[chore]: Bump github.com/gin-gonic/gin from 1.9.0 to 1.9.1 (#1855)

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-11-23 00:33:50 +01:00 · 2023-06-01 22:20:16 +01:00 · 2023-06-01 22:20:16 +01:00 · 55aacaf4b0
commit 55aacaf4b0
parent 23705b31e4
138 changed files with 27543 additions and 25484 deletions
--- a/go.mod
+++ b/go.mod
@ -24,7 +24,7 @@ require (
 	github.com/gin-contrib/cors v1.4.0
 	github.com/gin-contrib/gzip v0.0.6
 	github.com/gin-contrib/sessions v0.0.5
-	github.com/gin-gonic/gin v1.9.0
+	github.com/gin-gonic/gin v1.9.1
 	github.com/go-fed/httpsig v1.1.0
 	github.com/go-playground/form/v4 v4.2.0
 	github.com/go-playground/validator/v10 v10.14.0
@ -82,7 +82,7 @@ require (
 	codeberg.org/gruf/go-maps v1.0.3 // indirect
 	codeberg.org/gruf/go-pools v1.1.0 // indirect
 	github.com/aymerick/douceur v0.2.0 // indirect
-	github.com/bytedance/sonic v1.8.0 // indirect
+	github.com/bytedance/sonic v1.9.1 // indirect
 	github.com/cenkalti/backoff/v4 v4.2.1 // indirect
 	github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311 // indirect
 	github.com/cilium/ebpf v0.9.1 // indirect
@ -108,7 +108,7 @@ require (
 	github.com/go-playground/locales v0.14.1 // indirect
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-xmlfmt/xmlfmt v0.0.0-20211206191508-7fd73a941850 // indirect
-	github.com/goccy/go-json v0.10.0 // indirect
+	github.com/goccy/go-json v0.10.2 // indirect
 	github.com/godbus/dbus/v5 v5.0.4 // indirect
 	github.com/golang-jwt/jwt v3.2.2+incompatible // indirect
 	github.com/golang/geo v0.0.0-20210211234256-740aa86cb551 // indirect
@ -132,13 +132,13 @@ require (
 	github.com/klauspost/cpuid/v2 v2.2.4 // indirect
 	github.com/leodido/go-urn v1.2.4 // indirect
 	github.com/magiconair/properties v1.8.7 // indirect
-	github.com/mattn/go-isatty v0.0.18 // indirect
+	github.com/mattn/go-isatty v0.0.19 // indirect
 	github.com/minio/md5-simd v1.1.2 // indirect
 	github.com/minio/sha256-simd v1.0.1 // indirect
 	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
 	github.com/modern-go/reflect2 v1.0.2 // indirect
 	github.com/opencontainers/runtime-spec v1.0.2 // indirect
-	github.com/pelletier/go-toml/v2 v2.0.6 // indirect
+	github.com/pelletier/go-toml/v2 v2.0.8 // indirect
 	github.com/pkg/errors v0.9.1 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
 	github.com/quasoft/memstore v0.0.0-20191010062613-2bce066d2b0b // indirect
@ -154,7 +154,7 @@ require (
 	github.com/tdewolff/parse/v2 v2.6.5 // indirect
 	github.com/tmthrgd/go-hex v0.0.0-20190904060850-447a3041c3bc // indirect
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
-	github.com/ugorji/go/codec v1.2.9 // indirect
+	github.com/ugorji/go/codec v1.2.11 // indirect
 	github.com/uptrace/opentelemetry-go-extra/otelsql v0.1.21 // indirect
 	github.com/vmihailenco/msgpack/v5 v5.3.5 // indirect
 	github.com/vmihailenco/tagparser/v2 v2.0.0 // indirect
@ -162,7 +162,7 @@ require (
 	go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.14.0 // indirect
 	go.opentelemetry.io/otel/metric v0.37.0 // indirect
 	go.opentelemetry.io/proto/otlp v0.19.0 // indirect
-	golang.org/x/arch v0.0.0-20210923205945-b76863e36670 // indirect
+	golang.org/x/arch v0.3.0 // indirect
 	golang.org/x/mod v0.10.0 // indirect
 	golang.org/x/sys v0.8.0 // indirect
 	golang.org/x/tools v0.6.0 // indirect
--- a/go.sum
+++ b/go.sum
@ -102,8 +102,8 @@ github.com/aymerick/douceur v0.2.0/go.mod h1:wlT5vV2O3h55X9m7iVYN0TBM0NH/MmbLnd3
 github.com/buckket/go-blurhash v1.1.0 h1:X5M6r0LIvwdvKiUtiNcRL2YlmOfMzYobI3VCKCZc9Do=
 github.com/buckket/go-blurhash v1.1.0/go.mod h1:aT2iqo5W9vu9GpyoLErKfTHwgODsZp3bQfXjXJUxNb8=
 github.com/bytedance/sonic v1.5.0/go.mod h1:ED5hyg4y6t3/9Ku1R6dU/4KyJ48DZ4jPhfY1O2AihPM=
-github.com/bytedance/sonic v1.8.0 h1:ea0Xadu+sHlu7x5O3gKhRpQ1IKiMrSiHttPF0ybECuA=
-github.com/bytedance/sonic v1.8.0/go.mod h1:i736AoUSYt75HyZLoJW9ERYxcy6eaN6h4BZXU064P/U=
+github.com/bytedance/sonic v1.9.1 h1:6iJ6NqdoxCDr6mbY8h18oSO+cShGSMRGCEo7F2h0x8s=
+github.com/bytedance/sonic v1.9.1/go.mod h1:i736AoUSYt75HyZLoJW9ERYxcy6eaN6h4BZXU064P/U=
 github.com/cenkalti/backoff/v4 v4.2.1 h1:y4OZtCnogmCPw98Zjyt5a6+QwPLGkiQsYW5oUqylYbM=
 github.com/cenkalti/backoff/v4 v4.2.1/go.mod h1:Y3VNntkOUPxTVeUxJ/G5vcM//AlwfmyYozVcomhLiZE=
 github.com/census-instrumentation/opencensus-proto v0.2.1/go.mod h1:f6KPmirojxKA12rnyqOA5BBL4O983OfeGPqjHWSTneU=
@ -201,8 +201,8 @@ github.com/gin-contrib/sessions v0.0.5/go.mod h1:vYAuaUPqie3WUSsft6HUlCjlwwoJQs9
 github.com/gin-contrib/sse v0.1.0 h1:Y/yl/+YNO8GZSjAhjMsSuLt29uWRFHdHYUb5lYOV9qE=
 github.com/gin-contrib/sse v0.1.0/go.mod h1:RHrZQHXnP2xjPF+u1gW/2HnVO7nvIa9PG3Gm+fLHvGI=
 github.com/gin-gonic/gin v1.8.1/go.mod h1:ji8BvRH1azfM+SYow9zQ6SZMvR8qOMZHmsCuWR9tTTk=
-github.com/gin-gonic/gin v1.9.0 h1:OjyFBKICoexlu99ctXNR2gg+c5pKrKMuyjgARg9qeY8=
-github.com/gin-gonic/gin v1.9.0/go.mod h1:W1Me9+hsUSyj3CePGrd1/QrKJMSJ1Tu/0hFEH89961k=
+github.com/gin-gonic/gin v1.9.1 h1:4idEAncQnU5cB7BeOkPtxjfCSye0AAm1R0RVIqJ+Jmg=
+github.com/gin-gonic/gin v1.9.1/go.mod h1:hPrL7YrpYKXt5YId3A/Tnip5kqbEAP+KLuI3SUcPTeU=
 github.com/go-errors/errors v1.0.1/go.mod h1:f4zRHt4oKfwPJE5k8C9vpYG+aDHdBFUsgrm6/TyX73Q=
 github.com/go-errors/errors v1.0.2/go.mod h1:psDX2osz5VnTOnFWbDeWwS7yejl+uV3FEWEp4lssFEs=
 github.com/go-errors/errors v1.1.1/go.mod h1:psDX2osz5VnTOnFWbDeWwS7yejl+uV3FEWEp4lssFEs=
@ -239,8 +239,8 @@ github.com/go-test/deep v1.0.8 h1:TDsG77qcSprGbC6vTN8OuXp5g+J+b5Pcguhf7Zt61VM=
 github.com/go-xmlfmt/xmlfmt v0.0.0-20211206191508-7fd73a941850 h1:PSPmmucxGiFBtbQcttHTUc4LQ3P09AW+ldO2qspyKdY=
 github.com/go-xmlfmt/xmlfmt v0.0.0-20211206191508-7fd73a941850/go.mod h1:aUCEOzzezBEjDBbFBoSiya/gduyIiWYRP6CnSFIV8AM=
 github.com/goccy/go-json v0.9.7/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
-github.com/goccy/go-json v0.10.0 h1:mXKd9Qw4NuzShiRlOXKews24ufknHO7gx30lsDyokKA=
-github.com/goccy/go-json v0.10.0/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
+github.com/goccy/go-json v0.10.2 h1:CrxCmQqYDkv1z7lO7Wbh2HN93uovUHgrECaO5ZrCXAU=
+github.com/goccy/go-json v0.10.2/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
 github.com/godbus/dbus/v5 v5.0.4 h1:9349emZab16e7zQvpmsbtjc18ykshndd8y2PG3sgJbA=
 github.com/godbus/dbus/v5 v5.0.4/go.mod h1:xhWf0FNVPg57R7Z0UbKHbJfkEywrmjJnf7w5xrFpKfA=
 github.com/golang-jwt/jwt v3.2.1+incompatible/go.mod h1:8pz2t5EyA70fFQQSrl6XZXzqecmYZeUEB8OUGHkxJ+I=
@ -440,8 +440,8 @@ github.com/mattn/go-isatty v0.0.5/go.mod h1:Iq45c/XA43vh69/j3iqttzPXn0bhXyGjM0Hd
 github.com/mattn/go-isatty v0.0.7/go.mod h1:Iq45c/XA43vh69/j3iqttzPXn0bhXyGjM0Hdxcsrc5s=
 github.com/mattn/go-isatty v0.0.12/go.mod h1:cbi8OIDigv2wuxKPP5vlRcQ1OAZbq2CE4Kysco4FUpU=
 github.com/mattn/go-isatty v0.0.14/go.mod h1:7GGIvUiUoEMVVmxf/4nioHXj79iQHKdU27kJ6hsGG94=
-github.com/mattn/go-isatty v0.0.18 h1:DOKFKCQ7FNG2L1rbrmstDN4QVRdS89Nkh85u68Uwp98=
-github.com/mattn/go-isatty v0.0.18/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
+github.com/mattn/go-isatty v0.0.19 h1:JITubQf0MOLdlGRuRq+jtsDlekdYPia9ZFsB8h/APPA=
+github.com/mattn/go-isatty v0.0.19/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
 github.com/mattn/go-sqlite3 v2.0.3+incompatible h1:gXHsfypPkaMZrKbD5209QV9jbUTJKjyR5WD3HYQSd+U=
 github.com/microcosm-cc/bluemonday v1.0.24 h1:NGQoPtwGVcbGkKfvyYk1yRqknzBuoMiUrO6R7uFTPlw=
 github.com/microcosm-cc/bluemonday v1.0.24/go.mod h1:ArQySAMps0790cHSkdPEJ7bGkF2VePWH773hsJNSHf8=
@ -476,8 +476,8 @@ github.com/opencontainers/runtime-spec v1.0.2/go.mod h1:jwyrGlmzljRJv/Fgzds9SsS/
 github.com/orcaman/writerseeker v0.0.0-20200621085525-1d3f536ff85e h1:s2RNOM/IGdY0Y6qfTeUKhDawdHDpK9RGBdx80qN4Ttw=
 github.com/orcaman/writerseeker v0.0.0-20200621085525-1d3f536ff85e/go.mod h1:nBdnFKj15wFbf94Rwfq4m30eAcyY9V/IyKAGQFtqkW0=
 github.com/pelletier/go-toml/v2 v2.0.1/go.mod h1:r9LEWfGN8R5k0VXJ+0BkIe7MYkRdwZOjgMj2KwnJFUo=
-github.com/pelletier/go-toml/v2 v2.0.6 h1:nrzqCb7j9cDFj2coyLNLaZuJTLjWjlaz6nvTvIwycIU=
-github.com/pelletier/go-toml/v2 v2.0.6/go.mod h1:eumQOmlWiOPt5WriQQqoM5y18pDHwha2N+QD+EUNTek=
+github.com/pelletier/go-toml/v2 v2.0.8 h1:0ctb6s9mE31h0/lhu+J6OPmVeDxJn+kYnJc2jZR9tGQ=
+github.com/pelletier/go-toml/v2 v2.0.8/go.mod h1:vuYfssBdrU2XDZ9bYydBu6t+6a6PYNcZljzZR9VXg+4=
 github.com/pkg/diff v0.0.0-20210226163009-20ebb0f2a09e/go.mod h1:pJLUxLENpZxwdsKMEsNbx1VGcRFpLqf3715MtcvvzbA=
 github.com/pkg/errors v0.8.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4=
@ -588,8 +588,8 @@ github.com/twitchyliquid64/golang-asm v0.15.1 h1:SU5vSMR7hnwNxj24w34ZyCi/FmDZTkS
 github.com/twitchyliquid64/golang-asm v0.15.1/go.mod h1:a1lVb/DtPvCB8fslRZhAngC2+aY1QWCk3Cedj/Gdt08=
 github.com/ugorji/go v1.2.7/go.mod h1:nF9osbDWLy6bDVv/Rtoh6QgnvNDpmCalQV5urGCCS6M=
 github.com/ugorji/go/codec v1.2.7/go.mod h1:WGN1fab3R1fzQlVQTkfxVtIBhWDRqOviHU95kRgeqEY=
-github.com/ugorji/go/codec v1.2.9 h1:rmenucSohSTiyL09Y+l2OCk+FrMxGMzho2+tjr5ticU=
