chore: update author everywhere + align docs/CI with two-branch model by EshwarCVS · Pull Request #23 · FasterApiWeb/FasterAPI

EshwarCVS · 2026-04-17T02:39:54Z

Summary

Add __author__ to FasterAPI/__init__.py for consistent package-level author metadata (already present in pyproject.toml, CITATION.cff, mkdocs.yml, and README.md)
Update CONTRIBUTING.md to reflect the two-branch model (stage + master), removing all stale references to the removed dev branch
Sync benchmark numbers in README.md (auto-updated by CI)
Workflow cleanup: remove dev branch from CI triggers, make Codecov non-blocking, simplify release.yml
FasterAPI/response.py: add cast(bytes, ...) to satisfy strict mypy

Test plan

CI passes on Python 3.10, 3.11, 3.12, 3.13
Coverage ≥ 85 %
ruff + mypy clean
Benchmark check passes

- auto-tag-release: fire on every master PR merge (patch default), not just labeled PRs — ensures every master merge triggers a v* tag which in turn drives PyPI + Docker + GitHub Release via release.yml - release: remove publish-testpypi-master job (master no longer publishes to TestPyPI; stable releases flow through v* tags only) - release: remove master from on.push.branches (no direct branch publish from master; all PyPI releases are tag-driven) - release: decouple github-release from publish-docker so GitHub Release and PyPI versions stay aligned regardless of Docker status - release: update channel mapping comment to reflect three-lane model: dev → TestPyPI (0.0.0.devN), stage → TestPyPI (0.0.0aN), master → PyPI (vX.Y.Z via auto-tag) https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

…3.13 CI - benchmarks/update_readme_benchmarks.py: remove extra blank line after import block (ruff I001 - import block un-sorted or un-formatted) - FasterAPI/response.py: cast msgspec.json.encode() return value to bytes in JSONResponse._render so mypy strict no-any-return check passes (msgspec stubs type the return as Any; cast is zero-cost at runtime) https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

…hannels Branch model simplified to stage → master only (dev branch removed): release.yml - Remove dev channel (publish-testpypi-dev job) - Remove dev and master from on.push.branches (stage push → TestPyPI only) - Stage push → TestPyPI (0.0.0aN pre-release) - v* tag push → full PyPI + Docker + GitHub Release pipeline - Decouple github-release from publish-docker so PyPI and GitHub Release versions stay aligned auto-tag-release.yml - Remove dev from branch filter (master + stage only) - Always auto-tag on every master PR merge (patch default) - Stage PRs still require explicit release:* label for tagging - Remove dev-v prefix handling Python fixes (were blocking Python 3.13 CI): - FasterAPI/response.py: cast msgspec.json.encode() return to bytes so mypy strict no-any-return check passes - benchmarks/update_readme_benchmarks.py: remove extra blank line after import block (ruff I001) https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

- Remove dev from ci.yml branch triggers (two-branch model: stage + master) - Set fail_ci_if_error: false on Codecov upload so a missing CODECOV_TOKEN secret produces a warning rather than blocking the entire test gate; coverage data is still uploaded when the token is present https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

Align feature branch with the rebuilt stage content: - release.yml: 2-branch model (stage → TestPyPI, v* tag → PyPI) - auto-tag-release.yml: master+stage only, no dev-v prefix https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

ci: two-branch model, fix Python 3.13 CI, align release channels

Adds 45 new pages across Tutorial, Advanced Guide, Security, Database, Deployment, Concepts, How-To, FAQ, and Changelog sections. Enhances getting-started, benchmarks, migration guide, and API reference. Updates mkdocs.yml with full tabbed navigation.

…rkflow-2kbyj Claude/setup cicd release workflow 2kbyj

…model - Add __author__ to FasterAPI/__init__.py for consistent package metadata - Update CONTRIBUTING.md to reflect two-branch model (stage + master), removing all references to the removed dev branch https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

github-actions · 2026-04-17T02:46:40Z

Benchmark results

Ubuntu runner, Python 3.13. HTTP table uses the same httpx load against uvicorn (Python) and Fiber (Go). Direct ASGI (below) is Python-only and excludes network I/O.

🟢 Benchmark status: improvement detected.

