feat: UNSTRING解析增强 + 跨FD数值统一 + 文件I/O模块

cobol_testgen/core.py

@@ -1068,11 +1068,31 @@ class _BrParser:
         source_part = m.group(1).strip()
         targets_part = m.group(2).strip()
         source_vars = re.findall(r'[A-Z][A-Z0-9-]*', source_part)
-        targets = re.findall(r'[A-Z][A-Z0-9-]*', targets_part)
+        targets_clean = re.sub(r'\s+(DELIMITER|COUNT|TALLYING)\s+IN\s+[A-Z][A-Z0-9-]*', '', targets_part, flags=re.IGNORECASE)
+        targets = re.findall(r'[A-Z][A-Z0-9-]*', targets_clean)
         source_var = source_vars[0] if source_vars else ''
+
+        # Extract delimiter: DELIMITED BY <literal|identifier|SIZE>
+        delimiter = None
+        dm = re.search(r'DELIMITED\s+BY\s+(.+)', source_part, re.IGNORECASE)
+        if dm:
+            delim_raw = dm.group(1).strip()
+            if delim_raw.upper().startswith('SIZE'):
+                delimiter = None
+            elif delim_raw.startswith("'") or delim_raw.startswith('"'):
+                delimiter = delim_raw[1:-1] if len(delim_raw) >= 2 else None
+            else:
+                fid = re.match(r'[A-Z][A-Z0-9-]*', delim_raw, re.IGNORECASE)
+                delimiter = fid.group(0) if fid else None
+
         seq = BrSeq()
         for tgt in targets:
-            info = {'type': 'unstring_split', 'source_vars': [source_var], 'index': targets.index(tgt)}
+            info = {
+                'type': 'unstring_split',
+                'source_vars': [source_var],
+                'index': targets.index(tgt),
+                'delimiter': delimiter,
+            }
             self.assignments.setdefault(tgt, []).append(info)
             seq.add(Assign(tgt, info))
         return seq
@@ -1660,6 +1680,7 @@ def propagate_assignments(rec, assignments, fields, file_sec=None):
                 src_var = asgn.get('source_vars', [None])[0]
                 resolved_src = _resolve_subscript(src_var, rec) if src_var else None
                 idx = asgn.get('index', 0)
+                delimiter = asgn.get('delimiter')
                 if resolved_src and resolved_src not in rec:
                     children = _init_child_names(resolved_src, fields)
                     if children:
@@ -1667,10 +1688,14 @@ def propagate_assignments(rec, assignments, fields, file_sec=None):
                 if resolved_src and resolved_src in rec:
                     src_val = str(rec[resolved_src])
                     ftype = pi.get('type', 'unknown')
-                    if idx == 0:
-                        val = src_val
+                    if delimiter is not None:
+                        segments = src_val.split(delimiter)
+                        if idx < len(segments):
+                            val = segments[idx].strip()
+                        else:
+                            val = ' ' if ftype in ('alphanumeric', 'alphabetic') else '0'
                     else:
-                        val = ' ' if ftype in ('alphanumeric', 'alphabetic') else '0'
+                        val = src_val if idx == 0 else (' ' if ftype in ('alphanumeric', 'alphabetic') else '0')
                     if ftype in ('alphanumeric', 'alphabetic'):
                         val = val.ljust(pi.get('length', len(val)))[:pi.get('length', len(val))]
                     rec[resolved_tgt] = val