search

j2kun / riemann-divisor-sum

Code for the series "Searching for Riemann Hypothesis Counterexamples"
https://jeremykun.com/2020/09/11/searching-for-rh-counterexamples-setting-up-pytest/
21 stars 2 forks source link

Add a type-checking test #19

Closed j2kun closed 3 years ago

j2kun commented 3 years ago

This may be difficult to do if the gmp-based types get in the way, but at least we can do it partially

j2kun commented 3 years ago

First run of mypy reports 29 errors

riemann/types.py:97: error: Incompatible types in assignment (expression has type "None", variable has type "datetime")
riemann/types.py:98: error: Incompatible types in assignment (expression has type "None", variable has type "datetime")
riemann/types.py:122: error: Incompatible types in assignment (expression has type "None", variable has type "str")
riemann/superabundant.py:6: error: Cannot find implementation or library stub for module named 'gmpy2'
riemann/superabundant.py:7: error: Cannot find implementation or library stub for module named 'numba'
riemann/divisor.py:5: error: Cannot find implementation or library stub for module named 'numba'
riemann/divisor.py:5: note: See https://mypy.readthedocs.io/en/latest/running_mypy.html#missing-imports
riemann/divisor.py:43: error: No overload variant of "__setitem__" of "list" matches argument types "int", "RiemannDivisorSum"
riemann/divisor.py:43: note: Possible overload variants:
riemann/divisor.py:43: note:     def __setitem__(self, int, None) -> None
riemann/divisor.py:43: note:     def __setitem__(self, slice, Iterable[None]) -> None
riemann/divisor.py:47: error: Incompatible return value type (got "List[None]", expected "List[RiemannDivisorSum]")
riemann/search_strategy.py:71: error: Argument 1 of "max" is incompatible with supertype "SearchStrategy"; supertype defines the argument type as "List[SearchIndex]"
riemann/search_strategy.py:71: note: This violates the Liskov substitution principle
riemann/search_strategy.py:71: note: See https://mypy.readthedocs.io/en/stable/common_issues.html#incompatible-overrides
riemann/search_strategy.py:74: error: Argument 1 of "starting_from" is incompatible with supertype "SearchStrategy"; supertype defines the argument type as "SearchIndex"
riemann/search_strategy.py:74: note: This violates the Liskov substitution principle
riemann/search_strategy.py:74: note: See https://mypy.readthedocs.io/en/stable/common_issues.html#incompatible-overrides
riemann/search_strategy.py:95: error: "SearchIndex" has no attribute "n"
riemann/search_strategy.py:114: error: Argument 1 of "max" is incompatible with supertype "SearchStrategy"; supertype defines the argument type as "List[SearchIndex]"
riemann/search_strategy.py:114: note: This violates the Liskov substitution principle
riemann/search_strategy.py:114: note: See https://mypy.readthedocs.io/en/stable/common_issues.html#incompatible-overrides
riemann/search_strategy.py:123: error: Argument 1 of "starting_from" is incompatible with supertype "SearchStrategy"; supertype defines the argument type as "SearchIndex"
riemann/search_strategy.py:123: note: This violates the Liskov substitution principle
riemann/search_strategy.py:123: note: See https://mypy.readthedocs.io/en/stable/common_issues.html#incompatible-overrides
riemann/search_strategy.py:171: error: "SearchIndex" has no attribute "level"
riemann/search_strategy.py:173: error: "SearchIndex" has no attribute "level"
riemann/search_strategy.py:174: error: "SearchIndex" has no attribute "index_in_level"
riemann/search_strategy.py:179: error: "SearchIndex" has no attribute "level"
riemann/search_strategy.py:184: error: "SearchIndex" has no attribute "index_in_level"
riemann/search_strategy.py:185: error: "SearchIndex" has no attribute "index_in_level"
riemann/postgres_database.py:5: error: Cannot find implementation or library stub for module named 'psycopg2.extras'
riemann/postgres_database.py:5: error: Cannot find implementation or library stub for module named 'psycopg2'
riemann/postgres_database.py:6: error: Cannot find implementation or library stub for module named 'gmpy2'
riemann/postgres_database.py:129: error: Argument "largest_computed_n" to "SummaryStats" has incompatible type "None"; expected "RiemannDivisorSum"
riemann/postgres_database.py:130: error: Argument "largest_witness_value" to "SummaryStats" has incompatible type "None"; expected "RiemannDivisorSum"
riemann/postgres_database.py:192: error: Return type "Optional[SearchMetadata]" of "claim_next_search_block" incompatible with return type "SearchMetadata" in supertype "DivisorDb"
riemann/in_memory_database.py:75: error: Argument "largest_computed_n" to "SummaryStats" has incompatible type "None"; expected "RiemannDivisorSum"
riemann/in_memory_database.py:76: error: Argument "largest_witness_value" to "SummaryStats" has incompatible type "None"; expected "RiemannDivisorSum"
riemann/cleanup_stale_blocks.py:38: error: Argument 2 to "get_stale_blocks" has incompatible type "Optional[timedelta]"; expected "timedelta"
riemann/cleanup_stale_blocks.py:53: error: Argument 1 to "sleep" has incompatible type "Optional[int]"; expected "float"
Found 29 errors in 7 files (checked 13 source files)
j2kun commented 3 years ago

For numba, maybe see https://stackoverflow.com/questions/42310278/using-python-type-hints-with-numba

j2kun commented 3 years ago

For gmpy2, my first attempt at stubbing failed. Cf. https://github.com/python/typeshed/issues/5031

j2kun commented 3 years ago

Fixed by https://github.com/j2kun/riemann-divisor-sum/pull/24

Ignoring missing stubs for now