Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merge a few years worth of changes #377

Merged
merged 118 commits into from
May 20, 2024
Merged

Merge a few years worth of changes #377

merged 118 commits into from
May 20, 2024

Conversation

jeremy-murphy
Copy link
Contributor

OK, somehow the pandemic created a blind spot to my responsibility for this, so here we go. Next release will be a killer!

jan-grimo and others added 30 commits February 16, 2020 11:12
Isomorphism docs
- The expected concept for invariant functors is AdaptableUnaryFunction,
  not UnaryFunction
- Null vertices can appear in the isomorphism map output if both graphs
  contain disconnected vertices, add a note to that parameter's
  documentation
Isomorphism: Ignore vertex_max_invariant
- vertex_max_invariant and invariant2.max() are misnomers since what is
  expected is an upper exclusive bound on the possible invariant values,
  not their maximum value.
- The parameter can be ignored and the upper exclusive bound found
  cheaply at the start of test_isomorphism
- Removes the additional requirement of a nullary max member function on
  invariant2
…lts-overload

Add tests for the all defaults overload of `boykov_kolmogorov_max_flow`
e.g. [a,b] -> [(a,b)] to avoid confusion with C++23 array subscript operator[]
Fix compilation using C++23 by adding parentheses around comma operator inside square brackets
Don't run performance test in CI
exception.hpp uses BOOST_SYMBOL_VISIBLE, which is defined in boost/config.hpp. Without this, compiling just exception.hpp is broken.
Add missing include on boost/config.hpp
Fix r_c_shortest_paths example
The arguments are already required to be the representative of their
class. If we do not know the representatives, we are supposed to call
union_set instead, which first finds them.
Add missing break to switch
jeremy-murphy and others added 25 commits April 14, 2024 07:24
Invariant contiguous range requirement removal
- Vertex invariants for use in isomorphism algorithm must no longer have
  low upper bounds due to a hidden allocation linear in the maximum
  encountered vertex invariant.
- Vertex invariants must no longer be convertible to `size_t`, but can
  be any comparable and hashable types
- Build `unordered_map`-backed invariant multiplicity map efficiently
  from sorted vertex invariants
Avoid requiring invariant default-constructibility
- Refactor concept checking with boost type_traits
…riants

Isomorphism: Remove invariant contiguous range requirement
Fix labeled_graph typo in name adjacenct_vertices.
I confirmed with Noel that he was happy with this and told him
that the door is always open if he wants to come back.
Thanks for all your help, Noel!
This choice is based on feedback to the Boost.Graph user survey.
Works around invalidation of bundle property maps (see #373).

The `#if SEHE_UNSTABLE_PROPERTY_MAPS_FIXED` section is there to signal
my intent to investigate a generalized fix under that issue. It doubles
as literate documentation of the need for the workaround, so it's less
likely to bite the unwary.
The code was mostly fine (except for unhygienic `using namespace` in
places), but it was hard to see what was covered.

I've seperated fixtures (sample input + expected output) and this will
simplify invoking the ComparisonDriver (test_graph).
Non-keyword graphs never worked (!). This was uncovered because of
security issue #364.

parse_subgraph() incorrectly dealt with first_token in the case
where the `subgraph` keyword wasn't used.
Eliminating need for manual re-test after review updates PR #376
Fix and restore broken test_basic_csr_directed_graph
Refactoring graphviz_test for review
Fix security issue #364 and non-keyword subgraph parsing
@jzmaddock
Copy link
Contributor

Looks good to me, please do go ahead and merge!

@jeremy-murphy
Copy link
Contributor Author

Looks good to me, please do go ahead and merge!

Thanks, John!

@jeremy-murphy jeremy-murphy merged commit 198616d into master May 20, 2024
14 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.