v1.2.0 - BigQuery improvements, date/time improvements, nullable ints and more

Breaking changes:

• None

New features:

• Add Avro support for Redshift import #137 (vinceatbluelabs)
• Add support for BigQuery bulk export (to Avro, for now) #136 (vinceatbluelabs)
• Prefer pandas nullable integers for int fields #159 (cwegrzyn)
• Use dataset location for bigquery load job #162 (cwegrzyn)

Bug fixes / reliability improvements:

• Improve date/time hint handling based on better testing for:
  • Redshift #149 (vinceatbluelabs)
  • Pandas #148 (vinceatbluelabs)
  • Vertica #147 (vinceatbluelabs)
  • Postgres #146 (vinceatbluelabs)
  • MySQL #151 (vinceatbluelabs)
• Support more date/time hints during Pandas-based generation of CSV files #155 (vinceatbluelabs) (fixes #142 and #143)

Other updates:

• Loosen validation and mypy type requirements for date/time types, start increased testing #144 (vinceatbluelabs)
• Drop overly-complected Christmas tree tests #145 (vinceatbluelabs)
• Fix Pandas date formatting, add date/time integration tests #153 (vinceatbluelabs)
• MySQL time column to dataframe fix for Pandas 1.x, increase MySQL integration testing #152 (vinceatbluelabs)
• Ensure close() is called on result of pandas.read_sql(chunksize=...) #150 (vinceatbluelabs)
• Various dependency tweaks:
  • Loosen tenacity dependency #158 (cwegrzyn)
  • Restrict sqlalchemy to version 1.3.x #165 (cwegrzyn)
  • Allow newer PyYAML #164 (cwegrzyn)
  • Workaround for chardet dependency issue #154 (vinceatbluelabs)
• Tweak collections.abc import to resolve mypy issue #163 (cwegrzyn)