Improve summary parser

12 files changed, 168 insertions, 80 deletions

diff --git a/packages/gitbook-html/lib/dom.js b/packages/gitbook-html/lib/dom.js
index 2c2eaf7..df4de90 100644
--- a/packages/gitbook-html/lib/dom.js
+++ b/packages/gitbook-html/lib/dom.js
@@ -3,12 +3,18 @@ var cheerio = require('cheerio');
 
 // Parse an HTML string and return its content
 function parse(html) {
-    var $ = cheerio.load('<div>'+html+'</div>');
+    var $ = cheerio.load(html);
     var $el = $('html, body').first();
 
     return $el.length > 0? $el : $;
 }
 
+// Return main element
+function root($) {
+    var $el = $('html, body, > div').first();
+    return $el.length > 0? $el : $.root();
+}
+
 // Return text node of an element
 function textNode($el) {
     return _.reduce($el.children, function(text, e) {
@@ -19,5 +25,6 @@ function textNode($el) {
 
 module.exports = {
     parse: parse,
-    textNode: textNode
+    textNode: textNode,
+    root: root
 };
diff --git a/packages/gitbook-html/lib/glossary.js b/packages/gitbook-html/lib/glossary.js
index 648ba1a..9d3799b 100755
--- a/packages/gitbook-html/lib/glossary.js
+++ b/packages/gitbook-html/lib/glossary.js
@@ -7,9 +7,10 @@ function parseGlossary(html) {
 
     var entries = [];
 
-    $("h2").each(function() {
+    $('h2').each(function() {
         var $heading = $(this);
-        var $p = $heading.next();
+        var $next = $heading.next()
+        var $p =  $next.is('p')? $next.first() : $next.find('p').first();
 
         var entry = {};
 
diff --git a/packages/gitbook-html/lib/index.js b/packages/gitbook-html/lib/index.js
index a7c478c..0e67c94 100755
--- a/packages/gitbook-html/lib/index.js
+++ b/packages/gitbook-html/lib/index.js
@@ -1,8 +1,33 @@
+var _ = require('lodash');
 
-module.exports = {
+var htmlParser = {
     summary: require("./summary"),
     glossary: require("./glossary"),
     langs: require("./langs"),
     readme: require("./readme"),
     page: require("./page")
 };
+
+// Compose a function with a transform function for the first args
+function compose(toHTML, fn) {
+    return function() {
+        var args = _.toArray(arguments);
+        args[0] = toHTML(args[0]);
+
+        return fn.apply(null, args);
+    }
+}
+
+// Create a GitBook parser
+function createParser(toHTML) {
+    return {
+        summary: compose(toHTML, htmlParser.summary),
+        glossary: compose(toHTML, htmlParser.glossary),
+        langs: compose(toHTML, htmlParser.langs),
+        readme: compose(toHTML, htmlParser.readme),
+        page: compose(toHTML, htmlParser.page)
+    }
+}
+
+module.exports = htmlParser;
+module.exports.createParser = createParser;
diff --git a/packages/gitbook-html/lib/langs.js b/packages/gitbook-html/lib/langs.js
index 1042dcb..270a9f6 100755
--- a/packages/gitbook-html/lib/langs.js
+++ b/packages/gitbook-html/lib/langs.js
@@ -1,9 +1,9 @@
 var _ = require('lodash');
-var parseEntries = require('./summary').entries;
+var parseSummary = require('./summary');
 
 // HTML -> Languages
 function parseLangs(content) {
-    return parseEntries(content);
+    return parseSummary(content).parts[0].articles;
 }
 
 // Languages -> HTML
diff --git a/packages/gitbook-html/lib/page.js b/packages/gitbook-html/lib/page.js
index e687050..6f056fc 100755
--- a/packages/gitbook-html/lib/page.js
+++ b/packages/gitbook-html/lib/page.js
@@ -2,8 +2,10 @@ var Q = require('q');
 var _ = require('lodash');
 
 // HTML -> HTML
-function parsePage(src) {
-    return src;
+function parsePage(html) {
+    return {
+        content: html
+    };
 }
 
 module.exports = parsePage;
diff --git a/packages/gitbook-html/lib/summary.js b/packages/gitbook-html/lib/summary.js
index 1e2d63d..e71d6b5 100755
--- a/packages/gitbook-html/lib/summary.js
+++ b/packages/gitbook-html/lib/summary.js
@@ -1,14 +1,17 @@
 var _ = require('lodash');
 var dom = require('./dom');
 
+var SELECTOR_LIST = '.olist > ol, ol, ul';
+var SELECTOR_LINK = 'a, p > a';
+
+var BL = '\n';
 
 // parse a ul list and return list of chapters recursvely
 function parseList($ul, $) {
     var articles = [];
 
-    $ul.children('>li').each(function() {
+    $ul.children('li').each(function() {
         var article = {};
-
         var $li = $(this);
 
         // Get text for the entry
@@ -16,14 +19,14 @@ function parseList($ul, $) {
         article.title = $p.text() ||  dom.textNode($li.get(0));
 
