refactor: use project-named state file and store prompt/params directly
All checks were successful
Continuous Integration / Build Package (push) Successful in 48s
Continuous Integration / Lint, Check & Test (push) Successful in 1m1s

- State filename now derives from config: cards.bulkgen.yaml produces
  .cards.bulkgen-state.yaml instead of .bulkgen.state.yaml
- Store resolved prompt text and extra params directly in state file
  instead of hashing them, making state files human-readable
- Only file input contents remain hashed (SHA-256)
- Thread project_name through builder and CLI
- Remove hash_string() and _extra_hash() helpers
- Update .gitignore pattern to .*.bulkgen-state.yaml
This commit is contained in:
Konstantin Fickel 2026-02-15 13:56:12 +01:00
parent 870023865d
commit 0ecf1f0f9e
Signed by: kfickel
GPG key ID: A793722F9933C1A5
7 changed files with 98 additions and 82 deletions

View file

@ -147,6 +147,7 @@ async def _build_single_target(
async def run_build(
config: ProjectConfig,
project_dir: Path,
project_name: str,
target: str | None = None,
on_progress: ProgressCallback = _noop_callback,
) -> BuildResult:
@ -171,7 +172,7 @@ async def run_build(
raise ValueError(msg)
graph = get_subgraph_for_target(graph, target)
state = load_state(project_dir)
state = load_state(project_dir, project_name)
generations = get_build_order(graph)
target_names = set(config.targets)
@ -209,7 +210,7 @@ async def run_build(
)
_process_outcomes(outcomes, config, project_dir, state, result, on_progress)
save_state(state, project_dir)
save_state(state, project_dir, project_name)
return result

View file

@ -14,9 +14,21 @@ from bulkgen.builder import BuildEvent, BuildResult, run_build
from bulkgen.config import ProjectConfig, load_config
from bulkgen.graph import build_graph, get_build_order
from bulkgen.providers.registry import get_all_models
from bulkgen.state import state_filename
app = typer.Typer(name="bulkgen", help="AI artifact build tool.")
_CONFIG_SUFFIX = ".bulkgen.yaml"
def _project_name(config_path: Path) -> str:
"""Derive the project name from a config path.
``cards.bulkgen.yaml`` ``cards``
"""
name = config_path.name
return name.removesuffix(_CONFIG_SUFFIX)
def _find_config(directory: Path) -> Path:
"""Find the single ``*.bulkgen.yaml`` file in *directory*."""
@ -49,7 +61,7 @@ def _format_elapsed(seconds: float) -> str:
def _run_build(
config: ProjectConfig, project_dir: Path, target: str | None
config: ProjectConfig, project_dir: Path, project_name: str, target: str | None
) -> tuple[BuildResult, float]:
"""Run the async build with click-styled progress output and timing."""
@ -85,7 +97,7 @@ def _run_build(
start = time.monotonic()
result = asyncio.run(
run_build(config, project_dir, target, on_progress=on_progress)
run_build(config, project_dir, project_name, target, on_progress=on_progress)
)
elapsed = time.monotonic() - start
@ -102,10 +114,11 @@ def build(
project_dir = Path.cwd()
config_path = _find_config(project_dir)
config = load_config(config_path)
name = _project_name(config_path)
click.echo(click.style("bulkgen", fg="cyan", bold=True) + " building targets...\n")
result, elapsed = _run_build(config, project_dir, target)
result, elapsed = _run_build(config, project_dir, name, target)
# Summary
click.echo("")
@ -131,6 +144,7 @@ def clean() -> None:
project_dir = Path.cwd()
config_path = _find_config(project_dir)
config = load_config(config_path)
state_name = state_filename(_project_name(config_path))
removed = 0
for target_name in config.targets:
@ -140,10 +154,10 @@ def clean() -> None:
click.echo(click.style(" rm ", fg="red") + target_name)
removed += 1
state_path = project_dir / ".bulkgen.state.yaml"
state_path = project_dir / state_name
if state_path.exists():
state_path.unlink()
click.echo(click.style(" rm ", fg="red") + ".bulkgen.state.yaml")
click.echo(click.style(" rm ", fg="red") + state_name)
click.echo(click.style(f"\nCleaned {removed} artifact(s)", bold=True))

View file

@ -1,4 +1,4 @@
"""Incremental build state tracking via ``.bulkgen.state.yaml``."""
"""Incremental build state tracking via ``.<project>.bulkgen-state.yaml``."""
from __future__ import annotations
@ -8,16 +8,23 @@ from pathlib import Path
import yaml
from pydantic import BaseModel
STATE_FILENAME = ".bulkgen.state.yaml"
def state_filename(project_name: str) -> str:
"""Return the state filename for a given project name.
For a config file named ``cards.bulkgen.yaml`` the project name is
``cards`` and the state file is ``.cards.bulkgen-state.yaml``.
"""
return f".{project_name}.bulkgen-state.yaml"
class TargetState(BaseModel):
"""Recorded state of a single target from its last successful build."""
input_hashes: dict[str, str]
prompt_hash: str
prompt: str
model: str
extra_hash: str = ""
extra_params: dict[str, object] = {}
class BuildState(BaseModel):
@ -35,14 +42,9 @@ def hash_file(path: Path) -> str:
return h.hexdigest()
def hash_string(value: str) -> str:
"""Compute the SHA-256 hex digest of a string."""
return hashlib.sha256(value.encode("utf-8")).hexdigest()
def load_state(project_dir: Path) -> BuildState:
def load_state(project_dir: Path, project_name: str) -> BuildState:
"""Load build state from disk, returning empty state if the file is missing."""
state_path = project_dir / STATE_FILENAME
state_path = project_dir / state_filename(project_name)
if not state_path.exists():
return BuildState()
with state_path.open() as f:
@ -52,20 +54,13 @@ def load_state(project_dir: Path) -> BuildState:
return BuildState.model_validate(raw)
def save_state(state: BuildState, project_dir: Path) -> None:
def save_state(state: BuildState, project_dir: Path, project_name: str) -> None:
"""Persist build state to disk."""
state_path = project_dir / STATE_FILENAME
state_path = project_dir / state_filename(project_name)
with state_path.open("w") as f:
yaml.dump(state.model_dump(), f, default_flow_style=False, sort_keys=False)
def _extra_hash(params: dict[str, object]) -> str:
"""Hash extra target parameters (width, height, etc.) for change detection."""
if not params:
return ""
return hash_string(str(sorted(params.items())))
def is_target_dirty(
target_name: str,
*,
@ -98,10 +93,10 @@ def is_target_dirty(
if prev.model != model:
return True
if prev.prompt_hash != hash_string(resolved_prompt):
if prev.prompt != resolved_prompt:
return True
if prev.extra_hash != _extra_hash(extra_params):
if prev.extra_params != extra_params:
return True
for dep_path in dep_files:
@ -131,7 +126,7 @@ def record_target_state(
state.targets[target_name] = TargetState(
input_hashes=input_hashes,
prompt_hash=hash_string(resolved_prompt),
prompt=resolved_prompt,
model=model,
extra_hash=_extra_hash(extra_params),
extra_params=extra_params,
)