Operation on text randomly merges words #11588

Siguardius · 2018-11-07T14:52:06Z

Describe the bug
When interacting (bolding, inserting links) with text block that has pasted content with prior formatting, block will randomly merge a few words in the text.

To Reproduce
Steps to reproduce the behavior:

Copy & paste text into the text block directly from word editor (LibreOffice in my case)
Bold a word
Enjoy the bug

Expected behavior
A few worlds in the text block will be merged.

Screenshots
BEFORE:

AFTER:

Desktop (please complete the following information):

OS: macOS
Browser: Chrome
Version: 70.0.3538.77

Additional context

Present in latest Gutenberg release
Avoidable by pasting and matching the style

ellatrix · 2018-11-08T08:42:51Z

Hi @Siguardius! I cannot reproduce the problem, we'll need some more information. Could you share what was logged in the browser console when you paste the text from LibreOffice?

Siguardius · 2018-11-08T14:17:00Z

@iseulde This is everything that was in the console.
www.pixelophobia.pl-1541686251826.log

designsimply · 2018-11-08T17:06:45Z

Noting an additional report of this from a duplicate at #11626 and their testing steps for reference:

Paste text containing soft line breaks (from Word or from a PDF)

Edit the pasted text

All instances of soft line breaks are now deleted

swissspidy · 2018-11-08T17:09:35Z

I just noticed the same when copying content which was on separate lines because each line was actually a <div>. So the HTML looks a bit like this:

<div>Some Text <a href="https://example.com"><u>Some link</u></a></div><div>Some Text <a href="https://example.com"><u>Some link</u></a><br /></div><div>Some Text <a href="https://example.com"><u>Some link</u></a><br /></div><div>Some Text <a href="https://example.com"><u>Some link</u></a><br /></div><div>Some Text <a href="https://example.com"><u>Some link</u></a><br /></div>

The "Received plain text:" part in the console properly shows the content on individual lines.

The "Processed HTML piece:" section shows almost the identical HTML, but without the divs. Leading to the pasted content being on a single line.

Haldaug · 2018-11-12T11:20:35Z

Is it possible that the culprit of this bug could be these changes? #10019

Haldaug · 2018-11-12T14:47:05Z

This bug can also be reproduced by manually inserting soft line breaks with "shift+enter" when you are editing paragraphs as HTML:

Haldaug · 2018-11-13T13:45:32Z

I found a temporary fix for this. I added this code after line 218 in the raw-handling code:

piece = piece.replace( /\n/g, ' ' );

It replaces all new lines with spaces. This is probably not ideal, but at least it lets you paste content from Word without randomly merging words again.

ellatrix · 2018-11-15T09:01:36Z

Yes, we'll have to handle this in the raw handling and/or creating rich text values, maybe along with other non meaningful whitespace. I think what should happen is stripping leading and trailing non meaningful whitespace entirely, and collapsing multiple non meaningful white space characters into a single one.

ellatrix · 2018-11-15T09:39:52Z

See also https://medium.com/@patrickbrosset/when-does-white-space-matter-in-html-b90e8a7cdd33.

What we have to do in rich text values seems to be: replace [\n\t] with spaces, then remove trailing and leading spaces, and reduce any other recurring spaces to one.

designsimply · 2018-11-15T16:07:54Z

Tested and confirmed using WordPress 4.9.8 and Gutenberg 4.3 by copying content from a word processing app to Gutenberg. I noticed the missing spaces match up with the line breaks in the Received HTML after pasting content:

LibreOffice document I copied from:

^{Tested with LibreOffice 6.0.6.2 on macOS 10.13.6.}

WordPress post I pasted into:

^{Seen at http://alittletestblog.com/wp-admin/post.php?post=15818&action=edit running WordPress 4.9.8 and Gutenberg 4.3 using Firefox 63.0.1 on macOS 10.13.6.}

Console output for reference:

Received HTML:

 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html>
<head>
	<meta http-equiv="content-type" content="text/html; charset=utf-8"/>
	<title></title>
	<meta name="generator" content="LibreOffice 6.0.6.2 (MacOSX)"/>
	<style type="text/css">
		@page { margin: 0.79in }
		p { margin-bottom: 0.1in; direction: ltr; font-size: 11pt; line-height: 115%; text-align: left; orphans: 2; widows: 2 }
		a:link { color: #0000ff }
	</style>
</head>
<body lang="en-US" link="#0000ff" dir="ltr">
<p style="margin-bottom: 0in; font-weight: normal; line-height: 100%; page-break-after: avoid">
<font face="Arial, serif"><font size="3" style="font-size: 12pt"><span lang="en-CA">But
we </span></font></font><font face="Arial, serif"><font size="3" style="font-size: 12pt"><span lang="en-CA"><i>should</i></span></font></font><font face="Arial, serif"><font size="3" style="font-size: 12pt"><span lang="en-CA"><span style="font-style: normal">
know that mythology uses contradictory, or at least inconsistent,
versions of the same story, to express alternative perspectives in a
something else blah blah process, rather than “a lie”.</span></span></font></font></p>
<p style="margin-bottom: 0in; font-weight: normal; line-height: 100%">
<br/>

</p>
<p style="margin-bottom: 0in; font-weight: normal; line-height: 100%">
<font face="Arial, serif"><font size="3" style="font-size: 12pt"><span lang="en-CA"><span style="font-style: normal">The
importance of the death of Zeus is that the story emerges exactly from
that point in time and cultural transformation in which Zeus is also
born and at that time it was familiar for a vegetative god,
representing nature blooming in spring and dying in autumn, to die
and be re-born within the immortality of the eternal round of the
year or yearly daemon.</span></span></font></font></p>
</body>
</html>� 

Received plain text:

 But we should know that mythology uses contradictory, or at least inconsistent, versions of the same story, to express alternative perspectives in a something else blah blah process, rather than “a lie”.

The importance of the death of Zeus is that the story emerges exactly from that point in time and cultural transformation in which Zeus is also born and at that time it was familiar for a vegetative god, representing nature blooming in spring and dying in autumn, to die and be re-born within the immortality of the eternal round of the year or yearly daemon. index.js:50:160096

Processed HTML piece:
 <p>
But
we <em>should</em>
know that mythology uses contradictory, or at least inconsistent,
versions of the same story, to express alternative perspectives in a
something else blah blah process, rather than “a lie”.</p><p>
The
importance of the death of Zeus is that the story emerges exactly from
that point in time and cultural transformation in which Zeus is also
born and at that time it was familiar for a vegetative god,
representing nature blooming in spring and dying in autumn, to die
and be re-born within the immortality of the eternal round of the
year or yearly daemon.</p>

ElYi · 2018-11-19T04:27:54Z

EDIT: As mentioned in other threads, this seems to happen in the desktop version of word and not the browser-based version. EDIT EDIT: Going back into the article after publishing, destroys it again with no spaces when doing it this way.
This also happens with Word. Am using the latest version in Office 365. It's infuriating. Happens in the first para when I paste over a whole article, or any block that I subsequently click on. Adding links does it too. Horrible bug. Must affect so many people and there are so many similar reports that keep getting closed before they're fixed.

An authoring tool is software that creates an eLearning lesson or course. Once created, the course is usually packaged in a software wrapper before being distributed to clients who then distribute the course via a Learning Management System. They are often very expensive and usually require some training and expertise to use. This, in turn, means they are not particularly flexible or agile and that any courseware created with one is often tricky to update. The nature of the eLearning industry throws other spanners into the works, too: company training courseware is often produced on a one-size-fits-all basis with multiple companies subscribing to it. As such it tends to be rather boring as all modern interactive features must get stripped away to maximise compatibility and market size. What’s typically left is a boring, clunky, unengaging lesson that’s delivered via a bloated SCORM file. It’s not very effective. Fortunately, whether you want an altogether new way of creating courseware or simply a way to enhance what you’ve already got, microlearning can considerably improve matters as the following authoring tool examples demonstrate.

ellatrix · 2018-11-19T08:40:29Z

I'll work on a fix asap.

Haldaug · 2018-11-19T10:36:05Z

FYI, my fix from here still works: #11588 (comment)

ellatrix · 2018-11-20T09:58:02Z

Could anyone test #12093?

sae13 · 2018-11-24T10:28:13Z

Could anyone test #12093?

how I can test that?

http://j.mp/2SdQ0Tv

designsimply · 2018-11-25T22:50:04Z

Closed #12293 as a duplicate.

designsimply · 2018-11-25T23:55:01Z

I noticed that #12093 was superseded by #12166. I tested #12166 and left a comment with my findings (#12166 (comment)).

