Some sites use a method code of 'x', which was breaking the json parser, which interpret '_x' as a duplicate column following pandas merge convention.
This fix removes update_merge() entirely (I might regret this 💣 💥) and adds the functionality into the json parser itself. If nothing else, this should improve performance.
Some sites use a method code of 'x', which was breaking the json parser, which interpret '_x' as a duplicate column following pandas
merge
convention.This fix removes
update_merge()
entirely (I might regret this 💣 💥) and adds the functionality into the json parser itself. If nothing else, this should improve performance.