-github.com/ugorji/go/codec v1.2.9/go.mod h1:UNopzCgEMSXjBc6AOMqYvWC1ktqTAfzJZUZgYf6w6lg=
+github.com/ugorji/go/codec v1.2.11 h1:BMaWp1Bb6fHwEtbplGBGJ498wD+LKlNSl25MjdZY4dU=
+github.com/ugorji/go/codec v1.2.11/go.mod h1:UNopzCgEMSXjBc6AOMqYvWC1ktqTAfzJZUZgYf6w6lg=
 github.com/ulule/limiter/v3 v3.11.2 h1:P4yOrxoEMJbOTfRJR2OzjL90oflzYPPmWg+dvwN2tHA=
 github.com/ulule/limiter/v3 v3.11.2/go.mod h1:QG5GnFOCV+k7lrL5Y8kgEeeflPH3+Cviqlqa8SVSQxI=
 github.com/uptrace/bun v1.1.13 h1:IrxlIJHzCHFwmIzx66A9vi6qx8rHsHFiiT9LqlafHZw=
@ -671,8 +671,9 @@ go.uber.org/goleak v1.2.1 h1:NBol2c7O1ZokfZ0LEU9K6Whx/KnwvepVetCUhtKja4A=
 go.uber.org/multierr v1.1.0/go.mod h1:wR5kodmAFQ0UK8QlbwjlSNy0Z68gJhDJUG5sjR94q/0=
 go.uber.org/zap v1.9.1/go.mod h1:vwi/ZaCAaUcBkycHslxD9B2zi4UTXhF60s6SWpuDF0Q=
 go.uber.org/zap v1.10.0/go.mod h1:vwi/ZaCAaUcBkycHslxD9B2zi4UTXhF60s6SWpuDF0Q=
-golang.org/x/arch v0.0.0-20210923205945-b76863e36670 h1:18EFjUmQOcUvxNYSkA6jO9VAiXCnxFY6NyDX0bHDmkU=
 golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
+golang.org/x/arch v0.3.0 h1:02VY4/ZcO/gBOH6PUaoiptASxtXU10jazRCP865E97k=
+golang.org/x/arch v0.3.0/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
 golang.org/x/crypto v0.0.0-20190411191339-88737f569e3a/go.mod h1:WFFai1msRO1wXaEeE5yQxYXgSfI8pQAWXbQop6sCtWE=
 golang.org/x/crypto v0.0.0-20190510104115-cbcb75029529/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
--- a/vendor/github.com/bytedance/sonic/INTRODUCTION.md
+++ b/vendor/github.com/bytedance/sonic/INTRODUCTION.md
@ -1,48 +0,0 @@
-# Introduction to Sonic
-## Background
-According to the overall profiling of production services in Bytedance, we found that the overhead of JSON serialization and deserialization is unexpectedly high: the total is near to 10% CPU, and the extreme one accounts for more than 40% CPU. Therefore, **the performance of JSON lib is a key issue for the promotion of machine utilization**.
-
-## Research
-We conducted a series of surveys and benchmarks on open-sourced JSON libraries for Golang, but the result is disappointing: **no silver bullet**. First of all, no one can perform at least the top three across various business scenarios. Even the most widely used [json-iterator](https://github.com/json-iterator/go) will severely degrade in generic (no-schema) or big-volume JSON serialization and deserialization. Secondly, compared with other JSON libraries writing in other languages, their speed is generally much slower. For example, [Simdjson-go](https://github.com/minio/simdjson-go) has a 50% reduction in decoding performance compared to [simdjson](https://github.com/simdjson/simdjson). What's more, we barely found JSON libraries which provide API to modify the underlying values.
-
-Therefore, we decided to **develop a brand-new JSON library with high performance as well as wide applicability**.
-
-## Thinking
-Before starting our design, we need to figure out some questions:
-
-### Why is Json-iterator faster than Standard Library?
-First of all, the **schema-based processing mechanism** used by the standard library is commendable, in which the parser can obtain meta information in advance when scanning, thereby shortening the time of branch selection. However, its original implementation did not make good use of this mechanism, instead, **it spent a lot of time reflecting to obtain meta info of schema**. Meanwhile, The approach of json-iterator is: Interprete structure as field-by-field encoding and decoding functions, and then assembled and cached them, minimizing the performance loss cost by reflection. But does it work once and for all? No. In practical tests, we found that **the deeper and larger the input JSON got, the smaller the gap between json-iterator and other libraries gradually became** - eventually event got surpassed:
-![Scalability](introduction-1.png) 
-
-The reason is that **this implementation transforms into a large number of interface encapsulations and function calls**, followed by function-call losses:
-1. **Calling interface involves dynamic addressing of itab**
-2. **Assembly functions cannot be inlined**, while Golang's function-call performance is poor (no parameter-passing-by-register)
-
-#### Is there a way to avoid the function-call overhead of dynamic assembly?
-The first thing we thought about was code generation like [easyjson](https://github.com/mailru/easyjson). But it comes with **schema dependency and convenience losses**. To achieve a real drop-in replacement of the standard library, we turned to another technology - **[JIT](https://en.wikipedia.org/wiki/Jit) (just-in-time compiling)**. Because the compiled codec function is an integrated function, which can greatly reduce function calls while ensuring flexibility.
-
-### Why is Simdjson-go not fast enough?
-[SIMD](https://en.wikipedia.org/wiki/SIMD) (Single-Instruction-Multi-Data) is a special set of CPU instructions for the parallel processing of vectorized data. At present, it is supported by most CPUs and widely used in image processing and big data computing. Undoubtedly, SIMD is useful in JSON processing (itoa, char-search， and so on are all suitable scenarios). We can see that simdjson-go is very competitive in large JSON scenarios (>100KB). However, for some extremely small or irregular character strings, **the extra load operation required by SIMD will lead to performance degradation**. Therefore, we need to dedicate ourselves to branch predicting and decide which scenarios should use SIMD and which should not (for example, the string length is less than 16 bytes).
-
-The second problem comes from the Go compiler itself. In order to ensure the compilation speed, **Golang does very little optimization work during the compilation phase** and cannot directly use compiler backends such as [LLVM](https://en.wikipedia.org/wiki/LLVM) (Low-Level Virtual Machine) for optimization.
-
-So, **can some crucial calculation functions be written in another language with higher execution efficiency**?
-C/Clang is an ideal compilation tool (internal integration LLVM). But the key is how to embed the optimized assembly into Golang.
-
-### How to use Gjson well?
-We also found that [gjson](https://github.com/tidwall/gjson) has a huge advantage in single-key lookup scenarios. This is because its lookup is implemented by a **lazy-load mechanism**, which subtlely skips passing-by values and effectively reduces a lot of unnecessary parsing. Practical application has proved that making good use of this feature in product can indeed bring benefits. But when it comes to multi-key lookup, Gjson does worse event than std,  which is a side effect of its skipping mechanism - **searching for the same path leads to repeated parsing** (skip is also a lightweight parsing). Therefore, the accurate adaptation of practical scenarios is the key.
-
-## Design
-Based on the above questions, our design is easy to implement:
-
-1. Aiming at the function-call overhead cost by the codec dynamic-assembly, **`JIT` tech is used to assemble opcodes (asm) corresponding to the schema at runtime**, which is finally cached into the off-heap memory in the form of Golang functions.
-2. For practical scenarios where big data and small data coexist, we **use pre-conditional judgment** (string size, floating precision, etc.) **to combine `SIMD` with scalar instructions** to achieve the best adaptation.
-3. As for insufficiency in compiling optimization of go language, we decided to **use `C/Clang` to write and compile core computational functions**, and **developed a set of [asm2asm](https://github.com/chenzhuoyu/asm2asm) tools to translate the fully optimized x86 assembly into plan9** and finally load it into Golang runtime.
-4. Giving the big speed gap between parsing and skipping, the **`lazy-load` mechanism** is certainly used in our AST parser, but in **a more adaptive and efficient way to reduce the overhead of multiple-key queries**.
-![design](introduction-2.png)
-
-In detail,  we conducted some further optimization:
-1. Since the native-asm functions cannot be inlined in Golang, we found that its cost even exceeded the improvement brought by the optimization of the C compiler. So we reimplemented a set of lightweight function-calls in JIT:
-    - `Global-function-table + static offset` for calling instruction
-    - **Pass parameters using registers**
-2. `Sync.Map` was used to cache the codecs at first, but for our **quasi-static** (read far more than write), **fewer elements** (usually no more than a few dozen) scenarios, its performance is not optimal, so we reimplement a high-performance and concurrent-safe cache with `open-addressing-hash + RCU` tech.
--- a/vendor/github.com/bytedance/sonic/README.md
+++ b/vendor/github.com/bytedance/sonic/README.md
@ -1,5 +1,7 @@
 # Sonic

