rusteerio 0.0.0-stage → 0.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (206) hide show
  1. package/CONTRIBUTING.md +73 -0
  2. package/Cargo.lock +1051 -0
  3. package/Cargo.toml +36 -0
  4. package/LICENSE +21 -0
  5. package/README.md +252 -2
  6. package/RELEASING.md +138 -0
  7. package/THIRD_PARTY_NOTICES.md +7443 -0
  8. package/build.rs +3 -0
  9. package/dist/chunk-DKULX6VO.mjs +3185 -0
  10. package/dist/chunk-DKULX6VO.mjs.map +1 -0
  11. package/dist/chunk-OHPGEY3E.mjs +32 -0
  12. package/dist/chunk-OHPGEY3E.mjs.map +1 -0
  13. package/dist/index.cjs +3230 -0
  14. package/dist/index.cjs.map +1 -0
  15. package/dist/index.d.mts +113 -0
  16. package/dist/index.d.ts +113 -0
  17. package/dist/index.mjs +38 -0
  18. package/dist/index.mjs.map +1 -0
  19. package/dist/slim.cjs +2441 -0
  20. package/dist/slim.cjs.map +1 -0
  21. package/dist/slim.d.mts +11 -0
  22. package/dist/slim.d.ts +11 -0
  23. package/dist/slim.mjs +19 -0
  24. package/dist/slim.mjs.map +1 -0
  25. package/dist/types-BP6TpxE5.d.mts +532 -0
  26. package/dist/types-BP6TpxE5.d.ts +532 -0
  27. package/dist/utils.cjs +61 -0
  28. package/dist/utils.cjs.map +1 -0
  29. package/dist/utils.d.mts +13 -0
  30. package/dist/utils.d.ts +13 -0
  31. package/dist/utils.mjs +17 -0
  32. package/dist/utils.mjs.map +1 -0
  33. package/native/src/arena.rs +1781 -0
  34. package/native/src/attributes.rs +923 -0
  35. package/native/src/codec.rs +172 -0
  36. package/native/src/diagnostics.rs +1249 -0
  37. package/native/src/extraction.rs +226 -0
  38. package/native/src/gc.rs +166 -0
  39. package/native/src/hook_sink.rs +727 -0
  40. package/native/src/hook_stream.rs +513 -0
  41. package/native/src/lib.rs +2095 -0
  42. package/native/src/location.rs +392 -0
  43. package/native/src/runtime.rs +189 -0
  44. package/native/src/selector.rs +1287 -0
  45. package/native/src/sink.rs +261 -0
  46. package/native/src/stream.rs +334 -0
  47. package/native/src/tokenizer.rs +355 -0
  48. package/native/src/tracker.rs +232 -0
  49. package/native/vendor/README.md +25 -0
  50. package/native/vendor/html5ever-0.35.0/.cargo_vcs_info.json +6 -0
  51. package/native/vendor/html5ever-0.35.0/Cargo.lock +828 -0
  52. package/native/vendor/html5ever-0.35.0/Cargo.toml +85 -0
  53. package/native/vendor/html5ever-0.35.0/LICENSE-APACHE +201 -0
  54. package/native/vendor/html5ever-0.35.0/LICENSE-MIT +25 -0
  55. package/native/vendor/html5ever-0.35.0/README.md +60 -0
  56. package/native/vendor/html5ever-0.35.0/benches/html5ever.rs +86 -0
  57. package/native/vendor/html5ever-0.35.0/data/bench/lipsum-zh.html +19 -0
  58. package/native/vendor/html5ever-0.35.0/data/bench/lipsum.html +40 -0
  59. package/native/vendor/html5ever-0.35.0/data/bench/medium-fragment.html +24 -0
  60. package/native/vendor/html5ever-0.35.0/data/bench/small-fragment.html +7 -0
  61. package/native/vendor/html5ever-0.35.0/data/bench/strong.html +1 -0
  62. package/native/vendor/html5ever-0.35.0/data/bench/tiny-fragment.html +1 -0
  63. package/native/vendor/html5ever-0.35.0/examples/arena.rs +356 -0
