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 +1,238 @@
+"""Phase 8: CALL / SEARCH ALL 系测试。
+
+测试覆盖:
+  - CALL 参数传递逻辑（by reference / by value / by content）
+  - SEARCH ALL 二分查找逻辑（找到 / 未找到 / 重复键 / 空表）
+"""
+
+from __future__ import annotations
+
+from typing import Any
+
+
+# ── CALL 模拟 －
+
+
+def _call_by_reference(param: list) -> list:
+    """模拟 COBOL CALL BY REFERENCE: 修改外部变量。"""
+    param[0] = param[0] * 2
+    return param
+
+
+def _call_by_value(param: int) -> int:
+    """模拟 COBOL CALL BY VALUE: 传入副本。"""
+    return param * 2
+
+
+def _call_by_content(param: list) -> list:
+    """模拟 COBOL CALL BY CONTENT: 传入副本，不修改原始值。"""
+    copy = param.copy()
+    copy[0] = copy[0] * 2
+    return copy
+
+
+def _call_with_multiple(
+    a: int,
+    b: int,
+    c: str = "",
+) -> dict[str, Any]:
+    """模拟多参数 CALL。"""
+    return {"sum": a + b, "concat": c * 2}
+
+
+# ── SEARCH ALL 模拟 ──
+
+
+def _search_all(table: list[dict], key_field: str, target: Any) -> int | None:
+    """模拟 COBOL SEARCH ALL（二分查找）。
+
+    要求 table 已按 key_field 升序排列。
+
+    参数
+    ----------
+    table : list[dict]
+        已排序的表。
+    key_field : str
+        待查找的键字段名。
+    target : Any
+        目标值。
+
+    返回
+    -------
+    int | None
+        找到时返回下标；未找到返回 None。
+    """
+    lo, hi = 0, len(table) - 1
+    while lo <= hi:
+        mid = (lo + hi) // 2
+        val = table[mid][key_field]
+        if val == target:
+            return mid
+        elif val < target:
+            lo = mid + 1
+        else:
+            hi = mid - 1
+    return None
+
+
+def _search_all_duplicate_keys(
+    table: list[dict],
+    key_field: str,
+    target: Any,
+) -> list[int]:
+    """查找所有匹配的记录下标（处理重复键）。"""
+    indices: list[int] = []
+    first = _search_all(table, key_field, target)
+    if first is None:
+        return []
+    # 向前扫描
+    i = first
+    while i >= 0 and table[i][key_field] == target:
+        indices.append(i)
+        i -= 1
+    indices.reverse()
+    # 向后扫描
+    i = first + 1
+    while i < len(table) and table[i][key_field] == target:
+        indices.append(i)
+        i += 1
+    return indices
+
+
+# ── 测试: CALL ──
+
+
+class TestCallByReference:
+    """CALL BY REFERENCE 参数传递"""
+
+    def test_by_reference_modifies_original(self):
+        data = [5]
+        result = _call_by_reference(data)
+        assert data[0] == 10, "BY REFERENCE 应修改原始值"
+        assert result == [10]
+
+    def test_by_reference_string(self):
+        data = ["hello"]
+        _call_by_reference(data)
+        assert data[0] == "hellohello"
+
+
+class TestCallByValue:
+    """CALL BY VALUE 参数传递"""
+
+    def test_by_value_no_side_effect(self):
+        x = 5
+        result = _call_by_value(x)
+        assert x == 5, "BY VALUE 不应修改原始值"
+        assert result == 10
+
+    def test_by_value_zero(self):
+        assert _call_by_value(0) == 0
+
+    def test_by_value_negative(self):
+        assert _call_by_value(-3) == -6
+
+
+class TestCallByContent:
+    """CALL BY CONTENT 参数传递"""
+
+    def test_by_content_preserves_original(self):
+        data = [5]
+        result = _call_by_content(data)
+        assert data[0] == 5, "BY CONTENT 不应修改原始值"
+        assert result == [10]
+
+
+class TestCallMultipleParameters:
+    """多参数 CALL"""
+
+    def test_multiple_params(self):
+        result = _call_with_multiple(3, 4)
+        assert result["sum"] == 7
+
+    def test_multiple_params_with_string(self):
+        result = _call_with_multiple(1, 2, c="ab")
+        assert result["sum"] == 3
+        assert result["concat"] == "abab"
+
+    def test_multiple_params_default(self):
+        result = _call_with_multiple(10, 20)
+        assert result["concat"] == ""
+
+
+# ── 测试: SEARCH ALL ──
+
+
+class TestSearchAllFound:
+    """SEARCH ALL — 找到"""
+
+    def test_search_found_first(self):
+        table = [{"K": 1}, {"K": 3}, {"K": 5}, {"K": 7}]
+        idx = _search_all(table, "K", 1)
+        assert idx == 0
+
+    def test_search_found_last(self):
+        table = [{"K": 1}, {"K": 3}, {"K": 5}, {"K": 7}]
+        idx = _search_all(table, "K", 7)
+        assert idx == 3
+
+    def test_search_found_middle(self):
+        table = [{"K": 1}, {"K": 3}, {"K": 5}, {"K": 7}]
+        idx = _search_all(table, "K", 5)
+        assert idx == 2
+
+    def test_search_string_keys(self):
+        table = [{"K": "a"}, {"K": "b"}, {"K": "c"}, {"K": "d"}]
+        idx = _search_all(table, "K", "c")
+        assert idx == 2
+
+
+class TestSearchAllNotFound:
+    """SEARCH ALL — 未找到"""
+
+    def test_search_not_found(self):
+        table = [{"K": 1}, {"K": 3}, {"K": 5}]
+        idx = _search_all(table, "K", 4)
+        assert idx is None
+
+    def test_search_below_all(self):
+        table = [{"K": 10}, {"K": 20}]
+        idx = _search_all(table, "K", 5)
+        assert idx is None
+
+    def test_search_above_all(self):
+        table = [{"K": 10}, {"K": 20}]
+        idx = _search_all(table, "K", 25)
+        assert idx is None
+
+
+class TestSearchAllDuplicateKeys:
+    """SEARCH ALL — 重复键"""
+
+    def test_search_duplicate_keys(self):
+        table = [{"K": 1}, {"K": 2}, {"K": 2}, {"K": 2}, {"K": 3}]
+        indices = _search_all_duplicate_keys(table, "K", 2)
+        assert indices == [1, 2, 3]
+
+    def test_search_no_duplicate(self):
+        table = [{"K": 1}, {"K": 2}, {"K": 3}]
+        indices = _search_all_duplicate_keys(table, "K", 2)
+        assert indices == [1]
+
+
+class TestSearchAllEdgeCases:
+    """SEARCH ALL — 边界"""
+
+    def test_search_empty_table(self):
+        idx = _search_all([], "K", 1)
+        assert idx is None
+
+    def test_search_single_element_found(self):
+        table = [{"K": 42}]
+        idx = _search_all(table, "K", 42)
+        assert idx == 0
+
+    def test_search_single_element_not_found(self):
+        table = [{"K": 42}]
+        idx = _search_all(table, "K", 99)
+        assert idx is None