Python Tricks I Actually Reach For Daily

Python Tricks

Tuple Unpacking

Click to reveal

Python Tricks

Assign multiple values at once: `a, b = b, a`. Works for swaps, function returns, and dict entries. Zero overhead tuple creation.

Tuple Unpacking

Python Tricks

enumerate()

Click to reveal

Python Tricks

Loop with index and value: `for i, item in enumerate(items, start=1)`. Replaces error-prone `range(len(items))` pattern.

enumerate()

Python Tricks

zip()

Click to reveal

Python Tricks

Iterate multiple sequences in parallel: `for a, b in zip(list1, list2)`. Stops at shortest sequence by default.

zip()

Python Tricks

dict.get()

Click to reveal

Python Tricks

Safe dictionary access with default: `config.get('port', 8080)`. Returns default if key missing - no KeyError.

dict.get()

Python Tricks

Set lookup

Click to reveal

Python Tricks

Convert list to set for O(1) membership: `if item in allowed_set`. Much faster than list O(n) for repeated checks.

Set lookup

Python Tricks

Walrus operator

Click to reveal

Python Tricks

Assignment expression `:=` - assign and test in one: `if (n := len(items)) > 10: ...`. Avoids duplicate function calls.

Walrus operator

Python Tricks

dataclass

Click to reveal

Python Tricks

Decorator that auto-generates __init__, __repr__, __eq__: `@dataclass class Point: x: float; y: float`. Less boilerplate.

dataclass

Python Tricks

pathlib.Path

Click to reveal

Python Tricks

Object-oriented file paths: `base / 'config' / f'{env}.yaml'`. Cross-platform, chainable methods, reads/writes text directly.

pathlib.Path

Forget the clever golfed code that reads like riddles. These are the Python features I reach for in production code - the ones that reduce boilerplate, prevent actual bugs, and make other developers nod in approval rather than scrolling to Stack Overflow.

Audio Briefing• Voice: Google Cloud Journey-F

0:00 / --:--

The Core 7 (with real context)

1. Tuple Unpacking for Swaps

No temp variable, no ceremony:

python

a, b = b, a

This creates a tuple on the right, then destructures it on the left. Works in-place, zero overhead. The same pattern handles multi-values:

python

# Swap three values in one line
i, j, k = j, k, i

# Unpack a function's multiple return values
x_coord, y_coord, label = get_point()

# Swap dict entries
d = {'a': 1, 'b': 2}
d['a'], d['b'] = d['b'], d['a']  # Now {'a': 2, 'b': 1}

Interactive Quiz

What does `a, b = b, a` actually do under the hood?

2. enumerate() - Index + Value

Stop doing for i in range(len(items)). It's error-prone and reads backwards:

python

# ❌ Old way - off-by-one bugs hiding here
for i in range(len(colors)):
    print(i, colors[i])

# ✅ Python way - what you meant
for i, color in enumerate(colors):
    print(i, color)

Pro move: start at 1

enumerate(colors, start=1) gives 1-based numbering - useful for display, error messages, test cases.

Interactive Quiz

What's the benefit of `enumerate(items, start=1)`?

3. zip() - Loop Two Sequences in Parallel

Cleaner than manual index management when two sequences must stay in sync:

python

names    = ["Alice", "Bob", "Charlie"]
scores   = [85, 92, 78]
subjects = ["Math", "Science", "English"]

# Pairing two lists
for name, score in zip(names, scores):
print(f"{name}: {score}")

# Three-way zipping (tuples auto-unpacked)
for name, score, subject in zip(names, scores, subjects):
print(f"{name} ({subject}): {score}")

# Dictionaries from two lists
scores_dict = dict(zip(names, scores))
# {'Alice': 85, 'Bob': 92, 'Charlie': 78}

zip truncates to the shortest

If names has 3 items and scores has 5, only 3 pairs are produced. Use itertools.zip_longest if you want padding behavior.

Interactive Quiz

What happens when you zip two lists of different lengths?

