bpo-43798: Add source location attributes to alias (GH-25324)

* Add source location attributes to alias.
* Move alias star construction to pegen helper.

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

Grammar/python.gram

@@ -144,20 +144,20 @@ import_from[stmt_ty]:
 import_from_targets[asdl_alias_seq*]:
     | '(' a=import_from_as_names [','] ')' { a }
     | import_from_as_names !','
-    | '*' { (asdl_alias_seq*)_PyPegen_singleton_seq(p, CHECK(alias_ty, _PyPegen_alias_for_star(p))) }
+    | '*' { (asdl_alias_seq*)_PyPegen_singleton_seq(p, CHECK(alias_ty, _PyPegen_alias_for_star(p, EXTRA))) }
     | invalid_import_from_targets
 import_from_as_names[asdl_alias_seq*]:
     | a[asdl_alias_seq*]=','.import_from_as_name+ { a }
 import_from_as_name[alias_ty]:
     | a=NAME b=['as' z=NAME { z }] { _PyAST_alias(a->v.Name.id,
                                                (b) ? ((expr_ty) b)->v.Name.id : NULL,
-                                               p->arena) }
+                                               EXTRA) }
 dotted_as_names[asdl_alias_seq*]:
     | a[asdl_alias_seq*]=','.dotted_as_name+ { a }
 dotted_as_name[alias_ty]:
     | a=dotted_name b=['as' z=NAME { z }] { _PyAST_alias(a->v.Name.id,
                                                       (b) ? ((expr_ty) b)->v.Name.id : NULL,
-                                                      p->arena) }
+                                                      EXTRA) }
 dotted_name[expr_ty]:
     | a=dotted_name '.' b=NAME { _PyPegen_join_names_with_dot(p, a, b) }
     | NAME

Include/internal/pycore_ast.h

@@ -543,6 +543,10 @@ struct _keyword {
 struct _alias {
     identifier name;
     identifier asname;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
 };
 
 struct _withitem {
@@ -751,7 +755,9 @@ arg_ty _PyAST_arg(identifier arg, expr_ty annotation, string type_comment, int
 keyword_ty _PyAST_keyword(identifier arg, expr_ty value, int lineno, int
                           col_offset, int end_lineno, int end_col_offset,
                           PyArena *arena);
-alias_ty _PyAST_alias(identifier name, identifier asname, PyArena *arena);
+alias_ty _PyAST_alias(identifier name, identifier asname, int lineno, int
+                      col_offset, int end_lineno, int end_col_offset, PyArena
+                      *arena);
 withitem_ty _PyAST_withitem(expr_ty context_expr, expr_ty optional_vars,
                             PyArena *arena);
 match_case_ty _PyAST_match_case(expr_ty pattern, expr_ty guard, asdl_stmt_seq *

Lib/test/test_asdl_parser.py

@@ -62,7 +62,9 @@ class TestAsdlParser(unittest.TestCase):
         alias = self.types['alias']
         self.assertEqual(
             str(alias),
-            'Product([Field(identifier, name), Field(identifier, asname, opt=True)])')
+            'Product([Field(identifier, name), Field(identifier, asname, opt=True)], '
+            '[Field(int, lineno), Field(int, col_offset), '
+            'Field(int, end_lineno, opt=True), Field(int, end_col_offset, opt=True)])')
 
     def test_attributes(self):
         stmt = self.types['stmt']

Lib/test/test_ast.py

@@ -335,6 +335,26 @@ class AST_Tests(unittest.TestCase):
         mod.body[0].module = " __future__ ".strip()
         compile(mod, "<test>", "exec")
 
+    def test_alias(self):
+        im = ast.parse("from bar import y").body[0]
+        self.assertEqual(len(im.names), 1)
+        alias = im.names[0]
+        self.assertEqual(alias.name, 'y')
+        self.assertIsNone(alias.asname)
+        self.assertEqual(alias.lineno, 1)
+        self.assertEqual(alias.end_lineno, 1)
+        self.assertEqual(alias.col_offset, 16)
+        self.assertEqual(alias.end_col_offset, 17)
+
+        im = ast.parse("from bar import *").body[0]
+        alias = im.names[0]
+        self.assertEqual(alias.name, '*')
+        self.assertIsNone(alias.asname)
+        self.assertEqual(alias.lineno, 1)
+        self.assertEqual(alias.end_lineno, 1)
+        self.assertEqual(alias.col_offset, 16)
+        self.assertEqual(alias.end_col_offset, 17)
+
     def test_base_classes(self):
         self.assertTrue(issubclass(ast.For, ast.stmt))
         self.assertTrue(issubclass(ast.Name, ast.expr))
@@ -1037,7 +1057,8 @@ Module(
 
