Added Markdown headings extraction possibility

Readme.md

@@ -18,6 +18,29 @@ Metalsmith(__dirname)
   .build();
 ```
 
+### Enhanced mode (Markdown)
+
+Use an array and even decide, if it should already parse your Markdown files:
+
+```js
+var Metalsmith = require('metalsmith');
+var headings = require('headings');
+var markdown = require('metalsmith-markdown');
+
+Metalsmith(__dirname)
+  .use(headings({
+    mode: 'md',         // set to anything else or remove this property for HTML-mode
+    selectors: ['h1', 'h2'],
+  }))
+  .use(markdown())      // Invoke this function somewhere after 'headings'
+  .build();
+```
+
+Bear in mind to include this function somewhere before your `.use(markdown())` function.  
+Otherwise, `metalsmith` would compile your files into HTML before it would extract your headings.
+
+One other thing to mention would be that `mode: 'md'` orders your headings according to their position in the `selectors` array in the options object, whereas the *legacy* HTML-mode orders your headings according to their position in your files. You choose.
+
 ## License
 
   MIT

lib/index.js

@@ -1,41 +1,95 @@
 
 var cheerio = require('cheerio');
+var debug = require('debug')('metalsmith-headings');
 var extname = require('path').extname;
+var slug = require('slug');
 
 /**
  * Expose `plugin`.
  */
 
 module.exports = plugin;
 
+/**
+ * Filters selectors out of a Markdown file
+ * @param {Object} data 
+ * @param {string[]} selectors 
+ */
+function filterMarkdown(data, selectors) {
+  const headers = {
+    h1: /^#{1}\s+(.+?)$/gim,
+    h2: /^#{2}\s+(.+?)$/gim,
+    h3: /^#{3}\s+(.+?)$/gim,
+    h4: /^#{4}\s+(.+?)$/gim,
+    h5: /^#{5}\s+(.+?)$/gim,
+    h6: /^#{6}\s+(.+?)$/gim,
+  };
+  const regexes = selectors.map((selector) => headers[selector]).filter((entry) => entry !== null || typeof entry !== 'undefined');
+  const contents = data.contents.toString();
+  data.headings = [];
+
+  regexes.forEach((regex, index) => {
+    let found;
+    while ((found = regex.exec(contents)) !== null) {
+      if (found.index === regex.lastIndex) {
+        regex.lastIndex++;
+      }
+
+      if (found[1].length) {
+        data.headings.push({
+          id: slug(found[1]),
+          tag: selectors[index],
+          text: found[1],
+        });
+      }
+    }
+  });
+  return data;
+}
+
+/**
+ * Filters selectors out of an HTML file
+ * @param {Object} data the file object
+ * @param {string[]} selectors The selectors array
+ */
+function filterHTML(data, selectors) {
+  var contents = data.contents.toString();
+  var $ = cheerio.load(contents);
+  data.headings = [];
+  $(selectors.join(',')).each(function() {
+    data.headings.push({
+      id: $(this).attr('id'),
+      tag: $(this)[0].name,
+      text: $(this).text()
+    });
+  });
+  return data;
+}
+
 /**
  * Get the headings from any html files.
  *
  * @param {String or Object} options (optional)
  *   @property {Array} selectors
  */
-
-function plugin(options){
+function plugin(options) {
   if ('string' == typeof options) options = { selectors: [options] };
-  options = options || {};
+  options = options || { };
   var selectors = options.selectors || ['h2'];
 
   return function(files, metalsmith, done){
     setImmediate(done);
-    Object.keys(files).forEach(function(file){
-      if ('.html' != extname(file)) return;
+    Object.keys(files).forEach(function(file) {
       var data = files[file];
-      var contents = data.contents.toString();
-      var $ = cheerio.load(contents);
-      data.headings = [];
-
-        $(selectors.join(',')).each(function(){
-          data.headings.push({
-            id: $(this).attr('id'),
-            tag: $(this)[0].name,
-            text: $(this).text()
-          });
-      });
+      debug(`Currently processing: ${file}\nMode: ${options.mode || 'html (default)'}`);
+      if (options.mode !== 'md' && '.html' === extname(file)) {
+        data = filterHTML(data, selectors);
+      }
+      if (options.mode === 'md' && '.md' === extname(file)) {
+        data = filterMarkdown(data, selectors);
+      }
+      debug(`Found: ${data.headings.length} headings.`);
+      return;
     });
   };
 }

package.json

@@ -9,11 +9,13 @@
     "test": "mocha --reporter spec"
   },
   "dependencies": {
-    "cheerio": "^0.14.0"
+    "cheerio": "^0.14.0",
+    "debug": "^3.1.0",
+    "slug": "^0.9.1"
   },
   "devDependencies": {
-    "mocha": "1.x",
     "metalsmith": "0.x",
-    "metalsmith-markdown": "^0.2.1"
+    "metalsmith-markdown": "^0.2.1",
+    "mocha": "1.x"
   }
 }

test/index.js

@@ -19,6 +19,36 @@ describe('metalsmith-headings', function(){
       });
   });
 
+  it('should parse headings from Markdown only for h2', function(done){
+    Metalsmith('test/fixture')
+      .use(headings({ selectors: ['h2'], mode: 'md' }))
+      .use(markdown())
+      .build(function(err, files){
+        if (err) return done(err);
+        assert.deepEqual(files['index.html'].headings, [
+          { id: 'two-one', tag: 'h2', text: 'two one' },
+          { id: 'two-two', tag: 'h2', text: 'two two' }
+        ]);
+        done();
+      });
+  });
+
+  it('should parse headings from Markdown for h1 and h2', function(done){
+    Metalsmith('test/fixture')
+      .use(headings({ selectors: ['h1', 'h2'], mode: 'md' }))
+      .use(markdown())
+      .build(function(err, files){
+        if (err) return done(err);
+        assert.deepEqual(files['index.html'].headings, [
+          { id: 'one-one', tag: 'h1', text: 'one one' },
+          { id: 'one-two', tag: 'h1', text: 'one two' },
+          { id: 'two-one', tag: 'h2', text: 'two one' },
+          { id: 'two-two', tag: 'h2', text: 'two two' }
+        ]);
+        done();
+      });
+  });
+
   it('should accept a string shorthand', function(done){
     Metalsmith('test/fixture')
       .use(markdown())