Removing time strings from word document

[LBW]

I have videos that are transcribed into a word file. For each new speaker, there is a time stamp and their name.

This makes it very difficult to read. I would like to automatically remove the time stamps and replace them with a blank space.

Here is an example of the text

0:9:10.310 --> 0:9:18.20

Jane Doe

OK, it's doable only if we approach it as you know.

0:9:19.100 --> 0:9:27.180

Jane Doe

Thinking through what is it that actually ...

0:9:28.220 --> 0:10:2.730

Jane Doe

If you think about what is it that ...

I am extremely new, so as much details as possible would be greatly appreciated.

[schwibach]

If you have the input available as a string, it could be done.

If it's just inside a Word document it will not be possible to do just with Power Automate's tools.

[grantjenkins]

A few questions:

1. Will the transcript files you're trying to fix be copied to the same folder each time?
2. Will the transcript files be located in SharePoint or OneDrive?
3. Is there any other content in the Word Documents apart from the transcript text?
4. Do you have access to run the Premium connector: Populate a Microsoft Word Template?

[LBW]

Hi Grantjenkins,

First of all, thank you for your reply.

To answer your questions:

1. Yes

2. I can do either

3. No other content than the text

4. I believe so.

[grantjenkins]

Ok - I'll try to get something for you tomorrow (almost midnight for me where I am). I've got a working solution, but just want to streamline it slightly then will reply here with screenshots, details, etc. to you.

The output would be something like the following (note that I'm using Office dark mode).

[LBW]

Wow, this is great!!!!

[grantjenkins]

Ok - here we go 🙂

Note that there is a lot of complexity in how we build the flow, so it's important that you keep the default names (as I have) for all the actions you use.

All you will need to do is put your original document into a folder and it will create a new document in another folder.

In my OneDrive My files I've created a folder called Video Transcripts and within that I've created four more folders.

1. Input - This is where you will put your original transcript Word documents
2. Output - This is where the new transcript Word document will appear after the times have been removed
3. Template - This will hold the Word Template that we use to create a new Word document
4. Zip - This will hold the unzipped Word documents temporarily so that we can get the content from the document

Firstly, we'll create a new Word Template. We only have to do this once but required as we can't create a new Word document without a Template. To create a Word Template, you will need to create a new Word document in your Microsoft Word on your Desktop (not in the browser).

If you don't see Developer on the Toolbar (as shown below) then you'll need to add it.

To add it, go into File > Options > Customize Ribbon, then select Developer and click on OK.

     

In the Word document I've just typed Video Transcript and made the font larger. Then add a Plain Text Content Control by clicking on the icon shown.

We then want to give our Plain Text Content Control a nice name that we can refer to within Power Automate, and a couple of other settings. Make sure you've selected the Plain Text Content Control in the document then click on Properties. Then enter a Title and Tag (I've used Content for my example) and check the two boxes shown.

That's all we need to do with our Word Template. Now save it with whatever name you want. Then copy it into your Template folder we created before. I saved my template as Video Transcript Template.docx.

Ok, now onto our Power Automate (Flow). Below is the full flow. I'll go into each of the actions.

The trigger is set to run when a new file is added to our Input folder within our Video Transcripts folder.

A Word Document is actually made up of multiple files that are zipped up into a .docx file. To get to these files we need to unzip the Word Document. We use Extract archive to folder for this using the Path property from the trigger. I unzip it to the Zip folder. It will also create a folder within the Zip folder using the Id of the document which is unique so if the flow is running on more than one document, they won't affect each other.

Next, we get the content from the appropriate unzipped file using a Filter array with the following expression:

@And(startsWith(item()['Name'], 'document'),endsWith(item()['Name'], 'xml'))

Then we get the file content from the Filter array using the following expression:

first(body('Filter_array'))['Id']

Next, we extract out all the text that makes up our document content. This will include the time, person and what they spoke. We convert the file content to xml then run xpath across it using the following expression.

xpath(xml(outputs('Get_file_content')?['body']), '//*[name()=''w:t'']/text()')

Now we check to see if each line of text contains '-->' which is part of the time text and replaces it with a new line character. The expression we use here is:

if(contains(item(), '-->'), decodeUriComponent('%0A'), item())

We then join all the lines of text with a carriage return using the following expression:

decodeUriComponent('%0D')

This should give us the final content we want to add back into a new Word Document. We use Populate a Microsoft Word Template using the Word Template we created earlier and passing in the output from our Join. This won't actually give us a document, but it will give us the Content we need to create one in the next step.

Now, we create our new Word Document using Create file. The Folder Path will be our Output folder. And the Name will come from the Trigger action.

We now have our new Word Document, but still need to remove the Id folder we created when we unzipped the original Word Document. There isn't a Delete folder option in OneDrive or SharePoint (no idea why), but we can use the SharePoint Delete item action. But before we use this, we need to get the Id of the folder.

Get folder metadata will get us the Id of the temporary Zip/Id folder we created. Note that since this is a SharePoint action, you'll need to paste in your OneDrive URL for the Site Address. And for the File Identifier we'll need to build up the path to include the Id from the Trigger, so it gets that particular folder.

And finally, we can use Delete item, again pasting in our OneDrive URL, and using the ItemId from our Get folder metadata action.

Hopefully that gives you what you're after. A lot to setup - if you have any issues, let me know.

Document before the conversion:

Document after the conversion:

[LBW]

Sorry for taking a long time to reply. Is there any way to do this without the Premium version?

[grantjenkins]

Not that I know of - which is why I asked if you had access to the Word Premium connector 😉