  64. package/native/vendor/html5ever-0.35.0/examples/capi/tokenize.c +74 -0
  65. package/native/vendor/html5ever-0.35.0/examples/noop-tokenize.rs +47 -0
  66. package/native/vendor/html5ever-0.35.0/examples/noop-tree-builder.rs +132 -0
  67. package/native/vendor/html5ever-0.35.0/examples/print-tree-actions.rs +185 -0
  68. package/native/vendor/html5ever-0.35.0/examples/tokenize.rs +111 -0
  69. package/native/vendor/html5ever-0.35.0/src/driver.rs +148 -0
  70. package/native/vendor/html5ever-0.35.0/src/lib.rs +30 -0
  71. package/native/vendor/html5ever-0.35.0/src/macros.rs +34 -0
  72. package/native/vendor/html5ever-0.35.0/src/serialize/mod.rs +255 -0
  73. package/native/vendor/html5ever-0.35.0/src/tokenizer/char_ref/mod.rs +451 -0
  74. package/native/vendor/html5ever-0.35.0/src/tokenizer/interface.rs +99 -0
  75. package/native/vendor/html5ever-0.35.0/src/tokenizer/mod.rs +2305 -0
  76. package/native/vendor/html5ever-0.35.0/src/tokenizer/states.rs +97 -0
  77. package/native/vendor/html5ever-0.35.0/src/tree_builder/data.rs +170 -0
  78. package/native/vendor/html5ever-0.35.0/src/tree_builder/mod.rs +1848 -0
  79. package/native/vendor/html5ever-0.35.0/src/tree_builder/rules.rs +1494 -0
  80. package/native/vendor/html5ever-0.35.0/src/tree_builder/tag_sets.rs +114 -0
  81. package/native/vendor/html5ever-0.35.0/src/tree_builder/types.rs +91 -0
  82. package/native/vendor/html5ever-0.35.0/src/util/str.rs +52 -0
  83. package/native/vendor/html5gum-0.8.4/Cargo.lock +1006 -0
  84. package/native/vendor/html5gum-0.8.4/Cargo.toml +109 -0
  85. package/native/vendor/html5gum-0.8.4/LICENSE +19 -0
  86. package/native/vendor/html5gum-0.8.4/README.md +110 -0
  87. package/native/vendor/html5gum-0.8.4/benches/files/0050.html +2131 -0
  88. package/native/vendor/html5gum-0.8.4/benches/files/6D30.html +1004 -0
  89. package/native/vendor/html5gum-0.8.4/benches/files.rs +32 -0
  90. package/native/vendor/html5gum-0.8.4/benches/patterns.rs +31 -0
  91. package/native/vendor/html5gum-0.8.4/src/arrayvec.rs +29 -0
  92. package/native/vendor/html5gum-0.8.4/src/char_validator.rs +134 -0
  93. package/native/vendor/html5gum-0.8.4/src/emitters/callback.rs +676 -0
  94. package/native/vendor/html5gum-0.8.4/src/emitters/default.rs +207 -0
  95. package/native/vendor/html5gum-0.8.4/src/emitters/emitter.rs +344 -0
  96. package/native/vendor/html5gum-0.8.4/src/emitters/html5ever.rs +306 -0
  97. package/native/vendor/html5gum-0.8.4/src/emitters/mod.rs +25 -0
  98. package/native/vendor/html5gum-0.8.4/src/entities.rs +3342 -0
  99. package/native/vendor/html5gum-0.8.4/src/error.rs +96 -0
  100. package/native/vendor/html5gum-0.8.4/src/htmlstring.rs +130 -0
  101. package/native/vendor/html5gum-0.8.4/src/lib.rs +66 -0
  102. package/native/vendor/html5gum-0.8.4/src/machine.rs +2136 -0
  103. package/native/vendor/html5gum-0.8.4/src/machine_helper.rs +277 -0
  104. package/native/vendor/html5gum-0.8.4/src/read_helper.rs +244 -0
  105. package/native/vendor/html5gum-0.8.4/src/reader.rs +440 -0
