Does this close #5630 fully?

@jorisvandenbossche - OK, fixed

I think it fully closes #5630.

As a second thought, I'm not sure if it makes sense to add support to Quandl/python or pydatastream directly in web.DataReader. It will amount to

import quandl
return Quandl.get(...)

or

from pydatastream import Datastream
DWE = Datastream(username="...", password="...")
return DWE.fetch(...)

What do you think?

I also think including this in pandas is not needed. First you raise the point of the non-free data source. And if the packages have a nice interface to get these data as you show above, they can certainly be a package on their own. We are also thinking of rather moving some of the data readers included in pandas to separate packages, rather than including more of them (see #8961). It is very easy to install eg the quandl package.

Then, for this PR: we always ask to squash the commits into one commit (you can do this with git rebase -i upstream/master) for small changes. Can you do this?
If problems, just ask, or I can also do it on merge.

Just to be sure: after rebase I'm getting another branch (as the first one was already pushed to github).
Do I need to submit another PR from the new branch?

Normally, the command I gave does rebase your existing branch (doc_ecosystem_api), so you can just push it again to your fork, and this PR will automatically be updated (you will only have to 'force push': git push --force origin doc_ecosystem_api

Aha, force push! Thanks, I did not know about it

Done then

OK, thanks!

And if you are interested more in general in the data readers, certainly chime in on the linked issue #8961