Initial commit: ABAP indexing pipeline (ingest, parser, summarize, index, query, wiki)

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

tests/test_normalize.py

@@ -0,0 +1,43 @@
+"""Stage 1 검증 (계획서 §3) — 샘플 원본이 data/raw 에 있을 때만 실행."""
+import json
+from pathlib import Path
+
+import pytest
+
+from ingest.normalize import clean_text_field, normalize_raw_text, parse_collected_file
+
+ROOT = Path(__file__).resolve().parent.parent
+RAW = ROOT / "data" / "raw"
+
+
+def test_normalize_regex_removes_wrap_artifact():
+    raw = '{"A":"REPORT\n   zfir10070 MESSAGE-ID zfim01."}'
+    assert json.loads(normalize_raw_text(raw))["A"] == "REPORT zfir10070 MESSAGE-ID zfim01."
+
+
+def test_escaped_newline_preserved():
+    # 이스케이프된 \n(백슬래시+n)은 json.loads 후 진짜 줄바꿈이 되어야 한다
+    raw = '{"A":"LINE1\\nLINE2"}'
+    assert json.loads(normalize_raw_text(raw))["A"] == "LINE1\nLINE2"
+
+
+def test_clean_text_field():
+    assert clean_text_field("총계정원장    조회") == "총계정원장 조회"
+
+
+@pytest.mark.skipif(not (RAW / "ZFIR10070.txt").exists(), reason="샘플 원본 없음")
+def test_sample_program_source():
+    data = parse_collected_file((RAW / "ZFIR10070.txt").read_text(encoding="utf-8"))
+    assert clean_text_field(data["DESCRIPTION"]) == "총계정원장 조회"
+    f01 = next(i for i in data["INCLUDE_PROGRAM"] if "F01" in i["INCLUDE"])
+    form_lines = [
+        ln for ln in f01["SOURCE_CODE"].split("\n") if ln.strip().upper().startswith("FORM ")
+    ]
+    assert len(form_lines) == 73
+
+
+@pytest.mark.skipif(not (RAW / "ZFI01.txt").exists(), reason="샘플 원본 없음")
+def test_sample_package_list():
+    data = parse_collected_file((RAW / "ZFI01.txt").read_text(encoding="utf-8"))
+    assert isinstance(data, list) and len(data) > 900
+    assert data[0]["DEVCLASS"] == "ZFI01"