     def test_level_as_none(self):
         body = [ast.ImportFrom(module='time',
-                               names=[ast.alias(name='sleep')],
+                               names=[ast.alias(name='sleep',
+                                                lineno=0, col_offset=0)],
                                level=None,
                                lineno=0, col_offset=0)]
         mod = ast.Module(body, [])
@@ -1735,6 +1756,7 @@ class EndPositionTests(unittest.TestCase):
         ''').strip()
         imp = ast.parse(s).body[0]
         self._check_end_pos(imp, 3, 1)
+        self._check_end_pos(imp.names[2], 2, 16)
 
     def test_slices(self):
         s1 = 'f()[1, 2] [0]'
@@ -2095,8 +2117,8 @@ exec_results = [
 ('Module', [('Try', (1, 0, 4, 6), [('Pass', (2, 2, 2, 6))], [('ExceptHandler', (3, 0, 4, 6), ('Name', (3, 7, 3, 16), 'Exception', ('Load',)), None, [('Pass', (4, 2, 4, 6))])], [], [])], []),
 ('Module', [('Try', (1, 0, 4, 6), [('Pass', (2, 2, 2, 6))], [], [], [('Pass', (4, 2, 4, 6))])], []),
 ('Module', [('Assert', (1, 0, 1, 8), ('Name', (1, 7, 1, 8), 'v', ('Load',)), None)], []),
-('Module', [('Import', (1, 0, 1, 10), [('alias', 'sys', None)])], []),
+('Module', [('Import', (1, 0, 1, 10), [('alias', (1, 7, 1, 10), 'sys', None)])], []),
-('Module', [('ImportFrom', (1, 0, 1, 17), 'sys', [('alias', 'v', None)], 0)], []),
+('Module', [('ImportFrom', (1, 0, 1, 17), 'sys', [('alias', (1, 16, 1, 17), 'v', None)], 0)], []),
 ('Module', [('Global', (1, 0, 1, 8), ['v'])], []),
 ('Module', [('Expr', (1, 0, 1, 1), ('Constant', (1, 0, 1, 1), 1, None))], []),
 ('Module', [('Pass', (1, 0, 1, 4))], []),

Lib/test/test_peg_generator/test_c_parser.py

@@ -314,7 +314,7 @@ class TestCParser(TempdirManager, unittest.TestCase):
                             )
         simple_name[expr_ty]: NAME
         import_as_names_from[asdl_alias_seq*]: a[asdl_alias_seq*]=','.import_as_name_from+ { a }
-        import_as_name_from[alias_ty]: a=NAME 'as' b=NAME { _PyAST_alias(((expr_ty) a)->v.Name.id, ((expr_ty) b)->v.Name.id, p->arena) }
+        import_as_name_from[alias_ty]: a=NAME 'as' b=NAME { _PyAST_alias(((expr_ty) a)->v.Name.id, ((expr_ty) b)->v.Name.id, EXTRA) }
         """
         test_source = """
         for stmt in ("from a import b as c", "from . import a as b"):

Misc/NEWS.d/next/Core and Builtins/2021-04-10-00-01-43.bpo-43798.p_nJFM.rst

@@ -0,0 +1 @@
+:class:`ast.alias` nodes now include source location metadata attributes e.g. lineno, col_offset.

Parser/Python.asdl

@@ -124,6 +124,7 @@ module Python
 
     -- import name with optional 'as' alias.
     alias = (identifier name, identifier? asname)
+             attributes (int lineno, int col_offset, int? end_lineno, int? end_col_offset)
 
     withitem = (expr context_expr, expr? optional_vars)
 

