Refactor parser compilation units into specific components (GH-29676)

This commit is contained in:
Pablo Galindo Salgado 2021-11-21 01:08:50 +00:00 committed by GitHub
parent f7638dd0f9
commit c9c4444d9f
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
10 changed files with 1910 additions and 1870 deletions

View file

@ -23,6 +23,8 @@
#define PyPARSE_TYPE_COMMENTS 0x0040
#define PyPARSE_ASYNC_HACKS 0x0080
#define CURRENT_POS (-5)
typedef struct _memo {
int type;
void *node;
@ -114,6 +116,7 @@ typedef struct {
int is_keyword;
} KeywordOrStarred;
// Internal parser functions
#if defined(Py_DEBUG)
void _PyPegen_clear_memo_statistics(void);
PyObject *_PyPegen_get_memo_statistics(void);
@ -123,7 +126,6 @@ int _PyPegen_insert_memo(Parser *p, int mark, int type, void *node);
int _PyPegen_update_memo(Parser *p, int mark, int type, void *node);
int _PyPegen_is_memoized(Parser *p, int type, void *pres);
int _PyPegen_lookahead_with_name(int, expr_ty (func)(Parser *), Parser *);
int _PyPegen_lookahead_with_int(int, Token *(func)(Parser *, int), Parser *, int);
int _PyPegen_lookahead_with_string(int , expr_ty (func)(Parser *, const char*), Parser *, const char*);
@ -139,23 +141,24 @@ int _PyPegen_fill_token(Parser *p);
expr_ty _PyPegen_name_token(Parser *p);
expr_ty _PyPegen_number_token(Parser *p);
void *_PyPegen_string_token(Parser *p);
const char *_PyPegen_get_expr_name(expr_ty);
Py_ssize_t _PyPegen_byte_offset_to_character_offset(PyObject *line, Py_ssize_t col_offset);
// Error handling functions and APIs
typedef enum {
STAR_TARGETS,
DEL_TARGETS,
FOR_TARGETS
} TARGETS_TYPE;
int _Pypegen_raise_decode_error(Parser *p);
void _PyPegen_raise_tokenizer_init_error(PyObject *filename);
int _Pypegen_tokenizer_error(Parser *p);
void *_PyPegen_raise_error(Parser *p, PyObject *errtype, const char *errmsg, ...);
void *_PyPegen_raise_error_known_location(Parser *p, PyObject *errtype,
Py_ssize_t lineno, Py_ssize_t col_offset,
Py_ssize_t end_lineno, Py_ssize_t end_col_offset,
const char *errmsg, va_list va);
void *_PyPegen_dummy_name(Parser *p, ...);
void * _PyPegen_seq_last_item(asdl_seq *seq);
#define PyPegen_last_item(seq, type) ((type)_PyPegen_seq_last_item((asdl_seq*)seq))
void * _PyPegen_seq_first_item(asdl_seq *seq);
#define PyPegen_first_item(seq, type) ((type)_PyPegen_seq_first_item((asdl_seq*)seq))
#define CURRENT_POS (-5)
void _Pypegen_set_syntax_error(Parser* p, Token* last_token);
Py_LOCAL_INLINE(void *)
RAISE_ERROR_KNOWN_LOCATION(Parser *p, PyObject *errtype,
Py_ssize_t lineno, Py_ssize_t col_offset,
@ -170,10 +173,6 @@ RAISE_ERROR_KNOWN_LOCATION(Parser *p, PyObject *errtype,
va_end(va);
return NULL;
}
#define UNUSED(expr) do { (void)(expr); } while (0)
#define EXTRA_EXPR(head, tail) head->lineno, (head)->col_offset, (tail)->end_lineno, (tail)->end_col_offset, p->arena
#define EXTRA _start_lineno, _start_col_offset, _end_lineno, _end_col_offset, p->arena
#define RAISE_SYNTAX_ERROR(msg, ...) _PyPegen_raise_error(p, PyExc_SyntaxError, msg, ##__VA_ARGS__)
#define RAISE_INDENTATION_ERROR(msg, ...) _PyPegen_raise_error(p, PyExc_IndentationError, msg, ##__VA_ARGS__)
#define RAISE_SYNTAX_ERROR_KNOWN_RANGE(a, b, msg, ...) \
@ -182,6 +181,7 @@ RAISE_ERROR_KNOWN_LOCATION(Parser *p, PyObject *errtype,
RAISE_ERROR_KNOWN_LOCATION(p, PyExc_SyntaxError, (a)->lineno, (a)->col_offset, (a)->end_lineno, (a)->end_col_offset, msg, ##__VA_ARGS__)
#define RAISE_SYNTAX_ERROR_STARTING_FROM(a, msg, ...) \
RAISE_ERROR_KNOWN_LOCATION(p, PyExc_SyntaxError, (a)->lineno, (a)->col_offset, CURRENT_POS, CURRENT_POS, msg, ##__VA_ARGS__)
#define RAISE_SYNTAX_ERROR_INVALID_TARGET(type, e) _RAISE_SYNTAX_ERROR_INVALID_TARGET(p, type, e)
Py_LOCAL_INLINE(void *)
CHECK_CALL(Parser *p, void *result)
@ -207,6 +207,39 @@ CHECK_CALL_NULL_ALLOWED(Parser *p, void *result)
#define CHECK(type, result) ((type) CHECK_CALL(p, result))
#define CHECK_NULL_ALLOWED(type, result) ((type) CHECK_CALL_NULL_ALLOWED(p, result))
