hunterMakesPy

Utilities for converting mixed input to integers, calculating CPU limits, handling None values, importing code dynamically, and manipulating nested data. Also includes humpy_toolz, humpy_cytoolz, and humpy_tlz: typed forks of toolz and cytoolz with curried namespaces and a sandbox.

Validate, Convert, and Parse Inputs

Convert Mixed Input to Validated Integers

Accepts strings, floats, complex numbers, and binary data and returns a list of validated integers. Ambiguous or incompatible values produce descriptive errors.

from hunterMakesPy.parseParameters import intInnit

ports = intInnit(["8080", 443, "22"], "server_ports")
# Returns: [8080, 443, 22]

Calculate Concurrency Limits from Flexible Specifications

Pass 0.75 for 75% of CPUs, True for 1 CPU, 4 for exactly 4 CPUs, or -2 to reserve 2 CPUs.

from hunterMakesPy.parseParameters import defineConcurrencyLimit

workers = defineConcurrencyLimit(limit=0.75)  # 6 on an 8-core machine

Interpret Strings as Boolean or None

Parse string values such as "true", "false", or "none" into their Python equivalents without raising exceptions on mismatch.

from hunterMakesPy.parseParameters import oopsieKwargsie

oopsieKwargsie("True")   # Returns: True
oopsieKwargsie("none")   # Returns: None
oopsieKwargsie("hello")  # Returns: "hello"

Eliminate None with Runtime Validation

Convert Type | None return annotations to Type by raising at runtime if the value is None. Removes None-checking noise from downstream code and satisfies type checkers.

from hunterMakesPy import raiseIfNone

config = raiseIfNone(getConfig(), "Missing configuration")

Replace Ambiguous Numeric Literals with Semantic Names

Use decreasing instead of -1, inclusive for boundary adjustments, and zeroIndexed for index conversions. Intent becomes explicit and the code reads as prose.

from hunterMakesPy import decreasing, inclusive, zeroIndexed

rangeEnd = lengthSequence + inclusive
indexLast = lengthSequence - zeroIndexed
step = decreasing

Import Modules by Dot-Notation Path or File Path

Load functions or classes from "scipy.signal.windows" or "path/to/file.py" without manual module-loading boilerplate.

from hunterMakesPy.filesystemToolkit import importLogicalPath2Identifier

windowFunction = importLogicalPath2Identifier("scipy.signal.windows", "hann")

Create Directories Safely and Write Files with Formatting

Write to "deep/nested/path/file.txt" and parent directories are created automatically. writePython removes unused imports and sorts import statements before writing.

from hunterMakesPy.filesystemToolkit import writeStringToHere, writePython

writeStringToHere("content", "nested/dirs/file.txt")  # Creates dirs
writePython(sourceCode, "output/module.py")            # Formats, then writes

Extract Strings from Arbitrarily Nested Data

Recursively traverse dictionaries, lists, tuples, and sets and collect every string value into a flat list.

from hunterMakesPy.dataStructures import stringItUp

strings = stringItUp({"users": ["alice"], "config": {"host": "localhost"}})
# Returns: ["users", "alice", "config", "host", "localhost"]

Merge Dictionaries with List Values

Combine {"a": [1, 2]} and {"a": [3], "b": [4]} into {"a": [1, 2, 3], "b": [4]} with optional deduplication and sorting.

from hunterMakesPy.dataStructures import updateExtendPolishDictionaryLists

merged = updateExtendPolishDictionaryLists(
    {"servers": ["chicago"]},
    {"servers": ["tokyo", "chicago"]},
    destroyDuplicates=True,
)
# Returns: {"servers": ["chicago", "tokyo"]}

Compress NumPy Arrays to Self-Decoding Strings

Encode repetitive patterns and consecutive sequences using run-length encoding and Python range syntax. The resulting string evaluates back to the original data.

from hunterMakesPy.dataStructures import autoDecodingRLE

encoded = autoDecodingRLE(arrayTarget)

