mathiasandClaude Sonnet 5.5
ae02d4764c
- a source whose slug is taken by anything but this wing's own page gets a wing-qualified, stable slug (talk-homelab) instead of failing the ingest - an extracted entity/concept whose slug belongs to a different page is skipped with a warning, so one page can no longer lock a source out of re-ingest - in a wing only source/concept/entity are valid page types; the model (and so the clip) can no longer choose wiki/<type>/ directories Mutation-checked; stable-slug choice pinned directly because title resolution masks it at the Run level. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
236 lines
11 KiB
Go
236 lines
11 KiB
Go
package pipeline
|
|
|
|
import (
|
|
"context"
|
|
"os"
|
|
"path/filepath"
|
|
"strings"
|
|
"testing"
|
|
|
|
"github.com/stretchr/testify/assert"
|
|
"github.com/stretchr/testify/require"
|
|
)
|
|
|
|
func wingCfg(wing string, pages ...RawPage) Config {
|
|
return Config{
|
|
Wing: wing,
|
|
Schema: "# Schema",
|
|
Complete: func(context.Context, string, string) (string, error) {
|
|
return mustJSON(pages), nil
|
|
},
|
|
}
|
|
}
|
|
|
|
func sourceEntityConcept() []RawPage {
|
|
return []RawPage{
|
|
{Title: "Dark Factory Talk", Type: "source", Subtype: "video", Domain: "ai",
|
|
Content: "## Summary\n\nA talk.\n\n## Entities Mentioned\n\n[[Jev]]\n\n## Concepts Introduced or Reinforced\n\n[[Dark Factory]]\n"},
|
|
{Title: "Jev", Type: "entity", Subtype: "tool", Domain: "ai",
|
|
Content: "## Definition\n\nA decision model.\n\n## Sources\n"},
|
|
{Title: "Dark Factory", Type: "concept", Domain: "ai",
|
|
Content: "## Definition\n\nA fully automated build line.\n\n## Sources\n"},
|
|
}
|
|
}
|
|
|
|
func exists(dir, rel string) bool {
|
|
_, err := os.Stat(filepath.Join(dir, rel))
|
|
return err == nil
|
|
}
|
|
|
|
func read(t *testing.T, dir, rel string) string {
|
|
t.Helper()
|
|
b, err := os.ReadFile(filepath.Join(dir, rel))
|
|
require.NoError(t, err)
|
|
return string(b)
|
|
}
|
|
|
|
// With a wing, the SOURCE belongs to the wing; concepts and entities are the
|
|
// shared cross-wing layer (wiki/concepts, wiki/entities) because their slug is
|
|
// the graph's global identity and they are cited from many wings.
|
|
func TestRun_WithAWingTheSourceIsPerWingAndConceptsEntitiesAreShared(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "body of the talk", "dark-factory-talk", false)
|
|
require.NoError(t, err)
|
|
|
|
src := read(t, dir, "wiki/web-clip/sources/dark-factory-talk.md")
|
|
assert.Contains(t, src, "wing: web-clip")
|
|
assert.Contains(t, src, "hall: sources")
|
|
assert.True(t, exists(dir, "wiki/entities/jev.md"))
|
|
assert.True(t, exists(dir, "wiki/concepts/dark-factory.md"))
|
|
assert.NotContains(t, read(t, dir, "wiki/entities/jev.md"), "wing:", "a shared page belongs to no single wing")
|
|
assert.False(t, exists(dir, "wiki/web-clip/facts"), "no per-wing copy of a shared page")
|
|
assert.False(t, exists(dir, "wiki/sources"), "the retired source layout is not used")
|
|
}
|
|
|
|
func TestRun_WithAWingTheSharedPageLinksBackToTheSource(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "body", "dark-factory-talk", false)
|
|
require.NoError(t, err)
|
|
assert.Contains(t, read(t, dir, "wiki/entities/jev.md"), "[[dark-factory-talk|Dark Factory Talk]]")
|
|
}
|
|
|
|
func TestRun_TheSameEntityCitedFromTwoWingsIsOnePageWithBothSources(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "first body", "talk-one", false)
|
|
require.NoError(t, err)
|
|
// A wing must already exist to be named; create it the way a note would.
