feat(schema): stamp can:// ids across the symbol-table tree

Author: rahlk

codeanalyzer/core.py

@@ -17,6 +17,7 @@
     model_dump_json,
     model_validate_json,
 )
+from codeanalyzer.schema.assign_ids import assign_ids
 from codeanalyzer.schema.py_schema import PyCallEdge
 from codeanalyzer.semantic_analysis.call_graph import (
     filter_external_edges,
@@ -455,6 +456,8 @@ def analyze(self) -> PyApplication:
             .build()
         )
 
+        assign_ids(app, self.options.app_name or self.project_dir.name)
+
         # L3/L4 dataflow emission rebuilt on the v2 tree in Stage 3+
 
         # Save to cache

codeanalyzer/schema/assign_ids.py

@@ -0,0 +1,37 @@
+"""Walk the symbol-table tree and stamp every node with its can:// id."""
+from __future__ import annotations
+from typing import Dict
+from codeanalyzer.schema import ids
+from codeanalyzer.schema.py_schema import PyApplication, PyModule, PyClass, PyCallable
+
+
+def assign_ids(app: PyApplication, app_name: str) -> Dict[str, str]:
+    """Sets `.id` on the app + every module/class/callable. Returns a
+    `signature -> can://id` map for later stages (identity layer input)."""
+    app.id = ids.application_id(app_name); app.kind = "application"
+    sig_to_id: Dict[str, str] = {}
+
+    def do_callable(parent_id: str, c: PyCallable) -> None:
+        seg = ids.callable_sig_segment(c.name, [p.name for p in c.parameters])
+        c.id = ids.child_id(parent_id, seg)
+        sig_to_id[c.signature] = c.id
+        for ic in (c.inner_callables or {}).values():
+            do_callable(c.id, ic)
+        for icl in (c.inner_classes or {}).values():
+            do_class(c.id, icl)
+
+    def do_class(parent_id: str, cl: PyClass) -> None:
+        cl.id = ids.child_id(parent_id, cl.name); cl.kind = "class"
+        sig_to_id[cl.signature] = cl.id
+        for m in (cl.methods or {}).values():
+            do_callable(cl.id, m)
+        for ic in (cl.inner_classes or {}).values():
+            do_class(cl.id, ic)
+
+    for file_key, mod in app.symbol_table.items():
+        mod.id = ids.module_id(app_name, file_key); mod.kind = "module"
+        for fn in (mod.functions or {}).values():
+            do_callable(mod.id, fn)
+        for cl in (mod.classes or {}).values():
+            do_class(mod.id, cl)
+    return sig_to_id

test/test_v2_conformance.py

@@ -0,0 +1,15 @@
+from codeanalyzer.schema.assign_ids import assign_ids
+from codeanalyzer.schema.py_schema import PyApplication, PyModule, PyClass, PyCallable
+
+
+def test_ids_assigned_down_the_tree():
+    fn = PyCallable(name="hash", path="m.py", signature="m.Hasher.hash",
+                    parameters=[])
+    cl = PyClass(name="Hasher", signature="m.Hasher", methods={"hash": fn})
+    mod = PyModule(file_path="pkg/m.py", module_name="m", classes={"m.Hasher": cl})
+    app = PyApplication(symbol_table={"pkg/m.py": mod})
+    assign_ids(app, "myapp")
+    assert app.id == "can://python/myapp"
+    assert mod.id == "can://python/myapp/pkg/m.py"
+    assert cl.id == "can://python/myapp/pkg/m.py/Hasher"
+    assert fn.id == "can://python/myapp/pkg/m.py/Hasher/hash()"