X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=src%2Fmkws-handlebars.js;h=a8dc14292b44867f9ca8c5298692a71bc752ff56;hb=e01ba74a96c0523035b7328be28b5dda2755f97a;hp=6e93578a6b026f75ff8943093a163b28d9aabd48;hpb=d81d10eb466240b6e557379a4e298709a5086bf1;p=mkws-moved-to-github.git

diff --git a/src/mkws-handlebars.js b/src/mkws-handlebars.js
index 6e93578..a8dc142 100644
--- a/src/mkws-handlebars.js
+++ b/src/mkws-handlebars.js
@@ -5,11 +5,37 @@ Handlebars.registerHelper('mkws-json', function(obj) {
 });
 
 
-Handlebars.registerHelper('mkws-paragraphs', function(obj) {
+// This is intended to handle paragraphs from Wikipedia, hence the
+// rather hacky code to remove numbered references.
+//
+Handlebars.registerHelper('mkws-paragraphs', function(obj, nPara, nSent) {
   var acc = [];
-  for (var i = 0; i < obj.length; i++) {
-    acc.push('<p>', obj[i], '</p>');
+
+  // For some reason, Handlebars provides the value
+  // {"hash":{},"data":{}} for parameters that are not provided. So we
+  // have to be prepared for actual numbers, explicitly undefined
+  // values and this dumb magic value.
+  if (nPara === undefined || nPara.hasOwnProperty('hash') || nPara == 0 || nPara > obj.length) {
+    nPara = obj.length;
+  }
+  if (nSent === undefined || nSent.hasOwnProperty('hash') || nSent == 0) {
+    nSent = Infinity;
   }
+
+  for (var i = 0; i < nPara; i++) {
+    // Remove numbered references such as "[1,3,4]" from text
+    var text = obj[i].replace(/\[[0-9,]+\]/g, '');
+    // Next line from http://stackoverflow.com/questions/18914629/split-string-into-sentences-in-javascript
+    var sentences = text.replace(/([.?!])\s*(?=[A-Z])/g, "$1|").split("|");
+    if (sentences.length > nSent)
+      sentences.length = nSent;
+
+    acc.push('<p>', sentences.join(' '), '</p>');
+    nSent -= sentences.length;
+    if (nSent == 0)
+      break;
+  }
+
   return acc.join('');
 });