Pydra

- moving from Travis to Github Actions for CI
- fixing issues with `numpy.arrays` provided as input fields
- adding a combiner option for dotfiles and graphs plots
- refactoring the input classes (adding `FunctionSpec`)
- adding `MultiInpObj` and `MultiOutObj` for a possible types in `input/output_spec`
- fixing `copyfile_workflow`
- setting proper `Workflow.output_spec` using tasks' `output_spec`s
- setting `output_dir` as a parent directory for all files created using the output templates
- adding `requires` key to `output_spec`'s metadata in order to set conditions for creating a specific output field
- adding `input_spec` and `output_spec` as class attributes that could be rewritten in the init or in the child classes

- refactoring template formatting for `input_spec`
- fixing issues with input fields with extension (and using them in templates)
- adding simple validators to input spec (using `attr.validator)
- adding `create_dotfile` for workflows, that creates graphs as dotfiles (can convert to other formats if dot available)
- adding a simple user guide with `input_spec` description
- expanding docstrings for `State`, `audit` and `messanger`
- updating syntax to newer python

- refactoring the error handling by padra: improving raised errors, removing nodes from the workflow graph that can't be run
- refactoring of the `input_spec`: adapting better to the nipype interfaces
- switching from `pkg_resources.declare_namespace` to the `stdlib pkgutil.extend_path`
- moving readme to rst format

* Add ``pydra.tasks`` namespace package to enable separate packages of ``Task``s to be installed into ``pydra.tasks``.
* Raise error when task or workflow name conflicts with names of attributes, methods, or other tasks already added to workflow
* Mention ``requirements.txt`` in README

- removing the tutorial to a [separate repo](https://github.com/nipype/pydra-tutorial)
- adding windows tests to codecov
- accepting `None` as a valid output from a `FunctionTask`, also for function that returns multiple values
- fixing slurm error files
- adding `wf._connection` to `checksum`
- allowing for updates of `wf._connections`
- editing output, so it works with `numpy.arrays`
- removing `to_job` and pickling task instead (workers read the tasks and set the proper input, so the multile copies of the input are not kept in the memory)
- adding standalone function `load_and_run` that can load and run a task from a pickle file
- removing `create_pyscript` and simplifying the slurm worker
- improving error reports in errors flies
- fixing `make_class` so the `Output` is properly formatted

- fixing `hash_dir` function
- adding `get_available_cpus` to get the number of CPUs available to the current process or available on the system
- adding simple implementation for `BoshTask` that uses boutiques descriptor
- adding azure to CI
- fixing code for windows
- etelementry updates
- adding more verbose output for task `result` - returns values or indices for input fields
- adding an experimental implementation of Dask Worker (limited testing with ci)