  106. package/native/vendor/html5gum-0.8.4/src/span.rs +98 -0
  107. package/native/vendor/html5gum-0.8.4/src/state.rs +16 -0
  108. package/native/vendor/html5gum-0.8.4/src/testutils.rs +71 -0
  109. package/native/vendor/html5gum-0.8.4/src/tokenizer.rs +189 -0
  110. package/native/vendor/html5gum-0.8.4/src/utils.rs +88 -0
  111. package/native/vendor/markup5ever-0.35.0/.cargo_vcs_info.json +6 -0
  112. package/native/vendor/markup5ever-0.35.0/Cargo.toml +45 -0
  113. package/native/vendor/markup5ever-0.35.0/LICENSE-APACHE +201 -0
  114. package/native/vendor/markup5ever-0.35.0/LICENSE-MIT +25 -0
  115. package/native/vendor/markup5ever-0.35.0/interface/mod.rs +422 -0
  116. package/native/vendor/markup5ever-0.35.0/interface/tree_builder.rs +280 -0
  117. package/native/vendor/markup5ever-0.35.0/lib.rs +59 -0
  118. package/native/vendor/markup5ever-0.35.0/serialize.rs +77 -0
  119. package/native/vendor/markup5ever-0.35.0/util/buffer_queue.rs +333 -0
  120. package/native/vendor/markup5ever-0.35.0/util/smallcharset.rs +88 -0
  121. package/package.json +137 -4
  122. package/rusteerio.darwin-arm64.node +0 -0
  123. package/rusteerio.darwin-x64.node +0 -0
  124. package/rusteerio.linux-arm64-gnu.node +0 -0
  125. package/rusteerio.linux-arm64-musl.node +0 -0
  126. package/rusteerio.linux-x64-gnu.node +0 -0
  127. package/rusteerio.linux-x64-musl.node +0 -0
  128. package/rusteerio.win32-arm64-msvc.node +0 -0
  129. package/rusteerio.win32-x64-msvc.node +0 -0
  130. package/scripts/check-native.mjs +15 -0
  131. package/scripts/check-release.mjs +235 -0
  132. package/scripts/fix-declarations.mjs +27 -0
  133. package/scripts/generate-notices.mjs +176 -0
  134. package/scripts/notices/README.md +13 -0
  135. package/scripts/notices/cheerio-MIT.txt +21 -0
  136. package/scripts/notices/html5ever-AUTHORS.txt +20 -0
  137. package/scripts/notices/html5ever-COPYRIGHT.txt +8 -0
  138. package/scripts/notices/html5ever-LICENSE-APACHE.txt +201 -0
  139. package/scripts/notices/html5ever-LICENSE-MIT.txt +25 -0
  140. package/scripts/notices/napi-rs-MIT.txt +43 -0
  141. package/scripts/notices/neon-MIT.txt +19 -0
  142. package/scripts/notices/runtime-01c266bced4a434da005.txt +3 -0
  143. package/scripts/notices/runtime-0a54e647fe54104658b5.txt +14 -0
  144. package/scripts/notices/runtime-0b74dfa0bcee5c420c6b.txt +25 -0
  145. package/scripts/notices/runtime-0d542e0c8804e39aa7f3.txt +177 -0
  146. package/scripts/notices/runtime-0f96a83840e146e43c0e.txt +21 -0
  147. package/scripts/notices/runtime-123a331b5dbf04c30097.txt +25 -0
  148. package/scripts/notices/runtime-172020dbfd5b53a226df.txt +34 -0
  149. package/scripts/notices/runtime-209fbbe0ad52d9235e37.txt +26 -0
  150. package/scripts/notices/runtime-23f18e03dc49df91622f.txt +23 -0
  151. package/scripts/notices/runtime-268872b9816f90fd8e85.txt +220 -0
  152. package/scripts/notices/runtime-35242e7a83f69875e6ed.txt +176 -0
  153. package/scripts/notices/runtime-378f5840b258e2779c39.txt +25 -0
  154. package/scripts/notices/runtime-4108245a1f2df9d4e94d.txt +25 -0