@sae13 the main way to test pull requests would be to install a local development environment as outlined in https://github.com/WordPress/gutenberg/blob/master/CONTRIBUTING.md#getting-started and then use git to checkout the branch for the pull request that needs testing. Setting up a local development environment wouldn't be expected unless you are contributing code or very deep into testing though. 🙂

I also found another way to test but you need to be logged in to GitHub and know the URL (I think). Go to https://github.com/WordPress/gutenberg/tree/try/rich-text-newlines-to-space and click the "Clone or download" button. That will give you a version of the plugin running the try/rich-text-newlines-to-space branch and you can delete Gutenberg and install+activate the downloaded one in order to test.

gooma2 · 2018-12-07T16:46:18Z

Glad this is being worked on as it does this is every sentence pasted in. If you copy/paste to wordpad then it works fine, but having to do 2 steps with what was only 1 step before in Classic Mode seems counterproductive.

ellatrix · 2018-12-07T16:53:33Z

The fix will be shipped in WordPress 5.0.1.

sae13 · 2018-12-08T07:14:56Z

now button colors not working 🤣

MightyGadget · 2018-12-13T06:13:07Z

The fix will be shipped in WordPress 5.0.1.

Just to confirm that WordPress 5.0.1 has no fixed this. Copying from Word 2016 still has the problem

gooma2 · 2018-12-13T06:30:21Z

Oddly enough, it actually has fixed the issue for me. I'm also able to highlight blocks of words now unlike before. The only difference now is that if I try to edit an older article, it won't let me do it. It actually acts like the way 5.0 did when I tried to use Gutenberg.

…

On Thu, Dec 13, 2018 at 12:13 AM MightyGadget ***@***.***> wrote: The fix will be shipped in WordPress 5.0.1. Just to confirm that WordPress 5.0.1 has no fixed this. Copying from Word 2016 still has the problem — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#11588 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AYnSvDoxyw9HuUfnscoadUbB4XxLczZJks5u4fAGgaJpZM4YSlBe> .

swissspidy · 2018-12-13T06:46:55Z

@MightyGadget Since 5.0.1 is a security release only, it will be fixed 5.0.2 now. ETA is next week.

MatildaMarseillaise · 2018-12-13T22:00:25Z

I thought it was just me! I have upgraded to 5.0.1 and am finding that copying from Word into Wordpress draft article is also merging words.

Also, if I am scheduling a post rather than posting it immediately it doesn't seem to be doing the automatic spell check (which I would hope would pick up these merged words). Any idea where the spellcheck function is now?

gooma2 · 2018-12-14T15:14:41Z

I had that issue when updating to 5.0 but after 5.0.1 I was able to copy paste from Word into WP. Now it just won't let me edit older articles so back to Classic Editor. It'll get there.

kylelarkin · 2018-12-18T17:48:24Z

I can also confirm that this bug still exists in WordPress 5.0.1 with documents pasted from Word 2016. Copying text from a Word File into a Google doc and then into Gutenberg somehow bypasses the issue.

jersnav · 2018-12-19T05:08:55Z

It looks like some of my existing posts created before 5.0 are impacted by the missing space issue when the posts are converted to blocks. So the posts are correct until I go to edit the old posts and select "convert to blocks." At that point it seems like a few spaces in between words per paragraph disappear.

ChapeauDigital · 2018-12-19T09:34:04Z

I can also confirm that I get regular complains from my costumers, that pasting from Word is getting cropped of spaces between words.

designsimply · 2019-02-20T00:31:41Z

This issue should have been fixed in WordPress version 5.0.2 released on December 19, 2018. If you are still having trouble after updating to WordPress 5.0.2 or later, please leave a new comment there noting your WordPress version and any relevant details you think may be helpful (i.e. you are using a very old version of Word). Thank you!

desrosj added the [Type] Bug An existing feature does not function as intended label Nov 7, 2018

ellatrix added the [Feature] Paste label Nov 8, 2018

Haldaug mentioned this issue Nov 8, 2018

Soft line breaks are not replaced by space #11626

Closed

designsimply added the Needs Testing Needs further testing to be confirmed. label Nov 8, 2018

ellatrix mentioned this issue Nov 15, 2018

spaces between words are lost when OK post is edited #11853

Closed

ellatrix self-assigned this Nov 15, 2018

ellatrix added the [Feature] Rich Text Related to the Rich Text component that allows developers to render a contenteditable label Nov 15, 2018

ellatrix added this to the WordPress 5.0 milestone Nov 15, 2018

designsimply removed the Needs Testing Needs further testing to be confirmed. label Nov 15, 2018

designsimply mentioned this issue Nov 15, 2018

When you paste text from Microsoft word, it messes up all of the formatting #11834

