Tweet
Tweet

Slide 1

Slide 1 text

OF SENIOR ENGINEERS CODE REVIEW PYTHON Keith Yang

Slide 2

Slide 2 text

Agenda Six Key Points of Senior Python Code Review Supportive Environment and Tools around Pull-Request 1. 2. 2023 talk in PyCon TW

Slide 3

Slide 3 text

Anti-Agenda How-to Details Find A Good Job πŸ‘οΈ πŸ‘€ πŸ‘οΈβ€πŸ—¨οΈ πŸ”Ž πŸ” πŸ”¦ 1. 2.

Slide 4

Slide 4 text

Been there reviewed that? You've got a review

Slide 5

Slide 5 text

Pull-Request (PR) > A pull request is a method used in collaborative coding projects to propose code changes. > It allows developer to request code review, and then merge those changes into the main code repository once approved.

Slide 6

Slide 6 text

Keith Yang Lead Backend Engineer in iCHEF Taipei.py co-founder and co-organizer PyCon APAC 2015 Chairperson Lead Backend Engineer in iCHEF Taipei.py co-founder and co-organizer PyCon APAC 2015 Chairperson

Slide 7

Slide 7 text

No content

Slide 8

Slide 8 text

No content

Slide 9

Slide 9 text

6 Key Points of Senior Python Code Review Agenda 1

Slide 10

Slide 10 text

6 Key Points of Senior Python Code Review Agenda 1 9,999 From Years of Practice

Slide 11

Slide 11 text

Code Review

Slide 12

Slide 12 text

Quality Assurance Knowledge Sharing Finding and Fixing Bugs Early Code Review

Slide 13

Slide 13 text

... to submit a change to the kernel, ... can be daunting ... if not familiar with "the system" - Submitting patches guide (Linux kernel) git send-email

Slide 14

Slide 14 text

No content

Slide 15

Slide 15 text

No content

Slide 16

Slide 16 text

#Comment Code Readibilirty Make code easy to be understood If it's tricky, write comments to help understanding 1st key point

Slide 17

Slide 17 text

GUESSING GUESSING PURPOSE PURPOSE

Slide 18

Slide 18 text

#Comment Example 1

Slide 19

Slide 19 text

No content

Slide 20

Slide 20 text

#Comment Example 1

Slide 21

Slide 21 text

#Comment Example 2

Slide 22

Slide 22 text

Concise Code Descriptive for Human Purpose instead of Function 2nd key point

Slide 23

Slide 23 text

Code Example

Slide 24

Slide 24 text

Naming Descriptive for Human Purpose instead of Function Uniqueness v.s. Common easy to understand 3rd key point

Slide 25

Slide 25 text

Uniqueness "user" in users.py "member" in members.py "user" in crm.py "member" in crm.py Easiest to find and reason? a. b. c. d.

Slide 26

Slide 26 text

Uniqueness "user" in users.py "member" in members.py "user" in crm.py "member" in crm.py Easiest to find and reason? a. b. c. d. Tool: ripgrep

Slide 27

Slide 27 text

EXAMPLE Naming in member_logic.py, not member_utils.py

Slide 28

Slide 28 text

Code & Infra Readability Make code and module easy to understand If it's tricky, write comments and doc to help understanding

Slide 29

Slide 29 text

The Zen of Python The Zen of Python

Slide 30

Slide 30 text

python3 -m this The Zen of Python

Slide 31

Slide 31 text

Beautiful is better than ugly. Explicit is better than implicit. Simple is better than complex. Complex is better than complicated. Flat is better than nested. Sparse is better than dense. Readability counts. Special cases aren't special enough to break the rules. Although practicality beats purity. Errors should never pass silently. Unless explicitly silenced. In the face of ambiguity, refuse the temptation to guess. There should be one-- and preferably only one --obvious way to do it. Although that way may not be obvious at first unless you're Dutch. Now is better than never. Although never is often better than *right* now. If the implementation is hard to explain, it's a bad idea. If the implementation is easy to explain, it may be a good idea. Namespaces are one honking great idea -- let's do more of those! The Zen of Python, by Tim Peters python3 -m this

Slide 32

Slide 32 text

Beautiful is better than ugly. Explicit is better than implicit. Simple is better than complex. CODE CODE NAMING NAMING COMMENT COMMENT

Slide 33

Slide 33 text

CODE NAMING COMMENT Readability counts. ...easy to explain, it may be a good idea Namespaces are one honking great idea

Slide 34

Slide 34 text

6 Key Points of Senior Python Code Review, Part II

Slide 35

Slide 35 text

Unused Code Linting Tools: ruff, flake8, pylint Avoid writing unused: variables imports functions classes 4th key point

Slide 36

Slide 36 text

EXAMPLE Unused Code

Slide 37

Slide 37 text

Check Unused EXAMPLE by pre-commit

Slide 38

Slide 38 text

Senior Wrote Knew Common Python Errors Defaults for function arguments Modifying a list while iterating over it Lambda in for loop? MISTAKES 5th key point

Slide 39

Slide 39 text

Been there Done that πŸ₯²

Slide 40

Slide 40 text

Lambda Code EXAMPLE in ITERRATION Expecting [0, 1, 2, 3, 4]?

