This Power BI report lists reports from a couple of workspaces in my Power BI tenant.
A Fabric Notebook calls the Power BI REST API to get a list of all reports in each workspace, then for each report gets report metadata and exports an image of the first report page into a PNG file. It then resizes the PNG files to create smaller thumbnails, converts them into base64-encoded images split into small chunks, and saves the metadata and base64 chunks into a table in a Lakehouse.
The data is imported from the Lakehouse into the semantic model. A measure is used to concatenate the base64 chunks into an image URL, so the thumbnails can be displayed in the Button slicer and Image visuals.
🔥 This is an interactive Power BI report embedded into this page using Power BI Embedded
Power BI REST API end points used to prepare the data
https://learn.microsoft.com/en-us/rest/api/power-bi/groups/get-groups – returns a list of workspaces the user has access to.
https://learn.microsoft.com/en-us/rest/api/power-bi/reports/get-reports-in-group – returns a list of reports from the specified workspace.
https://learn.microsoft.com/de-de/rest/api/power-bi/reports/get-pages-in-group – returns a list of pages within the specified report from the specified workspace.
https://learn.microsoft.com/en-us/rest/api/power-bi/reports/export-to-file-in-group – exports the specified report from the specified workspace to the requested file format.
https://learn.microsoft.com/en-us/rest/api/power-bi/reports/get-export-to-file-status – returns the current status of the Export to File job for the specified report.
https://learn.microsoft.com/en-us/rest/api/power-bi/reports/get-file-of-export-to-file – returns the file from the Export to File job for the specified report.
https://learn.microsoft.com/en-us/rest/api/power-bi/datasets/get-refresh-history-in-group – returns the refresh history for the specified dataset from the specified workspace.
PySpark (Python) Notebook
https://github.com/avatorl/Fabric/tree/main/metareport
DAX
The Notebook creates a table with report metadata and a table with base64-encoded images split into maximum 8,000-character chunks. The chunks are needed as a workaround for Power Query (32K characters) and the Lakehouse SQL endpoint (8,000 characters) limitations.
The measure returns either a report thumbnail (base64-encoded PNG image) or an SVG image with the “No Image” label when there is no thumbnail available.
Image =VAR Base64Body =
( CONCATENATEX ( ‘Reports’, ‘Reports'[Value], “”, ‘Reports'[Attribute], ASC ) )
VAR _Result =
IF (
Base64Body = “”,
“data:image/svg+xml;utf8,<svg xmlns=’http://www.w3.org/2000/svg’><rect width=’100%’ height=’100%’ fill=’#f0f0f0’/>” & “<text x=’50%’ y=’50%’ text-anchor=’middle’ dominant-baseline=’middle’ fill=’#999′ font-family=’Arial, sans-serif’ font-size=’32’>No Image</text>” & “</svg>”,
“data:image/png;base64,” & Base64Body
)
RETURN
_Result
Change the data category of the measure to Image URL:
Add this measure as an image into one of the visuals that support embedded images (I used the Button slicer and Image visual):
