merge

Author: b41sh

CHANGELOG.md

@@ -1,38 +1,71 @@
+## [v0.4.4] - 2024-11-16
+
+### Fixed
+
+- Fix: panic when facing corrupted jsonb (#67)
+
+### Added
+
+- Bump fast-float2 v0.2.3  (#69)
+- Feat: add a function to parse jsonb only (#66)
+- Feat: support `object_delete` and `object_pick` function (#65)
+- Feat: support `object_insert` function (#64)
+- Feat: Support json array functions (#62)
+- Feat: add lazy value (#61)
+
+## [v0.4.3] - 2024-09-30
+
+### Fixed
+
+- Fix: Fix compare object value with different length panic (#59)
+
+## [v0.4.2] - 2024-09-19
+
+### Added
+
+- Feat: make `preserve_order` a default feature (#56)
+
+## [v0.4.1] - 2024-07-18
+
+### Fixed
+
+- Fix: Fix jsonpath selector unwrap panic. (#53)
+
 ## [v0.4.0] - 2024-05-17
 
 ### Fixed
 
-Fix: Fix get by keypath with null value. (#47)
-Fix: Handle invalid jsonb value to avoid panic in functions. (#46)
-Fix: Fix builder & concat container jentry len. (#43) 
+- Fix: Fix get by keypath with null value. (#47)
+- Fix: Handle invalid jsonb value to avoid panic in functions. (#46)
+- Fix: Fix builder & concat container jentry len. (#43)
 
 ### Added
 
-Feat: Support convert jsonb value to `serde_json` value. (#49) 
-Feat: Add `exists` filter expression. (48)` 
-Feat: Add `delete_by_keypath`. (#45)
-Feat: Add `delete_by_index` & `delete_by_name`. (#44)
-Feat: Add `concat` & improve `strip_nulls`. (#42)
-Feat: Add jsonpath predicate support. (#41) 
-Feat: Add `contains` api. (#40) 
-Feat: Add `exists_any_keys` & `exists_all_keys`. (#38) 
-Feat: Support parse key paths. (#37)
-Feat: Add `get_by_keypath`. (#36)
+- Feat: Support convert jsonb value to `serde_json` value. (#49)
+- Feat: Add `exists` filter expression. (48)`
+- Feat: Add `delete_by_keypath`. (#45)
+- Feat: Add `delete_by_index` & `delete_by_name`. (#44)
+- Feat: Add `concat` & improve `strip_nulls`. (#42)
+- Feat: Add jsonpath predicate support. (#41)
+- Feat: Add `contains` api. (#40)
+- Feat: Add `exists_any_keys` & `exists_all_keys`. (#38)
+- Feat: Support parse key paths. (#37)
+- Feat: Add `get_by_keypath`. (#36)
 
 ## [v0.3.0] - 2023-10-13
 
 ### Added
 
-Docs: Add more jsonb encoding format descriptions. (#34)
-Feat: Support `object_each` api. (#33)
-Feat: Support `path_exists` api. (#32)
-Feat: Support `type_of` api. (#31)
-Feat: Support `strip_nulls` api. (#30)
-Perf: Add benches for parser and `get_path`. (#29)
-Chore: Add check fmt and clippy. (#27)
-Feat: Support `to_pretty_string` api. (#26)
-Feat: Support `traverse_check_string` function. (#25)
-Feat: Improve json path selector using less memory. (#24)
+- Docs: Add more jsonb encoding format descriptions. (#34)
+- Feat: Support `object_each` api. (#33)
+- Feat: Support `path_exists` api. (#32)
+- Feat: Support `type_of` api. (#31)
+- Feat: Support `strip_nulls` api. (#30)
+- Perf: Add benches for parser and `get_path`. (#29)
+- Chore: Add check fmt and clippy. (#27)
+- Feat: Support `to_pretty_string` api. (#26)
+- Feat: Support `traverse_check_string` function. (#25)
+- Feat: Improve json path selector using less memory. (#24)
 
 ## [v0.2.3] - 2023-07-10
 
@@ -81,8 +114,14 @@ Feat: Improve json path selector using less memory. (#24)
 - Implement `JSONB` encodes and decodes.
 - Implemented a number of `JSONB` functions.
 
