Isitfit

- feat: ec2 cost analyze now gathers a per-month version of the same displayed single-column table
- still doesn't show up by default in the end result unless `--verbose` is turned on. WIP
- add `regions_set` and `regions_str` as well as service column
- add date start/end
- apply the same binning class to redshift data
- implement new display/email functions for the binned data
- bugfix: `--ndays` was not really having any effect.. fixed
- enh: include ndays in redis cache keys
- enh: `dt_end` in binned report's last column should be incremented by 1 day when ndays<64 (boto3/cloudwatch bug?)
- bugfix: redshift cost analyze to also take into account `ndays`
- bugfix: more empty-data checks
- feat, cost analyze: change binning frequency depending on the ndays option

- bugfix: add init file in isitfit.migrations

- enh: each click group/command to have its own usage stats with its local parameters
- enh: click `cli_core:invoke_without_command=False`
- enh: move isitfit command from utils to `cli.click_descendents`
- feat: ping matomo on uncaught click exceptions. In isitfit command and group
- bugfix: several temporary files were still getting saved into `/tmp`. Fixed
- feat: concatenate `isitfit cost optimize` tables from ec2 and redshift into one
- bugfix: better handling of case of no redshift data
- tests: functional tests for cost was still clearing the cache files from the old deprecated locations
- enh: get logger from `isitfit.utils` instead of calling logging all over the place
- bugfix: `--filter-tags` to filter the redshift clusters just like it does on ec2

- enh: prompt for email if user forgot to include it to share results by email
- enh: add `--verbose` flag
- enh: add a global progress bar, hide other progress bars unless `--verbose` or `--debug` requested
- enh: clean up usages of `logger.{info,warning,...}` and `click.echo`
- feat: merge displayed reports of ec2 and redshift `isitfit cost analyze` commands
- enh: move pipeline of `cost optimize` (ec2, redshift) from cli.cost to cost.service similar to pipeline of `cost analyze`
- feat: move the display step of `cost optimize` from 2 separate pipelines (ec2, redshift), to one aggregator pipeline
- enh: gather all `isitfit cost analyze` code related to `ec2` in a single file `ec2_analyze.py`
- enh: same for `optimize` and `ec2` in `ec2_optimize.py`
- bugfix: when no redshift clusters yield optimizations, had a bug in the display step
- enh: gather `ec2TagFilter` and `Ec2Common` into `ec2_common.py`
- enh: gather code for `redshift_analyze.py` and `redshift_common.py`
- enh: gather code for `redshift_optimize.py`
- enh: rename the cloudtrail iterator classes + add functional tests
- bugfix: cloudtrail data was missing the event name + had the wrong region field name
- feat: add option `isitfit cost --ndays {analyze,optimize}` for users who want to run the numbers on the most recent n days
- enh: when pinging matomo, send values of options like `ndays`
- bugfix: `save-details` option was using `logger.info` whereas it should use `click.echo`
- enh: cloudtrail empty dataframe from `all region` to not set index
- feat: add migrations module, which will take care of running maintenance tasks required between upgrades
- enh: usage stats, gather options passed to `isitfit cost`
- enh: do not check version upgrade if command is `isitfit version`
- bugfix: merged ec2-redshift pipeline to display "no optimizations from ec2" if not available
- enh: use `/tmp/isitfit/` for the iterator region cache instead of `~/.isitfit`, since it's an expirable piece of data anyway
- enh: use `/tmp/isitfit/` folder as temp dir instead of throwing everything in `/tmp` and cluttering it
- enh: use `/tmp/isitfit/` for the ec2instances.info cache
- bugfix: if pypi.python.org is unreachable, just skip it silently instead of throwing an exception
- bugfix: `isitfit cost analyze --help` was prompting for `ndays`. fixed
- enh: instead of prompting twice about sharing by email, just prompt once with blank to skip
- bugfix: upgrade `matomo_sdk_py` to set the visitor ID = user ID when tracking usage stats
- enh: ping matomo specifically on users sharing results by email

- enh: add `--save-details` to `isitfit cost analyze` to save details behind numbers of EC2 to CSV files

- bugfix: `isitfit cost optimize` was not recommending anything for `idle`