+English | [中文](README_ZH_CN.md)
+
 A blazingly fast JSON serializing &amp; deserializing library, accelerated by JIT (just-in-time compiling) and SIMD (single-instruction-multiple-data).

 ## Requirement
@ -76,14 +78,14 @@ BenchmarkSetOne_Parallel_Sjson-16                      18194 ns/op         715.7
 BenchmarkSetOne_Parallel_Jsoniter-16                   33560 ns/op         388.05 MB/s       45892 B/op        964 allocs/op
 ```
 - [Small](https://github.com/bytedance/sonic/blob/main/testdata/small.go) (400B, 11 keys, 3 layers)
-![small benchmarks](bench-small.png)
+![small benchmarks](./docs/imgs/bench-small.png)
 - [Large](https://github.com/bytedance/sonic/blob/main/testdata/twitter.json) (635KB, 10000+ key, 6 layers)
-![large benchmarks](bench-large.png)
+![large benchmarks](./docs/imgs/bench-large.png)

 See [bench.sh](https://github.com/bytedance/sonic/blob/main/bench.sh) for benchmark codes.

 ## How it works
-See [INTRODUCTION.md](INTRODUCTION.md).
+See [INTRODUCTION.md](./docs/INTRODUCTION.md).

 ## Usage

@ -104,28 +106,29 @@ err := sonic.Unmarshal(output, &data)
 Sonic supports decoding json from `io.Reader` or encoding objects into `io.`Writer`, aims at handling multiple values as well as reducing memory consumption.
 - encoder
 ```go
-import "github.com/bytedance/sonic/encoder"
-
 var o1 = map[string]interface{}{
    "a": "b",
 }
 var o2 = 1
 var w = bytes.NewBuffer(nil)
-var enc = encoder.NewStreamEncoder(w)
+var enc = sonic.ConfigDefault.NewEncoder(w)
 enc.Encode(o1)
 enc.Encode(o2)
-println(w.String()) // "{"a":"b"}\n1"
+fmt.Println(w.String())
+// Output:
+// {"a":"b"}
+// 1
 ```
 - decoder
 ```go
-import "github.com/bytedance/sonic/decoder"
-
 var o =  map[string]interface{}{}
 var r = strings.NewReader(`{"a":"b"}{"1":"2"}`)
-var dec = decoder.NewStreamDecoder(r)
+var dec = sonic.ConfigDefault.NewDecoder(r)
 dec.Decode(&o)
 dec.Decode(&o)
-fmt.Printf("%+v", o) // map[1:2 a:b]
+fmt.Printf("%+v", o)
+// Output:
+// map[1:2 a:b]
 ```

 ### Use Number/Use Int64
--- a/vendor/github.com/bytedance/sonic/README_ZH_CN.md
+++ b/vendor/github.com/bytedance/sonic/README_ZH_CN.md
@ -0,0 +1,382 @@
+# Sonic
+
+[English](README.md) | 中文
+
+一个速度奇快的 JSON 序列化/反序列化库，由 JIT （即时编译）和 SIMD （单指令流多数据流）加速。
+
+## 依赖
+
+- Go 1.15~1.20
+- Linux/MacOS/Windows
+- Amd64 架构
+
+## 特色
+
+- 运行时对象绑定，无需代码生成
+- 完备的 JSON 操作 API
+- 快，更快，还要更快！
+
+## 基准测试
+
+对于**所有大小**的 json 和**所有使用场景**， **Sonic 表现均为最佳**。
+- [中型](https://github.com/bytedance/sonic/blob/main/decoder/testdata_test.go#L19) (13kB, 300+ 键, 6 层)
+```powershell
+goversion: 1.17.1
+goos: darwin
+goarch: amd64
+cpu: Intel(R) Core(TM) i9-9880H CPU @ 2.30GHz
+BenchmarkEncoder_Generic_Sonic-16                      32393 ns/op         402.40 MB/s       11965 B/op          4 allocs/op
+BenchmarkEncoder_Generic_Sonic_Fast-16                 21668 ns/op         601.57 MB/s       10940 B/op          4 allocs/op
+BenchmarkEncoder_Generic_JsonIter-16                   42168 ns/op         309.12 MB/s       14345 B/op        115 allocs/op
+BenchmarkEncoder_Generic_GoJson-16                     65189 ns/op         199.96 MB/s       23261 B/op         16 allocs/op
+BenchmarkEncoder_Generic_StdLib-16                    106322 ns/op         122.60 MB/s       49136 B/op        789 allocs/op
+BenchmarkEncoder_Binding_Sonic-16                       6269 ns/op        2079.26 MB/s       14173 B/op          4 allocs/op
+BenchmarkEncoder_Binding_Sonic_Fast-16                  5281 ns/op        2468.16 MB/s       12322 B/op          4 allocs/op
+BenchmarkEncoder_Binding_JsonIter-16                   20056 ns/op         649.93 MB/s        9488 B/op          2 allocs/op
+BenchmarkEncoder_Binding_GoJson-16                      8311 ns/op        1568.32 MB/s        9481 B/op          1 allocs/op
+BenchmarkEncoder_Binding_StdLib-16                     16448 ns/op         792.52 MB/s        9479 B/op          1 allocs/op
+BenchmarkEncoder_Parallel_Generic_Sonic-16              6681 ns/op        1950.93 MB/s       12738 B/op          4 allocs/op
+BenchmarkEncoder_Parallel_Generic_Sonic_Fast-16         4179 ns/op        3118.99 MB/s       10757 B/op          4 allocs/op
+BenchmarkEncoder_Parallel_Generic_JsonIter-16           9861 ns/op        1321.84 MB/s       14362 B/op        115 allocs/op
+BenchmarkEncoder_Parallel_Generic_GoJson-16            18850 ns/op         691.52 MB/s       23278 B/op         16 allocs/op
+BenchmarkEncoder_Parallel_Generic_StdLib-16            45902 ns/op         283.97 MB/s       49174 B/op        789 allocs/op
+BenchmarkEncoder_Parallel_Binding_Sonic-16              1480 ns/op        8810.09 MB/s       13049 B/op          4 allocs/op
+BenchmarkEncoder_Parallel_Binding_Sonic_Fast-16         1209 ns/op        10785.23 MB/s      11546 B/op          4 allocs/op
+BenchmarkEncoder_Parallel_Binding_JsonIter-16           6170 ns/op        2112.58 MB/s        9504 B/op          2 allocs/op
+BenchmarkEncoder_Parallel_Binding_GoJson-16             3321 ns/op        3925.52 MB/s        9496 B/op          1 allocs/op
+BenchmarkEncoder_Parallel_Binding_StdLib-16             3739 ns/op        3486.49 MB/s        9480 B/op          1 allocs/op
+
+BenchmarkDecoder_Generic_Sonic-16                      66812 ns/op         195.10 MB/s       57602 B/op        723 allocs/op
+BenchmarkDecoder_Generic_Sonic_Fast-16                 54523 ns/op         239.07 MB/s       49786 B/op        313 allocs/op
+BenchmarkDecoder_Generic_StdLib-16                    124260 ns/op         104.90 MB/s       50869 B/op        772 allocs/op
+BenchmarkDecoder_Generic_JsonIter-16                   91274 ns/op         142.81 MB/s       55782 B/op       1068 allocs/op
+BenchmarkDecoder_Generic_GoJson-16                     88569 ns/op         147.17 MB/s       66367 B/op        973 allocs/op
+BenchmarkDecoder_Binding_Sonic-16                      32557 ns/op         400.38 MB/s       28302 B/op        137 allocs/op
+BenchmarkDecoder_Binding_Sonic_Fast-16                 28649 ns/op         455.00 MB/s       24999 B/op         34 allocs/op
+BenchmarkDecoder_Binding_StdLib-16                    111437 ns/op         116.97 MB/s       10576 B/op        208 allocs/op
+BenchmarkDecoder_Binding_JsonIter-16                   35090 ns/op         371.48 MB/s       14673 B/op        385 allocs/op
+BenchmarkDecoder_Binding_GoJson-16                     28738 ns/op         453.59 MB/s       22039 B/op         49 allocs/op
+BenchmarkDecoder_Parallel_Generic_Sonic-16             12321 ns/op        1057.91 MB/s       57233 B/op        723 allocs/op
+BenchmarkDecoder_Parallel_Generic_Sonic_Fast-16        10644 ns/op        1224.64 MB/s       49362 B/op        313 allocs/op
+BenchmarkDecoder_Parallel_Generic_StdLib-16            57587 ns/op         226.35 MB/s       50874 B/op        772 allocs/op
+BenchmarkDecoder_Parallel_Generic_JsonIter-16          38666 ns/op         337.12 MB/s       55789 B/op       1068 allocs/op
+BenchmarkDecoder_Parallel_Generic_GoJson-16            30259 ns/op         430.79 MB/s       66370 B/op        974 allocs/op
+BenchmarkDecoder_Parallel_Binding_Sonic-16              5965 ns/op        2185.28 MB/s       27747 B/op        137 allocs/op
+BenchmarkDecoder_Parallel_Binding_Sonic_Fast-16         5170 ns/op        2521.31 MB/s       24715 B/op         34 allocs/op
+BenchmarkDecoder_Parallel_Binding_StdLib-16            27582 ns/op         472.58 MB/s       10576 B/op        208 allocs/op
+BenchmarkDecoder_Parallel_Binding_JsonIter-16          13571 ns/op         960.51 MB/s       14685 B/op        385 allocs/op
+BenchmarkDecoder_Parallel_Binding_GoJson-16            10031 ns/op        1299.51 MB/s       22111 B/op         49 allocs/op
+
+BenchmarkGetOne_Sonic-16                                3276 ns/op        3975.78 MB/s          24 B/op          1 allocs/op
+BenchmarkGetOne_Gjson-16                                9431 ns/op        1380.81 MB/s           0 B/op          0 allocs/op
+BenchmarkGetOne_Jsoniter-16                            51178 ns/op         254.46 MB/s       27936 B/op        647 allocs/op
+BenchmarkGetOne_Parallel_Sonic-16                      216.7 ns/op       60098.95 MB/s          24 B/op          1 allocs/op
+BenchmarkGetOne_Parallel_Gjson-16                       1076 ns/op        12098.62 MB/s          0 B/op          0 allocs/op
+BenchmarkGetOne_Parallel_Jsoniter-16                   17741 ns/op         734.06 MB/s       27945 B/op        647 allocs/op
+BenchmarkSetOne_Sonic-16                               9571 ns/op         1360.61 MB/s        1584 B/op         17 allocs/op
+BenchmarkSetOne_Sjson-16                               36456 ns/op         357.22 MB/s       52180 B/op          9 allocs/op
+BenchmarkSetOne_Jsoniter-16                            79475 ns/op         163.86 MB/s       45862 B/op        964 allocs/op
+BenchmarkSetOne_Parallel_Sonic-16                      850.9 ns/op       15305.31 MB/s        1584 B/op         17 allocs/op
+BenchmarkSetOne_Parallel_Sjson-16                      18194 ns/op         715.77 MB/s       52247 B/op          9 allocs/op
+BenchmarkSetOne_Parallel_Jsoniter-16                   33560 ns/op         388.05 MB/s       45892 B/op        964 allocs/op
+```
+- [小型](https://github.com/bytedance/sonic/blob/main/testdata/small.go) (400B, 11 个键, 3 层)
+![small benchmarks](./docs/imgs/bench-small.png)
+- [大型](https://github.com/bytedance/sonic/blob/main/testdata/twitter.json) (635kB, 10000+ 个键, 6 层)
+![large benchmarks](./docs/imgs/bench-large.png)
+
+要查看基准测试代码，请参阅 [bench.sh](https://github.com/bytedance/sonic/blob/main/bench.sh) 。
+
+## 工作原理
+
+请参阅 [INTRODUCTION_ZH_CN.md](./docs/INTRODUCTION_ZH_CN.md).
+
+## 使用方式
+
+### 序列化/反序列化
+
+默认的行为基本上与 `encoding/json` 相一致，除了 HTML 转义形式（参见 [Escape HTML](https://github.com/bytedance/sonic/blob/main/README.md#escape-html)) 和 `SortKeys` 功能（参见 [Sort Keys](https://github.com/bytedance/sonic/blob/main/README.md#sort-keys)）**没有**遵循 [RFC8259](https://datatracker.ietf.org/doc/html/rfc8259) 。
+ ```go
+import "github.com/bytedance/sonic"
+
+var data YourSchema
+// Marshal
+output, err := sonic.Marshal(&data)
+// Unmarshal
