gh-110309: Prune empty constant in format specs (#110320)

Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>

diff --git a/Lib/test/test_fstring.py b/Lib/test/test_fstring.py
index 4f05a149a901b2132a0e8fe35542f354c7d2c426..dd8c2dd628ee13582802347da57fe1456e4e4ca8 100644 (file)
--- a/Lib/test/test_fstring.py
+++ b/Lib/test/test_fstring.py
@@ -514,6 +514,54 @@ x = (
         self.assertEqual(type(format_spec), ast.JoinedStr)
         self.assertEqual(len(format_spec.values), 0)
 
+    def test_ast_fstring_format_spec(self):
+        expr = "f'{1:{name}}'"
+
+        mod = ast.parse(expr)
+        self.assertEqual(type(mod), ast.Module)
+        self.assertEqual(len(mod.body), 1)
+
+        fstring = mod.body[0].value
+        self.assertEqual(type(fstring), ast.JoinedStr)
+        self.assertEqual(len(fstring.values), 1)
+
+        fv = fstring.values[0]
+        self.assertEqual(type(fv), ast.FormattedValue)
+
+        format_spec = fv.format_spec
+        self.assertEqual(type(format_spec), ast.JoinedStr)
+        self.assertEqual(len(format_spec.values), 1)
+
+        format_spec_value = format_spec.values[0]
+        self.assertEqual(type(format_spec_value), ast.FormattedValue)
+        self.assertEqual(format_spec_value.value.id, 'name')
+
+        expr = "f'{1:{name1}{name2}}'"
+
+        mod = ast.parse(expr)
+        self.assertEqual(type(mod), ast.Module)
+        self.assertEqual(len(mod.body), 1)
+
+        fstring = mod.body[0].value
+        self.assertEqual(type(fstring), ast.JoinedStr)
+        self.assertEqual(len(fstring.values), 1)
+
+        fv = fstring.values[0]
+        self.assertEqual(type(fv), ast.FormattedValue)
+
+        format_spec = fv.format_spec
+        self.assertEqual(type(format_spec), ast.JoinedStr)
+        self.assertEqual(len(format_spec.values), 2)
+
+        format_spec_value = format_spec.values[0]
+        self.assertEqual(type(format_spec_value), ast.FormattedValue)
+        self.assertEqual(format_spec_value.value.id, 'name1')
+
+        format_spec_value = format_spec.values[1]
+        self.assertEqual(type(format_spec_value), ast.FormattedValue)
+        self.assertEqual(format_spec_value.value.id, 'name2')
+
+
     def test_docstring(self):
         def f():
             f'''Not a docstring'''

diff --git a/Misc/NEWS.d/next/Core and Builtins/2023-10-03-23-26-18.gh-issue-110309.Y8nDOF.rst b/Misc/NEWS.d/next/Core and Builtins/2023-10-03-23-26-18.gh-issue-110309.Y8nDOF.rst
new file mode 100644 (file)
index 0000000..8304287
--- /dev/null
+++ b/Misc/NEWS.d/next/Core and Builtins/2023-10-03-23-26-18.gh-issue-110309.Y8nDOF.rst
@@ -0,0 +1 @@
+Remove unnecessary empty constant nodes in the ast of f-string specs.

diff --git a/Parser/action_helpers.c b/Parser/action_helpers.c
index 36e0750220a30d2013e91c45a8060e0a2cfe4f97..b8713a329d4ef64d11f06d415314dd7c5aef3766 100644 (file)
--- a/Parser/action_helpers.c
+++ b/Parser/action_helpers.c
@@ -998,18 +998,38 @@ _PyPegen_setup_full_format_spec(Parser *p, Token *colon, asdl_expr_seq *spec, in
         return NULL;
     }
 
-    // This is needed to keep compatibility with 3.11, where an empty format spec is parsed
-    // as an *empty* JoinedStr node, instead of having an empty constant in it.
-    if (asdl_seq_LEN(spec) == 1) {
-        expr_ty e = asdl_seq_GET(spec, 0);
-        if (e->kind == Constant_kind
-                && PyUnicode_Check(e->v.Constant.value)
-                && PyUnicode_GetLength(e->v.Constant.value) == 0) {
-            spec = _Py_asdl_expr_seq_new(0, arena);
+    // This is needed to keep compatibility with 3.11, where an empty format
+    // spec is parsed as an *empty* JoinedStr node, instead of having an empty
+    // constant in it.
+    Py_ssize_t n_items = asdl_seq_LEN(spec);
+    Py_ssize_t non_empty_count = 0;
+    for (Py_ssize_t i = 0; i < n_items; i++) {
+        expr_ty item = asdl_seq_GET(spec, i);
+        non_empty_count += !(item->kind == Constant_kind &&
+                             PyUnicode_CheckExact(item->v.Constant.value) &&
+                             PyUnicode_GET_LENGTH(item->v.Constant.value) == 0);
+    }
+    if (non_empty_count != n_items) {
+        asdl_expr_seq *resized_spec =
+            _Py_asdl_expr_seq_new(non_empty_count, p->arena);
+        if (resized_spec == NULL) {
+            return NULL;
+        }
+        Py_ssize_t j = 0;
+        for (Py_ssize_t i = 0; i < n_items; i++) {
+            expr_ty item = asdl_seq_GET(spec, i);
+            if (item->kind == Constant_kind &&
+                PyUnicode_CheckExact(item->v.Constant.value) &&
+                PyUnicode_GET_LENGTH(item->v.Constant.value) == 0) {
+                continue;
+            }
+            asdl_seq_SET(resized_spec, j++, item);
         }
+        assert(j == non_empty_count);
+        spec = resized_spec;
     }
-
-    expr_ty res = _PyAST_JoinedStr(spec, lineno, col_offset, end_lineno, end_col_offset, p->arena);
+    expr_ty res = _PyAST_JoinedStr(spec, lineno, col_offset, end_lineno,
+                                   end_col_offset, p->arena);
     if (!res) {
         return NULL;
     }