Merge pull request #1049 from DeusData/fix/988-py-alias-import

fix(py-lsp): resolve calls through aliased imports

Author: DeusData

internal/cbm/lsp/py_lsp.c

@@ -206,8 +206,19 @@ void py_lsp_bind_imports(PyLSPContext *ctx) {
             // resolution checks the registry to upgrade to MODULE / class
             // / function as appropriate.
             t = cbm_type_named(ctx->arena, qn);
+        } else if (strchr(qn, '.') != NULL) {
+            // Dotted path whose tail does NOT match the local name: an
+            // ALIASED binding — `from X import Y as Z` (Z names function/
+            // class X.Y) or `import a.b as z` (z names module a.b). The
+            // CBMImport shape cannot distinguish the two, but NAMED(qn)
+            // covers both: phase 6 upgrades it to the registered function/
+            // class for the from-import and to MODULE for the module alias.
+            // Binding MODULE here made `g()` calls on `from m import f as g`
+            // resolve as calls on a module — lsp=MISS, and the whole CALLS
+            // edge was lost (#988).
+            t = cbm_type_named(ctx->arena, qn);
         } else {
-            // `import X` / `import X as Y` — bind to MODULE(X).
+            // `import X` / `import X as Y` (single segment) — MODULE(X).
             t = cbm_type_module(ctx->arena, qn);
         }
         py_scope_bind(ctx, local, t);
@@ -1910,11 +1921,23 @@ static void py_emit_call_for(PyLSPContext *ctx, TSNode call_node) {
             py_scope_restore(ctx, saved);
             return;
         }
-        // Constructor call (ClassName())
+        // Constructor call (ClassName()) — or a call through any other
+        // NAMED scope binding (aliased imports land here too).
         const CBMType *in_scope = cbm_scope_lookup(ctx->current_scope, fname);
         if (!cbm_type_is_unknown(in_scope) && in_scope->kind == CBM_TYPE_NAMED) {
             const char *qn = in_scope->data.named.qualified_name;
-            py_emit_resolved_call(ctx, qn, "lsp_constructor", 0.85f);
+            const char *tail = strrchr(qn, '.');
+            const char *qn_short = tail ? tail + 1 : qn;
+            if (strcmp(qn_short, fname) != 0) {
+                /* Aliased binding (`from m import f as g; g()`): the textual
+                 * callee ("g") differs from the resolved QN's tail ("f"), so
+                 * the pass join would never match them. Stash the textual
+                 * name in `reason` under a join-gated strategy, mirroring
+                 * lsp_dict_dispatch (see lsp_resolve.h) (#988). */
+                py_emit_resolved_call_reason(ctx, qn, "lsp_import_alias", 0.85f, fname);
+            } else {
+                py_emit_resolved_call(ctx, qn, "lsp_constructor", 0.85f);
+            }
             return;
         }
         // Module-local function

src/pipeline/lsp_resolve.h

@@ -174,6 +174,7 @@ static inline const CBMResolvedCall *cbm_pipeline_find_lsp_resolution(
                    strcmp(rc->strategy, "lsp_method_ref_ctor") == 0 ||
                    strcmp(rc->strategy, "lsp_method_ref_ctor_synth") == 0 ||
                    strcmp(rc->strategy, "lsp_dict_dispatch") == 0 ||
+                   strcmp(rc->strategy, "lsp_import_alias") == 0 ||
                    strcmp(rc->strategy, "lsp_destructor") == 0 ||
                    strcmp(rc->strategy, "php_method_dynamic") == 0) &&
                   strcmp(cbm_lsp_bare_segment(rc->reason), call_short) == 0)) {

tests/test_lang_contract.c

@@ -1207,6 +1207,41 @@ TEST(contract_edge_imports_alias_resolves_real_file_issue767) {
     PASS();
 }
 
+/* True if a CALLS edge exists whose source QN ends with `src_suffix` and
+ * target QN ends with `tgt_suffix`. */
+static int calls_edge_between(cbm_store_t *store, const char *project, const char *src_suffix,
+                              const char *tgt_suffix) {
+    cbm_edge_t *edges = NULL;
+    int n = 0;
+    if (cbm_store_find_edges_by_type(store, project, "CALLS", &edges, &n) != CBM_STORE_OK)
+        return 0;
+    int found = 0;
+    size_t ssl = strlen(src_suffix);
+    size_t tsl = strlen(tgt_suffix);
+    for (int i = 0; i < n && !found; i++) {
+        cbm_node_t s, t;
+        if (cbm_store_find_node_by_id(store, edges[i].source_id, &s) != CBM_STORE_OK)
+            continue;
+        if (cbm_store_find_node_by_id(store, edges[i].target_id, &t) != CBM_STORE_OK) {
+            cbm_node_free_fields(&s);
+            continue;
+        }
+        const char *sq = s.qualified_name;
+        const char *tq = t.qualified_name;
+        if (sq && tq) {
+            size_t sql = strlen(sq);
+            size_t tql = strlen(tq);
+            if (sql >= ssl && strcmp(sq + sql - ssl, src_suffix) == 0 && tql >= tsl &&
+                strcmp(tq + tql - tsl, tgt_suffix) == 0)
+                found = 1;
+        }
+        cbm_node_free_fields(&s);
+        cbm_node_free_fields(&t);
+    }
+    cbm_store_free_edges(edges, n);
+    return found;
+}
+
 /* #999: URLs in CI/tooling configs (.pre-commit-config.yaml, .github/
  * workflows) are repository references for TOOLING, not endpoints this
  * service exposes. They must not mint __route__infra__ Route nodes — the
@@ -1327,6 +1362,55 @@ static int calls_edge_targets(cbm_store_t *store, const char *project, const cha
     return found;
 }
 
