See also commit 5e0f06d3.

Joseph Weston authored 9 years ago and Christoph Groth committed 9 years ago

The fact that this test was broken was not seen before because
in Python 2 dictionaries have the same sort-order on subsequent
program runs. This is not the case for Python 3.

This is useful when the source is held in a git repository but this
repository is not an usual kwant repository.

This still does not allow to import kwant under other names:
http://thread.gmane.org/gmane.comp.python.cython.user/13809

Once we depend on Python 3.5, and "proper" relative cimports are supported by
Cython, we could add a test for exclusive use of relative imports.

Conflicts:
	kwant/physics/leads.py

Joseph Weston authored 9 years ago and Christoph Groth committed 9 years ago

Before this commit, if there was a lead for which
the inter-cell hopping matrix was the zero matrix, then
kwant.physics.modes() returned arrays with uninitialized memory.

It was not previously made explicit anywhere in the docs on how
the leads were numbered. This commit adds a line to the docstring
of `kwant.builder.Builder.attach_lead` which explains the meaning
of the return value. It also adds a few lines to the first tutorial
explaining the numbering.

This undoes 00b0536c, reaching the aim of clearer error messages in a more
systematic way.