* Add zero assignments to local variables in modcc output.

* Move `cv_policy` classes to new source files.
* Remove `cv_policy_flag::single_root_cv`.
* Allow all supplied `cv_policy` makers to be restricted a domain specified by a region: the policy is then applied per component, with the boundary points of the completion of the components included.
* Expose `cv_policy` domain in public interface.
* Add `components` function for splitting an extent into disconnected components. Add comment describing what is meant by connected in this context.
* Rename `super` region function to `complete`.
* Add region expression functions `complement` and `difference`.
* Add new locset expression function `boundary` which returns the extrmal points of each component of a region.
* Add new locset expression `cboundary` which returns the extremal points of the completion of each component of the region, which is what we need to close off CVs generated by region-restricted CV policies.
* Add locset expression `support` which returns the multiset support of the points in a locset, i.e. all points in the locset but with multiplicity one.
* Replace hodgepodge of `join` expressions in `cv_policy` implementations with locset `support` and `sum`.
* Inline `mcable::prox_loc` and `mcable::dist_loc`.
* Add cv policy combinators +_and |.
* Add `util::foldl` which acts like `std::accumulate` for ranges, except it explicitly allows mutation of the sequence, and pass the accumulation value by rvalue.
* Add unit tests for mextent components, new locset expressions.
* Remove spherical soma cases from `cv_policy` testing.
* Extent `cv_policy` unit tests to cover new functionality.
* Rejig `cv_policy` unit tests to use the morphology testing predicates.
* Add `cv_policy` documentation to cable cell docs.

Fixes #1058.

* Add new API for the SIMD library that is compatible with the ARM 'sizeless' SVE vectors. Language restrictions prevent the use of the operator overload interface used up to this point for SIMD.
* Add `indirect_expressions` and `indirect_indexed_expressions` for describing memory reads/writes. `where_expressions` control masked access to simd vectors.
* Implement the SVE SIMD back-end in accordance with the alternate SIMD API.
* Retrieve vector width information from compiled mechanisms.
* Use alternate SIMD API in modcc-generated mechanisms.
* Add assertion in generated mechanism code that checks runtime vector width compatibility.

Fixess #1021. 

* Add C++ analogue to simple NEURON demo provided by LFPy project under example/lfp.
* Add plot script for rending example output to match that from NEURON demo.

Remove a redundant copy in favor of a swap operation for a gain in performance;
especially on GPU since copies are synchronous. Similarly, instead of solving the 
linear system into an intermediate array, write output directly into the target.

Here is the effect on the busyring benchmark (swapped pas -> hh) with 8192 cells on a
V100 GPU (time for model-run in seconds).
```
|----------+--------------------------------+------------------------------------|
| Baseline | fvm_lowered_cell: copy -> swap | matrix: solve + copy -> solve_into |
|----------+--------------------------------+------------------------------------|
|    2.230 |                          2.199 |                              2.129 |
|    2.231 |                          2.209 |                              2.132 |
|    2.225 |                          2.209 |                              2.136 |
|    2.227 |                          2.186 |                              2.130 |
|    2.220 |                          2.204 |                              2.133 |
|----------+--------------------------------+------------------------------------|
|     2.22 |                          2.186 |                              2.129 |
|----------+--------------------------------+------------------------------------|
```

Add a new catalogue comprising optimised version of some of the mechanisms commonly used in models of the Allen Cell Types Database, eg http://celltypes.brain-map.org/mouse/experiment/electrophysiology/323535089.
This made required extending modcc to some new NMODL features, as well as to expose some optimisations.

In addition clean up the default catalogue by moving test mechanisms to their proper place in the unit tests.

Update core library:
* support derivation of named mechanisms in catalogue from a string description of the parent.
* fix bug in radois_ge region expression that was finding region_lt.
* fix test for radius_ge region expression

Add support for mechanism, global properties and catalogue shenanigans to python wrapper

* Guard `event_generator` implicit constructor with type test.
* Fix linking error from use of `empty_generator` by replacing pointers to `no_event` with nullptr.
* Clarify `regular_generator` semantics in comment.
* Add unit test for checking.

Fixes #1066

* Add new `isometry` class representing a direct Euclidean isometry comprising a rotation and translation. Translations are always taken with respect to absolute coordinates, but rotations are composed with respect to intrinsic coordinates.
* Add new `place_pwlin` class, that takes a morphology and an isometry, and answers location queries: where in space does a particular `mlocation` lie?
* Split out (some of) the common code between `place_pwlin` and `embed_pwlin`.
* Add equality operator to `mpoint`.

