Gspread-dataframe

-------------------
- Bump to v4.0.0. [Robin Thomas]
- Removed test of prefix= param removed in Pandas 2.x. [Robin Thomas]
- Support drop_empty_rows and drop_empty_columns parameters, True by
default (62) [Robin Thomas]

Fixes 61.
- Fixes 54. Worksheet names that are valid cell references no longer
cause Sheets API errors. (58) [Robin Thomas]
- Corrected Travis CI badge url. [Robin Thomas]
- Keep test credentials out of sdist. [Robin Thomas]

-------------------
- Bump to v3.3.1. [Robin Thomas]
- Update gspread_dataframe.py (53) [Rulowizard]

Fixes 52. Update the variable WORKSHEET_MAX_CELL_COUNT from 5000000 to 10000000.
- Fixes 50. Take the row and col arguments into consideration when
resizing the worksheet (51) [Motin]
- Removed stray endline. [Robin Thomas]

-------------------
- Bump to v3.3.0. [Robin Thomas]
- Handle all cases of header row writing/reading (44) (47) [Robin
Thomas]

Fixes 44. Writes header rows properly for all cases of include_index, MultiIndex columns or index itself, and names for columns object levels or index levels.
- Added :param entries for row and col in docstring. [Robin Thomas]
- Remove 3.10, still not released on travis. [Robin Thomas]
- Remove pypy from build matrix, add 3.10. [Robin Thomas]
- Try default html sphnix builder for RTD. [Robin Thomas]
- Removed pandas version checker because pandas 0.24.0 was a LONG time
ago. [Robin Thomas]
- Try autodoc mock imports. [Robin Thomas]
- Push proper dir onto sys.path. [Robin Thomas]
- Try pushing onto sys.path to get automodule to work for RTD. [Robin
Thomas]
- Rename to yaml file. [Robin Thomas]
- Add readthedocs yml file. [Robin Thomas]

-------------------
- Bump to v3.2.2. [Robin Thomas]
- Fixes 40. Import Cell from main gspread package to avoid breakages
with gspread>=5.0.0. [Robin Thomas]

-------------------
- Bump to v3.2.1. [Robin Thomas]
- Fixes 37. [Robin Thomas]

Fixes 37. Discovered that Sheets API applies rowCount first, and THEN (38) checks cell limits, so that if new rowCount * existing colCount > 5000000, then worksheet exceeds limit of 5000000 cells and the whole resize operation is aborted. Solution is to determine if such a condition would occur and then issue the smaller columnCount first as a separate call to reduce
cell count. Full test coverage added.

-------------------
- Bump to v3.2.0. [Robin Thomas]
- Fixes 35. Use `to_numpy('object')` to perform needed type conversions
(36) [Robin Thomas]

Fixes 35. Use `to_numpy('object')` to perform needed type conversions (36) on frame and index values before attempting to update cell values in the worksheet. Now requires Pandas>=0.24.0 instead of >=0.14.0.