-
-[v0.2.2]: https://github.com/datafuselabs/jsonb/compare/v0.2.1...v0.2.2
-[v0.2.1]: https://github.com/datafuselabs/jsonb/compare/v0.2.0...v0.2.1
-[v0.2.0]: https://github.com/datafuselabs/jsonb/compare/v0.1.1...v0.2.0
-[v0.1.1]: https://github.com/datafuselabs/jsonb/compare/v0.1.0...v0.1.1
+[v0.4.4]: https://github.com/databendlabs/jsonb/compare/v0.4.3...v0.4.4
+[v0.4.3]: https://github.com/databendlabs/jsonb/compare/v0.4.2...v0.4.3
+[v0.4.2]: https://github.com/databendlabs/jsonb/compare/v0.4.1...v0.4.2
+[v0.4.1]: https://github.com/databendlabs/jsonb/compare/v0.4.0...v0.4.1
+[v0.4.0]: https://github.com/databendlabs/jsonb/compare/v0.3.0...v0.4.0
+[v0.3.0]: https://github.com/databendlabs/jsonb/compare/v0.2.3...v0.3.0
+[v0.2.3]: https://github.com/databendlabs/jsonb/compare/v0.2.2...v0.2.3
+[v0.2.2]: https://github.com/databendlabs/jsonb/compare/v0.2.1...v0.2.2
+[v0.2.1]: https://github.com/databendlabs/jsonb/compare/v0.2.0...v0.2.1
+[v0.2.0]: https://github.com/databendlabs/jsonb/compare/v0.1.1...v0.2.0
+[v0.1.1]: https://github.com/databendlabs/jsonb/compare/v0.1.0...v0.1.1

Cargo.toml

@@ -17,25 +17,23 @@ authors = ["Databend Authors <opensource@datafuselabs.com>"]
 categories = ["encoding"]
 description = "JSONB implement in Rust."
 edition = "2021"
-homepage = "https://github.com/datafuselabs/jsonb"
+homepage = "https://github.com/databendlabs/jsonb"
 keywords = ["json", "jsonb", "jsonpath"]
 license = "Apache-2.0"
 name = "jsonb"
-repository = "https://github.com/datafuselabs/jsonb"
-version = "0.4.0"
+repository = "https://github.com/databendlabs/jsonb"
+version = "0.4.4"
 rust-version = "1.77"
 
 [dependencies]
 byteorder = "1.5.0"
-fast-float = "0.2.0"
+fast-float2 = "0.2.3"
 itoa = "1.0"
 nom = "7.1.3"
-ordered-float = { version = "4.2", default-features = false }
+ordered-float = { version = "4.5", default-features = false }
 rand = { version = "0.8.5", features = ["small_rng"] }
 ryu = "1.0"
-serde_json = { version = "1.0", default-features = false, features = [
-  "preserve_order",
-] }
+serde_json = { version = "1.0", default-features = false, features = ["std"] }
 
 [dev-dependencies]
 goldenfile = "1.7"
@@ -45,6 +43,9 @@ simd-json = "0.13.10"
 mockalloc = "0.1.2"
 criterion = "0.5.1"
 
+[features]
+default = ["serde_json/preserve_order"]
+
 [[bench]]
 name = "parser"
 harness = false

benches/get_path.rs

@@ -27,7 +27,7 @@ fn jsonb_get(data: &[u8], paths: &[&str], expected: &str) {
     let mut result_data = vec![];
     let mut result_offsets = vec![];
 
