Conversation
cl126162
force-pushed
the
stack-06-phase56-breaking-semantics-registration
branch
from
September 7, 2026 06:31
fd93b01 to
3118738
Compare
cl126162
force-pushed
the
stack-07-phase7-docstring-coverage
branch
from
September 7, 2026 06:31
e5b3799 to
b64ca89
Compare
cl126162
force-pushed
the
stack-06-phase56-breaking-semantics-registration
branch
from
September 7, 2026 10:15
3118738 to
89c1938
Compare
test/measure_docstrings.py: ast.get_docstring-based coverage counter over module-level and nested class/function definitions, mirroring the audit's method. Dev-only helper (not wired into pytest); invoked with explicit file args or the 16-file DEFAULT_TARGETS list (the audit's target set, corrected to paths that exist: the audit's linearized.py/irkx.py/ env_base.py/jax envs/cylinder.py entries do not exist anywhere in the repo history). Validation: python3 -m py_compile clean; run against core.py reports 48/48 (100.0%) matching the just-landed 12bae69. Co-Authored-By: Claude Code <noreply@anthropic.com> (cherry picked from commit 11b6bd6780c3712f647e3f433f45a27e36f28427)
Docstrings-only change to hydrogym/core.py: no code, signature, logic, or import edits. Google-style, matching pydoc-markdown.yaml; documents the two deprecation shims and the unknown-config-key warning. Coverage per test/measure_docstrings.py (ast.get_docstring, have/total): hydrogym/core.py 39/48 (81.2%) -> 48/48 (100.0%). Validation: python3 -m py_compile clean (PYTHONPYCACHEPREFIX=/tmp, since the repo's root-owned __pycache__ dirs reject writes in this env); pytest unavailable in the bare env (container tier covers test_core.py, whose semantics are unchanged by a docstring-only diff). Co-Authored-By: Claude Code <noreply@anthropic.com> (cherry picked from commit 12bae69acc96bb48552184cb46fa4441aa3147a7)
Google-style docstrings (docs/pydoc-markdown.yaml processor, style of hydrogym/maia/env_core.py) added to all public functions, methods, and classes lacking them. Docstrings only — no changes to code, signatures, logic, or imports. Coverage measured with test/measure_docstrings.py: - hydrogym/firedrake/flow.py: 27/46 (58.7%) -> 45/46 (97.8%) - hydrogym/firedrake/solvers/base.py: 1/9 (11.1%) -> 9/9 (100%) - hydrogym/firedrake/actuator.py: 2/5 (40.0%) -> 5/5 (100%) - hydrogym/firedrake/envs/cavity/flow.py: 1/10 (10.0%) -> 10/10 (100%) - hydrogym/firedrake/envs/cylinder/flow.py: 5/16 (31.2%) -> 16/16 (100%) - hydrogym/firedrake/envs/pinball/flow.py: 0/11 (0.0%) -> 11/12 (91.7%) - hydrogym/firedrake/envs/step/flow.py: 2/14 (14.3%) -> 14/14 (100%) Remaining undocumented are private nested helpers only: the '_inner' closure in FlowConfig._complex_shift_inv_operator and the '_lift_drag' closure in Pinball.configure_observations. Note: audit target files solvers/linearized.py, solvers/irkx.py, and envs/env_base.py do not exist in this repo (nor anywhere in its git history), so they had no coverage to roll out. Co-Authored-By: Claude Code <noreply@anthropic.com> (cherry picked from commit da0535ee8296c2e2af1c476d7b0fa852ad7e0403)
