I have a requirement to read a CSV file from a Windows environment. This CSV file has the same format across all endpoints, and I need to create an analysis that can read the values and generate a report in Web Reports for the API calls.
Could someone refer me to any relevant scenarios on how to read the csv file in same same format?
In rows and columns format instead of single cell value?
Hello! As somewhat suggested above, it would be helpful for us if you’re able to provide an example of the content and structure of the file, along with an example desired output. You don’t have to provide a real-world example…just one that is representative.
On all Windows systems where SQL is installed, there is a CSV file named backupinfo_hostname.csv, where hostname is dynamically replaced with the computer’s hostname. The file contains backup information. I need an analysis that reads all these CSV files in the same format and produces a consolidated report with consistent rows and columns, so that I can use it in Web Reports as a consolidated report for all servers.
| RDBMS | Hostname | Servername | Instancename | DbVersion | Edition | srvCollation | DateKey | DBName | state_desc |
|---|---|---|---|---|---|---|---|---|---|
| MSSQL | XYZ | XYZ | MSSQLSERVER | 15.0.4043.16 | Enterprise Edition: Core-based Licensing (64-bit) | SQL_Latin1_General_CP1_CI_AS | 12/13/2025 12:00:00 AM | admin | ONLINE |
| MSSQL | XYZ | XYZ | MSSQLSERVER | 15.0.4043.16 | Enterprise Edition: Core-based Licensing (64-bit) | SQL_Latin1_General_CP1_CI_AS | 12/14/2025 12:00:00 AM | admin | ONLINE |
Because analysis results are 'unordered', it would be difficult to align rows so that one 'RDBMS' entry gets assigned to the correct 'Instancename' entry. Where we have to return multiple related things we usually provide a single result delimited with a character, like the pipe operator we use in the Installed Software analyses - where a result might be formatted as a single string "ProductName|ProductVersion|InstallDate"
You could either keep the CSV rows as-is (comma-separated), or convert the commas to pipe or other delimiter in your Analysis. Then in the Web Report, consider something like 'SQLBackupInfo" to be a single cell with all those things pipe-delimited or comma-delimited in a single cell.
Otherwise, you'd have to switch to something like a fully custom report, where you build your own HTML table and use javascript or relevance to extract the fields from the Analysis back into separate columns in the results. You'd have to decide whether it's worthwhile to try to maintain that much logic in the report.
Is it one file per server that is customized to that server alone?
So server XYZ would have one file with just XYZ data,
and server ABC would have one file with just ABC data?
And the file name and location is in the same spot on every individual server?
You might be looking at something as easy as “lines of file”
PowerBi Can be used to import data using API. This was tested earlier as well. From webreports it would need custom report as highlighted by Jason
one file for server, lines of file can be used but when API is used it is typical to structure the data.
Thank you, Jason. I was able to use one of the custom scenarios to fetch the data as expected and successfully download the CSV. However, when using API the results are not as expected.
Might need to use some delimiters.