-    jsonb::get_by_path(data, json_path, &mut result_data, &mut result_offsets);
+    jsonb::get_by_path(data, json_path, &mut result_data, &mut result_offsets).unwrap();
 
     let s = jsonb::as_str(&result_data).unwrap();
     assert_eq!(s, expected);

src/de.rs

@@ -40,19 +40,19 @@ use super::value::Value;
 /// `Number`, `String` and `Container`. They have three different decode methods.
 /// 1. `Null`, `True` and `False` can be obtained by `JEntry`, no extra work required.
 /// 2. `Number` and `String` has related `RawData`, `JEntry` store the length
-/// or offset of this data, the `Value` can be read out and then decoded.
+///    or offset of this data, the `Value` can be read out and then decoded.
 /// 3. `Container` is actually a nested `Array` or `Object` with the same structure,
-/// `JEntry` store the length or offset of the lower-level `Header`,
-/// from where the same decode process can begin.
-
-/// `RawData` is the encoded `Value`.
-/// `Number` is a variable-length `Decimal`, store both int and float value.
-/// `String` is the original string, can be borrowed directly without extra decode.
-/// `Array` and `Object` is a lower-level encoded `JSONB` value.
-/// The upper-level doesn't care about the specific content.
-/// Decode can be executed recursively.
-
-/// Decode `JSONB` Value from binary bytes.
+///    `JEntry` store the length or offset of the lower-level `Header`,
+///    from where the same decode process can begin.
+///
+///    `RawData` is the encoded `Value`.
+///    `Number` is a variable-length `Decimal`, store both int and float value.
+///    `String` is the original string, can be borrowed directly without extra decode.
+///    `Array` and `Object` is a lower-level encoded `JSONB` value.
+///    The upper-level doesn't care about the specific content.
+///    Decode can be executed recursively.
+///
+///    Decode `JSONB` Value from binary bytes.
 pub fn from_slice(buf: &[u8]) -> Result<Value<'_>, Error> {
     let mut decoder = Decoder::new(buf);
     match decoder.decode() {
@@ -62,6 +62,11 @@ pub fn from_slice(buf: &[u8]) -> Result<Value<'_>, Error> {
     }
 }
 
+pub fn parse_jsonb(buf: &[u8]) -> Result<Value<'_>, Error> {
+    let mut decoder = Decoder::new(buf);
+    decoder.decode()
+}
+
 #[repr(transparent)]
 pub struct Decoder<'a> {
     buf: &'a [u8],
@@ -111,13 +116,15 @@ impl<'a> Decoder<'a> {
             FALSE_TAG => Ok(Value::Bool(false)),
             STRING_TAG => {
                 let offset = jentry.length as usize;
-                let s = unsafe { std::str::from_utf8_unchecked(&self.buf[..offset]) };
+                let string = &self.buf.get(..offset).ok_or(Error::InvalidUtf8)?;
+                let s = unsafe { std::str::from_utf8_unchecked(string) };
                 self.buf = &self.buf[offset..];
                 Ok(Value::String(Cow::Borrowed(s)))
             }
             NUMBER_TAG => {
                 let offset = jentry.length as usize;
-                let n = Number::decode(&self.buf[..offset]);
+                let number = &self.buf.get(..offset).ok_or(Error::InvalidJsonbNumber)?;
+                let n = Number::decode(number)?;
                 self.buf = &self.buf[offset..];
                 Ok(Value::Number(n))
             }

src/error.rs

@@ -79,12 +79,15 @@ pub enum Error {
     InvalidJsonb,
     InvalidJsonbHeader,
     InvalidJsonbJEntry,
+    InvalidJsonbNumber,
 
     InvalidJsonPath,
     InvalidJsonPathPredicate,
     InvalidKeyPath,
 
     InvalidJsonType,
+    InvalidObject,
+    ObjectDuplicateKey,
 
     Syntax(ParseErrorCode, usize),
 }
@@ -109,3 +112,9 @@ impl From<std::str::Utf8Error> for Error {
         Error::InvalidUtf8
     }
 }
+
+impl From<nom::Err<nom::error::Error<&[u8]>>> for Error {
+    fn from(_error: nom::Err<nom::error::Error<&[u8]>>) -> Self {
+        Error::InvalidJsonb
+    }
+}