feat: Phase 2 complete — 13 Phases of COBOL type classification and test benchmark

P0.6: gcov infrastructure P1: extract_structure output expansion (11 new feature fields) P2: Confusion group rule engine (8 pairs + contradiction + backtrack) P3: 4-factor confidence calculation + quality gate update P4: 33+2 COBOL program type test samples (22 files, 7 categories) P5: parametrized/ test data generation engine P6: japanese_data.py lookup tables P7-10: Type-specific test suites (~159 parametrized tests) P11: Full classification pipeline (classify_program) + orchestrator integration P12: Documentation (module-interfaces, test-plan v3.0, coverage-matrix) Architecture decisions: - classification_pipeline/ merged to hina/pipeline/ - parametrized/ as independent module - japanese_data.py as root-level file - hina/__all__ only exports classify_program() Co-Authored-By: Claude <noreply@anthropic.com>
2026-06-19 23:51:55 +08:00
parent 63b5284715
commit bc1d56d1a4
129 changed files with 19378 additions and 261 deletions
@@ -0,0 +1,203 @@
+"""JP-01~10: japanese_data 模块 — 日文测试数据生成函数"""
+from __future__ import annotations
+
+import sys
+import os
+
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "..")))
+
+from japanese_data import (
+    FULLWIDTH_KATAKANA,
+    FULLWIDTH_HIRAGANA,
+    FULLWIDTH_DIGITS,
+    FULLWIDTH_ALPHA,
+    HALFWIDTH_KATAKANA,
+    SJIS_5C_PROBLEM,
+    SJIS_7C_PROBLEM,
+    WAREKI_BOUNDARIES,
+    generate_fullwidth_text,
+    generate_halfwidth_katakana,
+    generate_sjis_5c_problem,
+    generate_sjis_7c_problem,
+    generate_wareki_date,
+    generate_wareki_boundary,
+    generate_encoding_test_data,
+    select_data_type,
+)
+
+
+# ── JP-01~02: 查找表常量 ──
+
+
+def test_fullwidth_katakana_constants():
+    """JP-01: 全角片假名表不为空"""
+    assert len(FULLWIDTH_KATAKANA) > 0
+    assert "ア" in FULLWIDTH_KATAKANA
+    assert "ン" in FULLWIDTH_KATAKANA
+
+
+def test_fullwidth_hiragana_constants():
+    """全角平假名表不为空"""
+    assert len(FULLWIDTH_HIRAGANA) > 0
+    assert "あ" in FULLWIDTH_HIRAGANA
+    assert "ん" in FULLWIDTH_HIRAGANA
+
+
+def test_halfwidth_katakana_constants():
+    """半角片假名表不为空"""
+    assert len(HALFWIDTH_KATAKANA) > 0
+    assert "ｱ" in HALFWIDTH_KATAKANA
+
+
+def test_sjis_problem_constants():
+    """SJIS 5C/7C 问题文字表内容"""
+    assert "ソ" in SJIS_5C_PROBLEM
+    assert "本" in SJIS_7C_PROBLEM
+    assert len(SJIS_5C_PROBLEM) > 0
+    assert len(SJIS_7C_PROBLEM) > 0
+
+
+def test_wareki_boundaries():
+    """和历边界表含有平成条目"""
+    eras = [e[0] for e in WAREKI_BOUNDARIES]
+    assert "平成" in eras
+    assert "昭和" in eras
+
+
+# ── JP-03~05: generate_fullwidth_text ──
+
+
+def test_fullwidth_text_type():
+    """JP-03: generate_fullwidth_text 返回 str"""
+    field = {"pic_info": {"type": "national", "length": 10}}
+    result = generate_fullwidth_text(field)
+    assert isinstance(result, str)
+
+
+def test_fullwidth_text_length():
+    """JP-04: generate_fullwidth_text 返回指定长度"""
+    field = {"pic_info": {"type": "national", "length": 5}}
+    result = generate_fullwidth_text(field)
+    assert len(result) == 5
+
+
+def test_fullwidth_text_contents():
+    """JP-05: generate_fullwidth_text 内容来自全角片假名表"""
+    field = {"pic_info": {"type": "national", "length": 20}}
+    result = generate_fullwidth_text(field)
+    for ch in result:
+        assert ch in FULLWIDTH_KATAKANA, f"意外字符 {ch!r}"
+
+
+# ── JP-06~07: generate_halfwidth_katakana ──
+
+
+def test_halfwidth_katakana_type():
+    """JP-06: generate_halfwidth_katakana 返回 str"""
+    field = {"pic_info": {"type": "alphanumeric", "length": 10}}
+    result = generate_halfwidth_katakana(field)
+    assert isinstance(result, str)
+
+
+def test_halfwidth_katakana_length():
+    """JP-07: generate_halfwidth_katakana 返回指定长度"""
+    field = {"pic_info": {"type": "alphanumeric", "length": 8}}
+    result = generate_halfwidth_katakana(field)
+    assert len(result) == 8
+
+
+# ── JP-08: generate_sjis_5c_problem ──
+
+
+def test_sjis_5c_text():
+    """JP-08: generate_sjis_5c_problem 字符来自 5C 表"""
+    field = {"pic_info": {"type": "alphanumeric", "length": 6}}
+    result = generate_sjis_5c_problem(field)
+    assert isinstance(result, str)
+    assert len(result) == 6
+    for ch in result:
+        assert ch in SJIS_5C_PROBLEM, f"意外字符 {ch!r}"
+
+
+# ── JP-09: generate_sjis_7c_problem ──
+
+
+def test_sjis_7c_text():
+    """JP-09: generate_sjis_7c_problem 字符来自 7C 表"""
+    field = {"pic_info": {"type": "alphanumeric", "length": 5}}
+    result = generate_sjis_7c_problem(field)
+    assert isinstance(result, str)
+    assert len(result) == 5
+    for ch in result:
+        assert ch in SJIS_7C_PROBLEM, f"意外字符 {ch!r}"
+
+
+# ── JP-10: generate_wareki_date ──
+
+
+def test_wareki_date_format():
+    """JP-10: generate_wareki_date 返回格式 H050101"""
+    result = generate_wareki_date("H")
+    assert isinstance(result, str)
+    # 格式: 1 prefix + 2 year + 2 month + 2 day = 7
+    assert len(result) == 7
+    assert result[0] == "H"
+    # 年份 01-30, 月份 01-12, 日期 01-28
+    year_part = int(result[1:3])
+    month_part = int(result[3:5])
+    day_part = int(result[5:7])
+    assert 1 <= year_part <= 30
+    assert 1 <= month_part <= 12
+    assert 1 <= day_part <= 28
+
+
+# ── 边界值测试 ──
+
+
+def test_wareki_boundary_heisei():
+    """generate_wareki_boundary 平成返回(初日, 末日)"""
+    start, end = generate_wareki_boundary("平成")
+    assert isinstance(start, str)
+    assert isinstance(end, str)
+    assert start.startswith("H")
+    assert start == "H010108"
+
+
+def test_encoding_test_data_type():
+    """generate_encoding_test_data 返回 bytes 元组"""
+    src, tgt = generate_encoding_test_data()
+    assert isinstance(src, bytes)
+    assert isinstance(tgt, bytes)
+
+
+def test_select_data_type_national():
+    """select_data_type 对 PIC N 返回 japanese"""
+    field = {"pic_info": {"type": "national"}}
+    assert select_data_type(field) == "japanese"
+
+
+def test_select_data_type_numeric():
+    """select_data_type 对 PIC 9 返回 numeric"""
+    field = {"pic_info": {"type": "numeric", "digits": 5}}
+    assert select_data_type(field) == "numeric"
+
+
+def test_select_data_type_halfwidth():
+    """select_data_type 对 PIC X 返回 halfwidth"""
+    field = {"pic_info": {"type": "alphanumeric", "length": 10}}
+    assert select_data_type(field) == "halfwidth"
+
+
+# ── 默认参数测试 ──
+
+
+def test_wareki_date_default():
+    """generate_wareki_date 无参数默认令和"""
+    result = generate_wareki_date()
+    assert result[0] == "R"
+
+
+def test_wareki_boundary_default():
+    """generate_wareki_boundary 无参数默认平成"""
+    prev, new = generate_wareki_boundary()
+    assert new.startswith("H")