         // Parse link
-        var $a = $li.find('> a, > p > a');
+        var $a = $li.children(SELECTOR_LINK);
         if ($a.length > 0) {
             article.title = $a.first().text();
             article.path = $a.attr('href').replace(/\\/g, '/').replace(/^\/+/, '')
         }
 
         // Sub articles
-        var $sub = $li.children('> .olist > ol, > ol, > ul');
+        var $sub = $li.children(SELECTOR_LIST).first();
         article.articles = parseList($sub, $);
 
         articles.push(article);
@@ -32,55 +35,73 @@ function parseList($ul, $) {
     return articles;
 }
 
-// Return a list of entries in a div
-function parseEntries (html) {
+// HTML -> Summary
+function parseSummary(html) {
     var $ = dom.parse(html);
-    var chapters = parseList($("> ol, > ul").first(), $);
-    return chapters;
-}
+    var $root = dom.root($);
 
-// HTML -> Summary
-function parseSummary(src) {
-    var chapters = parseEntries(src);
+    var $lists = $root.children(SELECTOR_LIST);
+    var parts = [];
+
+    $lists.each(function() {
+        var $list = $(this);
+
+        parts.push({
+            articles: parseList($(SELECTOR_LIST).first(), $)
+        });
+    });
 
     return {
-        chapters: chapters
+        parts: parts
     };
 }
 
 // Summary -> HTML
-function summaryToText(summary) {
-    var bl = '\n';
-
-    var _base = function(article) {
-        if (article.path) {
-            return '<a href="'+article.path+'">'+article.title+'</a>';
-        } else {
-            return article.title;
-        }
-    };
+function textPrefix(d) {
+    return Array(d*4).join(' ');
+}
 
-    var convertArticle = function(article, d) {
-        var content = Array(d+2).join(' ') + '<li>' + _base(article);
+function articleToText(article, d) {
+    var prefix = textPrefix(d);
+    var content = prefix + '<li>';
 
-        if (article.articles.length > 0) {
-            content += convertArticles(article.articles, d);
-        }
-        return content + '</li>' + bl;
-    };
+    if (article.path) {
+        content += '<a href="'+article.path+'">'+article.title+'</a>';
+    } else {
+        content += article.title;
+    }
 
-    var convertArticles = function(articles, d) {
-        var content = '<ul>' + bl;
-        _.each(articles, function(_article) {
-            content += convertArticle(_article, d + 1);
-        });
-        return content + '<ul>' + bl;
+    if (article.articles.length > 0) {
+        content += BL + articlesToText(article.articles, d) + prefix;
     }
+    content +=  '</li>' + BL;
+
+    return content;
+}
+
+function articlesToText(articles, d) {
+    var prefix = textPrefix(d);
+    var content = prefix + '<ul>' + BL;
+    _.each(articles, function(_article) {
+        content += articleToText(_article, d + 1);
+    });
+    return content + '</ul>' + BL;
+}
+
+function partsToText(part) {
+    return articlesToText(part.articles, 0) + BL + BL;
+}
+
+function summaryToText(summary) {
+    var content = '<h1>Summary</h1>' + BL;
+
+    _.each(summary.parts, function(part) {
+        content += partsToText(part);
+    });
 
-    return '<h1>Summary</h1>'+ bl+bl + convertArticles(summary.chapters, 0) + bl;
+    return content + BL;
 };
 
 
 module.exports = parseSummary;
-module.exports.entries = parseEntries;
 module.exports.toText = summaryToText;
diff --git a/packages/gitbook-html/package.json b/packages/gitbook-html/package.json
index ff28ca6..a21259b 100644
--- a/packages/gitbook-html/package.json
+++ b/packages/gitbook-html/package.json
@@ -7,13 +7,13 @@
     "dependencies": {
         "q": "^1.1.2",
         "lodash": "^3.2.0",
-        "cheerio": "^0.19.0"
+        "cheerio": "^0.20.0 && >=0.20.0"
     },
     "devDependencies": {
         "mocha": "^2.3.2"
     },
     "scripts": {
-        "test": "export TESTING=true; mocha --reporter list --bail"
+        "test": "export TESTING=true; mocha --reporter spec --bail"
     },
     "repository": {
         "type": "git",
diff --git a/packages/gitbook-html/test/fixtures/SUMMARY.html b/packages/gitbook-html/test/fixtures/SUMMARY.html
index f469249..bae97f3 100755
--- a/packages/gitbook-html/test/fixtures/SUMMARY.html
+++ b/packages/gitbook-html/test/fixtures/SUMMARY.html
@@ -23,4 +23,16 @@
         </ul>
     </li>
     <li>Unfinished Chapter</li>
-</ul>
\ No newline at end of file
+</ul>
+
+<ul>
+    <li>
+        <a href="chapter-1/README.md">Chapter 1</a>
+    </li>
+</ul>
+
+<ul>
+    <li>
+        <a href="chapter-1/README.md">Chapter 1</a>
+    </li>
+</ul>
diff --git a/packages/gitbook-html/test/glossary.js b/packages/gitbook-html/test/glossary.js
index 250c6f1..8bd77d6 100755
--- a/packages/gitbook-html/test/glossary.js
+++ b/packages/gitbook-html/test/glossary.js
@@ -4,10 +4,14 @@ var assert = require('assert');
 
 var glossary = require('../').glossary;
 