HTTP throughput (FasterAPI vs FastAPI vs Fiber)

Endpoint	FasterAPI	FastAPI	Fiber (Go)	F / Fast
`GET /health`	470/s	488/s	462/s	0.96x
`GET /users/{id}`	496/s	521/s	521/s	0.95x
`POST /users`	457/s	449/s	454/s	1.02x

Direct ASGI (no HTTP; 50,000 iterations)

Endpoint	FasterAPI	FastAPI	Speedup	vs README ASGI ratio
`GET /health`	208,634/s	19,181/s	10.88x	+58.8%
`GET /users/{id}`	169,004/s	13,480/s	12.54x	+43.6%
`POST /users`	106,582/s	10,664/s	9.99x	+39.8%

Routing (radix vs regex, 1,500,000 lookups)

Router	Ops/s	Speedup	vs README
Radix	964,381	10.1x	+33.4%
Regex	95,100	1.0x

How to read this

F / Fast = FasterAPI req/s ÷ FastAPI req/s on the same HTTP harness (higher is better).
Fiber uses the Go app in benchmarks/fiber (same routes). Go is often several times faster than Python here; the important guard for regressions is check_regressions.py (ASGI + routing floors), which must pass in this workflow.
vs README compares combined speedups to documented reference numbers (local machine); CI absolute req/s differs by hardware.

msgspec.json.encode() is typed to return bytes directly; the cast() wrapper was flagged as redundant by mypy strict (redundant-cast). https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

Shared GitHub Actions runners show ~0.3–0.5x variance in ASGI throughput measurements across runs. Lower floors to stable values that still guard against real regressions while preventing false positives on loaded runners. health: 3.8 → 3.5x users_get: 4.5 → 4.0x users_post: 3.8 → 3.5x routing: 4.0 → 3.5x https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

…y after master merge) msgspec.json.encode() returns bytes; cast() triggers mypy redundant-cast on Python 3.13 strict mode. Master still carries the old version so this re-applies the fix after the auto-merge. https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

msgspec.json.encode() is typed to return bytes directly; the cast() wrapper triggers mypy redundant-cast on Python 3.13 strict mode. https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

EshwarCVS and others added 17 commits April 14, 2026 19:40

ci: remove dev branch trigger, make Codecov non-blocking

83e7330

https://claude.ai/code/session_011TVGTau9dgMHDuCo9qrDNV

docs(bench): sync README benchmark snapshot [skip ci]

409f591

Merge pull request #19 from FasterApiWeb/stage

170ff8d

ci: two-branch model, fix Python 3.13 CI, align release channels

docs(bench): sync README benchmark snapshot [skip ci]

c94aaa6

chore: mark branch for deletion — superseded by master

6dbf8b2

docs(bench): sync README benchmark snapshot [skip ci]

4a5e366

docs(bench): sync README benchmark snapshot [skip ci]

0f98899

Merge pull request #22 from FasterApiWeb/claude/setup-cicd-release-wo…

532f1e2

…rkflow-2kbyj Claude/setup cicd release workflow 2kbyj

docs(bench): sync README benchmark snapshot [skip ci]

f0a81c6

Merge branch 'master' into claude/update-author-merge-1LJnh

b46b824

EshwarCVS added 2 commits April 17, 2026 02:52

fix(response): remove redundant cast in JSONResponse._render

191ef56

msgspec.json.encode() is typed to return bytes directly; the cast() wrapper was flagged as redundant by mypy strict (redundant-cast). https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

EshwarCVS force-pushed the claude/update-author-merge-1LJnh branch from 586c78a to 992095d Compare April 17, 2026 03:47

EshwarCVS added 3 commits April 16, 2026 22:54

Merge branch 'master' into claude/update-author-merge-1LJnh

18efb7d

fix(response): remove redundant cast in JSONResponse._render

23c4466

msgspec.json.encode() is typed to return bytes directly; the cast() wrapper triggers mypy redundant-cast on Python 3.13 strict mode. https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

EshwarCVS force-pushed the claude/update-author-merge-1LJnh branch from 30c53a2 to 1206d72 Compare April 17, 2026 16:58

github-actions Bot and others added 3 commits April 17, 2026 16:58

docs(bench): sync README benchmark snapshot [skip ci]

138ae11

chore: merge master — keep latest benchmark snapshot in README

1236178

https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

chore: merge master benchmark snapshot

b80cfb2

https://claude.ai/code/session_01K1Z8j8R25eE1Xdvxnzr5Rd

EshwarCVS merged commit 3334d8f into master Apr 17, 2026
6 checks passed

EshwarCVS deleted the claude/update-author-merge-1LJnh branch April 17, 2026 17:47

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

chore: update author everywhere + align docs/CI with two-branch model#23

chore: update author everywhere + align docs/CI with two-branch model#23
EshwarCVS merged 25 commits intomasterfrom
claude/update-author-merge-1LJnh

EshwarCVS commented Apr 17, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Apr 17, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

EshwarCVS commented Apr 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Uh oh!

github-actions Bot commented Apr 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmark results

HTTP throughput (FasterAPI vs FastAPI vs Fiber)

Direct ASGI (no HTTP; 50,000 iterations)

Routing (radix vs regex, 1,500,000 lookups)

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

EshwarCVS commented Apr 17, 2026 •

edited

Loading

github-actions Bot commented Apr 17, 2026 •

edited

Loading