fix(csharp): extract spog org id from cluster path

[msrathore-db]

Summary

• Fixes silent workspace-context loss on SPOG/custom-URL hosts when the C# Databricks driver connects with an all-purpose-compute HiveServer2 path like sql/protocolv1/o/<workspace-id>/<cluster-id>.
• PropertyHelper.ParseOrgIdFromProperties now extracts the workspace ID from the /o/<wsid>/ path segment as a fallback when ?o=<wsid> is not present.
• Priority order matches the JDBC/Python/Go fixes: ?o= query parameter wins, then /sql/protocolv1/o/<wsid>/<cluster-id> path segment.

Why

For all-purpose cluster Thrift paths on SPOG, PoPP can route the Thrift request from the /o/<workspace-id>/ path segment. Other HTTP requests built from the same connection, including SEA/telemetry-style requests, do not naturally carry that path context. Without the x-databricks-org-id header they can be routed without workspace context and redirected to /login.

What changes

• Adds a compiled cluster-path regex matching (?:^|/)sql/protocolv1/o/(\d+)/[^/?]+.
• Applies the fallback to both SparkParameters.Path and AdbcOptions.Uri inputs.
• Adds unit coverage for cluster path extraction, leading slash, ?o= precedence, warehouse-path non-match, and URI path extraction.

Test plan

• dotnet test csharp/test/AdbcDrivers.Databricks.Tests.csproj --filter FullyQualifiedName~PropertyHelperTests
• dotnet build csharp/src/AdbcDrivers.Databricks.csproj
• Same SPOG cluster behavior has been live-tested on sibling JDBC/Python/Go fixes against peco.azuredatabricks.net all-purpose cluster path.

This pull request and its description were written with assistance from Claude Code.