topic Re: European letters not supported on column names on Notebook + Lakehouse combination in Pipelines

European letters not supported on column names on Notebook + Lakehouse combination

TimoRiikonen — Fri, 12 Jul 2024 09:09:33 GMT

Hello,

SQL syntax has supported pretty well European, in this case Scandinavian, letters for the past 40 years.

The SQL engine in Lakehouse works also well.
I can query like this without issues SELECT [Päivitetty] FROM table.

And I haven't run into any issues anywhere in Fabric with Scandinavian letters on columns before this.

Percent sign (%) and space were rejected on column name in dataflow, but Scandinavian letters were supported in there.

But when I run the same command on Notebook I get PARSE_SYNTAX_ERROR from pretty deep in Scala:

And so far I have only figured out if I wish to have an error from [ sign or ä letter.

Why has the standard [] embrasing been given up?

Is there any way to code Scandinavian letters on the query, or do I need to rename all columns using 7 bit ascii?

In Fabric column names appear directly on the report as-is if you don't change every single one of them manually in Semantic model. Or at least I haven't found any way. It is not nice to change every _ as a space, "pct" as "%" and decrypt all Scandinavian letters. Just for one report it will be dozens of manual changes. And if you are unlucky you will reset Semantic model at some point and have to redo this work from scratch.

Re: European letters not supported on column names on Notebook + Lakehouse combination

Anonymous — Mon, 15 Jul 2024 03:34:43 GMT

Hi TimoRiikonen

For special characters, you can try the:

%%sql SELECT `Päivitetty` FROM Query

Regards,

Nono Chen

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

Re: European letters not supported on column names on Notebook + Lakehouse combination

TimoRiikonen — Mon, 15 Jul 2024 05:40:01 GMT

Right!

I had tested [], "" and '', but the correct answer was ``.

This different syntax may confuse other people besides me, so you think of using the standard SQL syntax and [].

But this solves it for me, thank you.