generate

Canary Generate Pipeline

This module implements the generateion lifecycle for converting generator output into fully resolved test specs. The central orchestrator is the Generator object, which progresses through validation, resolution, and finally hook-driven post-processing.

The canary_generate(generator) function coordinates the entire process using Pluggy hooks. Plugins may observe or modify the generate at specific stages.

Flow Diagram

The following diagram illustrates the full lifecycle:

Generator(s)
  Produces [Un]resolved test specs
                    ↓
Generator(generators)
  Generate JobSpecIR from generator outputs
                    ↓
canary_generate(generators)
• pluginmanager.hook.canary_generatestart()
• generator.run()
  • validate(...)
  • resolve(...)
• pluginmanager.hook.canary_generate_modifiyitems()
• pluginmanager.hook.canary_generate_report()
→ generator.resolved_specs()

class Generator(generators: list[AbstractTestGenerator], workspace: Path, on_options: Iterable[str] = ())

Bases: object

run() → list[JobSpec]

validate(specs: list[JobSpecIR | JobSpec]) → None

static setup_parser(parser: Parser) → None

canary_generate_report(generator: Generator) → None

generate_from_one(file: AbstractTestGenerator, on_options: list[str] | None) → list[JobSpecIR | JobSpec]

generate_jobspecs(generators: list[AbstractTestGenerator], on_options: list[str]) → list[JobSpecIR | JobSpec]

generate_jobspecs_parallel(generators: list[AbstractTestGenerator], on_options: list[str]) → list[JobSpecIR | JobSpec]

worker_init(snapshot: dict[str, Any])

generate_jobspecs_serial(generators: list[AbstractTestGenerator], on_options: list[str]) → list[JobSpecIR | JobSpec]