Forum Discussion

Quique's avatar
Quique
Frequent Visitor
2 years ago
Solved

Run Script for all Lakehouse Files / Lakehouse files storage account key

I would like to use the Notebook to run the same script for all files stored in the Lakehouse files. How can this be done? If there´s a need to get the storage account key for lakehouse files, where can this be obtained?

General
Q&A Show & Tell
  • Anonymous's avatar
    Anonymous
    2 years ago
    You can easily list all files in a Lakehouse folder from your notebook. Just make sure to connect the lakehouse in the left pane first. that will then become your default lakehouse in the notebook
     
    file_path = f"/lakehouse/default/Files/yoursubfoldername_here"
    lst = os.listdir(file_path)
     
    then you can iterate through the files and apply the same logic to each file in a loop. 
     
    for file in lst:
        --do stuff
     
    So the trick is not to run the same notebook on each file, but to read out the files in the notebook and the run the same logic on each file. 
     
    you could create a pipeline that reads out the list of files in a folder and then call a notebook and pass the file path as a parameter, but that just adds unnecessary complexity.
     
    if you're looking for the fully qualified path to where your files are in the lakehouse, then you can grab it from the properties on the Files folder

     

4 Replies

  • Anonymous's avatar
    Anonymous
    Not applicable
    2 years ago
    You can easily list all files in a Lakehouse folder from your notebook. Just make sure to connect the lakehouse in the left pane first. that will then become your default lakehouse in the notebook
     
    file_path = f"/lakehouse/default/Files/yoursubfoldername_here"
    lst = os.listdir(file_path)
     
    then you can iterate through the files and apply the same logic to each file in a loop. 
     
    for file in lst:
        --do stuff
     
    So the trick is not to run the same notebook on each file, but to read out the files in the notebook and the run the same logic on each file. 
     
    you could create a pipeline that reads out the list of files in a folder and then call a notebook and pass the file path as a parameter, but that just adds unnecessary complexity.
     
    if you're looking for the fully qualified path to where your files are in the lakehouse, then you can grab it from the properties on the Files folder

     

    • Quique's avatar
      Quique
      Frequent Visitor
      2 years ago

      UPDATE: It worked now, thanks again! There were a couple of problems: the delta tables I was creating had blank spaces in their names. Also, when creating the delta tables, I changed to use the qualified path, instead of the relative path. 

    • BoSe's avatar
      BoSe
      Frequent Visitor
      2 years ago

      For me it only works when i use the /lakehouse/default/Files... Path.
      However when i try to use it with the abfs path, do get the following Error:

      FileNotFoundError: [Errno 2] No such file or directory: 'abfss://.../input'

       

      file_path = f"abfss://.../input"
lst = os.listdir(file_path)
lst

       

      Any idea what causes that issue?

  • Quique's avatar
    Quique
    Frequent Visitor
    2 years ago

    Thanks! I'm sure this works and almost worked for me, but for some reason I keep getting this error Spark_Ambiguous_MsSparkUtils_UseMountedPathFailure. 

    I'll keep checking.