Commit 2faa18b
authored
Preserve child wrapper identity in the native AST facade (#391)
Stacked on #381.
#381's review surfaced a real semantics regression in the lazy native
AST facade: every accessor on `WP_MySQL_Native_Parser_Node` calls into
Rust and returns a freshly constructed PHP wrapper, so
`get_first_child_node()` returns a different object every time.
`WP_Parser_Node` has always given callers stable child identity — attach
state to a child once, walk past it, walk back, the state is still there
— and the lazy native facade is meant to keep that contract intact. This
restores it.
A per-AST `WP_MySQL_Native_AST_Cache` is created lazily on the root and
shared by reference with every wrapper that gets interned through it.
Each accessor looks the returned wrapper's `native_node_index` up in the
cache and either returns the canonical instance or registers the new
one. `materialize_native_children()` pulls children through the same
cache so any mutation a caller made through `get_first_child_node()`
before the parent went through `append_child()` survives into
`$this->children` — same instance, same mutations.
Tokens are unchanged. The public token API has no mutators and no
callers in this repo rely on `WP_MySQL_Token` identity; if that becomes
a need we can extend the cache.
## What's in here
- `class-wp-mysql-native-ast-cache.php` — small holder, one per AST.
- Native node accessors run results through `intern()` / `intern_all()`
/ `intern_nodes()`.
- `materialize_native_children()` reuses the interned wrappers so prior
mutations don't get lost.
- Regression test covering same-instance reads, descendant/child
identity sharing, and the mutate-then-materialize edge case. Skips when
the native extension isn't loaded.
- `run-native-ast-walk-benchmark.php` parses the MySQL server suite,
walks each AST, and reports `parsed`, `walked_nodes`, `duration`,
`peak_mem`, and an identity-stability flag.
- `Native AST Walk Perf` workflow runs the benchmark on this PR and on
the PR base (`codex/native-lazy-ast-facade`) on the same runner, so the
identity-cache cost is measured apples-to-apples on every push.
## Performance
Benchmarked on CI against the no-cache baseline, same runner, same
corpus (69,567 queries, 4.8M walked nodes). Full numbers in PR comments
— links below.
| scenario | baseline (no cache) | this PR | delta |
|---|---|---|---|
| native parse only | 1.28s | 1.28s | 0% |
| native walk duration | 3.52s | 3.33s | **+5% (+0.19s)** |
| native walk qps | 19,766 | 20,884 | **−5%** |
| native walk peak memory | 50.0MB | 60.5MB | **+10.5MB (+21%)** |
| native walk identity stable | **FALSE** | true | regression fixed |
Hot-path optimization in `intern_all()` (cache reference hoisted out of
the loop, per-item logic inlined) plus a typed `intern_nodes()` fast
path for accessors that return only nodes brought the time penalty down
from an initial +17% to +5%. Memory delta is structural — one retained
PHP wrapper per visited node — and is the price of stable identity.
Native walk is still ~5× faster than the pure-PHP path while now
preserving `WP_Parser_Node` semantics.
- First measurement:
#391 (comment)
- After hot-path optimization:
#391 (comment)
## Test plan
- [x] PHPUnit suite passes on the pure-PHP path.
- [x] PHPUnit suite passes with the native extension loaded; the new
identity tests run instead of skipping.