expr_ty _PyPegen_get_invalid_target(expr_ty e, TARGETS_TYPE targets_type);
const char *_PyPegen_get_expr_name(expr_ty);
Py_LOCAL_INLINE(void *)
_RAISE_SYNTAX_ERROR_INVALID_TARGET(Parser *p, TARGETS_TYPE type, void *e)
{
expr_ty invalid_target = CHECK_NULL_ALLOWED(expr_ty, _PyPegen_get_invalid_target(e, type));
if (invalid_target != NULL) {
const char *msg;
if (type == STAR_TARGETS || type == FOR_TARGETS) {
msg = "cannot assign to %s";
}
else {
msg = "cannot delete %s";
}
return RAISE_SYNTAX_ERROR_KNOWN_LOCATION(
invalid_target,
msg,
_PyPegen_get_expr_name(invalid_target)
);
}
return RAISE_SYNTAX_ERROR("invalid syntax");
}
// Action utility functions
void *_PyPegen_dummy_name(Parser *p, ...);
void * _PyPegen_seq_last_item(asdl_seq *seq);
#define PyPegen_last_item(seq, type) ((type)_PyPegen_seq_last_item((asdl_seq*)seq))
void * _PyPegen_seq_first_item(asdl_seq *seq);
#define PyPegen_first_item(seq, type) ((type)_PyPegen_seq_first_item((asdl_seq*)seq))
#define UNUSED(expr) do { (void)(expr); } while (0)
#define EXTRA_EXPR(head, tail) head->lineno, (head)->col_offset, (tail)->end_lineno, (tail)->end_col_offset, p->arena
#define EXTRA _start_lineno, _start_col_offset, _end_lineno, _end_col_offset, p->arena
PyObject *_PyPegen_new_type_comment(Parser *, const char *);
Py_LOCAL_INLINE(PyObject *)
@ -248,13 +281,6 @@ INVALID_VERSION_CHECK(Parser *p, int version, char *msg, void *node)
arg_ty _PyPegen_add_type_comment_to_arg(Parser *, arg_ty, Token *);
PyObject *_PyPegen_new_identifier(Parser *, const char *);
Parser *_PyPegen_Parser_New(struct tok_state *, int, int, int, int *, PyArena *);
void _PyPegen_Parser_Free(Parser *);
mod_ty _PyPegen_run_parser_from_file_pointer(FILE *, int, PyObject *, const char *,
const char *, const char *, PyCompilerFlags *, int *, PyArena *);
void *_PyPegen_run_parser(Parser *);
mod_ty _PyPegen_run_parser_from_string(const char *, int, PyObject *, PyCompilerFlags *, PyArena *);
asdl_stmt_seq *_PyPegen_interactive_exit(Parser *);
asdl_seq *_PyPegen_singleton_seq(Parser *, void *);
asdl_seq *_PyPegen_seq_insert_in_front(Parser *, void *, asdl_seq *);
asdl_seq *_PyPegen_seq_append_to_end(Parser *, asdl_seq *, void *);
@ -295,40 +321,18 @@ asdl_seq *_PyPegen_join_sequences(Parser *, asdl_seq *, asdl_seq *);
int _PyPegen_check_barry_as_flufl(Parser *, Token *);
int _PyPegen_check_legacy_stmt(Parser *p, expr_ty t);
mod_ty _PyPegen_make_module(Parser *, asdl_stmt_seq *);
// Error reporting helpers
typedef enum {
STAR_TARGETS,
DEL_TARGETS,
FOR_TARGETS
} TARGETS_TYPE;
expr_ty _PyPegen_get_invalid_target(expr_ty e, TARGETS_TYPE targets_type);
#define RAISE_SYNTAX_ERROR_INVALID_TARGET(type, e) _RAISE_SYNTAX_ERROR_INVALID_TARGET(p, type, e)
Py_LOCAL_INLINE(void *)
_RAISE_SYNTAX_ERROR_INVALID_TARGET(Parser *p, TARGETS_TYPE type, void *e)
{
expr_ty invalid_target = CHECK_NULL_ALLOWED(expr_ty, _PyPegen_get_invalid_target(e, type));
if (invalid_target != NULL) {
const char *msg;
if (type == STAR_TARGETS || type == FOR_TARGETS) {
msg = "cannot assign to %s";
}
else {
msg = "cannot delete %s";
}
return RAISE_SYNTAX_ERROR_KNOWN_LOCATION(
invalid_target,
msg,
_PyPegen_get_expr_name(invalid_target)
);
}
return RAISE_SYNTAX_ERROR("invalid syntax");
}
void *_PyPegen_arguments_parsing_error(Parser *, expr_ty);
void *_PyPegen_nonparen_genexp_in_call(Parser *p, expr_ty args, asdl_comprehension_seq *comprehensions);
// Parser API
Parser *_PyPegen_Parser_New(struct tok_state *, int, int, int, int *, PyArena *);
void _PyPegen_Parser_Free(Parser *);
mod_ty _PyPegen_run_parser_from_file_pointer(FILE *, int, PyObject *, const char *,
const char *, const char *, PyCompilerFlags *, int *, PyArena *);
void *_PyPegen_run_parser(Parser *);
mod_ty _PyPegen_run_parser_from_string(const char *, int, PyObject *, PyCompilerFlags *, PyArena *);
asdl_stmt_seq *_PyPegen_interactive_exit(Parser *);
// Generated function in parse.c - function definition in python.gram
void *_PyPegen_parse(Parser *);