Forum Discussion
GraphQL mutation from Power Query
The design goal of Power Query M is to take data from a source and transform it into a tabular format that can be consumed by Excel or Power BI, or a Parquet-style dataflow. Nowhere in the design specifications is there a place for "talking back to the data source" (the available options are more like unintended consequences of some of the design decisions). Unlike SSIS and Informatica Power Query is more of a one-way street.
Your line
#"Method"="POST",will be ignored. Web.Contents switches from GET to POST in the presence of a Content payload.
I would recommend you read this article. Chris Webb's BI Blog: Why Is Power BI Running My SQL Query Twice? Chris Webb's BI Blog (crossjoin.co.uk)
It talks about SQL but the general idea is the same. Power Query cannot guarantee to run requests only once. That's not what it is designed for.
Interesting, thanks. When you say "That's not what it is designed for", are you suggesting the M language is not suitable for working with APIs? Or SQL Inserts/Updates for that matter? That's suprising- and quite a blow to my future plans. Surely these issues could be handled in a custom connector, no?
The design goal of Power Query M is to take data from a source and transform it into a tabular format that can be consumed by Excel or Power BI, or a Parquet-style dataflow. Nowhere in the design specifications is there a place for "talking back to the data source" (the available options are more like unintended consequences of some of the design decisions). Unlike SSIS and Informatica Power Query is more of a one-way street.
Truthfully, I've already been working on a VBA solution for this component of the process (ie the outbound traffic). I am consistently surprised by the power of M, and was just hoping to hear some magic. Thanks for chiming in on it, much appreicated.