+err := sonic.Unmarshal(output, &data)
+ ```
+
+### 流式输入输出
+
+Sonic 支持解码 `io.Reader` 中输入的 json，或将对象编码为 json 后输出至 `io.Writer`，以处理多个值并减少内存消耗。
+- 编码器
+```go
+var o1 = map[string]interface{}{
+    "a": "b",
+}
+var o2 = 1
+var w = bytes.NewBuffer(nil)
+var enc = sonic.ConfigDefault.NewEncoder(w)
+enc.Encode(o1)
+enc.Encode(o2)
+fmt.Println(w.String())
+// Output:
+// {"a":"b"}
+// 1
+```
+- 解码器
+```go
+var o =  map[string]interface{}{}
+var r = strings.NewReader(`{"a":"b"}{"1":"2"}`)
+var dec = sonic.ConfigDefault.NewDecoder(r)
+dec.Decode(&o)
+dec.Decode(&o)
+fmt.Printf("%+v", o)
+// Output:
+// map[1:2 a:b]
+```
+
+### 使用 `Number` / `int64`
+
+```go
+import "github.com/bytedance/sonic/decoder"
+
+var input = `1`
+var data interface{}
+
+// default float64
+dc := decoder.NewDecoder(input)
+dc.Decode(&data) // data == float64(1)
+// use json.Number
+dc = decoder.NewDecoder(input)
+dc.UseNumber()
+dc.Decode(&data) // data == json.Number("1")
+// use int64
+dc = decoder.NewDecoder(input)
+dc.UseInt64()
+dc.Decode(&data) // data == int64(1)
+
+root, err := sonic.GetFromString(input)
+// Get json.Number
+jn := root.Number()
+jm := root.InterfaceUseNumber().(json.Number) // jn == jm
+// Get float64
+fn := root.Float64()
+fm := root.Interface().(float64) // jn == jm
+ ```
+
+### 对键排序
+
+考虑到排序带来的性能损失（约 10% ）， sonic 默认不会启用这个功能。如果你的组件依赖这个行为（如 [zstd](https://github.com/facebook/zstd)) ，可以仿照下面的例子：
+```go
+import "github.com/bytedance/sonic"
+import "github.com/bytedance/sonic/encoder"
+
+// Binding map only
+m := map[string]interface{}{}
+v, err := encoder.Encode(m, encoder.SortMapKeys)
+
+// Or ast.Node.SortKeys() before marshal
+var root := sonic.Get(JSON)
+err := root.SortKeys()
+```
+
+### HTML 转义
+
+考虑到性能损失（约15%）， sonic 默认不会启用这个功能。你可以使用 `encoder.EscapeHTML` 选项来开启（与 `encoding/json.HTMLEscape` 行为一致）。
+```go
+import "github.com/bytedance/sonic"
+
+v := map[string]string{"&&":"<>"}
+ret, err := Encode(v, EscapeHTML) // ret == `{"\u0026\u0026":{"X":"\u003c\u003e"}}`
+```
+
+### 紧凑格式
+Sonic 默认将基本类型（ `struct` ， `map` 等）编码为紧凑格式的 JSON ，除非使用 `json.RawMessage` or `json.Marshaler` 进行编码： sonic 确保输出的 JSON 合法，但出于性能考虑，**不会**加工成紧凑格式。我们提供选项 `encoder.CompactMarshaler` 来添加此过程，
+
+### 打印错误
+
+如果输入的 JSON 存在无效的语法，sonic 将返回 `decoder.SyntaxError`，该错误支持错误位置的美化输出。
+```go
+import "github.com/bytedance/sonic"
+import "github.com/bytedance/sonic/decoder"
+
+var data interface{}
+err := sonic.UnmarshalString("[[[}]]", &data)
+if err != nil {
+    /* One line by default */
+    println(e.Error()) // "Syntax error at index 3: invalid char\n\n\t[[[}]]\n\t...^..\n"
+    /* Pretty print */
+    if e, ok := err.(decoder.SyntaxError); ok {
+        /*Syntax error at index 3: invalid char
+
+            [[[}]]
+            ...^..
+        */
+        print(e.Description())
+    } else if me, ok := err.(*decoder.MismatchTypeError); ok {
+        // decoder.MismatchTypeError is new to Sonic v1.6.0
+        print(me.Description())
+    }
+}
+```
+
+#### 类型不匹配 [Sonic v1.6.0]
+
+如果给定键中存在**类型不匹配**的值， sonic 会抛出 `decoder.MismatchTypeError` （如果有多个，只会报告最后一个），但仍会跳过错误的值并解码下一个 JSON 。
+```go
+import "github.com/bytedance/sonic"
+import "github.com/bytedance/sonic/decoder"
+
+var data = struct{
+    A int
+    B int
+}{}
+err := UnmarshalString(`{"A":"1","B":1}`, &data)
+println(err.Error())    // Mismatch type int with value string "at index 5: mismatched type with value\n\n\t{\"A\":\"1\",\"B\":1}\n\t.....^.........\n"
+fmt.Printf("%+v", data) // {A:0 B:1}
+```
+### `Ast.Node`
+
+Sonic/ast.Node 是完全独立的 JSON 抽象语法树库。它实现了序列化和反序列化，并提供了获取和修改通用数据的鲁棒的 API。
+
+#### 查找/索引
+
+通过给定的路径搜索 JSON 片段，路径必须为非负整数，字符串或 `nil` 。
+```go
+import "github.com/bytedance/sonic"
+
+input := []byte(`{"key1":[{},{"key2":{"key3":[1,2,3]}}]}`)
+
+// no path, returns entire json
+root, err := sonic.Get(input)
+raw := root.Raw() // == string(input)
+
+// multiple paths
+root, err := sonic.Get(input, "key1", 1, "key2")
+sub := root.Get("key3").Index(2).Int64() // == 3
+```
+**注意**：由于 `Index()` 使用偏移量来定位数据，比使用扫描的 `Get()` 要快的多，建议尽可能的使用 `Index` 。 Sonic 也提供了另一个 API， `IndexOrGet()` ，以偏移量为基础并且也确保键的匹配。
+
+#### 修改
+
+使用 ` Set()` / `Unset()` 修改 json 的内容
+```go
+import "github.com/bytedance/sonic"
+
+// Set
+exist, err := root.Set("key4", NewBool(true)) // exist == false
+alias1 := root.Get("key4")
+println(alias1.Valid()) // true
+alias2 := root.Index(1)
+println(alias1 == alias2) // true
+
+// Unset
+exist, err := root.UnsetByIndex(1) // exist == true
+println(root.Get("key4").Check()) // "value not exist"
+```
+
+#### 序列化
+要将 `ast.Node` 编码为 json ，使用 `MarshalJson()` 或者 `json.Marshal()` （必须传递指向节点的指针）
+```go
+import (
+    "encoding/json"
+    "github.com/bytedance/sonic"
+)
+
+buf, err := root.MarshalJson()
+println(string(buf))                // {"key1":[{},{"key2":{"key3":[1,2,3]}}]}
+exp, err := json.Marshal(&root)     // WARN: use pointer
+println(string(buf) == string(exp)) // true
+```
+
+#### APIs
+- 合法性检查： `Check()`, `Error()`, `Valid()`, `Exist()`
+- 索引： `Index()`, `Get()`, `IndexPair()`, `IndexOrGet()`, `GetByPath()`
+- 转换至 go 内置类型： `Int64()`, `Float64()`, `String()`, `Number()`, `Bool()`, `Map[UseNumber|UseNode]()`, `Array[UseNumber|UseNode]()`, `Interface[UseNumber|UseNode]()`
+- go 类型打包： `NewRaw()`, `NewNumber()`, `NewNull()`, `NewBool()`, `NewString()`, `NewObject()`, `NewArray()`
+- 迭代： `Values()`, `Properties()`, `ForEach()`, `SortKeys()`
+- 修改： `Set()`, `SetByIndex()`, `Add()`
+
+## 兼容性
+由于开发高性能代码的困难性， Sonic **不**保证对所有环境的支持。对于在不同环境中使用 Sonic 构建应用程序的开发者，我们有以下建议：
+
+- 在 **Mac M1** 上开发：确保在您的计算机上安装了 Rosetta 2，并在构建时设置 `GOARCH=amd64` 。 Rosetta 2 可以自动将 x86 二进制文件转换为 arm64 二进制文件，并在 Mac M1 上运行 x86 应用程序。
+- 在 **Linux arm64** 上开发：您可以安装 qemu 并使用 `qemu-x86_64 -cpu max` 命令来将 x86 二进制文件转换为 arm64 二进制文件。qemu可以实现与Mac M1上的Rosetta 2类似的转换效果。
+
+对于希望在不使用 qemu 下使用 sonic 的开发者，或者希望处理 JSON 时与 `encoding/JSON` 严格保持一致的开发者，我们在 `sonic.API` 中提供了一些兼容性 API
+- `ConfigDefault`: 在支持 sonic 的环境下 sonic 的默认配置（`EscapeHTML=false`，`SortKeys=false`等）。行为与具有相应配置的 `encoding/json` 一致，一些选项，如 `SortKeys=false` 将无效。
+- `ConfigStd`: 在支持 sonic 的环境下与标准库兼容的配置（`EscapeHTML=true`，`SortKeys=true`等）。行为与 `encoding/json` 一致。
+- `ConfigFastest`: 在支持 sonic 的环境下运行最快的配置（`NoQuoteTextMarshaler=true`）。行为与具有相应配置的 `encoding/json` 一致，某些选项将无效。
+
+## 注意事项
+
+### 预热
+由于 Sonic 使用 [golang-asm](https://github.com/twitchyliquid64/golang-asm) 作为 JIT 汇编器，这个库并不适用于运行时编译，第一次运行一个大型模式可能会导致请求超时甚至进程内存溢出。为了更好地稳定性，我们建议在运行大型模式或在内存有限的应用中，在使用 `Marshal()/Unmarshal()` 前运行 `Pretouch()`。
+```go
+import (
+    "reflect"
+    "github.com/bytedance/sonic"
+    "github.com/bytedance/sonic/option"
+)
+
+func init() {
+    var v HugeStruct
+
+    // For most large types (nesting depth <= option.DefaultMaxInlineDepth)
+    err := sonic.Pretouch(reflect.TypeOf(v))
+
+    // with more CompileOption...
+    err := sonic.Pretouch(reflect.TypeOf(v), 
+        // If the type is too deep nesting (nesting depth > option.DefaultMaxInlineDepth),
+        // you can set compile recursive loops in Pretouch for better stability in JIT.
+        option.WithCompileRecursiveDepth(loop),
+        // For a large nested struct, try to set a smaller depth to reduce compiling time.
+        option.WithCompileMaxInlineDepth(depth),
+    )
+}
+```
+
+### 拷贝字符串
+
+当解码 **没有转义字符的字符串**时， sonic 会从原始的 JSON 缓冲区内引用而不是复制到新的一个缓冲区中。这对 CPU 的性能方面很有帮助，但是可能因此在解码后对象仍在使用的时候将整个 JSON 缓冲区保留在内存中。实践中我们发现，通过引用 JSON 缓冲区引入的额外内存通常是解码后对象的 20% 至 80% ，一旦应用长期保留这些对象（如缓存以备重用），服务器所使用的内存可能会增加。我们提供了选项 `decoder.CopyString()` 供用户选择，不引用 JSON 缓冲区。这可能在一定程度上降低 CPU 性能。
+
+### 传递字符串还是字节数组？
+为了和 `encoding/json` 保持一致，我们提供了传递 `[]byte` 作为参数的 API ，但考虑到安全性，字符串到字节的复制是同时进行的，这在原始 JSON 非常大时可能会导致性能损失。因此，你可以使用 `UnmarshalString()` 和 `GetFromString()` 来传递字符串，只要你的原始数据是字符串，或**零拷贝类型转换**对于你的字节数组是安全的。我们也提供了 `MarshalString()` 的 API ，以便对编码的 JSON 字节数组进行**零拷贝类型转换**，因为 sonic 输出的字节始终是重复并且唯一的，所以这样是安全的。
+
+### 加速 `encoding.TextMarshaler`
+
+为了保证数据安全性， `sonic.Encoder` 默认会对来自 `encoding.TextMarshaler` 接口的字符串进行引用和转义，如果大部分数据都是这种形式那可能会导致很大的性能损失。我们提供了 `encoder.NoQuoteTextMarshaler` 选项来跳过这些操作，但你**必须**保证他们的输出字符串依照 [RFC8259](https://datatracker.ietf.org/doc/html/rfc8259) 进行了转义和引用。
+
+
+### 泛型的性能优化
+
+在 **完全解析**的场景下， `Unmarshal()` 表现得比 `Get()`+`Node.Interface()` 更好。但是如果你只有特定 JSON 的部分模式，你可以将 `Get()` 和 `Unmarshal()` 结合使用：
+```go
+import "github.com/bytedance/sonic"
+
+node, err := sonic.GetFromString(_TwitterJson, "statuses", 3, "user")
+var user User // your partial schema...
+err = sonic.UnmarshalString(node.Raw(), &user)
+```
+甚至如果你没有任何模式，可以用 `ast.Node` 代替 `map` 或 `interface` 作为泛型的容器：
+```go
+import "github.com/bytedance/sonic"
+
+root, err := sonic.GetFromString(_TwitterJson)
+user := root.GetByPath("statuses", 3, "user")  // === root.Get("status").Index(3).Get("user")
+err = user.Check()
+
+// err = user.LoadAll() // only call this when you want to use 'user' concurrently...
+go someFunc(user)
+```
+为什么？因为 `ast.Node` 使用 `array` 来存储其子节点：
+- 在插入（反序列化）和扫描（序列化）数据时，`Array` 的性能比 `Map` **好得多**；
+- **哈希**（`map[x]`）的效率不如**索引**（`array[x]`）高效，而 `ast.Node` 可以在数组和对象上使用索引；
+- 使用 `Interface()` / `Map()` 意味着 sonic 必须解析所有的底层值，而 `ast.Node` 可以**按需解析**它们。
+
+**注意**：由于 `ast.Node` 的惰性加载设计，其**不能**直接保证并发安全性，但你可以调用 `Node.Load()` / `Node.LoadAll()` 来实现并发安全。尽管可能会带来性能损失，但仍比转换成 `map` 或 `interface{}` 更为高效。
+
+## 社区
+
+Sonic 是 [CloudWeGo](https://www.cloudwego.io/) 下的一个子项目。我们致力于构建云原生生态系统。
--- a/vendor/github.com/bytedance/sonic/api.go
+++ b/vendor/github.com/bytedance/sonic/api.go
@ -169,9 +169,10 @@ func UnmarshalString(buf string, val interface{}) error {
 // and returns its representing ast.Node.
 //
 // Each path arg must be integer or string:
-//     - Integer means searching current node as array
-//     - String means searching current node as object
+//     - Integer is target index(>=0), means searching current node as array.
+//     - String is target key, means searching current node as object.
 //
+// 
 // Note, the api expects the json is well-formed at least,
 // otherwise it may return unexpected result.
 func Get(src []byte, path ...interface{}) (ast.Node, error) {
--- a/vendor/github.com/bytedance/sonic/ast/api_amd64.go
+++ b/vendor/github.com/bytedance/sonic/ast/api_amd64.go
@ -1,5 +1,20 @@
 // +build amd64,go1.15,!go1.21