4. dict.get() - Avoid KeyError

The difference between "missing key expected" and "missing key is a bug":

python

config = {"port": 3000, "host": "localhost"}

# Returns default if key missing - no exception
port     = config.get("port", 8080)     # 3000
debug    = config.get("debug", False)   # False (key missing)
host     = config.get("host", "0.0.0.0") # "localhost"

# None is the implicit default
value = config.get("nonexistent")       # None

Scenario	Use	Reason
Required config, missing = bug	<code>config["key"]</code>	Fail fast, surface error
Optional key, default OK	<code>config.get("key", default)</code>	Graceful fallback
Need to distinguish missing from None	<code>if "key" in config</code>	None is a valid value

Interactive Quiz

When should you use `config["key"]` instead of `config.get("key", default)`?

5. Set Lookups for O(1) Membership

Converting a list to a set before checking membership is one of the most common "hidden slowness" fixes:

python

allowed_actions = {"read", "write", "delete", "update"}

# O(1) - constant time regardless of list size
if action in allowed_actions:
execute(action)

# ❌ With a list, this is O(n) - scales poorly
allowed_actions_list = ["read", "write", ...]
if action in allowed_actions_list:  # Slower as list grows
execute(action)

Rule of thumb

If x in collection runs more than a handful of times per request and the collection has more than ~10 items, it should almost certainly be a set.

Interactive Quiz

Why is `item in some_set` faster than `item in some_list` for large collections?

6. Underscore `_` for Intentional Ignoring

Communicates "I'm not using this value on purpose":

python

# Unpacking - ignore the key, keep the value
_, result = parse_line(line)

# Loop - ignore the index
for _ in range(3):
print("repeat")

# Match with explicit wildcard (Python 3.10+)
match status:
case 200:
    handle_ok()
case 404:
    handle_missing()
case _:
    handle_other()  # "anything else"

# Unpacking partial - keep start/end, ignore middle
first, *_, last = [1, 2, 3, 4, 5]
# first=1, last=5

Interactive Quiz

What does `first, *_, last = [1, 2, 3, 4, 5]` produce?

Deep-Cut Techniques (Actually Useful)

python

# Without walrus
data = fetch_record()
if data:
process(data)

# With walrus :=
if data := fetch_record():
process(data)

When NOT to use walrus

Don't force it into existing clean code. If the assignment belongs outside the condition, put it there. The walrus shines in: loop guards, catch blocks, and reducing duplicated function calls.

Interactive Quiz

When is the walrus operator `:=` most valuable?

Python 3.12 added f-string debugging with = and multiline f-strings:

python

name, age = "Alice", 30

# Debug mode | shows expression + value
print(f"{name=} {age=}")  # name='Alice' age=30

# Multiline (no more backslash continuation)
report = f"""
Name:   {name}
Status: {"Adult" if age >= 18 else "Minor"}
Score:  {95.5:.1f}%
"""

Interactive Quiz

What does `f"{name=} {age=}"` output when name="Alice" and age=30?

Stop concatenating with os.path.join():

python

from pathlib import Path

base = Path("/app/data")

# Fluent chaining | works cross-platform
config = base / "config" / f"env_{ENV}.yaml"
logs   = base / "logs" / "app.log"

# Read/write without manual open()
text = config.read_text(encoding="utf-8")
config.write_text(new_text, encoding="utf-8")

# Existence checks
if config.exists() and config.is_file():
size = config.stat().st_size

# Iterate directory
for log_file in base.glob("**/*.log"):
print(log_file.name, log_file.stat().st_size)

Why pathlib over os.path?

Path uses operator overloading (/ joins paths) which reads naturally. Methods are chainable. The same code works on Windows and Linux without changes.

Interactive Quiz

What does `Path('/app') / 'data' / 'config.yaml'` produce?