This functionality is added to aid the presentation of the local field potential example.

Update `mech_vec.cpp` to work with the new API 
Print profiler summary in the ring example

* Make recipe return probes as a vector.
* Remove `probe_id` from `probe_info`.
* Rename `fvm_probe_info` to `fvm_probe_data` (everything was being called info, and it was getting confusing).
* Make `probe_association_map` specific to `mc_cell_group`/`fvm_lowered_cell`.
* Change probe_association_map representation: unordered map for probe_id -> tag; unordered multimap for probe_id -> fvm_probe_data. This allows multiple probes to be associated with the same probe id.
* Call sampler callback separately for each probe with the same probe_id.
* Replace location-based probes with locset equivalents.
* Add index for probes sharing a probe id.
* Bundle all probe metadata (id, tag, index, probe-specific meta) into `probe_metadata` struct for passing to sampler callbacks.
* Change simple_sampler to work on `trace_vector`, a vector of `trace_data`. The _i_th element is the data from probe with index i.
* Consolidate hash composition and `std::hash` specialization code in new header.
* Update python lib for new API.
* Update tests and examples for new recipe, internal probe, and simple_sampler APIs.
* Update docs to suit.

* Tidy synapse collation code.
* Further reduce synapse collation allocations.

Fix for #1055

* Add ranges in STATE blocks of the form FROM ... TO ...
* Unit test for the above
* Add type check for <..., ...> ranges
* Add units in procedure argument lists
* Unit test for the above.

Fixes #1037. 

* Add fake unmatchable spike events to event queue for each exact sample request, so that integration boundaries align with exact sample times.
* Update probe-demo for exact sampling.
* Add copy assignment for `cable_cell`.
* Extend max args for ARB_PP_FOREACH to 20.
* Add unit test for ARB_PP_FOREACH.
* Use ARB_PP_FOREACH to instantiate backend-specific probe unit tests. This also fixes the GPU probe tests that were using multicore backend.
* Add unit test that exercises exact sampling requests combined with regular events in a gap junction context.

This is a bit experimental. It builds arbor with `-fprofile-arcs -ftest-coverage` and creates a docker image with just the project binaries, lcov, gcov and generated `.gcno` files. This should still be reasonably small since there are no object files etc. When run on daint, `.gcna` files are produced (should be thread / process safe according to gcc's manual) inside of the container, and `lcov` is used to combine all those things into a single file with a random name in the mounted git project folder (hopefully this solves clashes with multiple nodes). After all tests are run, the combined reports are uploaded to codecov.io.

Fixes #589 and #730, providing cell-wide probes and a correct total trans-membrane current.
Fixes #822, providing spatial interpolation of membrane voltages as governed by cable resistivity and diameter.

* Add partition of point mechanism target vector by cell index to `fvm_mechanism_config`.
* Fix sign error in axial current interpolation.
* Replace `fvm_probe_info` with class that wraps a variant type capturing one of a set of backend-cellgroup probe translation representations: `fvm_probe_scalar`, `fvm_probe_interpolated`, `fvm_probe_multi`, `fvm_probe_weighted_multi`, `fvm_probe_membrance_currents`.
* Refactor `fvm_lowered_cell_impl::resolve_probe_address` so that it avoids the long chain of if-else tests on wrapped `any` type with an invocation of `util::any_visitor`. Each specific cable cell probe address representation is translated into one of the intermediate `fvm_probe_info` variants via an overload of `resolve_probe`.
* Add new non-scalar cell probe address types for all-of-cell values, including point mechanism state, density mechanism state, ion currents and concentrations, membrane voltage by CV cable, and total trans-mebrane currents. The latter are ultimately calculated by computing the net axial current fluxes in each CV.
* Add `util::any_ptr` field to the sampling function interface, pointing to constant-per-simulation-run metadata associated with a probe address. Underlying metadata type varies between probe types.
* Extend `simple_sampler` so that it optionally captures sampler metadata.
* Add special case to `simple_sampler` for vector-valued sample data, allowing it to translate from a pointer-pair range to a std::vector.
* Fix minor bugs in `util::variant`.
* Add to `mc_cell_group` the logic for translating each intermediate probe representation to a sampler callback invocation, reusing temporary scratch vectors across calls.
* Add unit comments to `embed_pwlin_data`.
* Add probe variety demonstration `probe-demo` to examples, covering most of the new cable cell probe classes.
* Add `util::any_visitor` runtime-type dependent dispatcher class.
* Add `util::overload` helper for generating overloaded functional objects from individual functions, lambdas or functionals.
* Extend probe unit tests in `test_probe.cpp` to cover new probe address types.
* Add unit tests for `any_visitor` and `overload`.

Travis CI is failing on some simd tests for `ARB_ARCH=native`
GCC appears to be setting incorrect flags in the Travis environment for `-march=native`. Using `-march=skylake-avx512/haswell/etc` fixes the errors. 

Add a simple re-write to eliminate division by a known constant.

* Replace HH model with optimised version.

* [WIP] Investigate the effect of __restrict__ specifier on GPU code.

* Fix range parsing and add unit test

* Set the default ARB_ARCH target to native.
* Tweak travis config to avoid gcc 6 bug that incorrectly implies amd64 target can do avx512.

* Add VECTORIZE configuration feature macro to config, record in python lib.
* Add  ARB_ARCH configuration string as preprocessor define and character string in config and version info respectively.

Add a `restrict` locset expression type that returns all locations in a locset that are in a specified region.

Also allow region and locset names with hyphens.

Fixes #1031 
Fixes #1032 

Fuse kernels `gather` and `vec_minus` into a single kernel `set_dt_impl` for a small
performance improvement.

Here is the effect on the busyring benchmark (swapped pas -> hh) with 8192 cells on a 
V100 GPU (time for `model-run` in seconds).

```
|----------+-------|
| Baseline | After |
|----------+-------|
|    2.318 | 2.314 |
|    2.335 | 2.307 |
|    2.345 | 2.315 |
|    2.333 | 2.306 |
|    2.331 | 2.320 |
|----------+-------|
|    2.318 | 2.306 |
|----------+-------|
```

* Use the safe container based jobs
* Copy the source into the container and DRY it bit

* Move .gitlab-ci.yml inside the repo
* Remove submodule with gitlab pipeline

Add `maxset` function that returns the most distal set of locations in a location list, similar to existing `minset` function.

Use `minset` and `maxset` consistently in `most_proximal` and `most_distal` locset expressions respectively.

Split Python wrapper changes and tutorial example from the Python documentation branch.

- extend `append` methods for `sample_tree` Python wrapper to accommodate a flatter representation of samples.
- add a tutorial example (will make more sense as part of larger python docs)

Combine badges on readme and move ci git submodule to `ext/`

* Add functions for determining CVs and weights for membrane voltage interpolation and axial current determination.
* Add unit tests for voltage interpolation and axial current interpolation.
* Accommodate mextent changes.

- Add option to compile Arbor for HIP/CUDA backend using Clang. 
- Add new CMake option `ARB_GPU_COMPILE_TYPE` to distinguish between three possible gpu platforms/builds: `cuda-nvcc`; `cuda-clang`; `hip-clang`
- Add gpu wrapper functions in `arbor/backends/gpu/gpu_api.hpp` and `arborenv/gpu_api.hpp` which call HIP/CUDA functions depending on the platform
- Rename functions and files: cuda -> gpu
- Add downgraded warp primitives for HIP
- Implement `uuid` workaround for HIP
- Set correct alignment and warp size for AMD gpus
- Update installation guide and docs.
- Update pip/setuptools to support new gpu targets

Fixes #833

* Remove interleaved matrix solver on GPU.
* Fix out of bound memory bug when initialising matrix for all back ends
* fix singed-vs-unsigned comparison warning 

Fixes #1014 
Fixes #1013

Fix build errors for the GPU tests introduced in #1011.

* Add Dockerfiles for MPI + CUDA test builds on Daint
* Add bors.toml for CI on Daint
* Add the separate ci repo as submodule for convenience
* Add a CI badge for master to the top of the readme

* Extend repertoire of cable cell probes.
* Add the following scalar (i.e. corresponding to just one raw value)
probe address types for cable cells:
   - `cell_probe_membrane_voltage`, replacing `cell_probe_address::membrane_voltage`;
   - `cell_probe_total_ionic_current_density`, replaing `cell_probe_address::membrane_current`;
   - `cell_probe_density_state`;
   - `cell_probe_point_state`;
   - `cell_probe_ion_current_density`;
   - `cell_prone_ion_int_concentration`;
   - `cell_prone_ion_ext_concentration`.
* Extend backend class interface with `mechanism_field_data` method, that returns a raw (device) pointer to the per-instance values of a mechanism state variable.
* Make unit test for probes backend-generic.
* Add unit tests for new probes.
* Modify `fixed_ica_current.mod` unit test NMODL module to be more generic in naming, so it can be better used in probe unit tests.
* Update examples for new cable cell probes.
* Update Python interface for new probe types