1
0
mirror of https://github.com/stedolan/jq.git synced 2024-05-11 05:55:39 +00:00

improve tonumber/0 performance by parsing input as number literal

Previously, the tonumber/0 filter parses the input as JSON values, but
this is less-performant on large non-number strings. Parsing the input
string as number literal fixes the performance issue. Also, this fix
changes the filter to reject numbers with white spaces.
This commit is contained in:
itchyny
2024-03-02 00:33:17 +09:00
committed by GitHub
parent 913b26469f
commit ce0e788ce2
2 changed files with 19 additions and 6 deletions

View File

@@ -43,6 +43,8 @@ void *alloca (size_t);
#include "locfile.h"
#include "jv_unicode.h"
#include "jv_alloc.h"
#include "jv_dtoa.h"
#include "jv_dtoa_tsd.h"
#include "jv_private.h"
#include "util.h"
@@ -464,11 +466,22 @@ static jv f_tonumber(jq_state *jq, jv input) {
return input;
}
if (jv_get_kind(input) == JV_KIND_STRING) {
jv parsed = jv_parse(jv_string_value(input));
if (!jv_is_valid(parsed) || jv_get_kind(parsed) == JV_KIND_NUMBER) {
jv_free(input);
return parsed;
const char* s = jv_string_value(input);
#ifdef USE_DECNUM
jv number = jv_number_with_literal(s);
if (jv_get_kind(number) == JV_KIND_INVALID) {
return type_error(input, "cannot be parsed as a number");
}
#else
char *end = 0;
double d = jvp_strtod(tsd_dtoa_context_get(), s, &end);
if (end == 0 || *end != 0) {
return type_error(input, "cannot be parsed as a number");
}
jv number = jv_number(d);
#endif
jv_free(input);
return number;
}
return type_error(input, "cannot be parsed as a number");
}

View File

@@ -2000,8 +2000,8 @@ null
2
.[] |= try tonumber
["1", "2a", "3", 4]
[1, 3, 4]
["1", "2a", "3", " 4 ", "5.67", ".89", "-876", "+5.43", 21]
[1, 3, 5.67, 0.89, -876, 5.43, 21]
# Also 1859, but from 2073
any(keys[]|tostring?;true)