Skip to content

Commit ebc765e

Browse files
tannewtdhalbert
authored andcommitted
Speed up JSON parsing with readinto
Get a chunk of data from readinto instead of a single byte. This speeds up the parsing by reducing the number of function calls. Fixes #3703
1 parent 42ca57f commit ebc765e

File tree

1 file changed

+26
-9
lines changed

1 file changed

+26
-9
lines changed

extmod/modujson.c

Lines changed: 26 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -57,6 +57,8 @@ typedef struct _ujson_stream_t {
5757
int errcode;
5858
mp_obj_t python_readinto[2 + 1];
5959
mp_obj_array_t bytearray_obj;
60+
size_t start;
61+
size_t end;
6062
byte cur;
6163
} ujson_stream_t;
6264

@@ -77,28 +79,43 @@ STATIC byte ujson_stream_next(ujson_stream_t *s) {
7779
return s->cur;
7880
}
7981

82+
// We read from an object's `readinto` method in chunks larger than the json
83+
// parser needs to reduce the number of function calls done.
84+
85+
#define CIRCUITPY_JSON_READ_CHUNK_SIZE 64
86+
8087
STATIC mp_uint_t ujson_python_readinto(mp_obj_t obj, void *buf, mp_uint_t size, int *errcode) {
8188
ujson_stream_t* s = obj;
82-
s->bytearray_obj.items = buf;
83-
s->bytearray_obj.len = size;
84-
*errcode = 0;
85-
mp_obj_t ret = mp_call_method_n_kw(1, 0, s->python_readinto);
86-
if (ret == mp_const_none) {
87-
*errcode = MP_EAGAIN;
88-
return MP_STREAM_ERROR;
89+
90+
if (s->start == s->end) {
91+
*errcode = 0;
92+
mp_obj_t ret = mp_call_method_n_kw(1, 0, s->python_readinto);
93+
if (ret == mp_const_none) {
94+
*errcode = MP_EAGAIN;
95+
return MP_STREAM_ERROR;
96+
}
97+
s->start = 0;
98+
s->end = mp_obj_get_int(ret);
8999
}
90-
return mp_obj_get_int(ret);
100+
101+
*((uint8_t *)buf) = ((uint8_t*) s->bytearray_obj.items)[s->start];
102+
s->start++;
103+
return 1;
91104
}
92105

93106
STATIC mp_obj_t _mod_ujson_load(mp_obj_t stream_obj, bool return_first_json) {
94107
const mp_stream_p_t *stream_p = mp_proto_get(MP_QSTR_protocol_stream, stream_obj);
95108
ujson_stream_t s;
109+
uint8_t character_buffer[CIRCUITPY_JSON_READ_CHUNK_SIZE];
96110
if (stream_p == NULL) {
111+
s.start = 0;
112+
s.end = 0;
97113
mp_load_method(stream_obj, MP_QSTR_readinto, s.python_readinto);
98114
s.bytearray_obj.base.type = &mp_type_bytearray;
99115
s.bytearray_obj.typecode = BYTEARRAY_TYPECODE;
116+
s.bytearray_obj.len = CIRCUITPY_JSON_READ_CHUNK_SIZE;
100117
s.bytearray_obj.free = 0;
101-
// len and items are set at read time
118+
s.bytearray_obj.items = character_buffer;
102119
s.python_readinto[2] = MP_OBJ_FROM_PTR(&s.bytearray_obj);
103120
s.stream_obj = &s;
104121
s.read = ujson_python_readinto;

0 commit comments

Comments
 (0)