Before dataclasses, classes with many attributes needed verbose init, repr, and eq. Dataclasses generate those automatically:

python

from dataclasses import dataclass, field
from typing import List

@dataclass
class Project:
name: str
slug: str
tags: List[str] = field(default_factory=list)
starred: bool = False

# ✅ Auto-generated __init__, __repr__, __eq__
p = Project(
name="Portfolio",
slug="portfolio",
tags=["nextjs", "tailwind"],
starred=True,
)
print(p)  # Project(name='Portfolio', slug='portfolio', ...)

# ✅ Frozen dataclass = immutable (no keyword mutability issues)
from dataclasses import dataclass

@dataclass(frozen=True)
class Point:
x: float
y: float

pt = Point(1.0, 2.0)
# pt.x = 3.0  # FrozenInstanceError

Feature	Regular Class	Dataclass
__init__	Manual boilerplate	Auto-generated
__repr__	Manual or ugly default	Clean: Class(field=value, ...)
__eq__	Identity only (unless manual)	Value equality
Immutability	Manual property setters	@dataclass(frozen=True)

Interactive Quiz

What does `@dataclass(frozen=True)` give you that a regular class doesn't?

Combine defaultdict with context managers for clean resource handling:

python

from collections import defaultdict
from contextlib import contextmanager
from threading import Lock

# Grouping | defaultdict eliminates "if key in dict" checks
grouped = defaultdict(list)
for item in items:
grouped[item.category].append(item)

# Custom context manager for timing
@contextmanager
def timer(label):
import time
start = time.perf_counter()
yield
elapsed = time.perf_counter() - start
print(f"{label}: {elapsed*1000:.2f}ms")

with timer("database query"):
results = db.query("SELECT * FROM users")

Interactive Quiz

What does `defaultdict(list)` eliminate?

Decision Reference

Problem	Solution	Why
Avoid KeyError on optional config	<code>dict.get(key, default)</code>	Graceful fallback, no try/except
Loop with index + value	<code>enumerate(items)</code>	Eliminates range(len(...)) anti-pattern
Pair two sequences	<code>zip(a, b)</code>	No manual index state
Fast membership check	<code>set</code>	O(1) vs O(n) for lists
Assign + test in one line	<code>:=</code> walrus	Remove duplicate function calls
Group by key	<code>defaultdict(list)</code>	No if key not in dict boilerplate
Typed record with methods	<code>@dataclass</code>	Auto-init/eq/repr, less boilerplate
Cross-platform paths	<code>pathlib.Path</code>	Fluent API, chainable

TL;DR

Start with these today

Minimum viable skill stack:

Use enumerate() and zip() - they eliminate 80% of loop bugs
Use dict.get() for optional configs - fail only where it's a real error
Use _ consistently - it signals intent to humans and linters
Profile before optimizing, but profile sets first - they're the fastest free win

Interactive Quiz

You have a list of 50,000 allowed actions and check `action in allowed_actions` 100 times per request. What's the single best optimization?

Python Tips Productivity

Python Fundamentals You Actually Need - Start here if you're new to Python
Modern Python Development Environment - Set up pyenv, Poetry, uv, and pyproject.toml
Python Practice: Automation Scripts - Real-world automation examples

Python Tricks

Tuple Unpacking

Click to reveal

Python Tricks

Assign multiple values at once: `a, b = b, a`. Works for swaps, function returns, and dict entries. Zero overhead tuple creation.

Tuple Unpacking

Python Tricks

enumerate()

Click to reveal

Python Tricks

Loop with index and value: `for i, item in enumerate(items, start=1)`. Replaces error-prone `range(len(items))` pattern.

enumerate()

Python Tricks

zip()

Click to reveal

Python Tricks

Iterate multiple sequences in parallel: `for a, b in zip(list1, list2)`. Stops at shortest sequence by default.

zip()

Python Tricks

dict.get()

Click to reveal

Python Tricks

Safe dictionary access with default: `config.get('port', 8080)`. Returns default if key missing - no KeyError.

dict.get()

Python Tricks

Set lookup

Click to reveal

Python Tricks

Convert list to set for O(1) membership: `if item in allowed_set`. Much faster than list O(n) for repeated checks.

