diff --git a/py/compile.c b/py/compile.c index 92736e22e5..f9e8de4573 100644 --- a/py/compile.c +++ b/py/compile.c @@ -2816,16 +2816,6 @@ STATIC void compile_node(compiler_t *comp, mp_parse_node_t pn) { case MP_PARSE_NODE_STRING: EMIT_ARG(load_const_str, arg); break; - case MP_PARSE_NODE_BYTES: - // only create and load the actual bytes object on the last pass - if (comp->pass != MP_PASS_EMIT) { - EMIT_ARG(load_const_obj, mp_const_none); - } else { - size_t len; - const byte *data = qstr_data(arg, &len); - EMIT_ARG(load_const_obj, mp_obj_new_bytes(data, len)); - } - break; case MP_PARSE_NODE_TOKEN: default: if (arg == MP_TOKEN_NEWLINE) { diff --git a/py/parse.c b/py/parse.c index ff639122cb..f18f03b3f1 100644 --- a/py/parse.c +++ b/py/parse.c @@ -388,9 +388,6 @@ void mp_parse_node_print(const mp_print_t *print, mp_parse_node_t pn, size_t ind case MP_PARSE_NODE_STRING: mp_printf(print, "str(%s)\n", qstr_str(arg)); break; - case MP_PARSE_NODE_BYTES: - mp_printf(print, "bytes(%s)\n", qstr_str(arg)); - break; default: assert(MP_PARSE_NODE_LEAF_KIND(pn) == MP_PARSE_NODE_TOKEN); mp_printf(print, "tok(%u)\n", (uint)arg); @@ -504,8 +501,8 @@ STATIC void push_result_token(parser_t *parser, uint8_t rule_id) { } else if (lex->tok_kind == MP_TOKEN_FLOAT_OR_IMAG) { mp_obj_t o = mp_parse_num_decimal(lex->vstr.buf, lex->vstr.len, true, false, lex); pn = make_node_const_object(parser, lex->tok_line, o); - } else if (lex->tok_kind == MP_TOKEN_STRING || lex->tok_kind == MP_TOKEN_BYTES) { - // Don't automatically intern all strings/bytes. doc strings (which are usually large) + } else if (lex->tok_kind == MP_TOKEN_STRING) { + // Don't automatically intern all strings. Doc strings (which are usually large) // will be discarded by the compiler, and so we shouldn't intern them. qstr qst = MP_QSTRnull; if (lex->vstr.len <= MICROPY_ALLOC_PARSE_INTERN_STRING_LEN) { @@ -517,14 +514,16 @@ STATIC void push_result_token(parser_t *parser, uint8_t rule_id) { } if (qst != MP_QSTRnull) { // qstr exists, make a leaf node - pn = mp_parse_node_new_leaf(lex->tok_kind == MP_TOKEN_STRING ? MP_PARSE_NODE_STRING : MP_PARSE_NODE_BYTES, qst); + pn = mp_parse_node_new_leaf(MP_PARSE_NODE_STRING, qst); } else { - // not interned, make a node holding a pointer to the string/bytes object - mp_obj_t o = mp_obj_new_str_copy( - lex->tok_kind == MP_TOKEN_STRING ? &mp_type_str : &mp_type_bytes, - (const byte *)lex->vstr.buf, lex->vstr.len); + // not interned, make a node holding a pointer to the string object + mp_obj_t o = mp_obj_new_str_copy(&mp_type_str, (const byte *)lex->vstr.buf, lex->vstr.len); pn = make_node_const_object(parser, lex->tok_line, o); } + } else if (lex->tok_kind == MP_TOKEN_BYTES) { + // make a node holding a pointer to the bytes object + mp_obj_t o = mp_obj_new_bytes((const byte *)lex->vstr.buf, lex->vstr.len); + pn = make_node_const_object(parser, lex->tok_line, o); } else { pn = mp_parse_node_new_leaf(MP_PARSE_NODE_TOKEN, lex->tok_kind); } diff --git a/py/parse.h b/py/parse.h index a6eb380047..9c327c28b6 100644 --- a/py/parse.h +++ b/py/parse.h @@ -39,15 +39,13 @@ struct _mp_lexer_t; // - xxxx...xx00: pointer to mp_parse_node_struct_t // - xx...xx0010: an identifier; bits 4 and above are the qstr // - xx...xx0110: a string; bits 4 and above are the qstr holding the value -// - xx...xx1010: a string of bytes; bits 4 and above are the qstr holding the value -// - xx...xx1110: a token; bits 4 and above are mp_token_kind_t +// - xx...xx1010: a token; bits 4 and above are mp_token_kind_t #define MP_PARSE_NODE_NULL (0) #define MP_PARSE_NODE_SMALL_INT (0x1) #define MP_PARSE_NODE_ID (0x02) #define MP_PARSE_NODE_STRING (0x06) -#define MP_PARSE_NODE_BYTES (0x0a) -#define MP_PARSE_NODE_TOKEN (0x0e) +#define MP_PARSE_NODE_TOKEN (0x0a) typedef uintptr_t mp_parse_node_t; // must be pointer size diff --git a/tests/cmdline/cmd_parsetree.py.exp b/tests/cmdline/cmd_parsetree.py.exp index ac16f6d88d..53d62a5dbd 100644 --- a/tests/cmdline/cmd_parsetree.py.exp +++ b/tests/cmdline/cmd_parsetree.py.exp @@ -18,7 +18,7 @@ [ 8] literal \.\+ [ 9] \(rule\|expr_stmt\)(5) (n=2) id(d) - bytes(bytes) +[ 9] literal \.\+ [ 10] \(rule\|expr_stmt\)(5) (n=2) id(e) [ 10] literal \.\+