Parser/parser.c

@@ -3310,6 +3310,15 @@ import_from_targets_rule(Parser *p)
     }
     asdl_alias_seq* _res = NULL;
     int _mark = p->mark;
+    if (p->mark == p->fill && _PyPegen_fill_token(p) < 0) {
+        p->error_indicator = 1;
+        D(p->level--);
+        return NULL;
+    }
+    int _start_lineno = p->tokens[_mark]->lineno;
+    UNUSED(_start_lineno); // Only used by EXTRA macro
+    int _start_col_offset = p->tokens[_mark]->col_offset;
+    UNUSED(_start_col_offset); // Only used by EXTRA macro
     { // '(' import_from_as_names ','? ')'
         if (p->error_indicator) {
             D(p->level--);
@@ -3377,7 +3386,16 @@ import_from_targets_rule(Parser *p)
         )
         {
             D(fprintf(stderr, "%*c+ import_from_targets[%d-%d]: %s succeeded!\n", p->level, ' ', _mark, p->mark, "'*'"));
-            _res = ( asdl_alias_seq * ) _PyPegen_singleton_seq ( p , CHECK ( alias_ty , _PyPegen_alias_for_star ( p ) ) );
+            Token *_token = _PyPegen_get_last_nonnwhitespace_token(p);
+            if (_token == NULL) {
+                D(p->level--);
+                return NULL;
+            }
+            int _end_lineno = _token->end_lineno;
+            UNUSED(_end_lineno); // Only used by EXTRA macro
+            int _end_col_offset = _token->end_col_offset;
+            UNUSED(_end_col_offset); // Only used by EXTRA macro
+            _res = ( asdl_alias_seq * ) _PyPegen_singleton_seq ( p , CHECK ( alias_ty , _PyPegen_alias_for_star ( p , EXTRA ) ) );
             if (_res == NULL && PyErr_Occurred()) {
                 p->error_indicator = 1;
                 D(p->level--);
@@ -3466,6 +3484,15 @@ import_from_as_name_rule(Parser *p)
     }
     alias_ty _res = NULL;
     int _mark = p->mark;
+    if (p->mark == p->fill && _PyPegen_fill_token(p) < 0) {
+        p->error_indicator = 1;
+        D(p->level--);
+        return NULL;
+    }
+    int _start_lineno = p->tokens[_mark]->lineno;
+    UNUSED(_start_lineno); // Only used by EXTRA macro
+    int _start_col_offset = p->tokens[_mark]->col_offset;
+    UNUSED(_start_col_offset); // Only used by EXTRA macro
     { // NAME ['as' NAME]
         if (p->error_indicator) {
             D(p->level--);
@@ -3481,7 +3508,16 @@ import_from_as_name_rule(Parser *p)
         )
         {
             D(fprintf(stderr, "%*c+ import_from_as_name[%d-%d]: %s succeeded!\n", p->level, ' ', _mark, p->mark, "NAME ['as' NAME]"));
-            _res = _PyAST_alias ( a -> v . Name . id , ( b ) ? ( ( expr_ty ) b ) -> v . Name . id : NULL , p -> arena );
+            Token *_token = _PyPegen_get_last_nonnwhitespace_token(p);
+            if (_token == NULL) {
+                D(p->level--);
+                return NULL;
+            }
+            int _end_lineno = _token->end_lineno;
+            UNUSED(_end_lineno); // Only used by EXTRA macro
+            int _end_col_offset = _token->end_col_offset;
+            UNUSED(_end_col_offset); // Only used by EXTRA macro
+            _res = _PyAST_alias ( a -> v . Name . id , ( b ) ? ( ( expr_ty ) b ) -> v . Name . id : NULL , EXTRA );
             if (_res == NULL && PyErr_Occurred()) {
                 p->error_indicator = 1;
                 D(p->level--);
@@ -3551,6 +3587,15 @@ dotted_as_name_rule(Parser *p)
     }
     alias_ty _res = NULL;
     int _mark = p->mark;
+    if (p->mark == p->fill && _PyPegen_fill_token(p) < 0) {
+        p->error_indicator = 1;
+        D(p->level--);
+        return NULL;
+    }
+    int _start_lineno = p->tokens[_mark]->lineno;
+    UNUSED(_start_lineno); // Only used by EXTRA macro
+    int _start_col_offset = p->tokens[_mark]->col_offset;
+    UNUSED(_start_col_offset); // Only used by EXTRA macro
     { // dotted_name ['as' NAME]
         if (p->error_indicator) {
             D(p->level--);
@@ -3566,7 +3611,16 @@ dotted_as_name_rule(Parser *p)
         )
         {
             D(fprintf(stderr, "%*c+ dotted_as_name[%d-%d]: %s succeeded!\n", p->level, ' ', _mark, p->mark, "dotted_name ['as' NAME]"));
-            _res = _PyAST_alias ( a -> v . Name . id , ( b ) ? ( ( expr_ty ) b ) -> v . Name . id : NULL , p -> arena );
+            Token *_token = _PyPegen_get_last_nonnwhitespace_token(p);
+            if (_token == NULL) {
+                D(p->level--);
+                return NULL;
+            }
+            int _end_lineno = _token->end_lineno;
+            UNUSED(_end_lineno); // Only used by EXTRA macro
+            int _end_col_offset = _token->end_col_offset;
+            UNUSED(_end_col_offset); // Only used by EXTRA macro
+            _res = _PyAST_alias ( a -> v . Name . id , ( b ) ? ( ( expr_ty ) b ) -> v . Name . id : NULL , EXTRA );
             if (_res == NULL && PyErr_Occurred()) {
                 p->error_indicator = 1;
                 D(p->level--);