Set lookup

Python Tricks

Walrus operator

Click to reveal

Python Tricks

Assignment expression `:=` - assign and test in one: `if (n := len(items)) > 10: ...`. Avoids duplicate function calls.

Walrus operator

Python Tricks

dataclass

Click to reveal

Python Tricks

Decorator that auto-generates __init__, __repr__, __eq__: `@dataclass class Point: x: float; y: float`. Less boilerplate.

dataclass

Python Tricks

pathlib.Path

Click to reveal

Python Tricks

Object-oriented file paths: `base / 'config' / f'{env}.yaml'`. Cross-platform, chainable methods, reads/writes text directly.

pathlib.Path

Audio Briefing• Voice: Google Cloud Journey-F

0:00 / --:--

The Core 7 (with real context)

1. Tuple Unpacking for Swaps

No temp variable, no ceremony:

python

a, b = b, a

This creates a tuple on the right, then destructures it on the left. Works in-place, zero overhead. The same pattern handles multi-values:

python

# Swap three values in one line
i, j, k = j, k, i

# Unpack a function's multiple return values
x_coord, y_coord, label = get_point()

# Swap dict entries
d = {'a': 1, 'b': 2}
d['a'], d['b'] = d['b'], d['a']  # Now {'a': 2, 'b': 1}

Interactive Quiz

What does `a, b = b, a` actually do under the hood?

2. enumerate() - Index + Value

Stop doing for i in range(len(items)). It's error-prone and reads backwards:

python

# ❌ Old way - off-by-one bugs hiding here
for i in range(len(colors)):
    print(i, colors[i])

# ✅ Python way - what you meant
for i, color in enumerate(colors):
    print(i, color)

Pro move: start at 1

enumerate(colors, start=1) gives 1-based numbering - useful for display, error messages, test cases.

Interactive Quiz

What's the benefit of `enumerate(items, start=1)`?

3. zip() - Loop Two Sequences in Parallel

Cleaner than manual index management when two sequences must stay in sync:

python

names    = ["Alice", "Bob", "Charlie"]
scores   = [85, 92, 78]
subjects = ["Math", "Science", "English"]

# Pairing two lists
for name, score in zip(names, scores):
print(f"{name}: {score}")

# Three-way zipping (tuples auto-unpacked)
for name, score, subject in zip(names, scores, subjects):
print(f"{name} ({subject}): {score}")

# Dictionaries from two lists
scores_dict = dict(zip(names, scores))
# {'Alice': 85, 'Bob': 92, 'Charlie': 78}

zip truncates to the shortest

If names has 3 items and scores has 5, only 3 pairs are produced. Use itertools.zip_longest if you want padding behavior.

Interactive Quiz

What happens when you zip two lists of different lengths?

4. dict.get() - Avoid KeyError

The difference between "missing key expected" and "missing key is a bug":

python

config = {"port": 3000, "host": "localhost"}

# Returns default if key missing - no exception
port     = config.get("port", 8080)     # 3000
debug    = config.get("debug", False)   # False (key missing)
host     = config.get("host", "0.0.0.0") # "localhost"

# None is the implicit default
value = config.get("nonexistent")       # None

Scenario	Use	Reason
Required config, missing = bug	<code>config["key"]</code>	Fail fast, surface error
Optional key, default OK	<code>config.get("key", default)</code>	Graceful fallback
Need to distinguish missing from None	<code>if "key" in config</code>	None is a valid value

Interactive Quiz

When should you use `config["key"]` instead of `config.get("key", default)`?

5. Set Lookups for O(1) Membership

Converting a list to a set before checking membership is one of the most common "hidden slowness" fixes:

python

allowed_actions = {"read", "write", "delete", "update"}

# O(1) - constant time regardless of list size
if action in allowed_actions:
execute(action)

# ❌ With a list, this is O(n) - scales poorly
allowed_actions_list = ["read", "write", ...]
if action in allowed_actions_list:  # Slower as list grows
execute(action)

Rule of thumb

If x in collection runs more than a handful of times per request and the collection has more than ~10 items, it should almost certainly be a set.

