Using flow to analyse document file names and apply term store metadata

[Anonymous]

I have an O365 platform, and in one of the site collections I have a document library with files that are uploaded using a standard naming convention: <Business Unit>_<ITIL Practice>_<Document Type>_<Document Description>. Examples below:

ITSM_KM_POLICY_KnowledgeManagementPolicy.docx
ITSM_CM_PROCESS_ChangeManagementProcessTemplate.pdf
ITSM_AM_PROCESS_ManagingOrganizationalAssets.pdf
ITSM_KM_GOV_KnowledgeManagementGovernanceFramework.pdf

The library has 3 managed metadata columns:

1) Business Unit
2) ITIL Process
3) Document Type

I also have a managed term store, with the following hierarchy:

Branch 1
Parent term 1: Business Unit
Child terms 1: ITSM, ITSB, ITSA

Branch 2
Parent term 2: ITIL Practice
Child terms 2: KM, CM, AM, EM

Branch 3
Parent term 3: Document Type
Child terms 3: Policy, Process, Gov, Training

I would like to set up a flow that runs whenever a file is added or modified in the library, and which compares each part of the filename with the relevant term store branch, and where it finds a match, to apply the matched term store value in the correct metadata column. Where it doesn't find a match, it leaves the value blank.

Any help would really be appreciated. Let me know if you need more information.

[v-zhos-msft · Accepted solution]

Hi @Anonymous ,

Where does the data of (Business Unit ,ITIL Process ,Document Type) stored, in SharePoint list or excel file?

I set the data via Compse action by myself on my side.

Then I have made a flow for your scenario, please check it for a reference.

The expression in the Compose is :

split(triggerBody()?['{FilenameWithExtension}'],'_')

The setting of Condition1~Condition3 are similar.

The expression in the Condition 1 is : outputs('Compose')?[0]

The expression in the Condition 2 is : outputs('Compose')?[1]

The expression in the Condition 3 is : outputs('Compose')?[2]

Best Regards,

Community Support Team _ Zhongys

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[Anonymous]

Thanks @v-zhos-msft, the logic of your flow makes sense to me and is actually super useful. However, the data of (Business Unit ,ITIL Process ,Document Type) is stored, not in a SharePoint list or excel file or hard-coded into the flow, but in the SharePoint term store. So I presume we'd need some sort of comparison clause here so that the flow:

1) Breaks down the file names into outputs
2) Searches the term store for matches to those outputs (or maybe just searches a parent term and all children, if searching over a whole term store isn't advisable)
3) When it finds a match, applies the term store value to the file properties (rather than applying a term locally\just in text form in the file properties, which would mean that we can't take advantage of SP's platform-wide reach when it comes to term stores to do things like sset up search refiners for terms when doing platform-wide searches).
4) When it doesn't find a match, leaving the column entry blank

[v-zhos-msft · Accepted solution]

Hi @Anonymous ,

Could you please share a screenshot of the data in your SharePoint term store.

I am afraid there is no way to get the data (Business Unit ,ITIL Process ,Document Type)from SharePoint term store via flow currently.

1 Add the data (Business Unit ,ITIL Process ,Document Type) in an array.

2 Split the file name.

3 Check whether the file name in the array.

4 Update the file metadata.

Best Regards,

Community Support Team _ Zhongys

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[Anonymous]

Thanks @v-zhos-msft, much appreciated 🙂