Parser/pegen.c

@@ -1555,8 +1555,8 @@ _PyPegen_seq_count_dots(asdl_seq *seq)
 
 /* Creates an alias with '*' as the identifier name */
 alias_ty
-_PyPegen_alias_for_star(Parser *p)
+_PyPegen_alias_for_star(Parser *p, int lineno, int col_offset, int end_lineno,
-{
+                        int end_col_offset, PyArena *arena) {
     PyObject *str = PyUnicode_InternFromString("*");
     if (!str) {
         return NULL;
@@ -1565,7 +1565,7 @@ _PyPegen_alias_for_star(Parser *p)
         Py_DECREF(str);
         return NULL;
     }
-    return _PyAST_alias(str, NULL, p->arena);
+    return _PyAST_alias(str, NULL, lineno, col_offset, end_lineno, end_col_offset, arena);
 }
 
 /* Creates a new asdl_seq* with the identifiers of all the names in seq */

Parser/pegen.h

@@ -237,7 +237,7 @@ asdl_seq *_PyPegen_seq_append_to_end(Parser *, asdl_seq *, void *);
 asdl_seq *_PyPegen_seq_flatten(Parser *, asdl_seq *);
 expr_ty _PyPegen_join_names_with_dot(Parser *, expr_ty, expr_ty);
 int _PyPegen_seq_count_dots(asdl_seq *);
-alias_ty _PyPegen_alias_for_star(Parser *);
+alias_ty _PyPegen_alias_for_star(Parser *, int, int, int, int, PyArena *);
 asdl_identifier_seq *_PyPegen_map_names_to_ids(Parser *, asdl_expr_seq *);
 CmpopExprPair *_PyPegen_cmpop_expr_pair(Parser *, cmpop_ty, expr_ty);
 asdl_int_seq *_PyPegen_get_cmpops(Parser *p, asdl_seq *);

Python/Python-ast.c

@@ -675,6 +675,12 @@ static const char * const keyword_fields[]={
     "value",
 };
 static PyObject* ast2obj_alias(struct ast_state *state, void*);
+static const char * const alias_attributes[] = {
+    "lineno",
+    "col_offset",
+    "end_lineno",
+    "end_col_offset",
+};
 static const char * const alias_fields[]={
     "name",
     "asname",
@@ -1707,9 +1713,15 @@ init_types(struct ast_state *state)
                                   alias_fields, 2,
         "alias(identifier name, identifier? asname)");
     if (!state->alias_type) return 0;
-    if (!add_attributes(state, state->alias_type, NULL, 0)) return 0;
+    if (!add_attributes(state, state->alias_type, alias_attributes, 4)) return
+        0;
     if (PyObject_SetAttr(state->alias_type, state->asname, Py_None) == -1)
         return 0;
+    if (PyObject_SetAttr(state->alias_type, state->end_lineno, Py_None) == -1)
+        return 0;
+    if (PyObject_SetAttr(state->alias_type, state->end_col_offset, Py_None) ==
+        -1)
+        return 0;
     state->withitem_type = make_type(state, "withitem", state->AST_type,
                                      withitem_fields, 2,
         "withitem(expr context_expr, expr? optional_vars)");
@@ -3271,7 +3283,8 @@ _PyAST_keyword(identifier arg, expr_ty value, int lineno, int col_offset, int
 }
 
 alias_ty
-_PyAST_alias(identifier name, identifier asname, PyArena *arena)
+_PyAST_alias(identifier name, identifier asname, int lineno, int col_offset,
+             int end_lineno, int end_col_offset, PyArena *arena)
 {
     alias_ty p;
     if (!name) {
@@ -3284,6 +3297,10 @@ _PyAST_alias(identifier name, identifier asname, PyArena *arena)
         return NULL;
     p->name = name;
     p->asname = asname;
+    p->lineno = lineno;
+    p->col_offset = col_offset;
+    p->end_lineno = end_lineno;
+    p->end_col_offset = end_col_offset;
     return p;
 }
 
