arg.startsWith('--')) {
     inputFile = arg;
-  } else if (arg.match(/^--format=/)) {
+  } else if (arg.startsWith('--format=')) {
     format = arg.replace(/^--format=/, '');
-  } else if (arg.match(/^--template=/)) {
+  } else if (arg.startsWith('--template=')) {
     template = arg.replace(/^--template=/, '');
-  } else if (arg.match(/^--node-version=/)) {
+  } else if (arg.startsWith('--node-version=')) {
     nodeVersion = arg.replace(/^--node-version=/, '');
+  } else if (arg.startsWith('--analytics=')) {
+    analytics = arg.replace(/^--analytics=/, '');
   }
 });
 
@@ -54,6 +57,7 @@ function next(er, input) {
           filename: inputFile,
           template: template,
           nodeVersion: nodeVersion,
+          analytics: analytics,
         },
 
         function(er, html) {

tools/doc/html.js

@@ -67,6 +67,7 @@ function toHTML(opts, cb) {
         filename: opts.filename,
         template: template,
         nodeVersion: nodeVersion,
+        analytics: opts.analytics,
       }, cb);
     });
   }
@@ -128,6 +129,13 @@ function render(opts, cb) {
       gtocData.replace('class="nav-' + id, 'class="nav-' + id + ' active')
     );
 
+    if (opts.analytics) {
+      template = template.replace(
+        '<!-- __TRACKING__ -->',
+        analyticsScript(opts.analytics)
+      );
+    }
+
     // content has to be the last thing we do with
     // the lexed tokens, because it's destructive.
     const content = marked.parser(lexed);
@@ -137,6 +145,23 @@ function render(opts, cb) {
   });
 }
 
+function analyticsScript(analytics) {
+  return `
+    <script src="assets/dnt_helper.js"></script>
+    <script>
+      if (!_dntEnabled()) {
+        (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;
+        i[r]=i[r]||function(){(i[r].q=i[r].q||[]).push(arguments)},
+        i[r].l=1*new Date();a=s.createElement(o),m=s.getElementsByTagName(o)[0];
+        a.async=1;a.src=g;m.parentNode.insertBefore(a,m)})(window,document,
+        'script','//www.google-analytics.com/analytics.js','ga');
+        ga('create', '${analytics}', 'auto');
+        ga('send', 'pageview');
+      }
+    </script>
+  `;
+}
+
 // handle general body-text replacements
 // for example, link man page references to the actual page
 function parseText(lexed) {