Interactive Quiz

Why is `item in some_set` faster than `item in some_list` for large collections?

6. Underscore `_` for Intentional Ignoring

Communicates "I'm not using this value on purpose":

python

# Unpacking - ignore the key, keep the value
_, result = parse_line(line)

# Loop - ignore the index
for _ in range(3):
print("repeat")

# Match with explicit wildcard (Python 3.10+)
match status:
case 200:
    handle_ok()
case 404:
    handle_missing()
case _:
    handle_other()  # "anything else"

# Unpacking partial - keep start/end, ignore middle
first, *_, last = [1, 2, 3, 4, 5]
# first=1, last=5

Interactive Quiz

What does `first, *_, last = [1, 2, 3, 4, 5]` produce?

Deep-Cut Techniques (Actually Useful)

python

# Without walrus
data = fetch_record()
if data:
process(data)

# With walrus :=
if data := fetch_record():
process(data)

When NOT to use walrus

Don't force it into existing clean code. If the assignment belongs outside the condition, put it there. The walrus shines in: loop guards, catch blocks, and reducing duplicated function calls.

Interactive Quiz

When is the walrus operator `:=` most valuable?

Python 3.12 added f-string debugging with = and multiline f-strings:

python

name, age = "Alice", 30

# Debug mode | shows expression + value
print(f"{name=} {age=}")  # name='Alice' age=30

# Multiline (no more backslash continuation)
report = f"""
Name:   {name}
Status: {"Adult" if age >= 18 else "Minor"}
Score:  {95.5:.1f}%
"""

Interactive Quiz

What does `f"{name=} {age=}"` output when name="Alice" and age=30?

Stop concatenating with os.path.join():

python

from pathlib import Path

base = Path("/app/data")

# Fluent chaining | works cross-platform
config = base / "config" / f"env_{ENV}.yaml"
logs   = base / "logs" / "app.log"

# Read/write without manual open()
text = config.read_text(encoding="utf-8")
config.write_text(new_text, encoding="utf-8")

# Existence checks
if config.exists() and config.is_file():
size = config.stat().st_size

# Iterate directory
for log_file in base.glob("**/*.log"):
print(log_file.name, log_file.stat().st_size)

Why pathlib over os.path?

Path uses operator overloading (/ joins paths) which reads naturally. Methods are chainable. The same code works on Windows and Linux without changes.

Interactive Quiz

What does `Path('/app') / 'data' / 'config.yaml'` produce?

Before dataclasses, classes with many attributes needed verbose init, repr, and eq. Dataclasses generate those automatically:

python

from dataclasses import dataclass, field
from typing import List

@dataclass
class Project:
name: str
slug: str
tags: List[str] = field(default_factory=list)
starred: bool = False

# ✅ Auto-generated __init__, __repr__, __eq__
p = Project(
name="Portfolio",
slug="portfolio",
tags=["nextjs", "tailwind"],
starred=True,
)
print(p)  # Project(name='Portfolio', slug='portfolio', ...)

# ✅ Frozen dataclass = immutable (no keyword mutability issues)
from dataclasses import dataclass

@dataclass(frozen=True)
class Point:
x: float
y: float

pt = Point(1.0, 2.0)
# pt.x = 3.0  # FrozenInstanceError

Feature	Regular Class	Dataclass
__init__	Manual boilerplate	Auto-generated
__repr__	Manual or ugly default	Clean: Class(field=value, ...)
__eq__	Identity only (unless manual)	Value equality
Immutability	Manual property setters	@dataclass(frozen=True)

Interactive Quiz

What does `@dataclass(frozen=True)` give you that a regular class doesn't?