@@ -4850,6 +4867,26 @@ ast2obj_alias(struct ast_state *state, void* _o)
     if (PyObject_SetAttr(result, state->asname, value) == -1)
         goto failed;
     Py_DECREF(value);
+    value = ast2obj_int(state, o->lineno);
+    if (!value) goto failed;
+    if (PyObject_SetAttr(result, state->lineno, value) < 0)
+        goto failed;
+    Py_DECREF(value);
+    value = ast2obj_int(state, o->col_offset);
+    if (!value) goto failed;
+    if (PyObject_SetAttr(result, state->col_offset, value) < 0)
+        goto failed;
+    Py_DECREF(value);
+    value = ast2obj_int(state, o->end_lineno);
+    if (!value) goto failed;
+    if (PyObject_SetAttr(result, state->end_lineno, value) < 0)
+        goto failed;
+    Py_DECREF(value);
+    value = ast2obj_int(state, o->end_col_offset);
+    if (!value) goto failed;
+    if (PyObject_SetAttr(result, state->end_col_offset, value) < 0)
+        goto failed;
+    Py_DECREF(value);
     return result;
 failed:
     Py_XDECREF(value);
@@ -9723,6 +9760,10 @@ obj2ast_alias(struct ast_state *state, PyObject* obj, alias_ty* out, PyArena*
     PyObject* tmp = NULL;
     identifier name;
     identifier asname;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
 
     if (_PyObject_LookupAttr(obj, state->name, &tmp) < 0) {
         return 1;
@@ -9750,7 +9791,60 @@ obj2ast_alias(struct ast_state *state, PyObject* obj, alias_ty* out, PyArena*
         if (res != 0) goto failed;
         Py_CLEAR(tmp);
     }
-    *out = _PyAST_alias(name, asname, arena);
+    if (_PyObject_LookupAttr(obj, state->lineno, &tmp) < 0) {
+        return 1;
+    }
+    if (tmp == NULL) {
+        PyErr_SetString(PyExc_TypeError, "required field \"lineno\" missing from alias");
+        return 1;
+    }
+    else {
+        int res;
+        res = obj2ast_int(state, tmp, &lineno, arena);
+        if (res != 0) goto failed;
+        Py_CLEAR(tmp);
+    }
+    if (_PyObject_LookupAttr(obj, state->col_offset, &tmp) < 0) {
+        return 1;
+    }
+    if (tmp == NULL) {
+        PyErr_SetString(PyExc_TypeError, "required field \"col_offset\" missing from alias");
+        return 1;
+    }
+    else {
+        int res;
+        res = obj2ast_int(state, tmp, &col_offset, arena);
+        if (res != 0) goto failed;
+        Py_CLEAR(tmp);
+    }
+    if (_PyObject_LookupAttr(obj, state->end_lineno, &tmp) < 0) {
+        return 1;
+    }
+    if (tmp == NULL || tmp == Py_None) {
+        Py_CLEAR(tmp);
+        end_lineno = 0;
+    }
+    else {
+        int res;
+        res = obj2ast_int(state, tmp, &end_lineno, arena);
+        if (res != 0) goto failed;
+        Py_CLEAR(tmp);
+    }
+    if (_PyObject_LookupAttr(obj, state->end_col_offset, &tmp) < 0) {
+        return 1;
+    }
+    if (tmp == NULL || tmp == Py_None) {
+        Py_CLEAR(tmp);
+        end_col_offset = 0;
+    }
+    else {
+        int res;
+        res = obj2ast_int(state, tmp, &end_col_offset, arena);
+        if (res != 0) goto failed;
+        Py_CLEAR(tmp);
+    }
+    *out = _PyAST_alias(name, asname, lineno, col_offset, end_lineno,
+                        end_col_offset, arena);
     return 0;
 failed:
     Py_XDECREF(tmp);