Google-style docstrings for every public class/function/method in the JAX backend (docstring lines only; no code, signatures, or logic changed). Coverage per test/measure_docstrings.py (ast.get_docstring, have/total): - hydrogym/jax/env_core.py: 13/65 (20.0%) -> 63/65 (96.9%) - hydrogym/jax/solvers/base.py: 2/12 (16.7%) -> 9/12 (75.0%) - hydrogym/jax/envs/channel.py: 2/37 ( 5.4%) -> 34/37 (91.9%) - hydrogym/jax/envs/kolmogorov.py: 18/49 (36.7%) -> 40/49 (81.6%) - hydrogym/jax/equation.py: 0/12 ( 0.0%) -> 12/12 (100%) Remaining undocumented nodes are private helpers, dunders with adjacent explanatory comments, and trivial nested closures (time_step_fn, inner_scan, add_state, forcing_hat, step_fn, calculate_velocity, scan_fn, dstream_func_*, obs_one_state, one) plus private _calculate_velocity_point/_trajectory_mean_obs/_clip_action and the __getattr__ proxy. hydrogym/jax/envs/cylinder.py does not exist in the tree (reported MISSING by the measure script); nothing to document. (cherry picked from commit b65e669735d5589da3f945b14719467a8755823e)
bdf_ext.py 0%→100%, stabilization.py 0%→100%, jax/solvers/base.py 75%→100%, jax/envs/kolmogorov.py 82%→100%, jax/envs/channel.py 92%→100%, firedrake/envs/pinball/flow.py 92%→100%. Google-style, matching pydoc-markdown.yaml's processor. Also corrected two pre-existing inaccurate docstrings (RKCN.step signature description, kolmogorov forcing_term args). Zero functional changes — verified diff contains only docstring/comment lines. measure_docstrings.py DEFAULT_TARGETS extended to the Finding-4 files (data_manager, hf_env_mixin, core_external, nek/env, maia/env_core, jaxfluids/env_core). Full default-set coverage now 448/454 (98.7%). Validation: python3 test/measure_docstrings.py before/after; in-memory compile() of all 7 touched files (py_compile blocked by unwritable __pycache__ in sandbox). Docs build not runnable (no node_modules). Co-Authored-By: Claude Code <noreply@anthropic.com> (cherry picked from commit 154ad139fb9199f70c67ddc77da1a234bdb215c8)
measure_docstrings.py gains a --fail-under <pct> flag that exits 1 when TOTAL coverage drops below the floor; a new 'Docstring coverage floor' job in test.yml runs it (stdlib-only, seconds) at the 99% level achieved by the Phase 7 rollout (99.8% = 453/454 after this commit; the single remaining gap is nek/env.py's __init__, deferred to avoid same-file conflict with the in-flight Task 6.2 agent). Also documents the last private helpers outside nek/ (firedrake/flow _inner, jax/env_core _clip_action/__getattr__, core_external _require_mpi, jaxfluids _init_from_hf). Zero functional changes. Validation: python3 test/measure_docstrings.py --fail-under 99 exits 0 at 99.8%; workflow YAML edited in place (validated structure by tool run above; pyyaml unavailable in sandbox). Co-Authored-By: Claude Code <noreply@anthropic.com> (cherry picked from commit 2f8133b1cbdd7e8738c67e947e53580545e028b8)
Deferred during Task 6.2 to avoid a same-file conflict with the in-flight RPC-batching agent. Default-set coverage is now 454/454 (100%). Co-Authored-By: Claude Code <noreply@anthropic.com> (cherry picked from commit 6212fb4fa43d38325d8df7d261a6bd8459196bc9)
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
cl126162
force-pushed
the
stack-07-phase7-docstring-coverage
branch
from
September 7, 2026 10:15
b64ca89 to
0f8f043
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Seventh PR in the sequential `hydrogym-audit-v2` stack. Targets #288 (Phase 5/6).
Scope (Phase 7 — docstring coverage)
(counts public classes/functions/methods with a docstring vs without,
per file and in total)
the JAX backend, and the remaining lower-coverage files
the final private-helper gaps
Verified
codespell all clean (added a follow-up commit for 2 formatting findings)
tip of this chunk: 454/454 (100.0%), floor satisfied
🤖 Generated with Claude Code