-var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/GLOSSARY.html'), 'utf8');
-var LEXED = glossary(CONTENT);
-
 describe('Glossary parsing', function () {
+    var LEXED;
+
+    before(function() {
+        var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/GLOSSARY.html'), 'utf8');
+        LEXED = glossary(CONTENT);
+    });
+
     it('should only get heading + paragraph pairs', function() {
         assert.equal(LEXED.length, 5);
     });
diff --git a/packages/gitbook-html/test/langs.js b/packages/gitbook-html/test/langs.js
index c51cf2d..6b5e00b 100755
--- a/packages/gitbook-html/test/langs.js
+++ b/packages/gitbook-html/test/langs.js
@@ -4,10 +4,14 @@ var assert = require('assert');
 
 var langs = require('../').langs;
 
-var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/LANGS.html'), 'utf8');
-var LEXED = langs(CONTENT);
-
 describe('Languages parsing', function () {
+    var LEXED;
+
+    before(function() {
+        var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/LANGS.html'), 'utf8');
+        LEXED = langs(CONTENT);
+    });
+
     it('should detect paths and titles', function() {
         assert.equal(LEXED[0].path,'en/');
         assert.equal(LEXED[0].title,'English');
diff --git a/packages/gitbook-html/test/readme.js b/packages/gitbook-html/test/readme.js
index 9d9ca29..f38f40b 100755
--- a/packages/gitbook-html/test/readme.js
+++ b/packages/gitbook-html/test/readme.js
@@ -4,11 +4,13 @@ var assert = require('assert');
 
 var readme = require('../').readme;
 
-
-var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/README.html'), 'utf8');
-var LEXED = readme(CONTENT);
-
 describe('Readme parsing', function () {
+    var LEXED;
+
+    before(function() {
+        var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/README.html'), 'utf8');
+        LEXED = readme(CONTENT);
+    });
 
     it('should contain a title', function() {
         assert(LEXED.title);
diff --git a/packages/gitbook-html/test/summary.js b/packages/gitbook-html/test/summary.js
index 8d686fc..4d06c32 100755
--- a/packages/gitbook-html/test/summary.js
+++ b/packages/gitbook-html/test/summary.js
@@ -4,37 +4,47 @@ var assert = require('assert');
 
 var summary = require('../').summary;
 
-var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/SUMMARY.html'), 'utf8');
-var LEXED = summary(CONTENT);
 describe('Summary parsing', function () {
+    var LEXED, PART;
+
+    before(function() {
+        var CONTENT = fs.readFileSync(path.join(__dirname, './fixtures/SUMMARY.html'), 'utf8');
+        LEXED = summary(CONTENT);
+        PART = LEXED.parts[0];
+    });
+
+    it('should detect parts', function() {
+        assert.equal(LEXED.parts.length, 3);
+    });
+
     it('should detect chapters', function() {
-        assert.equal(LEXED.chapters.length, 5);
+        assert.equal(PART.articles.length, 5);
     });
 
     it('should support articles', function() {
-        assert.equal(LEXED.chapters[0].articles.length, 2);
-        assert.equal(LEXED.chapters[1].articles.length, 0);
-        assert.equal(LEXED.chapters[2].articles.length, 0);
+        assert.equal(PART.articles[0].articles.length, 2);
+        assert.equal(PART.articles[1].articles.length, 0);
+        assert.equal(PART.articles[2].articles.length, 0);
     });
 
     it('should detect paths and titles', function() {
-        assert(LEXED.chapters[0].path);
-        assert(LEXED.chapters[1].path);
-        assert(LEXED.chapters[2].path);
-        assert(LEXED.chapters[3].path);
-        assert.equal(LEXED.chapters[4].path, null);
-
-        assert(LEXED.chapters[0].title);
-        assert(LEXED.chapters[1].title);
-        assert(LEXED.chapters[2].title);
-        assert(LEXED.chapters[3].title);
-        assert(LEXED.chapters[4].title);
+        assert(PART.articles[0].path);
+        assert(PART.articles[1].path);
+        assert(PART.articles[2].path);
+        assert(PART.articles[3].path);
+        assert.equal(PART.articles[4].path, null);
+
+        assert(PART.articles[0].title);
+        assert(PART.articles[1].title);
+        assert(PART.articles[2].title);
+        assert(PART.articles[3].title);
+        assert(PART.articles[4].title);
     });
 
     it('should normalize paths from .md', function() {
-        assert.equal(LEXED.chapters[0].path,'chapter-1/README.md');
-        assert.equal(LEXED.chapters[1].path,'chapter-2/README.md');
-        assert.equal(LEXED.chapters[2].path,'chapter-3/README.md');
+        assert.equal(PART.articles[0].path,'chapter-1/README.md');
+        assert.equal(PART.articles[1].path,'chapter-2/README.md');
+        assert.equal(PART.articles[2].path,'chapter-3/README.md');
     });
 
     it('should correctly convert it to text', function() {