+/*
+ * Copyright 2022 ByteDance Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */

 package ast

@ -17,10 +32,12 @@

 var typeByte = rt.UnpackEface(byte(0)).Type

+//go:nocheckptr
 func quote(buf *[]byte, val string) {
    *buf = append(*buf, '"')
    if len(val) == 0 {
        *buf = append(*buf, '"')
+        return
    }

    sp := rt.IndexChar(val, 0)
@ -99,7 +116,9 @@ func (self *Parser) skipFast() (int, types.ParsingError) {
 }

 func (self *Parser) getByPath(path ...interface{}) (int, types.ParsingError) {
-    start := native.GetByPath(&self.s, &self.p, &path)
+    fsm := types.NewStateMachine()
+    start := native.GetByPath(&self.s, &self.p, &path, fsm)
+    types.FreeStateMachine(fsm)
    runtime.KeepAlive(path)
    if start < 0 {
        return self.p, types.ParsingError(-start)
@ -107,7 +126,6 @@ func (self *Parser) getByPath(path ...interface{}) (int, types.ParsingError) {
    return start, 0
 }

-
 func (self *Searcher) GetByPath(path ...interface{}) (Node, error) {
    var err types.ParsingError
    var start int
@ -115,6 +133,13 @@ func (self *Searcher) GetByPath(path ...interface{}) (Node, error) {
    self.parser.p = 0
    start, err = self.parser.getByPath(path...)
    if err != 0 {
+        // for compatibility with old version
+        if err == types.ERR_NOT_FOUND {
+            return Node{}, ErrNotExist
+        }
+        if err == types.ERR_UNSUPPORT_TYPE {
+            panic("path must be either int(>=0) or string")
+        }
        return Node{}, self.parser.syntaxError(err)
    }

--- a/vendor/github.com/bytedance/sonic/ast/api_compat.go
+++ b/vendor/github.com/bytedance/sonic/ast/api_compat.go
@ -1,5 +1,21 @@
 // +build !amd64 go1.21

+/*
+ * Copyright 2022 ByteDance Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
 package ast

 import (
@ -24,8 +40,6 @@ func unquote(src string) (string, types.ParsingError) {
    return rt.Mem2Str(out), 0
 }

-
-
 func decodeBase64(src string) ([]byte, error) {
    return base64.StdEncoding.DecodeString(src)
 }
@ -53,7 +67,12 @@ func (self *Parser) skip() (int, types.ParsingError) {
 }

 func (self *Parser) skipFast() (int, types.ParsingError) {
-    return self.skip()
+    e, s := skipValueFast(self.s, self.p)
+    if e < 0 {
+        return self.p, types.ParsingError(-e)
+    }
+    self.p = e
+    return s, 0
 }

 func (self *Node) encodeInterface(buf *[]byte) error {
@ -70,17 +89,16 @@ func (self *Searcher) GetByPath(path ...interface{}) (Node, error) {

    var err types.ParsingError
    for _, p := range path {
-        switch p := p.(type) {
-        case int:
-            if err = self.parser.searchIndex(p); err != 0 {
+        if idx, ok := p.(int); ok && idx >= 0 {
+            if err = self.parser.searchIndex(idx); err != 0 {
                return Node{}, self.parser.ExportError(err)
            }
-        case string:
-            if err = self.parser.searchKey(p); err != 0 {
+        } else if key, ok := p.(string); ok {
+            if err = self.parser.searchKey(key); err != 0 {
                return Node{}, self.parser.ExportError(err)
            }
-        default:
-            panic("path must be either int or string")
+        } else {
+            panic("path must be either int(>=0) or string")
        }
    }

--- a/vendor/github.com/bytedance/sonic/ast/decode.go
+++ b/vendor/github.com/bytedance/sonic/ast/decode.go
@ -1,3 +1,19 @@
+/*
+ * Copyright 2022 ByteDance Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
 package ast

 import (
@ -24,6 +40,7 @@ func isSpace(c byte) bool {
    return (int(1<<c) & _blankCharsMask) != 0
 }

+//go:nocheckptr
 func skipBlank(src string, pos int) int {
    se := uintptr(rt.IndexChar(src, len(src)))
    sp := uintptr(rt.IndexChar(src, pos))
@ -77,6 +94,7 @@ func decodeFalse(src string, pos int) (ret int) {
    return -int(types.ERR_INVALID_CHAR)
 }

+//go:nocheckptr
 func decodeString(src string, pos int) (ret int, v string) {
    ret, ep := skipString(src, pos)
    if ep == -1 {
@ -112,6 +130,7 @@ func isDigit(c byte) bool {
    return c >= '0' && c <= '9'
 }

+//go:nocheckptr
 func decodeInt64(src string, pos int) (ret int, v int64, err error) {
    sp := uintptr(rt.IndexChar(src, pos))
    ss := uintptr(sp)
@ -161,6 +180,7 @@ func isNumberChars(c byte) bool {
    return (c >= '0' && c <= '9') || c == '+' || c == '-' || c == 'e' || c == 'E' || c == '.'
 }

+//go:nocheckptr
 func decodeFloat64(src string, pos int) (ret int, v float64, err error) {
    sp := uintptr(rt.IndexChar(src, pos))
    ss := uintptr(sp)
@ -255,6 +275,7 @@ func decodeValue(src string, pos int) (ret int, v types.JsonState) {
    }
 }

+//go:nocheckptr
 func skipNumber(src string, pos int) (ret int) {
    sp := uintptr(rt.IndexChar(src, pos))
    se := uintptr(rt.IndexChar(src, len(src)))
@ -281,7 +302,7 @@ func skipNumber(src string, pos int) (ret int) {
        } else if nextNeedDigit {
            return -int(types.ERR_INVALID_CHAR)
        } else if c == '.' {
-            if !lastIsDigit || pointer || sp == ss {
+            if !lastIsDigit || pointer || exponent || sp == ss {
                return -int(types.ERR_INVALID_CHAR)
            }
            pointer = true
@ -319,6 +340,7 @@ func skipNumber(src string, pos int) (ret int) {
    return int(uintptr(sp) - uintptr((*rt.GoString)(unsafe.Pointer(&src)).Ptr))
 }

+//go:nocheckptr
 func skipString(src string, pos int) (ret int, ep int) {
    if pos+1 >= len(src) {
        return -int(types.ERR_EOF), -1
@ -327,6 +349,7 @@ func skipString(src string, pos int) (ret int, ep int) {
    sp := uintptr(rt.IndexChar(src, pos))
    se := uintptr(rt.IndexChar(src, len(src)))

+    // not start with quote
    if *(*byte)(unsafe.Pointer(sp)) != '"' {
        return -int(types.ERR_INVALID_CHAR), -1
    }
@ -344,18 +367,16 @@ func skipString(src string, pos int) (ret int, ep int) {
        }
        sp += 1
        if c == '"' {
-            break
+            return int(uintptr(sp) - uintptr((*rt.GoString)(unsafe.Pointer(&src)).Ptr)), ep
        }
    }

-    if sp > se {
-        return -int(types.ERR_EOF), -1
-    }
-
    runtime.KeepAlive(src)
-    return int(uintptr(sp) - uintptr((*rt.GoString)(unsafe.Pointer(&src)).Ptr)), ep
+    // not found the closed quote until EOF
+    return -int(types.ERR_EOF), -1
 }

+//go:nocheckptr
 func skipPair(src string, pos int, lchar byte, rchar byte) (ret int) {
    if pos+1 >= len(src) {
        return -int(types.ERR_EOF)
@ -403,7 +424,7 @@ func skipPair(src string, pos int, lchar byte, rchar byte) (ret int) {
    return int(uintptr(sp) - uintptr((*rt.GoString)(unsafe.Pointer(&src)).Ptr))
 }

-func skipValue(src string, pos int) (ret int, start int) {
+func skipValueFast(src string, pos int) (ret int, start int) {
    pos = skipBlank(src, pos)
    if pos < 0 {
        return pos, -1
@ -428,3 +449,127 @@ func skipValue(src string, pos int) (ret int, start int) {
    }
    return ret, pos
 }
+
+func skipValue(src string, pos int) (ret int, start int) {
+    pos = skipBlank(src, pos)
+    if pos < 0 {
+        return pos, -1
+    }
+    switch c := src[pos]; c {
+    case 'n':
+        ret = decodeNull(src, pos)
+    case '"':
+        ret, _ = skipString(src, pos)
+    case '{':
+        ret, _ = skipObject(src, pos)
+    case '[':
+        ret, _ = skipArray(src, pos)
+    case 't':
+        ret = decodeTrue(src, pos)
+    case 'f':
+        ret = decodeFalse(src, pos)
+    case '-', '+', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9':
+        ret = skipNumber(src, pos)
+    default:
+        ret = -int(types.ERR_INVALID_CHAR)
+    }
+    return ret, pos
+}
+
+func skipObject(src string, pos int) (ret int, start int) {
+    start = skipBlank(src, pos)
+    if start < 0 {
+        return start, -1
+    }
+
+    if src[start] != '{' {
+        return -int(types.ERR_INVALID_CHAR), -1
+    }
+
+    pos = start + 1
+    pos = skipBlank(src, pos)
+    if pos < 0 {
+        return pos, -1
+    }
+    if src[pos] == '}' {
+        return pos + 1, start
+    }
+
+    for {
+        pos, _ = skipString(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+
+        pos = skipBlank(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+        if src[pos] != ':' {
+            return -int(types.ERR_INVALID_CHAR), -1
+        }
+
+        pos++
+        pos, _ = skipValue(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+
+        pos = skipBlank(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+        if src[pos] == '}' {
+            return pos + 1, start
+        }
+        if src[pos] != ',' {
+            return -int(types.ERR_INVALID_CHAR), -1
+        }
+
+        pos++
+        pos = skipBlank(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+
+    }
+}
+
+func skipArray(src string, pos int) (ret int, start int) {
+    start = skipBlank(src, pos)
+    if start < 0 {
+        return start, -1
+    }
+
+    if src[start] != '[' {
+        return -int(types.ERR_INVALID_CHAR), -1
+    }
+
+    pos = start + 1
+    pos = skipBlank(src, pos)
+    if pos < 0 {
+        return pos, -1
+    }
+    if src[pos] == ']' {
+        return pos + 1, start
+    }
+
+    for {
+        pos, _ = skipValue(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+
+        pos = skipBlank(src, pos)
+        if pos < 0 {
+            return pos, -1
+        }
+        if src[pos] == ']' {
+            return pos + 1, start
+        }
+        if src[pos] != ',' {
+            return -int(types.ERR_INVALID_CHAR), -1
+        }
+        pos++
+    }
+}
--- a/vendor/github.com/bytedance/sonic/ast/node.go
+++ b/vendor/github.com/bytedance/sonic/ast/node.go
@ -1541,13 +1541,19 @@ func (self *Node) toGenericObjectUseNode() (map[string]Node, error) {
    emptyObjectNode = Node{t: types.V_OBJECT}
 )

