Range -> extractContents -> unexpected result

[Alexander]

Steps to reproduce

1. Open CKEditor demo page - ​http://ckeditor.com/demo
2. Select any text
3. Open "dev tools"(F12) -> console
4. Try to execute next code

currentRange = CKEDITOR.instances.editor1.getSelection().getRanges()[0];
clonedHTML = range.cloneContents().getHtml();
extractedHTML = range.extractContents().getHtml();

Expected result

clonedHTML and extractedHTML must be equal

Actual result

clonedHTML and extractedHTML are different

Other details (browser, OS, CKEditor version, installed plugins)

Reproducible in every browser starting from CKeditor 4.5.0

[kkrzton]

Running this code in a console:

var currentRange = CKEDITOR.instances.editor1.getSelection().getRanges()[0],
    clonedHTML = currentRange.cloneContents().getHtml(),
    extractedHTML = currentRange.extractContents().getHtml();

clonedHTML === extractedHTML;

there are some cases for which html differs - when part of the list or table is selected or source footer, e.g:

clonedHTML: <p>had three parts:</p><ol><li><strong>Command Module</strong> with a ca</li></ol>

extractedHTML: <p>had three parts:</p><ol><li> with a ca<strong>Command Module</strong></li></ol>

[Tomasz Jakut]

Probably that problem is caused by bug in extractContents implementation, described in ​my comment for #13842 issue.

[Jakub Ś]

Could be related #17030

[Jakub Ś]

#17033 was marked as duplicate.

TC from that ticket:

1. Put attached sample into ckeditor/samples folder and open it
2. Apply same styles and selections as in attached video:
   • This is some [sample text.] - apply yellow marker
   • This is [some sample text.] - apply underline
   • [This is some sample text.] - press Print Extract Contents button

Expected result

This is <u>some <span class="marker"><strong>sample text</strong>.</span></u>

[kkrzton]

See ​https://github.com/ckeditor/ckeditor-dev/issues/644.