Python utility library with Rust-accelerated functions for text processing, iteration, dictionary sorting, and number parsing.
pip install libb-utilfrom libb import Setting, compose, attrdict, timeout
from libb import multikeysort, numify, parse
# Configuration with dot notation
config = Setting()
config.database.host = 'localhost'
config.lock()
# Function composition
add_then_multiply = compose(lambda x: x * 2, lambda x: x + 1)
result = add_then_multiply(5) # (5 + 1) * 2 = 12
# Attribute dictionary
d = attrdict(x=10, y=20)
print(d.x) # 10
# Timeout decorator
@timeout(5)
def slow_function():
passpip install libb-util
# With extras
pip install "libb-util[pandas,text,web,math]"| Extra | Description |
|---|---|
pandas |
DataFrame utilities with pyarrow |
text |
Text processing (ftfy, rapidfuzz, chardet) |
web |
Web frameworks (Flask, Twisted, web.py) |
math |
Matplotlib charting |
test |
Testing tools (pytest, asserts) |
docs |
Sphinx documentation |
| Function/Class | Description |
|---|---|
Setting() |
Hierarchical config with dot notation and locking |
attrdict(**kw) |
Dict with attribute access |
dictobj(**kw) |
Immutable dict with attribute access |
| Function | Description |
|---|---|
compose(*funcs) |
Right-to-left function composition |
timeout(seconds) |
Decorator to limit function execution time |
retry(tries, delay) |
Decorator for retry with exponential backoff |
memoize(func) |
Cache function results |
once(func) |
Execute function only once |
| Function | Description |
|---|---|
multikeysort(items, columns) |
Sort list of dicts by multiple keys |
numify(val, to=int) |
Convert string to number with format handling |
parse(s) |
Extract number from string |
# Rust-accelerated sorting
data = [{'name': 'Alice', 'age': 30}, {'name': 'Bob', 'age': 25}]
multikeysort(data, ['age', '-name']) # Sort by age asc, name desc
# Number parsing handles accounting formats
parse('1,200') # 1200
parse('(500)') # -500 (accounting negative)
parse('100.5%') # 100.5| Function | Description |
|---|---|
chunked(iterable, n) |
Split into chunks of size n |
collapse(*args) |
Recursively flatten nested lists/tuples ⚡ |
backfill(values) |
Fill None values with last non-None ⚡ |
backfill_iterdict(iterdict) |
Back-fill dicts with latest values per key ⚡ |
same_order(ref, comp) |
Check if elements appear in same order ⚡ |
unique(iterable) |
Unique items preserving order |
⚡ = Rust-accelerated
| Function | Description |
|---|---|
sanitize_vulgar_string(s) |
Replace vulgar fractions with decimals ⚡ |
uncamel(camel) |
Convert CamelCase to snake_case ⚡ |
underscore_to_camelcase(s) |
Convert snake_case to camelCase ⚡ |
normalize(text) |
Unicode normalization |
slugify(text) |
URL-safe slug |
strip_html(text) |
Remove HTML tags |
truncate(text, length) |
Truncate with ellipsis |
⚡ = Rust-accelerated
| Function | Description |
|---|---|
ensure_dir(path) |
Create directory if not exists |
safe_filename(name) |
Sanitize filename |
temp_file(**kw) |
Context manager for temp files |
atomic_write(path) |
Atomic file write context |
| Function | Description |
|---|---|
threaded(func) |
Run function in thread |
parallel_map(func, items) |
Parallel execution |
Semaphore(n) |
Limit concurrent access |
| Function | Description |
|---|---|
run_cmd(cmd) |
Execute shell command |
is_running(pid) |
Check if process exists |
kill_tree(pid) |
Kill process and children |
Set that maintains insertion order:
from libb import OrderedSet
s = OrderedSet([1, 2, 3, 2, 1])
list(s) # [1, 2, 3]Min/max heap implementations:
from libb import MinHeap, MaxHeap
h = MinHeap([3, 1, 4, 1, 5])
h.pop() # 1Full documentation at bissli.github.io/libb-util:
# Install dev dependencies
pip install -e ".[test,docs]"
# Build Rust extension
maturin develop
# Run tests
pytestSee LICENSE file.