Closed

ellatrix mentioned this issue Nov 20, 2018

RichText: Correct HTML formatting whitespace #12093

Closed

4 tasks

ellatrix mentioned this issue Nov 21, 2018

RichText: convert HTML formatting whitespace to spaces #12166

Merged

4 tasks

ellatrix modified the milestones: WordPress 5.0, 4.6 Nov 21, 2018

mtias modified the milestones: 4.6, WordPress 5.0 Nov 22, 2018

designsimply mentioned this issue Nov 25, 2018

Pasting from MS Word fuses words #12293

Closed

mtias modified the milestones: WordPress 5.0, WordPress 5.0.1 Nov 29, 2018

mtias added the [Status] In Progress Tracking issues with work in progress label Nov 29, 2018

youknowriad closed this as completed in #12166 Nov 30, 2018

designsimply mentioned this issue Dec 5, 2018

Pasting a long text from Word - brings along some errors. #11012

Closed

designsimply removed the [Status] In Progress Tracking issues with work in progress label Dec 12, 2018

designsimply mentioned this issue Dec 19, 2018

Copying from Word to blocks - Joining words #13028

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Operation on text randomly merges words #11588

Operation on text randomly merges words #11588

Siguardius commented Nov 7, 2018

ellatrix commented Nov 8, 2018

Siguardius commented Nov 8, 2018

designsimply commented Nov 8, 2018 •

edited

Loading

swissspidy commented Nov 8, 2018

Haldaug commented Nov 12, 2018

Haldaug commented Nov 12, 2018 •

edited

Loading

Haldaug commented Nov 13, 2018

ellatrix commented Nov 15, 2018

ellatrix commented Nov 15, 2018

designsimply commented Nov 15, 2018

ElYi commented Nov 19, 2018 •

edited

Loading

ellatrix commented Nov 19, 2018

Haldaug commented Nov 19, 2018

ellatrix commented Nov 20, 2018

sae13 commented Nov 24, 2018

designsimply commented Nov 25, 2018

designsimply commented Nov 25, 2018

gooma2 commented Dec 7, 2018

ellatrix commented Dec 7, 2018

sae13 commented Dec 8, 2018

MightyGadget commented Dec 13, 2018

gooma2 commented Dec 13, 2018 via email

swissspidy commented Dec 13, 2018

MatildaMarseillaise commented Dec 13, 2018

gooma2 commented Dec 14, 2018

kylelarkin commented Dec 18, 2018

jersnav commented Dec 19, 2018

ChapeauDigital commented Dec 19, 2018

designsimply commented Feb 20, 2019

Operation on text randomly merges words #11588

Operation on text randomly merges words #11588

Comments

Siguardius commented Nov 7, 2018

ellatrix commented Nov 8, 2018

Siguardius commented Nov 8, 2018

designsimply commented Nov 8, 2018 • edited Loading

swissspidy commented Nov 8, 2018

Haldaug commented Nov 12, 2018

Haldaug commented Nov 12, 2018 • edited Loading

Haldaug commented Nov 13, 2018

ellatrix commented Nov 15, 2018

ellatrix commented Nov 15, 2018

designsimply commented Nov 15, 2018

ElYi commented Nov 19, 2018 • edited Loading

ellatrix commented Nov 19, 2018

Haldaug commented Nov 19, 2018

ellatrix commented Nov 20, 2018

sae13 commented Nov 24, 2018

designsimply commented Nov 25, 2018

designsimply commented Nov 25, 2018

gooma2 commented Dec 7, 2018

ellatrix commented Dec 7, 2018

sae13 commented Dec 8, 2018

MightyGadget commented Dec 13, 2018

gooma2 commented Dec 13, 2018 via email

swissspidy commented Dec 13, 2018

MatildaMarseillaise commented Dec 13, 2018

gooma2 commented Dec 14, 2018

kylelarkin commented Dec 18, 2018

jersnav commented Dec 19, 2018

ChapeauDigital commented Dec 19, 2018

designsimply commented Feb 20, 2019

designsimply commented Nov 8, 2018 •

edited

Loading

Haldaug commented Nov 12, 2018 •

edited

Loading

ElYi commented Nov 19, 2018 •

edited

Loading