CHANGELOG

# [0.0.4]
- Only run CREATE TABLE statements once, don't track field types afterwards either.
- Run COPY INTO only for new staged files (effectively no difference given Snowflake tables skip already copied files).
- Produce item from Scrapy pipeline even if exporter raises an exception.

# [0.0.3]
- Start normalizing column/table identifiers in accordance with https://docs.snowflake.com/en/sql-reference/identifiers-syntax.html (unquoted object identifiers).
- Quote fields in SELECT from staged file.
- Fix broken `SnowflakeStageExporter.clear_stage(...)`.
- Fix `float` being mapped to "NUMBER" - map to "FLOAT" instead.
- `SnowflakeStageExporter.record_field_types` -> `SnowflakeStageExporter._record_field_types`.
- +`snowflake_stage_exporter.utils.normalize_identifier`.
- +`SnowflakeStageExporter.multitype = "VARIANT"`.

# [0.0.2]
- `SnowflakeStageExporter.export_item(...)` now returns expanded `table_path` used for passed item.
- `SnowflakeStageExporter.default_typemap` -> `SnowflakeStageExporter.typemap`.
- Prefixed most `SnowflakeStageExporter` instance variables with underscore to indicate them being intended as "protected".