Mark glossary terms in html parsed

1 files changed, 70 insertions, 1 deletions

diff --git a/lib/utils/page.js b/lib/utils/page.js
index 00e4757..8fa704d 100644
--- a/lib/utils/page.js
+++ b/lib/utils/page.js
@@ -4,6 +4,60 @@ var cheerio = require('cheerio');
 
 var links = require('./links');
 
+function replaceText($, el, search, replace, text_only ) {
+    return $(el).each(function(){
+        var node = this.firstChild,
+            val,
+            new_val,
+
+            // Elements to be removed at the end.
+            remove = [];
+
+        // Only continue if firstChild exists.
+        if ( node ) {
+
+            // Loop over all childNodes.
+            do {
+                // Only process text nodes.
+                if ( node.nodeType === 3 ) {
+
+                    // The original node value.
+                    val = node.nodeValue;
+
+                    // The new value.
+                    new_val = val.replace( search, replace );
+
+                    // Only replace text if the new value is actually different!
+                    if ( new_val !== val ) {
+
+                        if ( !text_only && /</.test( new_val ) ) {
+                            // The new value contains HTML, set it in a slower but far more
+                            // robust way.
+                            $(node).before( new_val );
+
+                            // Don't remove the node yet, or the loop will lose its place.
+                            remove.push( node );
+                        } else {
+                            // The new value contains no HTML, so it can be set in this
+                            // very fast, simple way.
+                            node.nodeValue = new_val;
+                        }
+                    }
+                }
+
+            } while ( node = node.nextSibling );
+        }
+
+        // Time to remove those elements!
+        remove.length && $(remove).remove();
+    });
+};
+
+function pregQuote( str ) {
+    return (str+'').replace(/([\\\.\+\*\?\[\^\]\$\(\)\{\}\=\!\<\>\|\:])/g, "\\$1");
+};
+
+
 // Adapt an html snippet to be relative to a base folder
 function normalizeHtml(src, options) {
 	var $ = cheerio.load(src);
@@ -38,9 +92,21 @@ function normalizeHtml(src, options) {
         $(this).attr("href", href);
     });
 
+    // Replace glossayr terms
+    _.each(options.glossary, function(term) {
+        var r =  new RegExp( "\\b(" + pregQuote(term.name.toLowerCase()) + ")\\b" , 'gi' );
+
+        $("*").each(function() {
+            replaceText($, this, r, function(match) {
+                return "<span class='glossary-term' data-glossary-term='"+term.id+"' title='"+term.description+"'>"+match+"</span>";
+            });
+        });
+    });
+
 	return $.html();
 };
 
+
 // Adapt page content to be relative to a base folder
 function normalizePage(sections, options) {
 	options = _.defaults(options || {}, {
@@ -51,7 +117,10 @@ function normalizePage(sections, options) {
 		base: "./",
 
 		// Directory parent from the html output
-		output: "./"
+		output: "./",
+
+        // Glossary terms
+        glossary: []
 	});
 
 	return _.map(sections, function(section) {