158
I have a document library some filename named like [filename] + [timestamp].
I'm trying to build a flow to check all filename, remain newly file, and delete old files.
My scenario:
- Collect all files name
- Split files name to [filename] [timestamp]
- If [filename] same Compare [timestamp]
- Delete old [timestamp] file
Please teach me how to implement this scenario.
Is it possible to use [send HTTP request to SharePoint] action?
Any information will be appreciated! Thank you!
Hi, @grantjenkins
I just test the flow and it works perfectly!!
Thank you for teaching me and explaining it patiently.
Thank you so much!!
Hopefully this will get what you're after.
I've got the following files in my Timestamps folder.
See full flow below. I'll go into each of the actions.
Get files (properties only) orders the results by Filename (FileLeafRef) in descending order so the most recent file (based on timestamp) are first. By default, Get items only returns the first 100 items. If you have more than 100 items then you can set the Top Count to 5000 (max) to retrieve up to 5000. The Order By expression is below.
Share
Report
All responses (
Answers (
FileLeafRef desc
Select gets a list of all the file names up to the timestamp (if no timestamp then it will return the full name).
slice(item()?['{Name}'], 0, lastIndexOf(item()?['{Name}'], '_'))
Apply to each iterates over the filenames using the following expression to remove duplicates.
union(body('Select'), body('Select'))
Filter array uses the following expression to retrieve the files that match the current file we are iterating over.
slice(item()?['{Name}'], 0, lastIndexOf(item()?['{Name}'], '_'))
Apply to each file to delete iterates over each of the files retrieved in the Filter array, except for the first file (the most recent file we want to keep). It uses the following expression.
skip(body('Filter_array'), 1)
Delete file uses the following expression so it deletes the current file we are iterating over.
item()?['{Identifier}']
After running the flow, we would have the following output.
Hi, @grantjenkins
Thank you for replied and correcting my mistake.
I altered my sample above, thank you so much.
1. I need to look at the entire timestamp including the year (12 string), and only keep the latest one.
2. BI file don't have duplicate file.
3. AWS.txt is a file but without timestamp, so in this case, I need to skip it.
From your example above, you aren't looking at the year in the filename. What if you had two files with the same name but a different year. For example "PA_Test_2023...", and "PA_Test_2024...". Would you consider these to be two different files and need to keep both, or would you rather look at the entire timestamp including the year, and only keep the latest one?
Also, your BI_Test file has AWS after the timestamp. Is this something that can also happen? And if so, could you also potentially have underscores _ after the timestamp?
#1
Michael E. Gernaey
566
Super User 2025 Season 1
#2
David_MA
516
Super User 2025 Season 1
#3
stampcoin
492
