cifparse: Debian missing unistd include

[smoe]

Addresses https://github.com/RosettaCommons/rosetta/issues/31, @roccomoretti already had a first look.

[smoe]

I am not sure about how to read this - looks like it failed but on https://b3.graylab.jhu.edu/revision/rosetta:commits/20355 there is no log for me to inspect, other than the statement that the build on the M1 was cancelled. The email I received stated

commit [rosetta:main:62095](http://b3.graylab.jhu.edu/revision/rosetta:main/62095) [[diff]](https://github.com/RosettaCommons/rosetta/compare/dffc8b947ea41f1346fb98df62436525a33bcd8e...3e4fc2129edccfa8b8b6e78b27ceb21fe4860b18) by Steffen Möller broke:
[gcc-9.gcc.serialization.python37.PyRosetta.notebook](http://benchmark.graylab.jhu.edu/test/823480)
[02.02-Working-with-Pose-Residues](http://benchmark.graylab.jhu.edu/sub_test/282820028)
[linux.clang.integration.release_debug_no_symbols](http://benchmark.graylab.jhu.edu/test/823337)
[small_molecule_lattice_dock](http://benchmark.graylab.jhu.edu/sub_test/282738919)

but a quick look then pointed to "pip" not being installed - which sounds more like a feature than a bug ;-)

[roccomoretti]

The email is for a different result (https://benchmark.graylab.jhu.edu/revision?branch=rosetta%3Amain&id=62095) -- That's for the results of merging PR #39 into main (we do post-merge tests on all PRs as well). A quick look shows that the issues are either pre-existing, or are likely something on the test server side of things, independent of the PR.

[smoe]

Thanks for clearing up my confusion about this unistd.h patch after the string.h one! I still think that also the unistd.h patch is fine, have no idea what triggered the cancellation of the M1 build. If there is anything I can do, not having the credentials for anything, please tell me.