PBI Desktop changes the case of text fields

[Anonymous]

I am trying to debug a problem that PBI Desktop is changing the case of text fields.

Attached is the data and the PBIX file.

In the Transform Data, the  BEE_UNIQUEID field has 2 distinct values, ending in either VFm or VfM,

In the Table visual, Both values end in VFm 

Further, this causes JOIN issues where one to many join is interpreted as Many to Many. Right now I would settle for an explanation of how to work around this aberrant behaviour.

[Anonymous]

The PBI engine being case-insensitive does not explain how it is changing the text field between the transform Data and the Table visual.   One of the BEE_UNIQUEIDS is being changed into the other. There is no join happening.  I can only understand this if the PBI engine is indexing the case-variant values into a (incorrect) case-invariant index, then substituting the incorrect value are rendering time. Why not just shows case-invariant all the time if that is what is actually happening?

[Anonymous]

Read the blog on case sensitivity, and it appears it has a major impact on Salesforce UIDs.  Been using PBI for 6 years and it took till I started with Salesforce data to discover this major flaw in PBI. Also does not explain why the UI continues to show case sensitivity when the engine is not processing the same characters. However, does explain why PBI often shows Many:Many joins when I know the data is 1:Many with case sensitivity. Thanks to Theo.

[TheoC · Accepted solution]

@Anonymous Hopefully it provided some insight. 

[TheoC]

@Anonymous 

Hopefully this article explains the challenges with case sensitivity https://blog.crossjoin.co.uk/2019/10/06/power-bi-and-case-sensitivity/ 

Case sensitivity is lost on the load which causes the challenge. Some workarounds are captured in Chris Webb's article.

All the best.

Theo

[Anonymous]

Any idea how to attach PBIX and CSV files?

[Anonymous]

PBIX and CSV Files are at

https://1drv.ms/u/s!AopNvIEzQnBagsMP2jFFpeoyK3OKSg?e=xHMJ5M