Slide 41

Slide 41 text

PEP8 Read the doc? Tools black, isort flake8, ruff pre-commit 6th key point

Slide 42

Slide 42 text

PEP8 Check Real-time in Vim EXAMPLE

Slide 43

Slide 43 text

Type Hint Does Python need type hint? extra key point

Slide 44

Slide 44 text

Type Hint But it can be used to find mistakes early Small Projects v.s. Large Enterprise Projects No. extra key point

Slide 45

Slide 45 text

Supportive Environment and Tools around Pull-Request Agenda 2

Slide 46

Slide 46 text

β€œOne essential characteristic of modern life is that we all depend on systems” -- The Checklist Manifesto: How to Get Things Right by Atul Gawande

Slide 47

Slide 47 text

Why to and why not Commit Message Pull-Request Description Document File Documentation Documentation Complete the Code Review

Slide 48

Slide 48 text

Purposes and Reasons Implement Plan Do This Bο»Ώ efore Coding Design Review Design Review Complete the Code Review

Slide 49

Slide 49 text

Template Rocks 🀘 Design Review From Years of Practice

Slide 50

Slide 50 text

Operation Review Scop User Stories Security Review Inspirations Success Criteria B Plan (if not working) Designs (API, Model, DB) Performance Review Alternatives Considered Template Example & Tips

Slide 51

Slide 51 text

SRE Design Review Template Pull-Request Template Project Time Evaluation Template ... Template Everywhere Template Everywhere

Slide 52

Slide 52 text

Document Tool Exmaple Document Tool Exmaple .TXT, .md, .rst Notion, Amplenote, ... Notion, Google Doc, CryptPad, ... Figma, Archi, ...

Slide 53

Slide 53 text

CI / CD Continuous Integration Continuous Deployment Code ⬇️ Build ⬇️ Test --- ⬇️--- Staging ⬇️ Production

Slide 54

Slide 54 text

CI / CD Replace Painful Routines Reduced Manual Errors Rollbacks and Recovery

Slide 55

Slide 55 text

CI in Pull-Request

Slide 56

Slide 56 text

CI / CD Tool Example CI/CD pipelines Jenkins GitHub Actions PipeRider.io (data, dbt) pre-commit flake8, ruff, black, isort

Slide 57

Slide 57 text

"Spot issues before they emerge" ~ PipeRider ( instead of γ€€γ€€γ€€γ€€γ€€γ€€γ€€ Hotfix on Production )

Slide 58

Slide 58 text

Code Coverage Untested Code Collaboration Surprising Errors

Slide 59

Slide 59 text

Code Coverage Tool Example pytest-cov (ο»Ώ coverage.py) Codecov (by Sentry) SonarQube (SonarCloud)

Slide 60

Slide 60 text

Code Coverage checked in Pull-Request

Slide 61

Slide 61 text

The Zen of Python The Zen of Python

Slide 62

Slide 62 text

Errors should never pass silently Unless explicitly silenced Special cases aren't special enough to break the rules DOCUMENT DOCUMENT TEMPLATE TEMPLATE CI / CD COVERAGE CI / CD COVERAGE

Slide 63

Slide 63 text

Errors should never pass silently Unless explicitly silenced Special cases aren't special enough to break the rules DOCUMENT DOCUMENT TEMPLATE TEMPLATE CI / CD COVERAGE CI / CD COVERAGE TYPE HINT TYPE HINT

Slide 64

Slide 64 text

DOCUMENT TEMPLATE CI / CD There should be one -- and preferably only one -- obvious way to do it. TYPE HINT

Slide 65

Slide 65 text

AI Tools GPT4 ChatGPT Bard Claude GitHub Copilot

Slide 66

Slide 66 text

AI Tools Prompt vs real experienced Domain Knowledge Identify what's right and wrong with official doc Experiment tools and grow with your team

Slide 67

Slide 67 text

Mentorship Design Review Weekly Meeting Sharing Fun and Pain Teamwork

Slide 68

Slide 68 text

Internet AI Books, Podcast, Video (ο»Ώ O'Reilly Learning) Communities & (Py)Cons Different Languages & Design Paradigms & Pitfalls def Learning(self) Taiwan /APAC

Slide 69

Slide 69 text

Search on Internet Reference vs Pragmatic Mentor & Boss Interview & Culture Care Framework Care Framework Ladder: Your Mileage May Vary

Slide 70

Slide 70 text

Summary From Comment to Naming: Explicit Common Errors Supportive Environment with Tools: Find Errors at Very First Place Design Template CI/CD, AI, Soft Skills Senior: More Helpful

Slide 71

Slide 71 text

OF SENIOR ENGINEERS CODE REVIEW PYTHON @keitheis @mstdn.io 𝕏 discord @keitheis @mstdn.io 𝕏 discord Keith Yang