case | example |
---|---|
camel | someText |
snake | some_text |
kebab | some-text |
pascal | SomeText |
constant | SOME_TEXT |
pip install casers
The examples are checked by pytest
>>> from casers import to_camel, to_snake, to_kebab
>>> to_camel("some_text") == "someText"
True
>>> to_snake("someText") == "some_text"
True
>>> to_kebab("someText") == "some-text"
True
>>> to_kebab("some_text") == "some-text"
True
pip install "casers[pydantic]"
The package supports for pydantic 2
>>> from casers.pydantic import CamelAliases
>>> class Model(CamelAliases):
... snake_case: str
>>> Model.model_validate({"snakeCase": "value"}).snake_case == "value"
True
>>> Model.model_validate_json('{"snakeCase": "value"}').snake_case == "value"
True
Apple M3 Pro
----------------------------------------------------------------------------------------------- benchmark: 5 tests -----------------------------------------------------------------------------------------------
Name (time in us) Min Max Mean StdDev Median IQR Outliers OPS (Kops/s) Rounds Iterations
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_to_camel_rust 2.4580 (1.0) 10.7919 (1.0) 2.5684 (1.0) 0.0955 (1.0) 2.5420 (1.0) 0.0410 (1.0) 2123;2123 389.3475 (1.0) 79208 1
test_to_camel_python_builtin 10.3328 (4.20) 90.1250 (8.35) 10.7271 (4.18) 0.8965 (9.39) 10.6669 (4.20) 0.2082 (5.08) 1182;1739 93.2215 (0.24) 57419 1
test_to_camel_rust_parallel 20.1249 (8.19) 102.2089 (9.47) 29.5715 (11.51) 4.0862 (42.79) 28.4170 (11.18) 4.8331 (117.94) 855;158 33.8163 (0.09) 4783 1
test_to_camel_python_builtin_parallel 36.4999 (14.85) 1,233.1251 (114.26) 39.9730 (15.56) 19.0205 (199.20) 38.1658 (15.01) 0.8328 (20.32) 95;1059 25.0169 (0.06) 8741 1
test_to_camel_pure_python 39.4580 (16.05) 212.9169 (19.73) 40.6741 (15.84) 3.1588 (33.08) 40.2501 (15.83) 0.4161 (10.15) 614;2145 24.5857 (0.06) 21878 1
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Legend:
Outliers: 1 Standard Deviation from Mean; 1.5 IQR (InterQuartile Range) from 1st Quartile and 3rd Quartile.
OPS: Operations Per Second, computed as 1 / Mean