Loss of data during conversion (text from old "links to attachments")

[François POURRIER]

Hi,
The pages created using the legacy editor have been automatically converted to the new cloud editor.
I have noticed a significant loss of information during this conversion, as most of the references on my pages to attachments contained text (such as hyperlinks) that precisely described the content of the attachment (which the attachment name does not do).
In the converted version, all this text has disappeared and been replaced by the attachment name.
This is a major issue, and not all pages can be reverted to the old format, if only to allow me to extract this text from the link tags so as not to lose this information during conversion.
Is there a solution?
More generally, this is a significant regression in the editor’s functionality.

[Kris Klima _K15t_]

Hi @François POURRIER 

I wonder if Find and Replace by Easy Apps might help.

You can replace parts in URLs...

(there's also Redirects from the same vendor)

Usefulness of these apps would depend on the nature of your links but perhaps there's something you can work with.

[François POURRIER]

Thanks, I'll take a look at Find and Replace / Redirect.

[Stavros_Rougas_EasyApps]

@Kris Klima _K15t_ thanks for mentioning our apps.

Salut @François POURRIER. Before I start a tip (warning). Changing the name of the URL that is visible and the underlying URL can be separate. This helps explain why the below can seem overly complicated. 

If you only want to change the text you can use 'Find and replace', it is part of our app Space Content Manager.

But the problem is you need to know the text, and it sounds like your problem is more complex or at least part of it is.

So 'Find and replace (URLs)', another part of Space Content Manager, is useful as well. You can change the underlying URL (in bulk) and the text (individually).

I bet using both would make sense.

Based on my experience I would use both initially as search tools. Do a bunch of queries and see what you are dealing with. I personally would look at some of the more complicated pages individually to help understand or confirm the underlying issue before doing a lot of bulk cleanup.

Everything has a preview so searching and refining your searches have no impact. It's counterintuitive but you can use a bulk editor like Space Content Manager to also see problems that are best changed page by page. Yet without such a tool knowing which pages can be hard to identify.

[Stavros_Rougas_EasyApps]

@François POURRIER our Redirection app is for pages, individually or in bulk to redirect to other Confluence pages or any URL for that matter.

[Stavros_Rougas_EasyApps]

@François POURRIER as I read all the comments I'm left wondering if attachments are impacted. We have an Attachment Manager app as well. Since everything comes with a 30 free trial you don't have much to lose.

Feel free to contact us. Add a link to this page, I'm happy to help partly because these use cases are how we learn to improve the app.

[Arkadiusz Wroblewski]

Hello and welcome to the Community @François POURRIER 

Yes, that is more than just a formatting issue.

If the conversion replaced your descriptive attachment link text with only the file name, then real content was lost.

I would first check Version history and see whether the pre-conversion version of the page is still available. If yes, that may be the easiest way to recover the missing text.

I would also avoid editing too many of those pages further until you know how much can still be restored.

Main questions for me would be:

how many pages are affected ?

do the old versions still exist in page history ?

If this affected many pages, I would raise it with Atlassian as well, because that is a real data-loss concern, not just an editor change...

[François POURRIER]

I have around 50/100 pages affected, with old version still in page history (but the version right before the conversion is not in the page history)

I've juste raised it with Atlassian.

[Mirek]

Welcome @François POURRIER 

It is not good that you are impacted by this legacy migration task...  Unfortunately, I am afraid that there is no easy / automatic way to restore this text after conversion. Your best options to try are to recover it from page history or older versions via the API, or manually re-add it. For large-scale impact, contacting Atlassian support or scripting a recovery using historical data may be necessary... 

How many pages you think might be impacted? 

[François POURRIER]

Thank you !

I believe there are between 50 and 100 pages affected.