- [x] `Native AST Walk Perf` workflow reports the cache cost vs. the PR
base on every push.1 parent 5275791 commit 2faa18b
11 files changed
Lines changed: 862 additions & 37 deletions
File tree
- .github/workflows
- packages
- mysql-on-sqlite
- src
- mysql
- native
- sqlite
- tests
- mysql/native
- tools
- php-ext-wp-mysql-parser/src
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
81 | 81 | | |
82 | 82 | | |
83 | 83 | | |
84 | | - | |
| 84 | + | |
| 85 | + | |
| 86 | + | |
| 87 | + | |
85 | 88 | | |
86 | 89 | | |
87 | 90 | | |
| |||
149 | 152 | | |
150 | 153 | | |
151 | 154 | | |
| 155 | + | |
| 156 | + | |
| 157 | + | |
| 158 | + | |
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
| 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 | + | |
| 62 | + | |
| 63 | + | |
| 64 | + | |
| 65 | + | |
| 66 | + | |
| 67 | + | |
| 68 | + | |
| 69 | + | |
| 70 | + | |
| 71 | + | |
| 72 | + | |
| 73 | + | |
| 74 | + | |
| 75 | + | |
| 76 | + | |
| 77 | + | |
| 78 | + | |
| 79 | + | |
| 80 | + | |
| 81 | + | |
| 82 | + | |
| 83 | + | |
| 84 | + | |
| 85 | + | |
| 86 | + | |
| 87 | + | |
| 88 | + | |
| 89 | + | |
| 90 | + | |
| 91 | + | |
| 92 | + | |
| 93 | + | |
| 94 | + | |
| 95 | + | |
| 96 | + | |
| 97 | + | |
| 98 | + | |
| 99 | + | |
| 100 | + | |
| 101 | + | |
| 102 | + | |
| 103 | + | |
| 104 | + | |
| 105 | + | |
| 106 | + | |
| 107 | + | |
| 108 | + | |
| 109 | + | |
| 110 | + | |
| 111 | + | |
| 112 | + | |
| 113 | + | |
| 114 | + | |
| 115 | + | |
| 116 | + | |
| 117 | + | |
| 118 | + | |
| 119 | + | |
| 120 | + | |
| 121 | + | |
| 122 | + | |
| 123 | + | |
| 124 | + | |
| 125 | + | |
| 126 | + | |
| 127 | + | |
| 128 | + | |
| 129 | + | |
| 130 | + | |
| 131 | + | |
| 132 | + | |
| 133 | + | |
| 134 | + | |
| 135 | + | |
| 136 | + | |
| 137 | + | |
| 138 | + | |
| 139 | + | |
| 140 | + | |
| 141 | + | |
| 142 | + | |
| 143 | + | |
| 144 | + | |
| 145 | + | |
| 146 | + | |
| 147 | + | |
| 148 | + | |
| 149 | + | |
| 150 | + | |
| 151 | + | |
| 152 | + | |
| 153 | + | |
| 154 | + | |
| 155 | + | |
| 156 | + | |
| 157 | + | |
| 158 | + | |
| 159 | + | |
| 160 | + | |
| 161 | + | |
| 162 | + | |
| 163 | + | |
| 164 | + | |
| 165 | + | |
| 166 | + | |
| 167 | + | |
| 168 | + | |
| 169 | + | |
| 170 | + | |
| 171 | + | |
| 172 | + | |
| 173 | + | |
| 174 | + | |
| 175 | + | |
| 176 | + | |
| 177 | + | |
| 178 | + | |
| 179 | + | |
| 180 | + | |
| 181 | + | |
| 182 | + | |
| 183 | + | |
| 184 | + | |
| 185 | + | |
| 186 | + | |
| 187 | + | |
| 188 | + | |
| 189 | + | |
| 190 | + | |
| 191 | + | |
| 192 | + | |
| 193 | + | |
| 194 | + | |
| 195 | + | |
| 196 | + | |
| 197 | + | |
| 198 | + | |
| 199 | + | |
| 200 | + | |
| 201 | + | |
| 202 | + | |
| 203 | + | |
| 204 | + | |
| 205 | + | |
| 206 | + | |
| 207 | + | |
| 208 | + | |
| 209 | + | |
| 210 | + | |
| 211 | + | |
| 212 | + | |
| 213 | + | |
| 214 | + | |
| 215 | + | |
| 216 | + | |
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
54 | 54 | | |
55 | 55 | | |
56 | 56 | | |
57 | | - | |
58 | | - | |
| 57 | + | |
| 58 | + | |
59 | 59 | | |
60 | 60 | | |
61 | 61 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
27 | 27 | | |
28 | 28 | | |
29 | 29 | | |
| 30 | + | |
30 | 31 | | |
31 | 32 | | |
32 | 33 | | |
| |||
Lines changed: 11 additions & 0 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
8 | 8 | | |
9 | 9 | | |
10 | 10 | | |
| 11 | + | |
| 12 | + | |
| 13 | + | |
| 14 | + | |
| 15 | + | |
| 16 | + | |
| 17 | + | |
| 18 | + | |
| 19 | + | |
| 20 | + | |
| 21 | + | |
11 | 22 | | |
12 | 23 | | |
13 | 24 | | |
| |||
Lines changed: 23 additions & 0 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
| 1 | + | |
| 2 | + | |
| 3 | + | |
| 4 | + | |
| 5 | + | |
| 6 | + | |
| 7 | + | |
| 8 | + | |
| 9 | + | |
| 10 | + | |
| 11 | + | |
| 12 | + | |
| 13 | + | |
| 14 | + | |
| 15 | + | |
| 16 | + | |
| 17 | + | |
| 18 | + | |
| 19 | + | |
| 20 | + | |
| 21 | + | |
| 22 | + | |
| 23 | + | |
0 commit comments