Configure Package Metadata at Runtime

PackageSettings reads pyproject.toml automatically to resolve the package name and installation path.

from hunterMakesPy import PackageSettings

settings = PackageSettings(identifierPackageFALLBACK="myPackage")

Reuse Test Suites for Custom Implementations

Import parameterized test generators to validate custom functions that match the expected signatures.

from hunterMakesPy.tests.test_parseParameters import PytestFor_intInnit

@pytest.mark.parametrize("test_name,test_func", PytestFor_intInnit(myFunction))
def test_my_integer_validator(test_name, test_func):
    test_func()

---

humpy_toolz: Typed Pure-Python Functional Utilities

humpy_toolz is a typed fork of toolz. It provides composable functions for iterators, dictionaries, and function composition with full type stubs and docstrings.

Dictionary Transformations without Mutation

Create new mappings by associating, dissociating, filtering, and mapping over keys or values. Nested access is supported through key-path sequences.

from humpy_toolz import merge, valmap, keyfilter, assoc

merged = merge({"a": 1}, {"b": 2})             # {"a": 1, "b": 2}
doubled = valmap(lambda x: x * 2, {"a": 1})    # {"a": 2}
evens = keyfilter(lambda k: k > 1, {1: "a", 2: "b"})  # {2: "b"}
updated = assoc({"x": 10}, "y", 20)            # {"x": 10, "y": 20}

Compose, Curry, and Thread Functions

Build transformation sequences from left or right, partially apply arguments with curry, and thread a value through a series of functions.

from humpy_toolz import compose_left, curry, pipe

increment = lambda x: x + 1
double = lambda x: x * 2
transform = compose_left(increment, double)
transform(3)  # 8

pipe(3, increment, double)  # 8

Slice, Group, and Deduplicate Iterators

Lazy operations over iterables: take, drop, partition, sliding_window, unique, interleave, groupby, frequencies, and more.

from humpy_toolz import take, frequencies, groupby, unique

list(take(3, range(100)))              # [0, 1, 2]
frequencies(["a", "b", "a", "c"])      # {"a": 2, "b": 1, "c": 1}
groupby(len, ["cat", "mouse", "dog"])  # {3: ["cat", "dog"], 5: ["mouse"]}
list(unique([1, 2, 1, 3]))            # [1, 2, 3]

Curried Namespace for Partial Application

Every function in humpy_toolz.curried accepts partial arguments and returns a new callable waiting for the rest.

from humpy_toolz.curried import map, filter, get

list(map(str.upper, ["hello", "world"]))  # ["HELLO", "WORLD"]
list(filter(lambda x: x > 2, [1, 2, 3, 4]))  # [3, 4]
list(map(get(0), [(1, 2), (3, 4)]))       # [1, 3]

Sandbox: Parallel Fold and Equality-Based Hashing

humpy_toolz.sandbox provides fold for unordered parallel reductions and EqualityHashKey for hashing otherwise-unhashable types.

from humpy_toolz.sandbox import fold, EqualityHashKey
from operator import add

fold(add, range(100), default=0)  # 4950

---

humpy_cytoolz: Cython-Accelerated Functional Utilities

humpy_cytoolz is a typed fork of cytoolz. It exposes the same API as humpy_toolz but the core modules (dicttoolz, functoolz, itertoolz, recipes, utils) are compiled as Cython extension modules for lower overhead on hot paths.

from humpy_cytoolz import groupby, curry, merge

Install with the cython build dependency to compile the extensions. If compilation is unavailable, use humpy_toolz directly.

---

humpy_tlz: Automatic Cython-or-Pure Dispatch

humpy_tlz mirrors the humpy_toolz API and imports from humpy_cytoolz when available, falling back to humpy_toolz otherwise. Use humpy_tlz in library code so callers benefit from Cython acceleration without requiring it.

from humpy_tlz import pipe, curry, groupby

---

My recovery