1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
|
Performance Notes
=================
Pending
-------
- codepoints and getbyte, setbyte, << for BSON_TYPE
- string.force_encoding("UTF-8").valid_encoding?
Top concerns
------------
- String - pure - cext@ - Tyler has experience, says UTF8 (with LATIN1 subset) is sufficient, note keys need special char check
- to_utf8_binary@
- to_bson_string@
- to_bson_cstring@
- append_bson_int32
- to_bson_key - rb_string_to_bson_key
- check_for_illegal_characters
- encode
- set_int32
- force_encoding
- to_bson
- Symbol
- to_bson
- to_bson_key
- rb_symbol_to_bson_key
- Binary
- rb_binary_to_bson
- Integer - sizing done twice for serialization - bson_type and to_bson
- discarded as not worthy
- new_hash_to_bson_hint
- new_hash_to_bson_integer
- Array to_bson - repeat above
TODO: Review
------------
- key optimization
- note threading concerns
- no safety limit needed for non-pathological use (review this)
- symbol ~ gain: 0.25 (36 --> 27) Xeon, gain: 0.34 (41 --> 27) Core 2
- string ~ gain: 0.15 (33 --> 28) Xeon, gain: 0.24 (39 --> 29) Core 2
- with safety limit, mutex overhead eats up the benefit on Xeon
- symbol ~ gain: 0.15 (41 --> 35) Core 2
- string ~ gain: 0.05 (39 --> 37) Core 2
- rb_float_to_bson ~ gain: 0.61 (15 --> 6, allocated: 3 --> 1) Core 2
Performance gains
-----------------
- catalog, extract techniques, tech talk
Driver notes
------------
Check for initial '$' or inclusion of '.' is purposely left to the driver.
See bench.rb: test_string_to_bson_key_mongodb
|
