Documentation
¶
Overview ¶
Package javascript runs and parses test output for JS/TS test runners — vitest, jest, playwright. Produces the SOTA structured runtimev0.TestResponse the codefly platform expects (see proto/codefly/services/runtime/v0/runtime.proto).
Three parsers, two schemas:
ParseJestVitestJSON — vitest and jest emit nearly-identical JSON (vitest copied jest's reporter shape). One parser handles both. `--reporter=json` (vitest) or `--json` (jest); both produce the same envelope.
ParsePlaywrightJSON — playwright emits its own shape with nested suites + per-test retry results. `--reporter=json`.
Output retention rule (load-bearing): same as the python+go runners. PASSED + SKIPPED cases get empty captured_output; FAILED + ERRORED cases carry failure detail up to the per-case cap. JS test runners' JSON envelopes already follow this pattern — failureMessages array is empty for passed tests, populated for failed.
All parsers feed StructuredTestRun, the same convertible-to-proto type the python and go parsers use. ToProtoResponse(runner, suiteName, duration) returns runtimev0.TestResponse with both the structured tree and the legacy flat fields populated.
Index ¶
Constants ¶
const MaxCapturedOutputBytesPerCase = 32 * 1024
MaxCapturedOutputBytesPerCase mirrors the Go-runner + python-runner caps. Vitest/jest's `failureMessages[]` are typically a few KB (assertion message + Node stack trace); 32KiB allows fairly deep stacks without bloating the response.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type LegacyTestSummary ¶
type LegacyTestSummary struct {
Run int32
Passed int32
Failed int32
Skipped int32
Coverage float32
Failures []string
}
LegacyTestSummary returns the flat shape pre-overhaul callers used. Computed from the structured tree — single source of truth.
func (*LegacyTestSummary) SummaryLine ¶
func (s *LegacyTestSummary) SummaryLine() string
SummaryLine returns "N passed, M failed, K skipped" — used by agent log lines.
type StructuredCase ¶
type StructuredCase struct {
Name string
FullName string
State runtimev0.TestCaseState
Duration time.Duration
File string
Line int32
Output string
Truncated bool
Failure *StructuredFailure
// Retries is non-empty when the runner re-ran this case
// (playwright `retries`). The case's top-level State reflects
// the FINAL attempt; per-attempt detail lives here.
Retries []*StructuredRetry
}
StructuredCase mirrors the python+go shape. Adds Retries for playwright's repeated-attempts-on-the-same-test feature.
type StructuredFailure ¶
type StructuredFailure struct {
Message string
Detail string
Kind runtimev0.TestFailureKind
}
StructuredFailure: same shape as the python parser's.
type StructuredRetry ¶
type StructuredRetry struct {
Attempt int32
State runtimev0.TestCaseState
Duration time.Duration
Failure *StructuredFailure
}
StructuredRetry is one attempt of a flaky test.
type StructuredSuite ¶
type StructuredSuite struct {
// Name is typically the file path; vitest/jest use absolute
// paths in their JSON.
Name string
File string
Duration time.Duration
Cases []*StructuredCase
}
StructuredSuite is one source-file's worth of cases.
type StructuredTestRun ¶
type StructuredTestRun struct {
// Suites — one per source file. Vitest/jest emit one
// `testResults` entry per file; playwright nests differently
// but we flatten to file-level for proto consistency.
Suites []*StructuredSuite
// CoveragePct is scraped from the runner's coverage summary
// (vitest --coverage prints `All files | NN.NN%`; jest's
// --coverage prints similar). 0 when not requested.
CoveragePct float32
// contains filtered or unexported fields
}
StructuredTestRun is the JS-side equivalent of the per-runner structured test types in core/runners/golang and core/runners/python. Building one is the parser's job; converting to proto is shared.
func ParseJestVitestJSON ¶
func ParseJestVitestJSON(rawJSON string, coverage float32) *StructuredTestRun
ParseJestVitestJSON parses the JSON envelope vitest/jest produce and returns a StructuredTestRun. Empty raw, malformed JSON, or envelope-with-no-suites all surface as an empty run — the caller uses the runErr to decide if the runner crashed.
coverage is scraped separately from the runner's terminal output (the JSON's `coverageMap` carries per-file marks but not the summary percentage); pass coverage as 0 when not measured.
func ParsePlaywrightJSON ¶
func ParsePlaywrightJSON(rawJSON string) *StructuredTestRun
ParsePlaywrightJSON parses Playwright's `--reporter=json` output and returns a StructuredTestRun. Coverage is not modeled — Playwright doesn't emit coverage in its standard JSON; users running coverage typically wrap with c8 or v8-coverage which produce separate artifacts.
func (*StructuredTestRun) LegacyTestSummary ¶
func (r *StructuredTestRun) LegacyTestSummary() *LegacyTestSummary
LegacyTestSummary builds the flat summary from the structured tree.
func (*StructuredTestRun) ToProtoResponse ¶
func (r *StructuredTestRun) ToProtoResponse(runner, suiteName string, duration time.Duration) *runtimev0.TestResponse
ToProtoResponse constructs the runtimev0.TestResponse with both structured (preferred) and legacy fields populated.
runner is "vitest" | "jest" | "playwright"; suiteName echoes TestRequest.suite. duration is the wall-clock for the whole run.
Source Files