ImportHTML: don't remove new lines from HTML before feeding it to cheerio

contentcollector: add tests for <pre> and paragraphs with multiple lines
2020-11-10 08:11:25 +01:00 · 2020-11-10 08:09:53 +01:00
2 changed files with 24 additions and 4 deletions
--- a/src/node/utils/ImportHtml.js
+++ b/src/node/utils/ImportHtml.js
@ -33,7 +33,7 @@ exports.setPadHTML = async (pad, html) => {
  rehype()
    .use(format, opts)
    .process(html, function(err, output){
-      html = String(output).replace(/(\r\n|\n|\r)/gm,"");
+      html = String(output);
  })

  var $ = cheerio.load(html);
--- a/tests/backend/specs/contentcollector.js
+++ b/tests/backend/specs/contentcollector.js
@ -1,8 +1,6 @@
-const Changeset        = require("../../../src/static/js/Changeset");
 const contentcollector = require("../../../src/static/js/contentcollector");
 const AttributePool    = require("../../../src/static/js/AttributePool");
 const cheerio          = require("../../../src/node_modules/cheerio");
-const util             = require('util');

 const tests = {
  nestedLi:{
@ -101,8 +99,30 @@ const tests = {
    expectedText: ["*should be 1","*should be 2","*should be 3"],
    noteToSelf: "<p></p>should create a line break but not break numbering -- This is what I can't get working!",
    disabled: true
+  },
+  multiLineParagraph:{
+    description: "A paragraph with multiple lines should not loose spaces when lines are combined",
+    html:`<html><body><p>
+а б в г ґ д е є ж з и і ї й к л м н о
+п р с т у ф х ц ч ш щ ю я ь</p>
+</body></html>`,
+    expectedLineAttribs: [ '+1t' ],
+    expectedText: ["а б в г ґ д е є ж з и і ї й к л м н о п р с т у ф х ц ч ш щ ю я ь"]
+  },
+  multiLineParagraphWithPre:{
+    description: "lines in preformatted text should be kept intact",
+    html:`<html><body><p>
+а б в г ґ д е є ж з и і ї й к л м н о
+<pre>multiple
+lines
+in
+pre
+</pre></p><p>п р с т у ф х ц ч ш щ ю я
+ь</p>
+</body></html>`,
+    expectedLineAttribs: [ '+12', '+8', '+5', '+2', '+3', '+r' ],
+    expectedText: ["а б в г ґ д е є ж з и і ї й к л м н о ", "multiple", "lines", "in", "pre", "п р с т у ф х ц ч ш щ ю я ь"]
  }
-
 }

 describe(__filename, function() {
Author	SHA1	Message	Date
webzwo0i	25d87260e5	ImportHTML: don't remove new lines from HTML before feeding it to cheerio	2020-11-10 08:11:25 +01:00
webzwo0i	de2f6fc55b	contentcollector: add tests for <pre> and paragraphs with multiple lines	2020-11-10 08:09:53 +01:00