-// NewRaw creates a node of raw json, and decides its type by first char.
+// NewRaw creates a node of raw json.
+// If the input json is invalid, NewRaw returns a error Node.
 func NewRaw(json string) Node {
-    if json == "" {
-        panic("empty json string")
+    parser := NewParser(json)
+    start, err := parser.skip()
+    if err != 0 {
+        return *newError(err, err.Message()) 
    }
-    it := switchRawType(json[0])
-    return newRawNode(json, it)
+    it := switchRawType(parser.s[start])
+    if it == _V_NONE {
+        return Node{}
+    }
+    return newRawNode(parser.s[start:parser.p], it)
 }

 // NewAny creates a node of type V_ANY if any's type isn't Node or *Node, 
--- a/vendor/github.com/bytedance/sonic/ast/parser.go
+++ b/vendor/github.com/bytedance/sonic/ast/parser.go
@ -350,7 +350,7 @@ func (self *Parser) searchKey(match string) types.ParsingError {

        /* skip value */
        if key != match {
-            if _, err = self.skip(); err != 0 {
+            if _, err = self.skipFast(); err != 0 {
                return err
            }
        } else {
@ -398,7 +398,7 @@ func (self *Parser) searchIndex(idx int) types.ParsingError {
    for i := 0; i < idx; i++ {

        /* decode the value */
-        if _, err = self.skip(); err != 0 {
+        if _, err = self.skipFast(); err != 0 {
            return err
        }

--- a/vendor/github.com/bytedance/sonic/bench-large.png
+++ b/vendor/github.com/bytedance/sonic/bench-large.png
--- a/vendor/github.com/bytedance/sonic/bench-small.png
+++ b/vendor/github.com/bytedance/sonic/bench-small.png
--- a/vendor/github.com/bytedance/sonic/decoder/decoder_amd64.go
+++ b/vendor/github.com/bytedance/sonic/decoder/decoder_amd64.go
@ -0,0 +1,66 @@
+// +build amd64,go1.15,!go1.21
+
+/*
+* Copyright 2023 ByteDance Inc.
+*
+* Licensed under the Apache License, Version 2.0 (the "License");
+* you may not use this file except in compliance with the License.
+* You may obtain a copy of the License at
+*
+*     http://www.apache.org/licenses/LICENSE-2.0
+*
+* Unless required by applicable law or agreed to in writing, software
+* distributed under the License is distributed on an "AS IS" BASIS,
+* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+* See the License for the specific language governing permissions and
+* limitations under the License.
+*/
+
+package decoder
+
+import (
+    `github.com/bytedance/sonic/internal/decoder`
+)
+
+// Decoder is the decoder context object
+type Decoder = decoder.Decoder
+
+type MismatchTypeError = decoder.MismatchTypeError
+
+// Options for decode.
+type Options = decoder.Options
+
+const (
+    OptionUseInt64         Options = decoder.OptionUseInt64
+    OptionUseNumber        Options = decoder.OptionUseNumber
+    OptionUseUnicodeErrors Options = decoder.OptionUseUnicodeErrors
+    OptionDisableUnknown   Options = decoder.OptionDisableUnknown
+    OptionCopyString       Options = decoder.OptionCopyString
+    OptionValidateString   Options = decoder.OptionValidateString
+)
+
+// StreamDecoder is the decoder context object for streaming input.
+type StreamDecoder = decoder.StreamDecoder
+
+type SyntaxError = decoder.SyntaxError
+
+var (
+    // NewDecoder creates a new decoder instance.
+    NewDecoder = decoder.NewDecoder
+
+    // NewStreamDecoder adapts to encoding/json.NewDecoder API.
+    //
+    // NewStreamDecoder returns a new decoder that reads from r.
+    NewStreamDecoder = decoder.NewStreamDecoder
+
+    // Pretouch compiles vt ahead-of-time to avoid JIT compilation on-the-fly, in
+    // order to reduce the first-hit latency.
+    //
+    // Opts are the compile options, for example, "option.WithCompileRecursiveDepth" is
+    // a compile option to set the depth of recursive compile for the nested struct type.
+    Pretouch = decoder.Pretouch
+    
+    // Skip skips only one json value, and returns first non-blank character position and its ending position if it is valid.
+    // Otherwise, returns negative error code using start and invalid character position using end
+    Skip = decoder.Skip
+)
--- a/vendor/github.com/bytedance/sonic/decoder/decoder_compat.go
+++ b/vendor/github.com/bytedance/sonic/decoder/decoder_compat.go
@ -0,0 +1,196 @@
+// +build !amd64 go1.21
+
+/*
+* Copyright 2023 ByteDance Inc.
+*
+* Licensed under the Apache License, Version 2.0 (the "License");
+* you may not use this file except in compliance with the License.
+* You may obtain a copy of the License at
+*
+*     http://www.apache.org/licenses/LICENSE-2.0
+*
+* Unless required by applicable law or agreed to in writing, software
+* distributed under the License is distributed on an "AS IS" BASIS,
+* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+* See the License for the specific language governing permissions and
+* limitations under the License.
+*/
+
+package decoder
+
+import (
+     `encoding/json`
+     `bytes`
+     `reflect`
+     `github.com/bytedance/sonic/internal/native/types`
+     `github.com/bytedance/sonic/option`
+     `io`
+)
+
+const (
+     _F_use_int64 = iota
+     _F_use_number
+     _F_disable_urc
+     _F_disable_unknown
+     _F_copy_string
+     _F_validate_string
+
+     _F_allow_control = 31
+)
+
+type Options uint64
+
+const (
+     OptionUseInt64         Options = 1 << _F_use_int64
+     OptionUseNumber        Options = 1 << _F_use_number
+     OptionUseUnicodeErrors Options = 1 << _F_disable_urc
+     OptionDisableUnknown   Options = 1 << _F_disable_unknown
+     OptionCopyString       Options = 1 << _F_copy_string
+     OptionValidateString   Options = 1 << _F_validate_string
+)
+
+func (self *Decoder) SetOptions(opts Options) {
+     if (opts & OptionUseNumber != 0) && (opts & OptionUseInt64 != 0) {
+         panic("can't set OptionUseInt64 and OptionUseNumber both!")
+     }
+     self.f = uint64(opts)
+}
+
+
+// Decoder is the decoder context object
+type Decoder struct {
+     i int
+     f uint64
+     s string
+}
+
+// NewDecoder creates a new decoder instance.
+func NewDecoder(s string) *Decoder {
+     return &Decoder{s: s}
+}
+
+// Pos returns the current decoding position.
+func (self *Decoder) Pos() int {
+     return self.i
+}
+
+func (self *Decoder) Reset(s string) {
+     self.s = s
+     self.i = 0
+     // self.f = 0
+}
+
+// NOTE: api fallback do nothing
+func (self *Decoder) CheckTrailings() error {
+     pos := self.i
+     buf := self.s
+     /* skip all the trailing spaces */
+     if pos != len(buf) {
+         for pos < len(buf) && (types.SPACE_MASK & (1 << buf[pos])) != 0 {
+             pos++
+         }
+     }
+
+     /* then it must be at EOF */
+     if pos == len(buf) {
+         return nil
+     }
+
+     /* junk after JSON value */
+     return nil
+}
+
+
+// Decode parses the JSON-encoded data from current position and stores the result
+// in the value pointed to by val.
+func (self *Decoder) Decode(val interface{}) error {
+    r := bytes.NewBufferString(self.s)
+   dec := json.NewDecoder(r)
+   if (self.f | uint64(OptionUseNumber)) != 0  {
+       dec.UseNumber()
+   }
+   if (self.f | uint64(OptionDisableUnknown)) != 0  {
+       dec.DisallowUnknownFields()
+   }
+   return dec.Decode(val)
+}
+
+// UseInt64 indicates the Decoder to unmarshal an integer into an interface{} as an
+// int64 instead of as a float64.
+func (self *Decoder) UseInt64() {
+     self.f  |= 1 << _F_use_int64
+     self.f &^= 1 << _F_use_number
+}
+
+// UseNumber indicates the Decoder to unmarshal a number into an interface{} as a
+// json.Number instead of as a float64.
+func (self *Decoder) UseNumber() {
+     self.f &^= 1 << _F_use_int64
+     self.f  |= 1 << _F_use_number
+}
+
+// UseUnicodeErrors indicates the Decoder to return an error when encounter invalid
+// UTF-8 escape sequences.
+func (self *Decoder) UseUnicodeErrors() {
+     self.f |= 1 << _F_disable_urc
+}
+
+// DisallowUnknownFields indicates the Decoder to return an error when the destination
+// is a struct and the input contains object keys which do not match any
+// non-ignored, exported fields in the destination.
+func (self *Decoder) DisallowUnknownFields() {
+     self.f |= 1 << _F_disable_unknown
+}
+
+// CopyString indicates the Decoder to decode string values by copying instead of referring.
+func (self *Decoder) CopyString() {
+     self.f |= 1 << _F_copy_string
+}
+
+// ValidateString causes the Decoder to validate string values when decoding string value 
+// in JSON. Validation is that, returning error when unescaped control chars(0x00-0x1f) or
+// invalid UTF-8 chars in the string value of JSON.
+func (self *Decoder) ValidateString() {
+     self.f |= 1 << _F_validate_string
+}
+
+// Pretouch compiles vt ahead-of-time to avoid JIT compilation on-the-fly, in
+// order to reduce the first-hit latency.
+//
+// Opts are the compile options, for example, "option.WithCompileRecursiveDepth" is
+// a compile option to set the depth of recursive compile for the nested struct type.
+func Pretouch(vt reflect.Type, opts ...option.CompileOption) error {
+     return nil
+}
+
+type StreamDecoder struct {
+   r       io.Reader
+   buf     []byte
+   scanp   int
+   scanned int64
+   err     error
+   Decoder
+}
+
+// NewStreamDecoder adapts to encoding/json.NewDecoder API.
+//
+// NewStreamDecoder returns a new decoder that reads from r.
+func NewStreamDecoder(r io.Reader) *StreamDecoder {
+   return &StreamDecoder{r : r}
+}
+
+// Decode decodes input stream into val with corresponding data. 
+// Redundantly bytes may be read and left in its buffer, and can be used at next call.
+// Either io error from underlying io.Reader (except io.EOF) 
+// or syntax error from data will be recorded and stop subsequently decoding.
+func (self *StreamDecoder) Decode(val interface{}) (err error) {
+   dec := json.NewDecoder(self.r)
+   if (self.f | uint64(OptionUseNumber)) != 0  {
+       dec.UseNumber()
+   }
+   if (self.f | uint64(OptionDisableUnknown)) != 0  {
+       dec.DisallowUnknownFields()
+   }
+   return dec.Decode(val)
+}
+
--- a/vendor/github.com/bytedance/sonic/encoder/encoder_amd64.go
+++ b/vendor/github.com/bytedance/sonic/encoder/encoder_amd64.go
@ -0,0 +1,108 @@
+// +build amd64,go1.15,!go1.21
+
+/*
+ * Copyright 2023 ByteDance Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package encoder
+
+import (
+    `github.com/bytedance/sonic/internal/encoder`
+)
+
+
+// Encoder represents a specific set of encoder configurations.
+type Encoder = encoder.Encoder
+
+// StreamEncoder uses io.Writer as input.
+type StreamEncoder = encoder.StreamEncoder
+
+// Options is a set of encoding options.
+type Options = encoder.Options
+
+const (
+    // SortMapKeys indicates that the keys of a map needs to be sorted
+    // before serializing into JSON.
+    // WARNING: This hurts performance A LOT, USE WITH CARE.
+    SortMapKeys Options = encoder.SortMapKeys
+
+    // EscapeHTML indicates encoder to escape all HTML characters
+    // after serializing into JSON (see https://pkg.go.dev/encoding/json#HTMLEscape).
+    // WARNING: This hurts performance A LOT, USE WITH CARE.
