jsoncompat/Justfile at main · ostrowr/jsoncompat · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
help:
  @just --list

check:
  @echo "[just] checking Rust code …"
  cargo fmt --all
  @echo "[just] building Python extension …"
  if [ "$(uname)" = "Darwin" ]; then \
    cargo rustc -p jsoncompat_py --lib -- -C link-arg=-undefined -C link-arg=dynamic_lookup; \
  else \
    cargo build -p jsoncompat_py; \
  fi
  cargo clippy --workspace --all-features --all-targets --locked -- -D warnings
  cargo test --workspace --exclude jsoncompat_py --all-features --locked
  @echo "[just] checking Python code …"
  env -u VIRTUAL_ENV -u UV_DEFAULT_INDEX -u UV_INDEX -u UV_INDEX_URL -u UV_EXTRA_INDEX_URL uv run --no-config --project pybindings --all-extras --group benchmark --locked --with pyright==1.1.408 pyright
  @echo "[just] checking TypeScript code …"
  pnpm --prefix web/jsoncompatdotcom run ci
  pnpm --prefix web/jsoncompatdotcom run build

regen-dataclasses-fixtures:
  JSONCOMPAT_UPDATE_DATACLASSES_FIXTURES=1 cargo test --test dataclasses_fixtures -- --exact dataclass_snapshots_are_up_to_date_for_all_sample_schemas

bench:
  @echo "[just] running Rust benchmarks …"
  cargo bench --workspace --exclude jsoncompat_py --all-features --bench '*' --locked

bench-check:
  @echo "[just] smoke-checking Rust benchmarks …"
  cargo bench --workspace --exclude jsoncompat_py --all-features --bench '*' --locked -- --test

python_bench_command := "env -u VIRTUAL_ENV -u UV_DEFAULT_INDEX -u UV_INDEX -u UV_INDEX_URL -u UV_EXTRA_INDEX_URL PYTHONHASHSEED=0 JSONCOMPAT_NATIVE_PROFILE=release uv run --no-config --project pybindings --all-extras --group benchmark --locked python"

_build-python-release:
  @echo "[just] building release Python extension for representative timings …"
  if [ "$(uname)" = "Darwin" ]; then \
    cargo rustc --release -p jsoncompat_py --lib -- -C link-arg=-undefined -C link-arg=dynamic_lookup; \
  else \
    cargo build --release -p jsoncompat_py; \
  fi

[private]
_verify-dataclass-fixtures:
  @echo "[just] verifying checked-in generated dataclass fixtures are current …"
  cargo test --test dataclasses_fixtures -- --exact dataclass_snapshots_are_up_to_date_for_all_sample_schemas

[private]
_python-bench-provenance profile:
  @mkdir -p target/python-benchmark
  @git status -sb > "target/python-benchmark/provenance-{{profile}}.txt"
  @git rev-parse HEAD >> "target/python-benchmark/provenance-{{profile}}.txt"
  @cat "target/python-benchmark/provenance-{{profile}}.txt"

[private]
_python-bench-runtime iterations repeats profile: _verify-dataclass-fixtures
  @echo "[just] benchmarking generated Python dataclasses against Pydantic v2 …"
  @mkdir -p target/python-benchmark
  {{python_bench_command}} pybindings/bench_dataclasses_runtime.py --iterations {{iterations}} --repeats {{repeats}} > "target/python-benchmark/runtime-{{profile}}.txt"
  @cat "target/python-benchmark/runtime-{{profile}}.txt"

[private]
_python-bench-startup repeats profile: _verify-dataclass-fixtures
  @echo "[just] benchmarking cold generated-model startup against Pydantic v2 …"
  @mkdir -p target/python-benchmark
  {{python_bench_command}} pybindings/bench_dataclasses_startup.py --repeats {{repeats}} > "target/python-benchmark/startup-{{profile}}.txt"
  @cat "target/python-benchmark/startup-{{profile}}.txt"

[private]
_python-bench-scale depth fanout iterations repeats profile: _verify-dataclass-fixtures
  @echo "[just] benchmarking a large recursive dataclass graph against Pydantic v2 …"
  @mkdir -p target/python-benchmark
  {{python_bench_command}} pybindings/bench_dataclasses_scaling.py --depth {{depth}} --fanout {{fanout}} --iterations {{iterations}} --repeats {{repeats}} > "target/python-benchmark/scaling-{{profile}}.txt"
  @cat "target/python-benchmark/scaling-{{profile}}.txt"