+/* #988: `from m import f as g; g()` produced NO CALLS edge — the Python LSP
+ * bound the alias as MODULE("m.f") (the from-style heuristic keys on the
+ * local name matching the module-path tail, which an alias never does), so
+ * the call missed, and the registry fallback did not cover the minimal
+ * two-file shape either. The alias must resolve exactly like the plain
+ * import. The other import forms are pinned alongside as invariance guards
+ * (all resolve on main today and must keep resolving). */
+TEST(contract_edge_python_aliased_import_call_resolves_issue988) {
+    LangProj lp;
+    static const LangFile f[] = {
+        {"m.py", "def f(x):\n    return x + 1\n"},
+        {"pkg/__init__.py", ""},
+        {"pkg/dm.py", "def h(x):\n    return x\n"},
+        {"caller_alias.py", "from m import f as g\n"
+                            "\n"
+                            "def use_alias(x):\n"
+                            "    return g(x)\n"},
+        {"caller_plain.py", "from m import f\n"
+                            "\n"
+                            "def use_plain(x):\n"
+                            "    return f(x)\n"},
+        {"caller_modalias.py", "import m as mm\n"
+                               "\n"
+                               "def use_modalias(x):\n"
+                               "    return mm.f(x)\n"},
+        {"caller_dotalias.py", "import pkg.dm as dz\n"
+                               "\n"
+                               "def use_dotalias(x):\n"
+                               "    return dz.h(x)\n"}};
+    cbm_store_t *store = lang_index_files(&lp, f, 7);
+    ASSERT_TRUE(store != NULL);
+    int alias = calls_edge_between(store, lp.project, ".use_alias", ".m.f");
+    int plain = calls_edge_between(store, lp.project, ".use_plain", ".m.f");
+    int modalias = calls_edge_between(store, lp.project, ".use_modalias", ".m.f");
+    int dotalias = calls_edge_between(store, lp.project, ".use_dotalias", ".pkg.dm.h");
+    if (!alias || !plain || !modalias || !dotalias) {
+        fprintf(stderr,
+                "  [988] FAIL alias=%d plain=%d modalias=%d dotalias=%d (all import forms "
+                "must produce the CALLS edge)\n",
+                alias, plain, modalias, dotalias);
+    }
+    ASSERT_TRUE(alias); /* the #988 bug: aliased from-import call lost */
+    ASSERT_TRUE(plain);
+    ASSERT_TRUE(modalias);
+    ASSERT_TRUE(dotalias);
+    lang_cleanup(&lp, store);
+    PASS();
+}
+
 /* #871: a CommonJS require() binding shadowed call resolution. `const doThing
  * = require("../bs/doThing")` emitted a Variable def for `doThing` in the
  * importing module, so the call `doThing(...)` resolved to that same-module
@@ -1572,6 +1656,7 @@ SUITE(lang_contract) {
     RUN_TEST(contract_edge_workspaces_imports_issue408);
     RUN_TEST(contract_edge_imports_alias_no_phantom_folder_edge_issue767);
     RUN_TEST(contract_edge_imports_alias_resolves_real_file_issue767);
+    RUN_TEST(contract_edge_python_aliased_import_call_resolves_issue988);
     RUN_TEST(contract_edge_no_infra_routes_from_ci_configs_issue999);
     RUN_TEST(contract_edge_infra_routes_from_deploy_configs_still_minted);
     RUN_TEST(contract_edge_commonjs_require_call_resolves_issue871);

tests/test_py_lsp.c

@@ -192,16 +192,12 @@ TEST(pylsp_import_from_aliased) {
     const char *qns[] = {"pathlib.Path"};
     bind_imports_into_ctx(&ctx, &a, &reg, locals, qns, 1);
     const CBMType *t = py_lsp_lookup_in_scope(&ctx, "P");
-    /* Aliased name doesn't end module_qn with .P, so this binds as MODULE.
-     * Phase 6 registry lookup will downgrade to NAMED if registry has no
-     * matching module entry. Both behaviors are acceptable for v1; the
-     * test asserts the entry exists with the correct QN. */
-    ASSERT(t->kind == CBM_TYPE_NAMED || t->kind == CBM_TYPE_MODULE);
-    if (t->kind == CBM_TYPE_NAMED) {
-        ASSERT_STR_EQ(t->data.named.qualified_name, "pathlib.Path");
-    } else {
-        ASSERT_STR_EQ(t->data.module.module_qn, "pathlib.Path");
-    }
+    /* #988: an aliased from-import MUST bind NAMED (phase 6 upgrades it to
+     * the registered function/class/module). The earlier MODULE binding made
+     * `g()` calls on `from m import f as g` resolve as calls on a module —
+     * lsp=MISS and the CALLS edge was lost. */
+    ASSERT_EQ(t->kind, CBM_TYPE_NAMED);
+    ASSERT_STR_EQ(t->data.named.qualified_name, "pathlib.Path");
     cbm_arena_destroy(&a);
     PASS();
 }