Add `exactly1` option for disjunctions and fix bugs

[pulsipher]

Closes #72. Some minor bug fixes and changes that came up include:

• Have deletion methods for disjunctions and disjunct constraints remove all their mappings
• Remove unnecessary GDPData(args...) method
• Include the macro docstrings in the docs
• Moving reformulation specific methods
• Fixing the reformulation methods for exactly constraints to better handle assumptions on function type (more work is needed)
• Improved error messaging for incorrect keyword arguments to @disjunction
• Fixed some test functions to use @test on the first call of a method/macro to be tested which leads to more helpful error messages when it doesn't work (more work is needed to do this with all tests)

[codecov[bot]]

Codecov Report

Attention: 2 lines in your changes are missing coverage. Please review.

Comparison is base (3af182e) 98.91% compared to head (50cdc2a) 98.84%.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #90 +/- ## ========================================== - Coverage 98.91% 98.84% -0.07% ========================================== Files 10 10 Lines 921 952 +31 ========================================== + Hits 911 941 +30 - Misses 10 11 +1 ``` | [Files](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez) | Coverage Δ | | |---|---|---| | [src/bigm.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL2JpZ20uamw=) | `100.00% <100.00%> (ø)` | | | [src/constraints.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL2NvbnN0cmFpbnRzLmps) | `98.83% <100.00%> (+0.19%)` | :arrow_up: | | [src/datatypes.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL2RhdGF0eXBlcy5qbA==) | `100.00% <ø> (ø)` | | | [src/hull.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL2h1bGwuamw=) | `100.00% <100.00%> (ø)` | | | [src/logic.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL2xvZ2ljLmps) | `99.25% <100.00%> (+<0.01%)` | :arrow_up: | | [src/macros.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL21hY3Jvcy5qbA==) | `96.89% <ø> (+0.77%)` | :arrow_up: | | [src/reformulate.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL3JlZm9ybXVsYXRlLmps) | `100.00% <100.00%> (ø)` | | | [src/model.jl](https://app.codecov.io/gh/hdavid16/DisjunctiveProgramming.jl/pull/90?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Hector+Perez#diff-c3JjL21vZGVsLmps) | `93.33% <33.33%> (-6.67%)` | :arrow_down: |

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[hdavid16]

I think we should use a different name for this keyword argument since it is not exactly 1 for Nested disjunctions.

[pulsipher]

I think we should use a different name for this keyword argument since it is not exactly 1 for Nested disjunctions.

Any suggestions?