|
|
require.NoError(t, os.MkdirAll(filepath.Join(dir, "wiki/homelab/facts"), 0o755))
|
|
|
|
again := []RawPage{
|
|
{Title: "Second Talk", Type: "source", Subtype: "video", Content: "## Summary\n\nMore.\n\n## Entities Mentioned\n\n[[The Jev]]\n"},
|
|
{Title: "The Jev", Type: "entity", Subtype: "tool", Content: "## Definition\n\nStill a decision model.\n\n## Sources\n"},
|
|
}
|
|
_, err = Run(context.Background(), wingCfg("homelab", again...), dir, "second body", "talk-two", false)
|
|
require.NoError(t, err)
|
|
|
|
files, err := filepath.Glob(filepath.Join(dir, "wiki/entities/*.md"))
|
|
require.NoError(t, err)
|
|
assert.Len(t, files, 1, "one entity page, matched by title with the leading article ignored")
|
|
jev := read(t, dir, "wiki/entities/jev.md")
|
|
assert.Contains(t, jev, "talk-one")
|
|
assert.Contains(t, jev, "talk-two")
|
|
assert.False(t, exists(dir, "wiki/homelab/facts/jev.md"), "no second copy in the second wing")
|
|
}
|
|
|
|
// A source slug is the graph's identity too: a second wing must not take over
|
|
// the slug of a page that lives elsewhere. Two wings dropping a file of the same
|
|
// name (notes.md, readme.md) is ordinary, so the second source gets a distinct,
|
|
// stable slug instead of failing.
|
|
func TestRun_ASourceWhoseSlugIsTakenElsewhereGetsADistinctStableSlug(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "body", "talk", false)
|
|
require.NoError(t, err)
|
|
require.NoError(t, os.MkdirAll(filepath.Join(dir, "wiki/homelab/facts"), 0o755))
|
|
|
|
_, err = Run(context.Background(), wingCfg("homelab", sourceEntityConcept()...), dir, "different body", "talk", false)
|
|
require.NoError(t, err)
|
|
assert.True(t, exists(dir, "wiki/web-clip/sources/talk.md"), "the first wing's page is untouched")
|
|
assert.True(t, exists(dir, "wiki/homelab/sources/talk-homelab.md"), "the second gets a wing-qualified slug")
|
|
assert.False(t, exists(dir, "wiki/homelab/sources/talk.md"))
|
|
|
|
_, err = Run(context.Background(), wingCfg("homelab", sourceEntityConcept()...), dir, "different body, edited", "talk", false)
|
|
require.NoError(t, err)
|
|
files, _ := filepath.Glob(filepath.Join(dir, "wiki/homelab/sources/*.md"))
|
|
assert.Len(t, files, 1, "re-ingesting finds the same page, it does not make talk-homelab-2")
|
|
}
|
|
|
|
func TestRun_ASourceWhoseSlugIsTakenByARetiredPageOrASharedPageGetsADistinctSlug(t *testing.T) {
|
|
dir := t.TempDir()
|
|
require.NoError(t, os.MkdirAll(filepath.Join(dir, "wiki/sources"), 0o755))
|
|
require.NoError(t, os.WriteFile(filepath.Join(dir, "wiki/sources/talk.md"), []byte("---\ntitle: Talk\n---\nold"), 0o644))
|
|
require.NoError(t, os.MkdirAll(filepath.Join(dir, "wiki/entities"), 0o755))
|
|
require.NoError(t, os.WriteFile(filepath.Join(dir, "wiki/entities/other.md"), []byte("---\ntitle: Other\n---\nx"), 0o644))
|
|
|
|
_, err := Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "body", "talk", false)
|
|
require.NoError(t, err)
|
|
assert.True(t, exists(dir, "wiki/web-clip/sources/talk-web-clip.md"))
|
|
assert.Equal(t, "---\ntitle: Talk\n---\nold", read(t, dir, "wiki/sources/talk.md"), "the retired page is untouched")
|
|
}
|
|
|
|
// The other direction: a model-extracted entity or concept whose slug is a
|
|
// source's slug in some wing would lock that source out of re-ingest. The page
|
|
// is skipped with a warning; the ingest itself succeeds.
|
|
func TestRun_AnEntityWhoseSlugIsASourceElsewhereIsSkippedNotWritten(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg("web-clip", []RawPage{
|
|
{Title: "Jev", Type: "source", Subtype: "video", Content: "## Summary\n\nA clip titled Jev.\n"},
|
|
}...), dir, "body", "jev", false)
|
|
require.NoError(t, err)
|
|
require.NoError(t, os.MkdirAll(filepath.Join(dir, "wiki/homelab/facts"), 0o755))
|
|
|
|
res, err := Run(context.Background(), wingCfg("homelab", sourceEntityConcept()...), dir, "another body", "other-talk", false)
|
|
require.NoError(t, err)
|
|
assert.False(t, exists(dir, "wiki/entities/jev.md"), "the entity would take the source's slug")
|
|
assert.Contains(t, strings.Join(res.Warnings, "\n"), "jev")
|
|
|
|
_, err = Run(context.Background(), wingCfg("web-clip", []RawPage{
|
|
{Title: "Jev", Type: "source", Subtype: "video", Content: "## Summary\n\nEdited.\n"},
|
|
}...), dir, "body, edited", "jev", false)
|
|
require.NoError(t, err, "the first wing can still re-ingest its own source")
|
|
}
|
|
|
|
// The type comes from the model, so from the clip: in a wing it may only be one
|
|
// of the three the layout has a place for.