+    EscapeHTML Options = encoder.EscapeHTML
+
+    // CompactMarshaler indicates that the output JSON from json.Marshaler
+    // is always compact and needs no validation
+    CompactMarshaler Options = encoder.CompactMarshaler
+
+    // NoQuoteTextMarshaler indicates that the output text from encoding.TextMarshaler
+    // is always escaped string and needs no quoting
+    NoQuoteTextMarshaler Options = encoder.NoQuoteTextMarshaler
+
+    // NoNullSliceOrMap indicates all empty Array or Object are encoded as '[]' or '{}',
+    // instead of 'null'
+    NoNullSliceOrMap Options = encoder.NoNullSliceOrMap
+
+    // ValidateString indicates that encoder should validate the input string
+    // before encoding it into JSON.
+    ValidateString Options = encoder.ValidateString
+
+    // CompatibleWithStd is used to be compatible with std encoder.
+    CompatibleWithStd Options = encoder.CompatibleWithStd
+)
+
+
+var (
+    // Encode returns the JSON encoding of val, encoded with opts.
+    Encode = encoder.Encode
+
+    // EncodeInto is like Encode but uses a user-supplied buffer instead of allocating a new one.
+    EncodeIndented = encoder.EncodeIndented
+
+    // EncodeIndented is like Encode but applies Indent to format the output.
+    // Each JSON element in the output will begin on a new line beginning with prefix
+    // followed by one or more copies of indent according to the indentation nesting.
+    EncodeInto = encoder.EncodeInto
+
+    // HTMLEscape appends to dst the JSON-encoded src with <, >, &, U+2028 and U+2029
+    // characters inside string literals changed to \u003c, \u003e, \u0026, \u2028, \u2029
+    // so that the JSON will be safe to embed inside HTML <script> tags.
+    // For historical reasons, web browsers don't honor standard HTML
+    // escaping within <script> tags, so an alternative JSON encoding must
+    // be used.
+    HTMLEscape = encoder.HTMLEscape
+
+    // Pretouch compiles vt ahead-of-time to avoid JIT compilation on-the-fly, in
+    // order to reduce the first-hit latency.
+    //
+    // Opts are the compile options, for example, "option.WithCompileRecursiveDepth" is
+    // a compile option to set the depth of recursive compile for the nested struct type.
+    Pretouch = encoder.Pretouch
+
+    // Quote returns the JSON-quoted version of s.
+    Quote = encoder.Quote
+
+    // Valid validates json and returns first non-blank character position,
+    // if it is only one valid json value.
+    // Otherwise returns invalid character position using start.
+    //
+    // Note: it does not check for the invalid UTF-8 characters.
+    Valid = encoder.Valid
+
+    // NewStreamEncoder adapts to encoding/json.NewDecoder API.
+    //
+    // NewStreamEncoder returns a new encoder that write to w.
+    NewStreamEncoder = encoder.NewStreamEncoder
+)
--- a/vendor/github.com/bytedance/sonic/encoder/encoder_compat.go
+++ b/vendor/github.com/bytedance/sonic/encoder/encoder_compat.go
@ -0,0 +1,234 @@
+// +build !amd64 go1.21
+
+/*
+* Copyright 2023 ByteDance Inc.
+*
+* Licensed under the Apache License, Version 2.0 (the "License");
+* you may not use this file except in compliance with the License.
+* You may obtain a copy of the License at
+*
+*     http://www.apache.org/licenses/LICENSE-2.0
+*
+* Unless required by applicable law or agreed to in writing, software
+* distributed under the License is distributed on an "AS IS" BASIS,
+* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+* See the License for the specific language governing permissions and
+* limitations under the License.
+*/
+
+package encoder
+
+import (
+   `io`
+    `bytes`
+    `encoding/json`
+    `reflect`
+
+    `github.com/bytedance/sonic/option`
+)
+
+// Options is a set of encoding options.
+type Options uint64
+
+const (
+    bitSortMapKeys          = iota
+    bitEscapeHTML          
+    bitCompactMarshaler
+    bitNoQuoteTextMarshaler
+    bitNoNullSliceOrMap
+    bitValidateString
+
+    // used for recursive compile
+    bitPointerValue = 63
+)
+
+const (
+    // SortMapKeys indicates that the keys of a map needs to be sorted 
+    // before serializing into JSON.
+    // WARNING: This hurts performance A LOT, USE WITH CARE.
+    SortMapKeys          Options = 1 << bitSortMapKeys
+
+    // EscapeHTML indicates encoder to escape all HTML characters 
+    // after serializing into JSON (see https://pkg.go.dev/encoding/json#HTMLEscape).
+    // WARNING: This hurts performance A LOT, USE WITH CARE.
+    EscapeHTML           Options = 1 << bitEscapeHTML
+
+    // CompactMarshaler indicates that the output JSON from json.Marshaler 
+    // is always compact and needs no validation 
+    CompactMarshaler     Options = 1 << bitCompactMarshaler
+
+    // NoQuoteTextMarshaler indicates that the output text from encoding.TextMarshaler 
+    // is always escaped string and needs no quoting
+    NoQuoteTextMarshaler Options = 1 << bitNoQuoteTextMarshaler
+
+    // NoNullSliceOrMap indicates all empty Array or Object are encoded as '[]' or '{}',
+    // instead of 'null'
+    NoNullSliceOrMap     Options = 1 << bitNoNullSliceOrMap
+
+    // ValidateString indicates that encoder should validate the input string
+    // before encoding it into JSON.
+    ValidateString       Options = 1 << bitValidateString
+  
+    // CompatibleWithStd is used to be compatible with std encoder.
+    CompatibleWithStd Options = SortMapKeys | EscapeHTML | CompactMarshaler
+)
+
+// Encoder represents a specific set of encoder configurations.
+type Encoder struct {
+    Opts Options
+    prefix string
+    indent string
+}
+
+// Encode returns the JSON encoding of v.
+func (self *Encoder) Encode(v interface{}) ([]byte, error) {
+    if self.indent != "" || self.prefix != "" { 
+        return EncodeIndented(v, self.prefix, self.indent, self.Opts)
+    }
+    return Encode(v, self.Opts)
+}
+
+// SortKeys enables the SortMapKeys option.
+func (self *Encoder) SortKeys() *Encoder {
+    self.Opts |= SortMapKeys
+    return self
+}
+
+// SetEscapeHTML specifies if option EscapeHTML opens
+func (self *Encoder) SetEscapeHTML(f bool) {
+    if f {
+        self.Opts |= EscapeHTML
+    } else {
+        self.Opts &= ^EscapeHTML
+    }
+}
+
+// SetValidateString specifies if option ValidateString opens
+func (self *Encoder) SetValidateString(f bool) {
+    if f {
+        self.Opts |= ValidateString
+    } else {
+        self.Opts &= ^ValidateString
+    }
+}
+
+// SetCompactMarshaler specifies if option CompactMarshaler opens
+func (self *Encoder) SetCompactMarshaler(f bool) {
+    if f {
+        self.Opts |= CompactMarshaler
+    } else {
+        self.Opts &= ^CompactMarshaler
+    }
+}
+
+// SetNoQuoteTextMarshaler specifies if option NoQuoteTextMarshaler opens
+func (self *Encoder) SetNoQuoteTextMarshaler(f bool) {
+    if f {
+        self.Opts |= NoQuoteTextMarshaler
+    } else {
+        self.Opts &= ^NoQuoteTextMarshaler
+    }
+}
+
+// SetIndent instructs the encoder to format each subsequent encoded
+// value as if indented by the package-level function EncodeIndent().
+// Calling SetIndent("", "") disables indentation.
+func (enc *Encoder) SetIndent(prefix, indent string) {
+    enc.prefix = prefix
+    enc.indent = indent
+}
+
+// Quote returns the JSON-quoted version of s.
+func Quote(s string) string {
+    /* check for empty string */
+    if s == "" {
+        return `""`
+    }
+
+    out, _ := json.Marshal(s)
+    return string(out)
+}
+
+// Encode returns the JSON encoding of val, encoded with opts.
+func Encode(val interface{}, opts Options) ([]byte, error) {
+   return json.Marshal(val)
+}
+
+// EncodeInto is like Encode but uses a user-supplied buffer instead of allocating
+// a new one.
+func EncodeInto(buf *[]byte, val interface{}, opts Options) error {
+   if buf == nil {
+       panic("user-supplied buffer buf is nil")
+   }
+   w := bytes.NewBuffer(*buf)
+   enc := json.NewEncoder(w)
+   enc.SetEscapeHTML((opts & EscapeHTML) != 0)
+   err := enc.Encode(val)
+   *buf = w.Bytes()
+   return err
+}
+
+// HTMLEscape appends to dst the JSON-encoded src with <, >, &, U+2028 and U+2029
+// characters inside string literals changed to \u003c, \u003e, \u0026, \u2028, \u2029
+// so that the JSON will be safe to embed inside HTML <script> tags.
+// For historical reasons, web browsers don't honor standard HTML
+// escaping within <script> tags, so an alternative JSON encoding must
+// be used.
+func HTMLEscape(dst []byte, src []byte) []byte {
+   d := bytes.NewBuffer(dst)
+   json.HTMLEscape(d, src)
+   return d.Bytes()
+}
+
+// EncodeIndented is like Encode but applies Indent to format the output.
+// Each JSON element in the output will begin on a new line beginning with prefix
+// followed by one or more copies of indent according to the indentation nesting.
+func EncodeIndented(val interface{}, prefix string, indent string, opts Options) ([]byte, error) {
+   w := bytes.NewBuffer([]byte{})
+   enc := json.NewEncoder(w)
+   enc.SetEscapeHTML((opts & EscapeHTML) != 0)
+   enc.SetIndent(prefix, indent)
+   err := enc.Encode(val)
+   out := w.Bytes()
+   return out, err
+}
+
+// Pretouch compiles vt ahead-of-time to avoid JIT compilation on-the-fly, in
+// order to reduce the first-hit latency.
+//
+// Opts are the compile options, for example, "option.WithCompileRecursiveDepth" is
+// a compile option to set the depth of recursive compile for the nested struct type.
+func Pretouch(vt reflect.Type, opts ...option.CompileOption) error {
+   return nil
+}
+
+// Valid validates json and returns first non-blank character position,
+// if it is only one valid json value.
+// Otherwise returns invalid character position using start.
+//
+// Note: it does not check for the invalid UTF-8 characters.
+func Valid(data []byte) (ok bool, start int) {
+   return json.Valid(data), 0
+}
+
+// StreamEncoder uses io.Writer as 
+type StreamEncoder struct {
+   w io.Writer
+   Encoder
+}
+
+// NewStreamEncoder adapts to encoding/json.NewDecoder API.
+//
+// NewStreamEncoder returns a new encoder that write to w.
+func NewStreamEncoder(w io.Writer) *StreamEncoder {
+   return &StreamEncoder{w: w}
+}
+
+// Encode encodes interface{} as JSON to io.Writer
+func (enc *StreamEncoder) Encode(val interface{}) (err error) {
+   jenc := json.NewEncoder(enc.w)
+   jenc.SetEscapeHTML((enc.Opts & EscapeHTML) != 0)
+   jenc.SetIndent(enc.prefix, enc.indent)
+   err = jenc.Encode(val)
+   return err
+}
--- a/vendor/github.com/bytedance/sonic/internal/decoder/asm.s
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/asm.s
--- a/vendor/github.com/bytedance/sonic/internal/decoder/assembler_amd64_go116.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/assembler_amd64_go116.go
@ -319,6 +319,7 @@ func (self *_Assembler) compile() {
    _OP_dismatch_err     : (*_Assembler)._asm_OP_dismatch_err,
    _OP_go_skip          : (*_Assembler)._asm_OP_go_skip,
    _OP_add              : (*_Assembler)._asm_OP_add,
+    _OP_check_empty      : (*_Assembler)._asm_OP_check_empty,
 }

