1berry/ruby
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

string.c: Improve fstring_hash to reduce collisions

Browse Source 
`rb_str_hash` doesn't include the encoding for ASCII only strings
because ASCII only strings are equal regardless of their encoding.

But in the case if the `fstring_table`, two identical ASCII strings
with different encodings aren't equal.

Given it's common to have both `:foo` (or `def foo`) and `"foo"`
in the same source code, this causes a lot of collisions in the
`fstring_table`.
This commit is contained in:
Jean Boussier 2025-03-07 21:38:30 +01:00
parent eae0fe37c0
commit a14d9b8d57
Notes: git 2025-03-08 09:56:21 +00:00 
Merged: https://github.com/ruby/ruby/pull/12881
1 changed files with 7 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
string.c
View File

@ -366,13 +366,17 @@ static VALUE register_fstring(VALUE str, bool copy, bool force_precompute_hash);
static st_index_t static st_index_t
fstring_hash(VALUE str) fstring_hash(VALUE str)
{ {
st_index_t h;
if (FL_TEST_RAW(str, STR_FAKESTR)) { if (FL_TEST_RAW(str, STR_FAKESTR)) {
// register_fstring precomputes the hash and stores it in capa for fake strings // register_fstring precomputes the hash and stores it in capa for fake strings
return (st_index_t)RSTRING(str)->as.heap.aux.capa; h = (st_index_t)RSTRING(str)->as.heap.aux.capa;
} }
else { else {
return rb_str_hash(str); h = rb_str_hash(str);
} }
// rb_str_hash doesn't include the encoding for ascii only strings, so
// we add it to avoid common collisions between `:sym.name` (ASCII) and `"sym"` (UTF-8)
return rb_hash_end(rb_hash_uint32(h, (uint32_t)ENCODING_GET_INLINED(str)));
} }
#else #else
#define fstring_hash rb_str_hash #define fstring_hash rb_str_hash
@ -460,7 +464,7 @@ fstr_update_callback(st_data_t *key, st_data_t *value, st_data_t data, int exist
STR_SET_LEN(new_str, RSTRING_LEN(str)); STR_SET_LEN(new_str, RSTRING_LEN(str));
TERM_FILL(RSTRING_END(new_str), TERM_LEN(str)); TERM_FILL(RSTRING_END(new_str), TERM_LEN(str));
rb_enc_copy(new_str, str); rb_enc_copy(new_str, str);
str_store_precomputed_hash(new_str, fstring_hash(str)); str_store_precomputed_hash(new_str, str_do_hash(str));
} }
else { else {
new_str = str_new(rb_cString, RSTRING(str)->as.heap.ptr, RSTRING(str)->len); new_str = str_new(rb_cString, RSTRING(str)->as.heap.ptr, RSTRING(str)->len);