Ryan May

Would be good to integrate with xarray. Relevant links: https://github.com/pydata/xarray/blob/master/xarray/core/variable.py#L1038 https://github.com/pydata/xarray/blob/master/xarray/core/alignment.py#L183 https://github.com/pydata/xarray/blob/master/xarray/core/indexing.py#L183

Since Python uses the C API, and only the C API, to write the file, there should not be a way to create a file that does this. (At least...

Latest was done in #140. Leaving this open for best and full collection.

Have at it! I believe @akrherz was open to adapting the REST API for the service if we found something lacking.

No worries! This looks like a nice start. The biggest things I would want to see is that this return things as a Pandas `DataFrame`, rather than doing things with...

The easy fix for this, based on the [pandas docs](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_fwf.html) is to pass `infer_rows=200`.

100% chance this is because `.datasets` is effectively a `OrderedDict` where we override `__getitem__`. Easiest fix is probably to also override `__iter__` to iterate over `values()`.

Hrmmm... now that I think about this with just a bit more thought, this is standard `dict` behavior. For this code: ```python d = {'a': 1, 'b':2, 'c':3} print(type(d['a'])) #...

👍 to updating the docstring. I can't think of anything other place unless we add some proper narrative documentation about `TDSCatalog` or an example explicitly about working with `.datasets`.

For reference: - [httpio](https://github.com/barneygale/httpio) - [s3io](https://github.com/malexer/s3io) - Stdlib `io` module has [BufferedIOBase](https://docs.python.org/3/library/io.html#io.BufferedIOBase)