Combine defaultdict with context managers for clean resource handling:

python

from collections import defaultdict
from contextlib import contextmanager
from threading import Lock

# Grouping | defaultdict eliminates "if key in dict" checks
grouped = defaultdict(list)
for item in items:
grouped[item.category].append(item)

# Custom context manager for timing
@contextmanager
def timer(label):
import time
start = time.perf_counter()
yield
elapsed = time.perf_counter() - start
print(f"{label}: {elapsed*1000:.2f}ms")

with timer("database query"):
results = db.query("SELECT * FROM users")

Interactive Quiz

What does `defaultdict(list)` eliminate?

Decision Reference

Problem	Solution	Why
Avoid KeyError on optional config	<code>dict.get(key, default)</code>	Graceful fallback, no try/except
Loop with index + value	<code>enumerate(items)</code>	Eliminates range(len(...)) anti-pattern
Pair two sequences	<code>zip(a, b)</code>	No manual index state
Fast membership check	<code>set</code>	O(1) vs O(n) for lists
Assign + test in one line	<code>:=</code> walrus	Remove duplicate function calls
Group by key	<code>defaultdict(list)</code>	No if key not in dict boilerplate
Typed record with methods	<code>@dataclass</code>	Auto-init/eq/repr, less boilerplate
Cross-platform paths	<code>pathlib.Path</code>	Fluent API, chainable

TL;DR

Start with these today

Minimum viable skill stack:

Use enumerate() and zip() - they eliminate 80% of loop bugs
Use dict.get() for optional configs - fail only where it's a real error
Use _ consistently - it signals intent to humans and linters
Profile before optimizing, but profile sets first - they're the fastest free win

Interactive Quiz

You have a list of 50,000 allowed actions and check `action in allowed_actions` 100 times per request. What's the single best optimization?

Python Tips Productivity

Python Fundamentals You Actually Need - Start here if you're new to Python
Modern Python Development Environment - Set up pyenv, Poetry, uv, and pyproject.toml
Python Practice: Automation Scripts - Real-world automation examples

The Core 7 (with real context)

1. Tuple Unpacking for Swaps

2. enumerate() - Index + Value

Pro move: start at 1

3. zip() - Loop Two Sequences in Parallel

zip truncates to the shortest

4. dict.get() - Avoid KeyError

When to use .get() vs direct access

5. Set Lookups for O(1) Membership

Rule of thumb

6. Underscore `_` for Intentional Ignoring

Deep-Cut Techniques (Actually Useful)

7. Walrus Operator `:=` Assignment Expressions

When NOT to use walrus

8. f-strings from Python 3.12+

9. pathlib | The Modern File Path Library

Why pathlib over os.path?

10. dataclasses | Less Boilerplate, Same Result

Bonus: Context Managers | With Defaultdict

Decision Reference

TL;DR

Start with these today

Related Posts

You Might Also Like

Stay Updated

Free In-Browser Developer Tools

Related Articles

The Python Fundamentals You Actually Need to Start

Python's Underscores: Conventions, Not Access Modifiers

A Practical pytest Tutorial That Actually Fixes Early Gotchas

The Core 7 (with real context)

1. Tuple Unpacking for Swaps

2. enumerate() - Index + Value

Pro move: start at 1

3. zip() - Loop Two Sequences in Parallel

zip truncates to the shortest

4. dict.get() - Avoid KeyError

When to use .get() vs direct access

5. Set Lookups for O(1) Membership

Rule of thumb

6. Underscore `_` for Intentional Ignoring

Deep-Cut Techniques (Actually Useful)

7. Walrus Operator `:=` Assignment Expressions

When NOT to use walrus

8. f-strings from Python 3.12+

9. pathlib | The Modern File Path Library

Why pathlib over os.path?

10. dataclasses | Less Boilerplate, Same Result

Bonus: Context Managers | With Defaultdict

Decision Reference

TL;DR

Start with these today

Related Posts

You Might Also Like

Stay Updated

Free In-Browser Developer Tools

Related Articles

The Python Fundamentals You Actually Need to Start

Python's Underscores: Conventions, Not Access Modifiers

A Practical pytest Tutorial That Actually Fixes Early Gotchas