|
|
func TestRun_WithAWingAModelChosenTypeCannotWriteOutsideTheLayout(t *testing.T) {
|
|
dir := t.TempDir()
|
|
require.NoError(t, os.MkdirAll(filepath.Join(dir, "wiki/alpha/facts"), 0o755))
|
|
pages := append(sourceEntityConcept(),
|
|
RawPage{Title: "Planted", Type: "sources", Content: "## x\n\nplanted\n"},
|
|
RawPage{Title: "Loose", Type: "alpha", Content: "## x\n\nloose\n"},
|
|
RawPage{Title: "Sneaky", Type: "homelab", Content: "## x\n\nsneaky\n"})
|
|
res, err := Run(context.Background(), wingCfg("web-clip", pages...), dir, "body", "talk", false)
|
|
require.NoError(t, err)
|
|
assert.False(t, exists(dir, "wiki/sources/planted.md"))
|
|
assert.False(t, exists(dir, "wiki/alpha/loose.md"))
|
|
assert.False(t, exists(dir, "wiki/homelab"))
|
|
assert.Contains(t, strings.Join(res.Warnings, "\n"), "invalid type")
|
|
}
|
|
|
|
func TestRun_ReIngestingTheSameSourceIntoItsOwnWingIsNotACollision(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "body one", "talk", false)
|
|
require.NoError(t, err)
|
|
_, err = Run(context.Background(), wingCfg("web-clip", sourceEntityConcept()...), dir, "body two, edited", "talk", false)
|
|
require.NoError(t, err)
|
|
}
|
|
|
|
func TestRun_WithAWingRejectsANameThatIsNotAPlainWing(t *testing.T) {
|
|
for _, bad := range []string{"../escape", "a/b", "Radar", "has space", "."} {
|
|
t.Run(bad, func(t *testing.T) {
|
|
dir := t.TempDir()
|
|
_, err := Run(context.Background(), wingCfg(bad, sourceEntityConcept()...), dir, "body", "talk", false)
|
|
require.Error(t, err)
|
|
assert.Len(t, listAll(t, dir), 1, "nothing is written for a rejected wing (only the temp dir itself exists)")
|
|
})
|
|
}
|
|
}
|
|
|
|
func listAll(t *testing.T, dir string) []string {
|
|
t.Helper()
|
|
var out []string
|
|
_ = filepath.WalkDir(dir, func(p string, _ os.DirEntry, _ error) error { out = append(out, p); return nil })
|
|
return out
|
|
}
|
|
|
|
// Title resolution can mask a wrong slug choice at the Run level, so the
|
|
// stability of the choice is pinned directly: the page this wing made earlier
|
|
// under the qualified slug is "its own" and is found again.
|
|
func TestResolveSourceSlug(t *testing.T) {
|
|
dir := t.TempDir()
|
|
for _, rel := range []string{
|
|
"wiki/web-clip/sources/talk.md",
|
|
"wiki/homelab/sources/talk-homelab.md",
|
|
"wiki/entities/kube.md",
|
|
} {
|
|
require.NoError(t, os.MkdirAll(filepath.Dir(filepath.Join(dir, rel)), 0o755))
|
|
require.NoError(t, os.WriteFile(filepath.Join(dir, rel), []byte("---\ntitle: x\n---\n"), 0o644))
|
|
}
|
|
tests := []struct {
|
|
name, wing, base, want string
|
|
}{
|
|
{"free slug is used as is", "homelab", "fresh", "fresh"},
|
|
{"own page under the plain slug", "web-clip", "talk", "talk"},
|
|
{"taken elsewhere: wing-qualified", "tapir", "talk", "talk-tapir"},
|
|
{"own earlier qualified page is found again", "homelab", "talk", "talk-homelab"},
|
|
{"taken by a shared page", "tapir", "kube", "kube-tapir"},
|
|
}
|
|
for _, tt := range tests {
|
|
t.Run(tt.name, func(t *testing.T) {
|
|
got, err := resolveSourceSlug(dir, tt.wing, tt.base)
|
|
require.NoError(t, err)
|
|
assert.Equal(t, tt.want, got)
|
|
})
|
|
}
|
|
}
|
|
|
|
func TestResolveSourceSlug_AFurtherClashGetsANumber(t *testing.T) {
|
|
dir := t.TempDir()
|
|
for _, rel := range []string{"wiki/a/sources/talk.md", "wiki/entities/talk-b.md"} {
|
|
require.NoError(t, os.MkdirAll(filepath.Dir(filepath.Join(dir, rel)), 0o755))
|
|
require.NoError(t, os.WriteFile(filepath.Join(dir, rel), []byte("x"), 0o644))
|
|
}
|
|
got, err := resolveSourceSlug(dir, "b", "talk")
|
|
require.NoError(t, err)
|
|
assert.Equal(t, "talk-b-2", got)
|
|
}
|