hyperguild/ingestion/internal/watcher/acceptance_test.go

package watcher

import (
	"context"
	"fmt"
	"os"
	"path/filepath"
	"regexp"
	"strings"
	"sync/atomic"
	"testing"

	"github.com/stretchr/testify/assert"
	"github.com/stretchr/testify/require"

	"github.com/mathiasbq/hyperguild/ingestion/internal/pipeline"
	"github.com/mathiasbq/hyperguild/ingestion/internal/wing"
)

// These tests are the sprint's goal (hyperguild#209) stated as behaviour, with
// the real pipeline and real file layout and only the model stubbed: a person
// drops a clip into raw/ and does nothing else.

func acceptanceCfg(t *testing.T) (Config, string, *atomic.Int32) {
	t.Helper()
	brainDir := t.TempDir()
	var calls atomic.Int32
	// Each call extracts a source with its own title, as real clips have: two
	// sources with one title are deliberately merged into one page by the
	// pipeline's title resolution.
	pagesFor := func(n int32) []pipeline.RawPage {
		return []pipeline.RawPage{
			{Title: fmt.Sprintf("Dark Factory Talk %d", n), 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", Content: "## Definition\n\nA decision model.\n\n## Sources\n"},
			{Title: "Dark Factory", Type: "concept", Content: "## Definition\n\nA fully automated build line.\n\n## Sources\n"},
		}
	}
	return Config{
		BrainDir: brainDir,
		Resolver: wing.Router{BrainDir: brainDir, Default: "web-clip", ClientBlock: regexp.MustCompile(`(?i)acme-client`)},
		Pipeline: pipeline.Config{
			Schema: "# Schema",
			Complete: func(context.Context, string, string) (string, error) {
				return mustJSONPages(pagesFor(calls.Add(1))), nil
			},
		},
	}, brainDir, &calls
}

// topLevel lists the regular files directly in raw/: the inbox.
func topLevel(t *testing.T, brainDir string) []string {
	t.Helper()
	entries, err := os.ReadDir(filepath.Join(brainDir, "raw"))
	require.NoError(t, err)
	var out []string
	for _, e := range entries {
		if e.Type().IsRegular() {
			out = append(out, e.Name())
		}
	}
	return out
}

func TestAcceptance_ADroppedClipLandsInWebClipAndTheInboxIsEmpty(t *testing.T) {
	cfg, brainDir, calls := acceptanceCfg(t)
	put(t, brainDir, "dark-factory-talk.md", "Transcript of a talk about dark factories.")

	require.Empty(t, processDir(context.Background(), cfg, "2026-10-20"))

	assert.True(t, has(brainDir, "wiki/web-clip/sources/dark-factory-talk.md"), "the clip is a page in web-clip")
	assert.True(t, has(brainDir, "wiki/entities/jev.md"), "its entity is in the shared layer")
	assert.True(t, has(brainDir, "wiki/concepts/dark-factory.md"), "its concept is in the shared layer")
	assert.Empty(t, topLevel(t, brainDir), "raw/ holds nothing unprocessed")
	assert.False(t, has(brainDir, "raw/needs-review"), "and nothing waits for a human")
	assert.Equal(t, int32(1), calls.Load(), "one extraction, no classification step")
}

func TestAcceptance_TheSameEntityFromTwoClipsIsOnePage(t *testing.T) {
	cfg, brainDir, _ := acceptanceCfg(t)
	put(t, brainDir, "talk-one.md", "First talk.")
	require.Empty(t, processDir(context.Background(), cfg, "2026-10-20"))
	put(t, brainDir, "talk-two.md", "Second talk, different words.")
	require.Empty(t, processDir(context.Background(), cfg, "2026-10-20"))

	files, err := filepath.Glob(filepath.Join(brainDir, "wiki", "entities", "*.md"))
	require.NoError(t, err)
	assert.Len(t, files, 1)
	b, _ := os.ReadFile(filepath.Join(brainDir, "wiki/entities/jev.md"))
	assert.True(t, strings.Contains(string(b), "talk-one") && strings.Contains(string(b), "talk-two"), "both sources are cited")
}

func TestAcceptance_AClipThatLooksLikeClientMaterialIsParkedNotFiled(t *testing.T) {
	cfg, brainDir, calls := acceptanceCfg(t)
	put(t, brainDir, "kickoff.md", "Notes from the ACME-CLIENT kickoff call.")

	require.Empty(t, processDir(context.Background(), cfg, "2026-10-20"))

	assert.Zero(t, calls.Load(), "not even extracted")
	assert.False(t, has(brainDir, "wiki/web-clip"), "never filed into a shared wing")
	assert.True(t, has(brainDir, "raw/needs-review/kickoff.md"), "waits for the owner")
	assert.Empty(t, topLevel(t, brainDir))
}

func TestAcceptance_ASubfolderNamingAnExistingWingPlacesThereAndATypoParks(t *testing.T) {
	cfg, brainDir, _ := acceptanceCfg(t)
	require.NoError(t, os.MkdirAll(filepath.Join(brainDir, "wiki", "homelab", "facts"), 0o755))
	put(t, brainDir, "homelab/ops-note.md", "A note about the cluster.")
	put(t, brainDir, "hoemlab/typo-note.md", "A misfiled note.")

	require.Empty(t, processDir(context.Background(), cfg, "2026-10-20"))

	assert.True(t, has(brainDir, "wiki/homelab/sources/ops-note.md"))
	assert.False(t, has(brainDir, "wiki/hoemlab"), "a typo never creates a wing")
	assert.True(t, has(brainDir, "raw/needs-review/typo-note.md"))
}