From 4be2aae66db4c86d3b6eb20378c0632cece31f26 Mon Sep 17 00:00:00 2001 From: NB-076 Date: Sun, 21 Jun 2026 16:13:58 +0800 Subject: [PATCH] =?UTF-8?q?fix:=20=E7=94=9F=E4=BA=A7=E7=BA=A7=20COBOL=20?= =?UTF-8?q?=E7=A8=8B=E5=BA=8F=E8=A7=A3=E6=9E=90=20=E2=80=94=20COPY=20+=20O?= =?UTF-8?q?CCURS=20TO=20+=20FD=20=E4=BF=AE=E5=A4=8D?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 对抗性测试发现的生产程序解析缺陷和修复: 缺陷1: COPY 语句从未被预处理（18 个月 bug） - resolve_copybooks() 在 main() CLI 中调用但在 extract_structure() 路径中从未被调用 - 修复: preprocess() 函数头部调用 resolve_copybooks() - 不可解析的 COPY 行被移除（避免 Lark 在 FD 块内遇到无法识别的指令） 缺陷2: Lark 语法的 fd 规则要求 data_item+ (至少一个记录) - 生产程序 FD 可以通过 COPY 引入记录定义 - COPY 被移除后 FD 内无 data_item 导致 Lark 崩溃 - 修复: fd 改为 data_item* (零或多个) 缺陷3: OCCURS 1 TO 100 TIMES（变量范围表） - 语法只支持 OCCURS INT TIMES，不支持 OCCURS 1 TO 100 TIMES - 修复: occurs_clause 增加 'TO' INT 可选部分 效果: 4 个生产程序中 2 个成功解析（CRDVAL, GENDATA） - 剩余 2 个（CRDCALC, CRDRPT）因固定格式续行限制未修复 全回归: 767 passed（0 new failures） --- cobol_testgen/grammar.lark | 6 +++--- cobol_testgen/read.py | 10 ++++++++-- tests/cobol_testgen/test_read.py | 5 +++-- 3 files changed, 14 insertions(+), 7 deletions(-) diff --git a/cobol_testgen/grammar.lark b/cobol_testgen/grammar.lark index b367c1c..c91f2f1 100644 --- a/cobol_testgen/grammar.lark +++ b/cobol_testgen/grammar.lark @@ -1,7 +1,7 @@ start: data_div_content data_div_content: (file_section | working_storage | linkage)* file_section: "FILE" "SECTION" DOT (fd | sd)+ -fd: "FD" NAME FD_SUFFIX data_item+ +fd: "FD" NAME FD_SUFFIX data_item* sd: "SD" NAME FD_SUFFIX data_item* FD_SUFFIX: /(?:"[^"]*"|'[^']*'|[^.])*\./ working_storage: "WORKING-STORAGE" "SECTION" DOT data_item* @@ -22,13 +22,13 @@ value_literal: INT | SIGNED_NUMBER | STRING | SQSTRING | "LOW-VALUE" | "LOW-VALUES" SQSTRING: /'[^']*'/ redefines_clause: "REDEFINES" NAME -occurs_clause: "OCCURS" INT "TIMES"? ("DEPENDING" "ON" NAME)? key_clause? indexed_clause? +occurs_clause: "OCCURS" INT ("TO" INT)? "TIMES"? ("DEPENDING" "ON" NAME)? key_clause? indexed_clause? key_clause: ("ASCENDING" | "DESCENDING") "KEY" "IS"? NAME (","? NAME)* indexed_clause: "INDEXED" "BY" NAME (","? NAME)* usage_clause: USAGE_VAL USAGE_VAL: "COMP" | "COMP-3" | "COMP-5" | "BINARY" | "PACKED-DECIMAL" | "DISPLAY" LEVEL: /0[1-9]|[1-4][0-9]|49|77|88/ -NAME: /[A-Z][A-Z0-9-]*/ +NAME: /[A-Z][A-Z0-9-]*/i PICTURE_STRING: /[0-9A-Z()+,\-*\/V]+/i INT: /[0-9]+/ DOT: /\./ diff --git a/cobol_testgen/read.py b/cobol_testgen/read.py index e3d4568..03bb24d 100644 --- a/cobol_testgen/read.py +++ b/cobol_testgen/read.py @@ -27,6 +27,10 @@ def _is_fixed_format(source: str) -> bool: def preprocess(source: str) -> str: + # COPY 预处理：展开或移除 COPY 语句 + # Lark 语法不支持 COPY（这是预处理指令），必须在解析前处理 + source = resolve_copybooks(source, '.') + fixed = _is_fixed_format(source) lines = [] for raw_line in source.splitlines(): @@ -110,10 +114,12 @@ def resolve_copybooks(source: str, source_dir: str) -> str: re.escape(old.strip()), new.strip(), cb, flags=re.IGNORECASE ) - result.append(f' * COPY {name}') + # 展开 COPYBOOK 内容，不添加注释行（避免 Lark 在 FD 块内看到注释） result.append(cb) else: - result.append(line) + # COPY 未找到时完全跳过（预处理指令，Lark 不应处理） + # 该行可能在 FD/SD 块内，保留会破坏 Lark 解析 + pass else: result.append(line) return '\n'.join(result) diff --git a/tests/cobol_testgen/test_read.py b/tests/cobol_testgen/test_read.py index 77f6838..f08fc13 100644 --- a/tests/cobol_testgen/test_read.py +++ b/tests/cobol_testgen/test_read.py @@ -92,11 +92,12 @@ def test_resolve_copybooks_found(): def test_resolve_copybooks_not_found(): - """COPY 文件不存在时返回含 NOT FOUND 或 NOTEXIST 的文本""" + """COPY 不可解析时移除该行（预处理器指令，Lark 不应处理）""" with tempfile.TemporaryDirectory() as tmp: src = " COPY NOTEXIST.\n" result = resolve_copybooks(src, tmp) - assert "NOT FOUND" in result or "NOTEXIST" in result.upper() + # COPY 被移除（无残留） + assert "NOTEXIST" not in result.upper() def test_resolve_copybooks_no_copy():