[private]
_python-bench-scale-profile depth fanout iterations repeats profile: _verify-dataclass-fixtures
  @echo "[just] profiling checked and trusted recursive JSON deserialization …"
  @mkdir -p target/python-benchmark
  {{python_bench_command}} pybindings/bench_dataclasses_scaling.py --depth {{depth}} --fanout {{fanout}} --iterations {{iterations}} --repeats {{repeats}} --profile > "target/python-benchmark/profile-{{profile}}.txt"
  @cat "target/python-benchmark/profile-{{profile}}.txt"

[private]
_python-bench-fixtures iterations repeats profile: _verify-dataclass-fixtures
  @echo "[just] benchmarking fixture JSON -> generated dataclass -> JSON; comparing semantically equivalent Pydantic v2 models …"
  {{python_bench_command}} pybindings/bench_fixture_models.py --iterations {{iterations}} --repeats {{repeats}} --profile "{{profile}}" --reuse-models --results "target/python-fixture-benchmark/results-{{profile}}.json"

[private]
_python-bench-fixtures-limited iterations repeats limit profile: _verify-dataclass-fixtures
  @echo "[just] smoke-benchmarking fixture JSON -> generated dataclass -> JSON for the first {{limit}} schemas …"
  {{python_bench_command}} pybindings/bench_fixture_models.py --iterations {{iterations}} --repeats {{repeats}} --limit {{limit}} --profile "{{profile}}" --reuse-models --results "target/python-fixture-benchmark/results-{{profile}}.json"

# Benchmark the representative small generated-model graph.
python-bench iterations="10000" repeats="5": _build-python-release (_python-bench-provenance "manual") (_python-bench-runtime iterations repeats "manual")

# Benchmark fresh-interpreter import and first-use costs.
python-bench-startup repeats="25": _build-python-release (_python-bench-provenance "startup") (_python-bench-startup repeats "startup")

# Benchmark the recursive graph at a caller-selected size.
python-bench-scale depth="5" fanout="4" iterations="100" repeats="5": _build-python-release (_python-bench-provenance "manual") (_python-bench-scale depth fanout iterations repeats "manual")

# Benchmark JSON -> generated dataclass -> JSON for every supported fixture and compare semantically equivalent Pydantic peers.
python-bench-fixtures iterations="200" repeats="5": _build-python-release (_python-bench-provenance "manual") (_python-bench-fixtures iterations repeats "manual")

# Quickly smoke-test all four Python benchmark surfaces; fixture E2E timings and ratios cover the first 50 schemas.
python-bench-quick: _build-python-release (_python-bench-provenance "quick") (_python-bench-startup "10" "quick") (_python-bench-runtime "2000" "3" "quick") (_python-bench-scale "4" "3" "25" "3" "quick") (_python-bench-fixtures-limited "25" "3" "50" "quick")

# Run the standard repeatable suite, including every fixture E2E round trip and semantically equivalent Pydantic comparison.
python-bench-standard: _build-python-release (_python-bench-provenance "standard") (_python-bench-startup "25" "standard") (_python-bench-runtime "10000" "5" "standard") (_python-bench-scale "5" "4" "100" "5" "standard") (_python-bench-fixtures "200" "5" "standard")

# Run a high-sample suite over every fixture E2E round trip, with Pydantic comparisons, for release performance reports.
python-bench-full: _build-python-release (_python-bench-provenance "full") (_python-bench-startup "100" "full") (_python-bench-runtime "50000" "10" "full") (_python-bench-scale "6" "4" "100" "10" "full") (_python-bench-fixtures "1000" "10" "full")

# Profile checked and trusted deserialization on the standard recursive graph.
python-bench-profile depth="5" fanout="4" iterations="100" repeats="5": _build-python-release (_python-bench-provenance "profile") (_python-bench-scale-profile depth fanout iterations repeats "standard")

# ---- Basic python smoke test ----

python-demo:
  env -u VIRTUAL_ENV uv run --reinstall-package jsoncompat examples/python/basic/demo.py

# ---- Basic javascript smoke test ----

wasm-demo:
  @command -v wasm-pack >/dev/null 2>&1 || (echo "error: wasm-pack not found (install with 'cargo install wasm-pack --locked')" >&2 && exit 1)
  @command -v uv >/dev/null 2>&1 || (echo "error: uv not found" >&2 && exit 1)
  @echo "[just] building wasm package for the Web target …"
  @wasm-pack build wasm --target web --release
  @echo "[just] serving example at http://localhost:8000/examples/wasm/demo.html …"
  @echo "Press Ctrl+C to stop."
  @env -u VIRTUAL_ENV uv run python -m http.server 8000

release version="patch":
  @echo "[just] releasing {{version}} (dry run)"
  cargo release {{version}} --workspace