lore-engine-poc/workers/lore-extractor/main_test.go

package main

import (
	"encoding/json"
	"strings"
	"testing"
)

// ── repairJSON ────────────────────────────────────────────────────────────────

func TestRepairJSON(t *testing.T) {
	tests := []struct {
		input string
		want  string
	}{
		{`{"a": 1,}`, `{"a": 1}`},
		{`[1, 2,]`, `[1, 2]`},
		{`{"a": 1, "b": 2}`, `{"a": 1, "b": 2}`}, // already valid, unchanged
		{`{"entities": [],}`, `{"entities": []}`},
	}
	for _, tt := range tests {
		got := repairJSON(tt.input)
		if got != tt.want {
			t.Errorf("repairJSON(%q) = %q, want %q", tt.input, got, tt.want)
		}
	}
}

// ── stripFences ───────────────────────────────────────────────────────────────

func TestStripFences(t *testing.T) {
	tests := []struct {
		name  string
		input string
		want  string
	}{
		{"json fence", "```json\n{}\n```", "{}"},
		{"plain fence", "```\n{}\n```", "{}"},
		{"no fence", "{}", "{}"},
		{"leading whitespace with fence", "  ```json\n{}\n```  ", "{}"},
		{"fence with content", "```json\n{\"a\":1}\n```", `{"a":1}`},
	}
	for _, tt := range tests {
		t.Run(tt.name, func(t *testing.T) {
			got := stripFences(tt.input)
			if got != tt.want {
				t.Errorf("stripFences(%q) = %q, want %q", tt.input, got, tt.want)
			}
		})
	}
}

// ── fixUnicodeEscapes ─────────────────────────────────────────────────────────

func TestFixUnicodeEscapes(t *testing.T) {
	t.Run("valid escape is preserved", func(t *testing.T) {
		input := `{"name": "Théron"}`
		got := fixUnicodeEscapes(input)
		if got != input {
			t.Errorf("fixUnicodeEscapes modified valid escape: got %q", got)
		}
		// Must still be valid JSON
		if err := json.Unmarshal([]byte(got), &map[string]any{}); err != nil {
			t.Errorf("result not valid JSON: %v", err)
		}
	})

	t.Run("invalid hex char is stripped", func(t *testing.T) {
		// \u201g — 'g' is not a hex digit
		input := `{"name": "test\u201gname"}`
		got := fixUnicodeEscapes(input)
		if strings.Contains(got, `\u`) {
			t.Errorf("fixUnicodeEscapes left invalid escape in: %q", got)
		}
		// Result must be parseable JSON
		if err := json.Unmarshal([]byte(got), &map[string]any{}); err != nil {
			t.Errorf("result not valid JSON after fix: %v — got: %q", err, got)
		}
	})

	t.Run("truncated escape at end is dropped", func(t *testing.T) {
		input := `{"name": "ab\u00"}`
		got := fixUnicodeEscapes(input)
		if strings.Contains(got, `\u`) {
			t.Errorf("incomplete escape still present: %q", got)
		}
	})
}

// ── coerceString ─────────────────────────────────────────────────────────────

func TestCoerceString(t *testing.T) {
	tests := []struct {
		name  string
		input string
		want  string
	}{
		{"plain string", `"Theron Ashveil"`, "Theron Ashveil"},
		{"name object", `{"name": "The Iron Council"}`, "The Iron Council"},
		{"single-element array", `["Thornwall Keep"]`, "Thornwall Keep"},
		{"empty string", `""`, ""},
		{"nested array with object", `[{"name": "Deep Keep"}]`, "Deep Keep"},
	}
	for _, tt := range tests {
		t.Run(tt.name, func(t *testing.T) {
			got := coerceString(json.RawMessage(tt.input))
			if got != tt.want {
				t.Errorf("coerceString(%q) = %q, want %q", tt.input, got, tt.want)
			}
		})
	}
}

// ── salvageEntities ───────────────────────────────────────────────────────────

func TestSalvageEntities(t *testing.T) {
	t.Run("recovers entities when relations are malformed", func(t *testing.T) {
		// Simulate a response where relations block is broken
		raw := `{"entities": [{"name": "Gromm", "type": "Person"}], "relations": [BROKEN`
		result := salvageEntities(raw)
		if result == nil {
			t.Fatal("expected salvaged result, got nil")
		}
		if len(result.Entities) != 1 || result.Entities[0].Name != "Gromm" {
			t.Errorf("unexpected entities: %+v", result.Entities)
		}
		if len(result.Relations) != 0 {
			t.Errorf("expected empty relations, got %d", len(result.Relations))
		}
	})

	t.Run("returns nil when entities are also malformed", func(t *testing.T) {
		raw := `{"entities": [BROKEN], "relations": []}`
		result := salvageEntities(raw)
		if result != nil {
			t.Errorf("expected nil, got %+v", result)
		}
	})
}