This represents a substantial and breaking set of changes to the Doltpy package. The main complaints with Doltpy historically have been:
muddling the Dolt SQL server interface with the Doltpy CLI wrapper
confused and poorly thought out write paths
verbose logging
lack of documentation
The steps taken to address this in this PR are as follows:
remove SQL server considerations completely from Dolt, renaming doltpy.core to doltpy.cli to indicate the narrow goal of that package
doltpy.core.{read, write} are now doltpy.cli.{read, write}, no longer use the Dolt SQL server, and are organized as write_pandas, write_rows, write_columns, and read interface is write_pandas, write_rows and write_columns
there is a new package called doltpy.sql that provides a identical read and write interfaces, though going through the server, as well as tools for managing the Dolt SQL server process in Python without interacting with the CLI
This represents a substantial and breaking set of changes to the Doltpy package. The main complaints with Doltpy historically have been:
The steps taken to address this in this PR are as follows:
Dolt, renamingdoltpy.coretodoltpy.clito indicate the narrow goal of that packagedoltpy.core.{read, write}are nowdoltpy.cli.{read, write}, no longer use the Dolt SQL server, and are organized aswrite_pandas,write_rows,write_columns, and read interface iswrite_pandas,write_rowsandwrite_columnsdoltpy.sqlthat provides a identical read and write interfaces, though going through the server, as well as tools for managing the Dolt SQL server process in Python without interacting with the CLIOpen Items:
doltpy.sqlDoltHubContext(how?)