Filtering files through 'Get files in folder ' action in PAD

[Anonymous]

currently I am working on files using power automate desktop, so I have the folder having 3000+ files(all pdfs) and out of it, I need to retrieve only 187 files (all pdfs) by filtering it by only the first 9 characters of the files (for ex: ABC-1234). the full file names is ABC-1234 v4.0.pdf. 

This first nine characters will be unique for each file. How can I give the filter option in the get files in folder action. Or is there any options available to retrieve the files by partial names ?

Please suggest me. Your help is much needed.

 

I want to give it as in screenshot for 187 files in a single field. Looking forward for the solution. 

[Lipsitz]

Loop will do it

[Anonymous]

out of the full name, can we filter only by initial some files of the files. Because I tried if loop, that is if the folder contains ABC-1234, then copy that file, but it is not working. 

In Get files in folder, we need to give the full file name in order to retrieve, but I cannot give the full file name, because the version may change frequently. So, I should be able to filter the files only by initial nine values of the file.

How do I do that?

[fraenK]

You first get all files in a folder as a list and then use for each to go through each item in the list and check if the filename contains your filter criteria. If yes, then do whatever you want to do. Really simple. 

[fraenK]

Alternative: if the file names follow a common scheme, then you can convert the list to text and extract the matching file paths using a regular expression. Should be much faster than for each.

[Anonymous]

I tried exactly what you mentioned, I have created 100 files in the list and checked each file using 'for each'. If it contains the specific file name, it should move to the specific folder. Likewise, there are 7 folder and subfolders. So according to the file name, it should go to the specific folder. Approximately it takes 1 minute per file to check the entire 100 files list.  100 files it takes 100 mins which is too long I feel.

Is there any solution? or is there any option to search for the file name in the folder generally as a whole(not as a each item) like what we do manually, just go to the folder and search for the file name.

[fraenK]

See faster alternative above. But I doubt that it takes 1min per file. Don't run it in edit mode!

[Anonymous]

It takes 1 min because, I have created nearly 25 separate lists for 25 folders so each file has to check with 25 lists and do the necessary action.

Now, I tried like this, is there any possibility to provide more than one value in the second operand of the 'if' action using AND operator, then my job will be easy.

and how to check whether it is running in edit mode!

[Anonymous]

and how to check whether it is running in edit mode!

[fraenK]

Why 25 lists? Get files in folder works for subfolders as well. Why don't you use a regular expression for checking the filename? Edit mode is when you open the flow for editing

[Anonymous]

sure, I will check the regular expression

[Anonymous]

using the regular expression is little difficult as the file name has more than one hyphens.

Let me ask you one anything, is there any way to filter with only partial names using 'Get files in folder' action

instead of *. For ex: the file name is ABC-1234-v.4.0 HTM - xxxxxxxx xxxx xxxx xxx.pdf and wants to filter it by only ABC-1234.

[fraenK]

Please post a few filename examples

[Anonymous]

ABC-1234-v.4.0 HTM - xxxxx x x x x xxxxxx.pdf

REW-4437-v.1.0 xxxxxx xxxx xxx x x x x x x.pdf

CVF-2345-v.1.0 xxxx x x x x  xxxxxx  x.pdf

LKJ-1470-v.1.0 HTM - xxx xxxxx   xxxxx xx x.pdf

The above is the file name examples and version number will be changing every time so should filter by first nine values of the file. ie ABC-1234.pdf, REW-4437.pdf, CVF-2345.pdf, LKJ-1470.pdf.

[fraenK]

You can easily check those file names using a regular expression:

\w{3}-\d{4}-v\.\d+\.\d+.+\.pdf

[fraenK]

For regular expression you need use parse text with the filename: https://docs.microsoft.com/en-us/power-automate/desktop-flows/actions-reference/text#parsetext

For many files this might take some time.

It might be faster to store the files list as text and then use parse text with a regular expression for the full path. The output is a list of all files which you want to process.

[burque505]

@fraenK's regex is robust - it will handle version numbers like v.303.34 as well. In the hope it may be of some use to you, here's a screenshot of my little test flow (using @fraenK's regex).

I find (and seem to recall seeing mention of this on StackOverflow) that ParseText can do strange things to the result list. For example, in the flow above I end up with (out of 6 files, not a large sample to be sure) with this (second message box):

Not nice 😀

'Remove item from list' is certainly handy, as it doesn't require a loop to remove all occurrences of the offending item. The empty string, as I believe @fraenK has noted elsewhere, is

%""%

Not very intuitive, eh?

And that cleans up the list.

[Anonymous]

I am able to get the files that matches particular pattern and it takes app 30 mins to complete.

Now I have list of files, from that I need to again segregate files to a specific folders.

For ex: the file

ABC-1234-v.4.0 HTM - xxxxx x x x x xxxxxx.pdf should go to first folder

REW-4437-v.1.0 xxxxxx xxxx xxx x x x x x x.pdf should go to the 1.1 folder

CVF-2345-v.1.0 xxxx x x x x  xxxxxx  x.pdf should go to the 2 folder

LKJ-1470-v.1.0 HTM - xxx xxxxx   xxxxx xx x.pdf should go to the 2.1 folder

Will the parse text work for this?

I want to search like in the screenshot. ie more than one file in 'Text to find' in 'parse text' 

[burque505]

Might I ask if the only letter prefixes are these?

• ABC
• REW
• CVF
• LKJ

If so, you might try splitting up your list again into four lists, 1 for Folder 1, 1 for Folder 1.1, one for Folder 2, one for Folder 2.1.

The ParseText regex should be simple for each list, since you already have only PDFs that match certain criteria in your first list, so something like this for Folder 1 (not tested):

ABC.+

with a similar regex for each folder.

You could run each regex on your list, yielding four new lists.

Each new list should then contain only those files that are supposed to go to the destination folder (if I am guessing correctly ABC*.pdf files go to Folder 1, REW*.pdf files go to Folder 1.1, CVF*.pdf  files go to Folder 2, and LKJ*.pdf files go to Folder 2.1) so further processing should be easy with a number of ways to do it (don't know about fast, though).

[Anonymous]

it is like ABC-1234 should go to 1 folder and ABC-1245 should go to the another folder. May be the first three letters are same but it has unique different numbers along with it. So it should check for the particular letters and numbers combination and accordingly it should move to the specific folder. like wise I have 25 ( folders and subfolders)