  155. package/scripts/notices/runtime-4455bf75a91154108304.txt +21 -0
  156. package/scripts/notices/runtime-553fffcd9b1cb158bc3e.txt +13 -0
  157. package/scripts/notices/runtime-62c7a1e35f56406896d7.txt +176 -0
  158. package/scripts/notices/runtime-6485b8ed310d3f0340bf.txt +25 -0
  159. package/scripts/notices/runtime-69fef7b0f322a6555415.txt +201 -0
  160. package/scripts/notices/runtime-6df43f6f4b5d4587f3d8.txt +187 -0
  161. package/scripts/notices/runtime-799e9ca9d179295ef372.txt +24 -0
  162. package/scripts/notices/runtime-7b63ecd5f1902af1b637.txt +25 -0
  163. package/scripts/notices/runtime-7bc7d5de2a0b793d4e1a.txt +21 -0
  164. package/scripts/notices/runtime-7e12e5df4bae12cb2158.txt +24 -0
  165. package/scripts/notices/runtime-861399f8c21c042b1105.txt +12 -0
  166. package/scripts/notices/runtime-8ada45cd9f843acf64e4.txt +201 -0
  167. package/scripts/notices/runtime-90eb64f0279b0d9432ac.txt +12 -0
  168. package/scripts/notices/runtime-a60eea817514531668d7.txt +201 -0
  169. package/scripts/notices/runtime-a7c936ff1ed8fa340172.txt +25 -0
  170. package/scripts/notices/runtime-ab6eec6caf0fa5775e41.txt +275 -0
  171. package/scripts/notices/runtime-b1181a40b2a7b25cf66f.txt +19 -0
  172. package/scripts/notices/runtime-b71bd43a069ca0641a9e.txt +25 -0
  173. package/scripts/notices/runtime-c16f8dcf1a368b83be78.txt +201 -0
  174. package/scripts/notices/runtime-c2cfccb812fe482101a8.txt +21 -0
  175. package/scripts/notices/runtime-e99d88d232bf57d70f0f.txt +25 -0
  176. package/scripts/notices/runtime-f32fb3b417a194167cfa.txt +9 -0
  177. package/scripts/notices/runtime-f5062c9a188d81dfe66b.txt +39 -0
  178. package/scripts/notices/runtime-f9070a832833fa7488fd.txt +4 -0
  179. package/scripts/notices/runtime-fab3dd6bdab226f1c086.txt +373 -0
  180. package/scripts/notices/runtime-fba3318d0c600380177e.txt +62 -0
  181. package/scripts/notices/runtime-ff8f68cb076caf8cefe7.txt +25 -0
  182. package/scripts/notices/runtime-ff92bed461f50338dd70.txt +74 -0
  183. package/scripts/notices/runtime.json +1632 -0
  184. package/scripts/notices/sources.json +555 -0
  185. package/scripts/notices/xray-MIT.txt +21 -0
  186. package/src/attributes.ts +271 -0
  187. package/src/codec.ts +74 -0
  188. package/src/dom.ts +87 -0
  189. package/src/extract-plan.ts +97 -0
  190. package/src/extract.ts +130 -0
  191. package/src/hooks.ts +220 -0
  192. package/src/index.ts +223 -0
  193. package/src/io.ts +491 -0
  194. package/src/native.ts +135 -0
  195. package/src/nodes.ts +290 -0
  196. package/src/parse.ts +53 -0
  197. package/src/parser-types.ts +157 -0
  198. package/src/router.ts +327 -0
  199. package/src/selection.ts +566 -0
  200. package/src/slim.ts +12 -0
  201. package/src/transport.ts +104 -0
  202. package/src/types.ts +186 -0
  203. package/src/utils.ts +29 -0
  204. package/src/workers.ts +21 -0
  205. package/tsconfig.json +15 -0
  206. package/tsup.config.ts +16 -0
@@ -0,0 +1,226 @@
1
+ //! Evaluate callback-free extraction maps on native worker threads.
2
+
3
+ use napi::bindgen_prelude::*;
4
+ use serde_json::{json, Map, Value};
5
+
6
+ use crate::NativeDocument;
7
+
8
+ fn invalid(message: &str) -> Error {
9
+ Error::new(Status::InvalidArg, message.to_owned())
10
+ }
11
+
12
+ fn entries(value: &Value) -> Result<&[Value]> {
13
+ value
14
+ .get("entries")
15
+ .and_then(Value::as_array)
16
+ .map(Vec::as_slice)
17
+ .ok_or_else(|| invalid("Extraction plan requires entries"))
18
+ }
19
+
20
+ fn extract(
21
+ document: &NativeDocument,
22
+ plan: &[Value],
23
+ contexts: Option<&[u32]>,
24
+ depth: usize,
25
+ ) -> Result<Value> {
26
+ if depth > 64 {
27
+ return Err(invalid("Extraction plan nesting exceeds 64 levels"));
28
+ }
29
+ let mut result = Map::new();
30
+ for entry in plan {
31
+ let key = entry
32
+ .get("key")
33
+ .and_then(Value::as_str)
34
+ .ok_or_else(|| invalid("Extraction entry requires a key"))?;
35
+ let selector = entry
36
+ .get("selector")
37
+ .and_then(Value::as_str)
38
+ .ok_or_else(|| invalid("Extraction entry requires a selector"))?;
39
+ let all = entry.get("all").and_then(Value::as_bool).unwrap_or(false);
40
+ let value = entry
41
+ .get("value")
42
+ .ok_or_else(|| invalid("Extraction entry requires a value"))?;
43
+ let ids = document.select_ids(selector, contexts)?;
44
+ let read = |id: u32| -> Result<Value> {
45
+ match value.get("kind").and_then(Value::as_str) {
46
+ Some("map") => extract(document, entries(value)?, Some(&[id]), depth + 1),
47
+ Some("property") => {
48
+ let name = value
49
+ .get("name")
50
+ .and_then(Value::as_str)
51
+ .ok_or_else(|| invalid("Extraction property requires a name"))?;
52
+ if name == "textContent" {
53
+ return Ok(Value::String(document.text_ids(&[id])));
54
+ }
55
+ if matches!(name, "id" | "class" | "title" | "alt")
56
+ || name.starts_with("data-")
57
+ || name.starts_with("aria-")
58
+ || matches!(name, "href" | "src")
59
+ && document.options.get("baseURI").is_none()
60
+ {
61
+ return Ok(document
62
+ .attribute(id, name)
63
+ .map(Value::String)
64
+ .unwrap_or(Value::Null));
65
+ }
66
+ document.ensure_arena()?;
67
+ let mut arena = document.arena.borrow_mut();
68
+ let arena = arena
69
+ .as_mut()
70
+ .ok_or_else(|| invalid("Missing native document arena"))?;
71
+ if matches!(name, "name" | "type" | "nodeType") {
72
+ return Ok(match arena.node(id) {
73
+ Some(node) if name == "name" => {
74
+ node.name.clone().map(Value::String).unwrap_or(Value::Null)
75
+ }
76
+ Some(node) if name == "type" => Value::String(node.kind.clone()),
77
+ Some(_) => json!(1),
78
+ None => Value::Null,
79
+ });
80
+ }
81
+ crate::attributes::execute(
82
+ arena,
83
+ "prop",
84
+ &json!({ "ids": [id], "name": name }),
85
+ )?
86
+ .ok_or_else(|| invalid("Unsupported extraction property"))
87
+ }
88
+ _ => Err(invalid("Unknown extraction value kind")),
89
+ }
90
+ };
91
+ if all {
92
+ let mut values = Vec::new();
93
+ for id in ids {
94
+ match read(id)? {
95
+ Value::Null => {}
96
+ Value::Array(items) => values.extend(items),
97
+ item => values.push(item),
98
+ }
99
+ }
100
+ result.insert(key.into(), Value::Array(values));
101
+ } else if let Some(id) = ids.first() {
102
+ let item = read(*id)?;
103
+ if !item.is_null() {
104
+ result.insert(key.into(), item);
105
+ }
106
+ }
107
+ }
108
+ Ok(Value::Object(result))
109
+ }
110
+
111
+ pub(crate) fn execute(document: &NativeDocument, plan: &Value) -> Result<Value> {
112
+ if plan.get("version").and_then(Value::as_u64) != Some(1) {
113
+ return Err(invalid("Unsupported extraction plan version"));
114
+ }
115
+ extract(document, entries(plan)?, None, 0)
116
+ }
117
+
118
+ fn extract_arena(
119
+ arena: &mut crate::arena::DomArena,
120
+ plan: &[Value],
121
+ contexts: &[u32],
122
+ depth: usize,
123
+ ) -> Result<Value> {
124
+ if depth > 64 {
125
+ return Err(invalid("Extraction plan nesting exceeds 64 levels"));
126
+ }
127
+ let mut result = Map::new();
128
+ for entry in plan {
129
+ let key = entry
130
+ .get("key")
131
+ .and_then(Value::as_str)
132
+ .ok_or_else(|| invalid("Extraction entry requires a key"))?;
133
+ let selector = entry
134
+ .get("selector")
135
+ .and_then(Value::as_str)
136
+ .ok_or_else(|| invalid("Extraction entry requires a selector"))?;
137
+ let all = entry.get("all").and_then(Value::as_bool).unwrap_or(false);
138
+ let value = entry
139
+ .get("value")
140
+ .ok_or_else(|| invalid("Extraction entry requires a value"))?;
141
+ let ids = crate::selector::select(arena, selector, Some(contexts))?;
142
+ let mut read = |id| -> Result<Value> {
143
+ match value.get("kind").and_then(Value::as_str) {
144
+ Some("map") => extract_arena(arena, entries(value)?, &[id], depth + 1),
145
+ Some("property") => {
146
+ let name = value
147
+ .get("name")
148
+ .and_then(Value::as_str)
149
+ .ok_or_else(|| invalid("Extraction property requires a name"))?;
150
+ if matches!(name, "name" | "type" | "nodeType") {
151
+ return Ok(match arena.node(id) {
152
+ Some(node) if name == "name" => {
153
+ node.name.clone().map(Value::String).unwrap_or(Value::Null)
154
+ }
155
+ Some(node) if name == "type" => Value::String(node.kind.clone()),
156
+ Some(_) => json!(1),
157
+ None => Value::Null,
158
+ });
159
+ }
160
+ crate::attributes::execute(arena, "prop", &json!({"ids":[id],"name":name}))?
161
+ .ok_or_else(|| invalid("Unsupported extraction property"))
162
+ }
163
+ _ => Err(invalid("Unknown extraction value kind")),
164
+ }
165
+ };
166
+ if all {
167
+ let mut values = Vec::new();
168
+ for id in ids {
169
+ match read(id)? {
170
+ Value::Null => {}
171
+ Value::Array(items) => values.extend(items),
172
+ value => values.push(value),
173
+ }
174
+ }
175
+ result.insert(key.into(), Value::Array(values));
176
+ } else if let Some(id) = ids.first() {
177
+ let value = read(*id)?;
178
+ if !value.is_null() {
179
+ result.insert(key.into(), value);
180
+ }
181
+ }
182
+ }
183
+ Ok(Value::Object(result))
184
+ }
185
+
186
+ /** Use the caller's active options and contexts for pure prefixes around callbacks. */
187
+ pub(crate) fn execute_arena(
188
+ arena: &mut crate::arena::DomArena,
189
+ plan: &Value,
190
+ contexts: &[u32],
191
+ ) -> Result<Value> {
192
+ if plan.get("version").and_then(Value::as_u64) != Some(1) {
193
+ return Err(invalid("Unsupported extraction plan version"));
194
+ }
195
+ extract_arena(arena, entries(plan)?, contexts, 0)
196
+ }
197
+
198
+ #[cfg(test)]
199
+ mod tests {
200
+ use super::*;
201
+
202
+ #[test]
203
+ fn nested_maps_preserve_native_property_and_missing_value_semantics() {
204
+ let document = NativeDocument::parse_options(
205
+ "<section><a href='/page'>A<b>B</b></a><input checked></section><section><p>C</p></section>",
206
+ true,
207
+ json!({ "baseURI": "https://example.test/base/" }),
208
+ None,
209
+ );
210
+ let plan = json!({"version":1,"entries":[
211
+ {"key":"rows","selector":"section","all":true,"value":{"kind":"map","entries":[
212
+ {"key":"link","selector":"a","all":false,"value":{"kind":"property","name":"href"}},
213
+ {"key":"content","selector":"a","all":false,"value":{"kind":"property","name":"innerHTML"}},
214
+ {"key":"checked","selector":"input","all":true,"value":{"kind":"property","name":"checked"}},
215
+ {"key":"missing","selector":".absent","all":false,"value":{"kind":"property","name":"textContent"}}
216
+ ]}}
217
+ ]});
218
+ assert_eq!(
219
+ execute(&document, &plan).unwrap(),
220
+ json!({"rows":[
221
+ {"link":"https://example.test/page","content":"A<b>B</b>","checked":[true]},
222
+ {"checked":[]}
223
+ ]})
224
+ );
225
+ }
226
+ }
@@ -0,0 +1,166 @@
1
+ //! Reclaim detached native components that have no live JavaScript handles.
2
+ use std::collections::HashMap;
3
+
4
+ use crate::arena::DomArena;
5
+
6
+ #[derive(Debug)]
7
+ pub(crate) struct GcStats {
8
+ pub live: u32,
9
+ pub slots: u32,
10
+ pub retained: u32,
11
+ pub reclaimed: u32,
12
+ pub free: Vec<u32>,
13
+ }
14
+
15
+ #[cfg(test)]
16
+ pub(crate) fn collect(arena: &mut DomArena, pins: &HashMap<u32, usize>) -> GcStats {
17
+ collect_managed(arena, pins, true)
18
+ }
19
+
20
+ pub(crate) fn collect_managed(
21
+ arena: &mut DomArena,
22
+ pins: &HashMap<u32, usize>,
23
+ keep_root: bool,
24
+ ) -> GcStats {
25
+ let mut marked = vec![false; arena.nodes.len()];
26
+ let mut stack = if keep_root { vec![0] } else { Vec::new() };
27
+ stack.extend(
28
+ pins.iter()
29
+ .filter_map(|(id, count)| (*count > 0).then_some(*id)),
30
+ );
31
+ // Following both directions marks the entire parent/child component. A
32
+ // held descendant keeps its detached parent and siblings navigable. The
33
+ // visited bitmap also bounds work for cyclic or inconsistent user graphs.
34
+ while let Some(id) = stack.pop() {
35
+ let Some(node) = arena.nodes.get(id as usize) else {
36
+ continue;
37
+ };
38
+ if marked[id as usize] || node.kind == "released" {
39
+ continue;
40
+ }
41
+ marked[id as usize] = true;
42
+ stack.extend(node.parent);
43
+ stack.extend(node.children.iter().copied());
44
+ }
45
+
46
+ let retained = pins
47
+ .iter()
48
+ .filter(|(id, count)| **count > 0 && marked.get(**id as usize) == Some(&true))
49
+ .count() as u32;
50
+ let mut reclaimed = 0;
51
+ for (index, node) in arena.nodes.iter_mut().enumerate() {
52
+ if marked[index] || node.kind == "released" {
53
+ continue;
54
+ }
55
+ reclaimed += 1;
56
+ node.kind = "released".into();
57
+ node.name = None;
58
+ node.data = None;
59
+ node.attribs = Default::default();
60
+ node.children = Vec::new();
61
+ node.extras = Default::default();
62
+ node.parent = None;
63
+ node.prev = None;
64
+ node.next = None;
65
+ }
66
+ while arena
67
+ .nodes
68
+ .last()
69
+ .is_some_and(|node| node.kind == "released")
70
+ {
71
+ arena.nodes.pop();
72
+ }
73
+ if arena.nodes.capacity() > arena.nodes.len().saturating_mul(2).saturating_add(64) {
74
+ arena.nodes.shrink_to_fit();
75
+ }
76
+ let free = arena
77
+ .nodes
78
+ .iter()
79
+ .filter(|node| node.kind == "released")
80
+ .map(|node| node.id)
81
+ .collect::<Vec<_>>();
82
+ GcStats {
83
+ live: arena.nodes.len() as u32 - free.len() as u32,
84
+ slots: arena.nodes.len() as u32,
85
+ retained,
86
+ reclaimed,
87
+ free,
88
+ }
89
+ }
90
+
91
+ #[cfg(test)]
92
+ mod tests {
93
+ use super::*;
94
+ use crate::arena::DomNode;
95
+ use serde_json::json;
96
+
97
+ fn fixture() -> DomArena {
98
+ let mut nodes = vec![DomNode::new("root")];
99
+ nodes.extend((0..5).map(|_| DomNode::new("tag")));
100
+ nodes[0].children = vec![1, 5];
101
+ nodes[2].children = vec![3, 4];
102
+ nodes[2].name = Some("detached".into());
103
+ nodes[3].data = Some("payload".repeat(100));
104
+ nodes[3].attribs.insert("key".into(), json!("value"));
105
+ nodes[3].extras.insert("data".into(), json!({"held":true}));
106
+ DomArena::from_records(nodes, json!({}))
107
+ }
108
+
109
+ #[test]
110
+ fn held_detached_child_preserves_its_parent_and_sibling() {
111
+ let mut arena = fixture();
112
+ let stats = collect(&mut arena, &HashMap::from([(3, 2)]));
113
+ assert_eq!(
114
+ (stats.live, stats.slots, stats.retained, stats.reclaimed),
115
+ (6, 6, 1, 0)
116
+ );
117
+ assert!(stats.free.is_empty());
118
+ assert_eq!(arena.nodes[3].parent, Some(2));
119
+ assert_eq!(arena.nodes[3].next, Some(4));
120
+ assert_eq!(arena.nodes[2].children, vec![3, 4]);
121
+ }
122
+
123
+ #[test]
124
+ fn unpinned_payloads_are_cleared_and_trailing_slots_trimmed() {
125
+ let mut arena = fixture();
126
+ let stats = collect(&mut arena, &HashMap::new());
127
+ assert_eq!((stats.live, stats.slots, stats.reclaimed), (3, 6, 3));
128
+ assert_eq!(stats.free, vec![2, 3, 4]);
129
+ let node = &arena.nodes[3];
130
+ assert_eq!(node.kind, "released");
131
+ assert!(node.data.is_none() && node.name.is_none());
132
+ assert!(node.attribs.is_empty() && node.children.is_empty() && node.extras.is_empty());
133
+ assert!(node.parent.is_none() && node.prev.is_none() && node.next.is_none());
134
+ arena.detach(5);
135
+ let stats = collect(&mut arena, &HashMap::from([(5, 0)]));
136
+ assert_eq!(
137
+ (stats.live, stats.slots, stats.retained, stats.reclaimed),
138
+ (2, 2, 0, 1)
139
+ );
140
+ assert!(stats.free.is_empty());
141
+ assert_eq!(collect(&mut arena, &HashMap::new()).reclaimed, 0);
142
+ }
143
+
144
+ #[test]
145
+ fn cyclic_components_are_marked_once() {
146
+ let mut arena = fixture();
147
+ arena.nodes[2].parent = Some(3);
148
+ arena.nodes[3].children.push(2);
149
+ let stats = collect(&mut arena, &HashMap::from([(3, 1)]));
150
+ assert_eq!((stats.live, stats.retained, stats.reclaimed), (6, 1, 0));
151
+ let stats = collect(&mut arena, &HashMap::new());
152
+ assert_eq!(stats.reclaimed, 3);
153
+ }
154
+
155
+ #[test]
156
+ fn managed_detached_handle_does_not_retain_the_original_document() {
157
+ let mut arena = fixture();
158
+ let stats = collect_managed(&mut arena, &HashMap::from([(3, 1)]), false);
159
+ assert_eq!((stats.live, stats.retained, stats.reclaimed), (3, 1, 3));
160
+ assert_eq!(arena.nodes[0].kind, "released");
161
+ assert_eq!(arena.nodes[1].kind, "released");
162
+ assert_eq!(arena.nodes[2].name.as_deref(), Some("detached"));
163
+ assert_eq!(arena.nodes[3].parent, Some(2));
164
+ assert_eq!(arena.nodes[3].next, Some(4));
165
+ }
166
+ }