fix(bash): link executed scripts across files

graphify/extractors/bash.py

@@ -62,6 +62,7 @@ def add_edge(src: str, tgt: str, relation: str, line: int,
     add_edge(file_nid, entry_nid, "contains", 1)
 
     _BASH_SOURCE_COMMANDS = frozenset({"source", "."})
+    _BASH_SCRIPT_RUNNERS = frozenset({"bash", "sh", "zsh", "ksh", "dash"})
     # Parent node types that mean a contained command is part of a substitution
     # or expansion, not a real function call. Token-level filtering misses
     # these because `$(build)` exposes `build` as a child command whose name
@@ -161,11 +162,11 @@ def walk(node, parent_nid: str) -> None:
                 cmd_name_node = node.children[0]
             if cmd_name_node:
                 cmd = literal(cmd_name_node)
+                args = [c for c in node.children
+                        if c.type in ("word", "string", "concatenation")
+                        and c != cmd_name_node]
                 if cmd in _BASH_SOURCE_COMMANDS and cmd not in defined_functions:
                     # find the path argument (first word after command name)
-                    args = [c for c in node.children
-                            if c.type in ("word", "string", "concatenation")
-                            and c != cmd_name_node]
                     if args:
                         raw = _read_text(args[0], source).strip().strip("'\"")
                         line = node.start_point[0] + 1
@@ -184,6 +185,23 @@ def walk(node, parent_nid: str) -> None:
                             if tgt_nid:
                                 add_edge(file_nid, tgt_nid, "imports", line,
                                          context="import")
+                elif cmd and cmd not in defined_functions:
+                    raw = cmd if cmd.endswith(".sh") else None
+                    if cmd in _BASH_SCRIPT_RUNNERS and args:
+                        raw = literal(args[0])
+                    if raw and raw.endswith(".sh"):
+                        resolved = (path.parent / raw).resolve()
+                        if resolved.is_file():
+                            target_path = resolved
+                            if not path.is_absolute():
+                                try:
+                                    target_path = resolved.relative_to(Path.cwd().resolve())
+                                except ValueError:
+                                    pass
+                            caller_nid = entry_nid if parent_nid == file_nid else parent_nid
+                            add_edge(caller_nid, _make_id(str(target_path)) + "__entry",
+                                     "calls", node.start_point[0] + 1,
+                                     context="script_invocation")
             return
 
         if t == "declaration_command":

tests/test_extract.py

@@ -1210,6 +1210,86 @@ def test_extract_bash_emits_source_imports_from(tmp_path):
     assert import_edges[0].get("context") == "import"
 
 
+@pytest.mark.parametrize("command", ["./helpers.sh", "bash ./helpers.sh"])
+def test_extract_bash_emits_script_invocation_calls(tmp_path, command):
+    helpers = tmp_path / "helpers.sh"
+    helpers.write_text("#!/bin/bash\necho helper\n", encoding="utf-8")
+    script = tmp_path / "deploy.sh"
+    script.write_text(f"#!/bin/bash\n{command}\n", encoding="utf-8")
+
+    result = extract_bash(script)
+    invocation = [
+        edge for edge in result["edges"]
+        if edge.get("relation") == "calls" and edge.get("context") == "script_invocation"
+    ]
+
+    assert invocation == [{
+        "source": _make_id(str(script)) + "__entry",
+        "target": _make_id(str(helpers.resolve())) + "__entry",
+        "relation": "calls",
+        "confidence": "EXTRACTED",
+        "source_file": str(script),
+        "source_location": "L2",
+        "weight": 1.0,
+        "context": "script_invocation",
+    }]
+
+
+def test_extract_bash_skips_missing_and_shadowed_script_invocations(tmp_path):
+    helpers = tmp_path / "helpers.sh"
+    helpers.write_text("#!/bin/bash\necho helper\n", encoding="utf-8")
+    script = tmp_path / "deploy.sh"
+    script.write_text(
+        "#!/bin/bash\n"
+        "bash() { echo custom; }\n"
+        "bash ./helpers.sh\n"
+        "./missing.sh\n",
+        encoding="utf-8",
+    )
+
+    result = extract_bash(script)
+
+    assert not any(edge.get("context") == "script_invocation" for edge in result["edges"])
+
+
+def test_extract_bash_skips_dynamic_script_invocation(tmp_path):
+    helpers = tmp_path / "helpers.sh"
+    helpers.write_text("#!/bin/bash\necho helper\n", encoding="utf-8")
+    script = tmp_path / "deploy.sh"
+    script.write_text('#!/bin/bash\nbash "./$SCRIPT.sh"\n', encoding="utf-8")
+
+    result = extract_bash(script)
+
+    assert not any(edge.get("context") == "script_invocation" for edge in result["edges"])
+
+
+def test_extract_bash_relative_script_invocation_targets_existing_entrypoint(tmp_path, monkeypatch):
+    monkeypatch.chdir(tmp_path)
+    helpers = Path("helpers.sh")
+    helpers.write_text("#!/bin/bash\necho helper\n", encoding="utf-8")
+    script = Path("deploy.sh")
+    script.write_text("#!/bin/bash\n./helpers.sh\n", encoding="utf-8")
+
+    result = extract([script, helpers], cache_root=tmp_path, parallel=False)
+    node_ids = {node["id"] for node in result["nodes"]}
+    invocation = next(edge for edge in result["edges"] if edge.get("context") == "script_invocation")
+
+    assert invocation["target"] in node_ids
+
+
+def test_extract_bash_attributes_script_invocation_to_function(tmp_path):
+    helpers = tmp_path / "helpers.sh"
+    helpers.write_text("#!/bin/bash\necho helper\n", encoding="utf-8")
+    script = tmp_path / "deploy.sh"
+    script.write_text("#!/bin/bash\ndeploy() { bash ./helpers.sh; }\n", encoding="utf-8")
+
+    result = extract_bash(script)
+    deploy = next(node for node in result["nodes"] if node["label"] == "deploy()")
+    invocation = next(edge for edge in result["edges"] if edge.get("context") == "script_invocation")
+
+    assert invocation["source"] == deploy["id"]
+
+
 def test_extract_bash_no_self_loops():
     result = extract_bash(FIXTURES / "sample.sh")
     for e in result["edges"]: