61 regex perf (besok#62)

* add rgex bench * fix inter * init impl * add config * fix complains --------- Co-authored-by: Boris Zhguchev <[email protected]>
xMAC94x · Feb 19, 2024 · a07c7b6 · a07c7b6
1 parent 668b626
commit a07c7b6
Show file tree

Hide file tree

Showing 11 changed files with 405 additions and 86 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -38,4 +38,8 @@
 * **`0.3.5`**
   * add `!` negation operation in filters
   * allow using  () in filters
+* **`0.5`**
+  * add config for jsonpath
+  * add an option to add a regex cache for boosting performance
+
 
diff --git a/Cargo.toml b/Cargo.toml
@@ -1,7 +1,7 @@
 [package]
 name = "jsonpath-rust"
 description = "The library provides the basic functionality to find the set of the data according to the filtering query."
-version = "0.4.0"
+version = "0.5.0"
 authors = ["BorisZhguchev <[email protected]>"]
 edition = "2018"
 license-file = "LICENSE"
@@ -17,6 +17,12 @@ regex = "1"
 pest = "2.0"
 pest_derive = "2.0"
 thiserror = "1.0.50"
+lazy_static = "1.4"
+once_cell = "1.19.0"
 
 [dev-dependencies]
-lazy_static = "1.0"
+criterion = "0.5.1"
+
+[[bench]]
+name = "regex_bench"
+harness = false
diff --git a/README.md b/README.md
@@ -389,7 +389,50 @@ fn test() {
 ** If the value has been modified during the search, there is no way to find a path of a new value. 
 It can happen if we try to find a length() of array, for in stance.**
 
+## Configuration
 
+The JsonPath provides a wat to configure the search by using `JsonPathConfig`.
+
+```rust
+pub fn main() {
+  let cfg = JsonPathConfig::new(RegexCache::Implemented(DefaultRegexCacheInst::default()));
+}
+```
+
+### Regex cache
+The configuration provides an ability to use a regex cache to improve the [performance](https://github.com/besok/jsonpath-rust/issues/61)
+
+To instantiate the cache needs to use `RegexCache` enum with the implementation of the trait `RegexCacheInst`.
+Default implementation `DefaultRegexCacheInst` uses `Arc<Mutex<HashMap<String,Regex>>>`.
+The pair of Box<Value> or Value  and config can be used:
+```rust
+pub fn main(){
+  let cfg = JsonPathConfig::new(RegexCache::Implemented(DefaultRegexCacheInst::default()));
+  let json = Box::new(json!({
+            "author":"abcd(Rees)",
+        }));
+
+  let _v = (json, cfg).path("$.[?(@.author ~= '.*(?i)d\\(Rees\\)')]")
+          .expect("the path is correct");
+
+
+}
+```
+or using `JsonPathFinder` :
+
+```rust
+fn main() {
+    let cfg = JsonPathConfig::new(RegexCache::Implemented(DefaultRegexCacheInst::default()));
+    let finder = JsonPathFinder::from_str_with_cfg(
+        r#"{"first":{"second":[{"active":1},{"passive":1}]}}"#,
+        "$.first.second[?(@.active)]",
+        cfg,
+    ).unwrap();
+    let slice_of_data: Vec<&Value> = finder.find_slice();
+    let js = json!({"active":1});
+    assert_eq!(slice_of_data, vec![JsonPathValue::Slice(&js, "$.first.second[0]".to_string())]);
+}
+```
 
 ## The structure
 

diff --git a/benches/regex_bench.rs b/benches/regex_bench.rs
@@ -0,0 +1,40 @@
+use criterion::{black_box, criterion_group, criterion_main, Criterion};
+use jsonpath_rust::path::config::cache::{DefaultRegexCacheInst, RegexCache};
+use jsonpath_rust::path::config::JsonPathConfig;
+use jsonpath_rust::{JsonPathFinder, JsonPathInst, JsonPathQuery};
+use once_cell::sync::Lazy;
+use serde_json::{json, Value};
+use std::str::FromStr;
+
+fn regex_perf_test_with_cache(cfg: JsonPathConfig) {
+    let json = Box::new(json!({
+        "author":"abcd(Rees)",
+    }));
+
+    let _v = (json, cfg)
+        .path("$.[?(@.author ~= '.*(?i)d\\(Rees\\)')]")
+        .expect("the path is correct");
+}
+
+fn regex_perf_test_without_cache() {
+    let json = Box::new(json!({
+        "author":"abcd(Rees)",
+    }));
+
+    let _v = json
+        .path("$.[?(@.author ~= '.*(?i)d\\(Rees\\)')]")
+        .expect("the path is correct");
+}
+
+pub fn criterion_benchmark(c: &mut Criterion) {
+    let cfg = JsonPathConfig::new(RegexCache::Implemented(DefaultRegexCacheInst::default()));
+    c.bench_function("regex bench without cache", |b| {
+        b.iter(|| regex_perf_test_without_cache())
+    });
+    c.bench_function("regex bench with cache", |b| {
+        b.iter(|| regex_perf_test_with_cache(cfg.clone()))
+    });
+}
+
+criterion_group!(benches, criterion_benchmark);
+criterion_main!(benches);
diff --git a/src/lib.rs b/src/lib.rs
@@ -116,6 +116,7 @@
 
 use crate::parser::model::JsonPath;
 use crate::parser::parser::parse_json_path;
+use crate::path::config::JsonPathConfig;
 use crate::path::{json_path_instance, PathInstance};
 use serde_json::Value;
 use std::convert::TryInto;
@@ -182,8 +183,12 @@ impl FromStr for JsonPathInst {
 }
 
 impl JsonPathInst {
-    pub fn find_slice<'a>(&'a self, value: &'a Value) -> Vec<JsonPtr<'a, Value>> {
-        json_path_instance(&self.inner, value)
+    pub fn find_slice<'a>(
+        &'a self,
+        value: &'a Value,
+        cfg: JsonPathConfig,
+    ) -> Vec<JsonPtr<'a, Value>> {
+        json_path_instance(&self.inner, value, cfg)
             .find(JsonPathValue::from_root(value))
             .into_iter()
             .filter(|v| v.has_value())
@@ -224,13 +229,27 @@ impl JsonPathQuery for Box<Value> {
     }
 }
 
+impl JsonPathQuery for (Box<Value>, JsonPathConfig) {
+    fn path(self, query: &str) -> Result<Value, String> {
+        let p = JsonPathInst::from_str(query)?;
+        Ok(JsonPathFinder::new_with_cfg(self.0, Box::new(p), self.1).find())
+    }
+}
+
 impl JsonPathQuery for Value {
     fn path(self, query: &str) -> Result<Value, String> {
         let p = JsonPathInst::from_str(query)?;
         Ok(JsonPathFinder::new(Box::new(self), Box::new(p)).find())
     }
 }
 
+impl JsonPathQuery for (Value, JsonPathConfig) {
+    fn path(self, query: &str) -> Result<Value, String> {
+        let p = JsonPathInst::from_str(query)?;
+        Ok(JsonPathFinder::new_with_cfg(Box::new(self.0), Box::new(p), self.1).find())
+    }
+}
+
 /// just to create a json path value of data
 /// Example:
 ///  - json_path_value(&json) = `JsonPathValue::Slice(&json)`
@@ -294,6 +313,7 @@ type JsPathStr = String;
 pub(crate) fn jsp_idx(prefix: &str, idx: usize) -> String {
     format!("{}[{}]", prefix, idx)
 }
+
 pub(crate) fn jsp_obj(prefix: &str, key: &str) -> String {
     format!("{}.['{}']", prefix, key)
 }
@@ -337,7 +357,7 @@ impl<'a, Data: Clone + Debug + Default> JsonPathValue<'a, Data> {
 }
 
 impl<'a, Data> JsonPathValue<'a, Data> {
-    fn only_no_value(input: &Vec<JsonPathValue<'a, Data>>) -> bool {
+    fn only_no_value(input: &[JsonPathValue<'a, Data>]) -> bool {
         !input.is_empty() && input.iter().filter(|v| v.has_value()).count() == 0
     }
     fn map_vec(data: Vec<(&'a Data, JsPathStr)>) -> Vec<JsonPathValue<'a, Data>> {
@@ -407,12 +427,26 @@ impl<'a, Data> JsonPathValue<'a, Data> {
 pub struct JsonPathFinder {
     json: Box<Value>,
     path: Box<JsonPathInst>,
+    cfg: JsonPathConfig,
 }
 
 impl JsonPathFinder {
     /// creates a new instance of [JsonPathFinder]
     pub fn new(json: Box<Value>, path: Box<JsonPathInst>) -> Self {
-        JsonPathFinder { json, path }
+        JsonPathFinder {
+            json,
+            path,
+            cfg: JsonPathConfig::default(),
+        }
+    }
+
+    pub fn new_with_cfg(json: Box<Value>, path: Box<JsonPathInst>, cfg: JsonPathConfig) -> Self {
+        JsonPathFinder { json, path, cfg }
+    }
+
+    /// sets a cfg with a new one
+    pub fn set_cfg(&mut self, cfg: JsonPathConfig) {
+        self.cfg = cfg
     }
 
     /// updates a path with a new one
@@ -440,10 +474,15 @@ impl JsonPathFinder {
         let path = Box::new(JsonPathInst::from_str(path)?);
         Ok(JsonPathFinder::new(json, path))
     }
+    pub fn from_str_with_cfg(json: &str, path: &str, cfg: JsonPathConfig) -> Result<Self, String> {
+        let json = serde_json::from_str(json).map_err(|e| e.to_string())?;
+        let path = Box::new(JsonPathInst::from_str(path)?);
+        Ok(JsonPathFinder::new_with_cfg(json, path, cfg))
+    }
 
     /// creates an instance to find a json slice from the json
     pub fn instance(&self) -> PathInstance {
-        json_path_instance(&self.path.inner, &self.json)
+        json_path_instance(&self.path.inner, &self.json, self.cfg.clone())
     }
     /// finds a slice of data in the set json.
     /// The result is a vector of references to the incoming structure.
@@ -494,6 +533,7 @@ impl JsonPathFinder {
 
 #[cfg(test)]
 mod tests {
+    use crate::path::config::JsonPathConfig;
     use crate::JsonPathQuery;
     use crate::JsonPathValue::{NoValue, Slice};
     use crate::{jp_v, JsonPathFinder, JsonPathInst, JsonPathValue};
@@ -1194,7 +1234,7 @@ mod tests {
         let query = JsonPathInst::from_str("$..book[?(@.author size 10)].title")
             .expect("the path is correct");
 
-        let results = query.find_slice(&json);
+        let results = query.find_slice(&json, JsonPathConfig::default());
         let v = results.first().expect("to get value");
 
         // V can be implicitly converted to &Value
@@ -1257,7 +1297,7 @@ mod tests {
             v,
             vec![Slice(
                 &json!({"second":{"active": 1}}),
-                "$.['first']".to_string()
+                "$.['first']".to_string(),
             )]
         );
 
@@ -1271,7 +1311,7 @@ mod tests {
             v,
             vec![Slice(
                 &json!({"second":{"active": 1}}),
-                "$.['first']".to_string()
+                "$.['first']".to_string(),
             )]
         );
 
@@ -1285,7 +1325,7 @@ mod tests {
             v,
             vec![Slice(
                 &json!({"second":{"active": 1}}),
-                "$.['first']".to_string()
+                "$.['first']".to_string(),
             )]
         );
 
@@ -1299,7 +1339,7 @@ mod tests {
             v,
             vec![Slice(
                 &json!({"second":{"active": 1}}),
-                "$.['first']".to_string()
+                "$.['first']".to_string(),
             )]
         );
     }

diff --git a/src/path/config.rs b/src/path/config.rs
@@ -0,0 +1,16 @@
+pub mod cache;
+
+use crate::path::config::cache::RegexCache;
+
+/// Configuration to adjust the jsonpath search
+#[derive(Clone, Default)]
+pub struct JsonPathConfig {
+    /// cache to provide
+    pub regex_cache: RegexCache,
+}
+
+impl JsonPathConfig {
+    pub fn new(regex_cache: RegexCache) -> Self {
+        Self { regex_cache }
+    }
+}