 func (self *_Assembler) instr(v *_Instr) {
@ -458,6 +459,7 @@ func (self *_Assembler) call_vf(fn obj.Addr) {
 var (
    _V_stackOverflow              = jit.Imm(int64(uintptr(unsafe.Pointer(&stackOverflow))))
    _I_json_UnsupportedValueError = jit.Itab(_T_error, reflect.TypeOf(new(json.UnsupportedValueError)))
+    _I_json_MismatchTypeError     = jit.Itab(_T_error, reflect.TypeOf(new(MismatchTypeError)))
 )

 func (self *_Assembler) type_error() {
@ -471,7 +473,12 @@ func (self *_Assembler) type_error() {


 func (self *_Assembler) mismatch_error() {
-    self.Link(_LB_mismatch_error)               // _type_error:
+    self.Link(_LB_mismatch_error)                     // _type_error:
+    self.Emit("MOVQ", _VAR_et, _ET)                   // MOVQ _VAR_et, ET
+    self.Emit("MOVQ", _VAR_ic, _EP)                   // MOVQ _VAR_ic, EP
+    self.Emit("MOVQ", _I_json_MismatchTypeError, _AX) // MOVQ _I_json_MismatchTypeError, AX
+    self.Emit("CMPQ", _ET, _AX)                       // CMPQ ET, AX
+    self.Sjmp("JE"  , _LB_error)                      // JE _LB_error
    self.Emit("MOVQ", _ARG_sp, _AX)
    self.Emit("MOVQ", _AX, jit.Ptr(_SP, 0))     // MOVQ    AX, (SP)
    self.Emit("MOVQ", _ARG_sl, _CX)
@ -1128,9 +1135,16 @@ func (self *_Assembler) decode_dynamic(vt obj.Addr, vp obj.Addr) {
    self.call_go(_F_decodeTypedPointer)         // CALL_GO decodeTypedPointer
    self.Emit("MOVQ" , jit.Ptr(_SP, 64), _ET)   // MOVQ    64(SP), ET
    self.Emit("MOVQ" , jit.Ptr(_SP, 72), _EP)   // MOVQ    72(SP), EP
-    self.Emit("TESTQ", _ET, _ET)                // TESTQ   ET, ET
-    self.Sjmp("JNZ"  , _LB_error)               // JNZ     _error
    self.Emit("MOVQ" , jit.Ptr(_SP, 56), _IC)   // MOVQ    56(SP), IC
+    self.Emit("TESTQ", _ET, _ET)                // TESTQ   ET, ET
+    self.Sjmp("JE", "_decode_dynamic_end_{n}")  // JE, _decode_dynamic_end_{n}
+    self.Emit("MOVQ", _I_json_MismatchTypeError, _AX) // MOVQ _I_json_MismatchTypeError, AX
+    self.Emit("CMPQ",  _ET, _AX)                // CMPQ ET, AX
+    self.Sjmp("JNE" , _LB_error)                // JNE  LB_error
+    self.Emit("MOVQ", _EP, _VAR_ic)             // MOVQ EP, VAR_ic
+    self.Emit("MOVQ", _ET, _VAR_et)             // MOVQ ET, VAR_et
+    self.Link("_decode_dynamic_end_{n}")
+    
 }

 /** OpCode Assembler Functions **/
@ -1161,6 +1175,8 @@ func (self *_Assembler) decode_dynamic(vt obj.Addr, vp obj.Addr) {

 var (
    _F_FieldMap_GetCaseInsensitive obj.Addr
+    _Empty_Slice = make([]byte, 0)
+    _Zero_Base = int64(uintptr(((*rt.GoSlice)(unsafe.Pointer(&_Empty_Slice))).Ptr))
 )

 const (
@ -1487,18 +1503,21 @@ func (self *_Assembler) _asm_OP_map_init(_ *_Instr) {
 func (self *_Assembler) _asm_OP_map_key_i8(p *_Instr) {
    self.parse_signed(int8Type, "", p.vi())                                                 // PARSE     int8
    self.range_signed(_I_int8, _T_int8, math.MinInt8, math.MaxInt8)     // RANGE     int8
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                              // MAPASSIGN int8, mapassign, st.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_i16(p *_Instr) {
    self.parse_signed(int16Type, "", p.vi())                                                     // PARSE     int16
    self.range_signed(_I_int16, _T_int16, math.MinInt16, math.MaxInt16)     // RANGE     int16
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                                  // MAPASSIGN int16, mapassign, st.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_i32(p *_Instr) {
    self.parse_signed(int32Type, "", p.vi())                                                     // PARSE     int32
    self.range_signed(_I_int32, _T_int32, math.MinInt32, math.MaxInt32)     // RANGE     int32
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)                                  // MAPASSIGN int32, mapassign, st.Iv
    } else {
@ -1508,6 +1527,7 @@ func (self *_Assembler) _asm_OP_map_key_i32(p *_Instr) {

 func (self *_Assembler) _asm_OP_map_key_i64(p *_Instr) {
    self.parse_signed(int64Type, "", p.vi())                                 // PARSE     int64
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)              // MAPASSIGN int64, mapassign, st.Iv
    } else {
@ -1519,18 +1539,21 @@ func (self *_Assembler) _asm_OP_map_key_i64(p *_Instr) {
 func (self *_Assembler) _asm_OP_map_key_u8(p *_Instr) {
    self.parse_unsigned(uint8Type, "", p.vi())                                   // PARSE     uint8
    self.range_unsigned(_I_uint8, _T_uint8, math.MaxUint8)  // RANGE     uint8
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                  // MAPASSIGN uint8, vt.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_u16(p *_Instr) {
    self.parse_unsigned(uint16Type, "", p.vi())                                       // PARSE     uint16
    self.range_unsigned(_I_uint16, _T_uint16, math.MaxUint16)   // RANGE     uint16
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                      // MAPASSIGN uint16, vt.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_u32(p *_Instr) {
    self.parse_unsigned(uint32Type, "", p.vi())                                       // PARSE     uint32
    self.range_unsigned(_I_uint32, _T_uint32, math.MaxUint32)   // RANGE     uint32
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)                      // MAPASSIGN uint32, vt.Iv
    } else {
@ -1540,6 +1563,7 @@ func (self *_Assembler) _asm_OP_map_key_u32(p *_Instr) {

 func (self *_Assembler) _asm_OP_map_key_u64(p *_Instr) {
    self.parse_unsigned(uint64Type, "", p.vi())                                       // PARSE     uint64
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)                      // MAPASSIGN uint64, vt.Iv
    } else {
@ -1552,11 +1576,13 @@ func (self *_Assembler) _asm_OP_map_key_f32(p *_Instr) {
    self.parse_number(float32Type, "", p.vi())                     // PARSE     NUMBER
    self.range_single()                     // RANGE     float32
    self.Emit("MOVSS", _X0, _VAR_st_Dv)     // MOVSS     X0, st.Dv
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Dv)  // MAPASSIGN ${p.vt()}, mapassign, st.Dv
 }

 func (self *_Assembler) _asm_OP_map_key_f64(p *_Instr) {
    self.parse_number(float64Type, "", p.vi())                     // PARSE     NUMBER
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Dv)  // MAPASSIGN ${p.vt()}, mapassign, st.Dv
 }

@ -1621,6 +1647,24 @@ func (self *_Assembler) _asm_OP_slice_init(p *_Instr) {
    self.Emit("MOVQ" , _AX, jit.Ptr(_VP, 8))        // MOVQ    AX, 8(VP)
 }

+func (self *_Assembler) _asm_OP_check_empty(p *_Instr) {
+    rbracket := p.vb()
+    if rbracket == ']' {
+        self.check_eof(1)
+        self.Emit("LEAQ", jit.Ptr(_IC, 1), _AX)                              // LEAQ    1(IC), AX
+        self.Emit("CMPB", jit.Sib(_IP, _IC, 1, 0), jit.Imm(int64(rbracket))) // CMPB    (IP)(IC), ']'
+        self.Sjmp("JNE" , "_not_empty_array_{n}")                            // JNE     _not_empty_array_{n}
+        self.Emit("MOVQ", _AX, _IC)                                          // MOVQ    AX, IC
+        self.StorePtr(_Zero_Base, jit.Ptr(_VP, 0), _AX)                      // MOVQ    $zerobase, (VP)
+        self.Emit("PXOR" , _X0, _X0)                                         // PXOR    X0, X0
+        self.Emit("MOVOU", _X0, jit.Ptr(_VP, 8))                             // MOVOU   X0, 8(VP)
+        self.Xjmp("JMP" , p.vi())                                            // JMP     {p.vi()}
+        self.Link("_not_empty_array_{n}")
+    } else {
+        panic("only implement check empty array here!")
+    }
+}
+
 func (self *_Assembler) _asm_OP_slice_append(p *_Instr) {
    self.Emit("MOVQ" , jit.Ptr(_VP, 8), _AX)            // MOVQ    8(VP), AX
    self.Emit("CMPQ" , _AX, jit.Ptr(_VP, 16))           // CMPQ    AX, 16(VP)
@ -1640,12 +1684,34 @@ func (self *_Assembler) _asm_OP_slice_append(p *_Instr) {
    self.WriteRecNotAX(8, _DI, jit.Ptr(_VP, 0), true, true)// MOVQ    DI, (VP)
    self.Emit("MOVQ" , _AX, jit.Ptr(_VP, 8))            // MOVQ    AX, 8(VP)
    self.Emit("MOVQ" , _SI, jit.Ptr(_VP, 16))           // MOVQ    SI, 16(VP)
+
+    // because growslice not zero memory {oldcap, newlen} when append et not has ptrdata.
+    // but we should zero it, avoid decode it as random values.
+    if rt.UnpackType(p.vt()).PtrData == 0 {
+        self.Emit("SUBQ" , _AX, _SI)                        // MOVQ    AX, SI
+    
+        self.Emit("ADDQ" , jit.Imm(1), jit.Ptr(_VP, 8))     // ADDQ    $1, 8(VP)
+        self.Emit("MOVQ" , _DI, _VP)                        // MOVQ    DI, VP
+        self.Emit("MOVQ" , jit.Imm(int64(p.vlen())), _CX)   // MOVQ    ${p.vlen()}, CX
+        self.From("MULQ" , _CX)                             // MULQ    CX
+        self.Emit("ADDQ" , _AX, _VP)                        // ADDQ    AX, VP
+
+        self.Emit("MOVQ" , _SI, _AX)                        // MOVQ    SI, AX
+        self.From("MULQ" , _CX)                             // MULQ    CX
+        self.Emit("MOVQ" , _AX, jit.Ptr(_SP, 8))            // MOVQ    AX, 8(SP)
+
+        self.Emit("MOVQ" , _VP, jit.Ptr(_SP, 0))            // MOVQ    VP, (SP)
+        self.mem_clear_fn(true)                             // CALL_GO memclr{Has,NoHeap}
+        self.Sjmp("JMP", "_append_slice_end_{n}")           // JMP    _append_slice_end_{n}
+    }
+
    self.Link("_index_{n}")                             // _index_{n}:
    self.Emit("ADDQ" , jit.Imm(1), jit.Ptr(_VP, 8))     // ADDQ    $1, 8(VP)
    self.Emit("MOVQ" , jit.Ptr(_VP, 0), _VP)            // MOVQ    (VP), VP
    self.Emit("MOVQ" , jit.Imm(int64(p.vlen())), _CX)   // MOVQ    ${p.vlen()}, CX
    self.From("MULQ" , _CX)                             // MULQ    CX
    self.Emit("ADDQ" , _AX, _VP)                        // ADDQ    AX, VP
+    self.Link("_append_slice_end_{n}")
 }

 func (self *_Assembler) _asm_OP_object_skip(_ *_Instr) {
@ -1787,10 +1853,14 @@ func (self *_Assembler) lspace(subfix string) {
 }

 func (self *_Assembler) _asm_OP_match_char(p *_Instr) {
+    self.match_char(p.vb())
+}
+
+func (self *_Assembler) match_char(char byte) {
    self.check_eof(1)
-    self.Emit("CMPB", jit.Sib(_IP, _IC, 1, 0), jit.Imm(int64(p.vb())))  // CMPB (IP)(IC), ${p.vb()}
-    self.Sjmp("JNE" , _LB_char_0_error)                                 // JNE  _char_0_error
-    self.Emit("ADDQ", jit.Imm(1), _IC)                                  // ADDQ $1, IC
+    self.Emit("CMPB", jit.Sib(_IP, _IC, 1, 0), jit.Imm(int64(char)))  // CMPB (IP)(IC), ${p.vb()}
+    self.Sjmp("JNE" , _LB_char_0_error)                               // JNE  _char_0_error
+    self.Emit("ADDQ", jit.Imm(1), _IC)                                // ADDQ $1, IC
 }

 func (self *_Assembler) _asm_OP_check_char(p *_Instr) {
--- a/vendor/github.com/bytedance/sonic/internal/decoder/assembler_amd64_go117.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/assembler_amd64_go117.go
@ -315,6 +315,7 @@ func (self *_Assembler) compile() {
    _OP_dismatch_err     : (*_Assembler)._asm_OP_dismatch_err,
    _OP_go_skip          : (*_Assembler)._asm_OP_go_skip,
    _OP_add              : (*_Assembler)._asm_OP_add,
+    _OP_check_empty      : (*_Assembler)._asm_OP_check_empty,
    _OP_debug            : (*_Assembler)._asm_OP_debug,
 }

@ -348,8 +349,8 @@ func (self *_Assembler) epilogue() {
    self.Emit("MOVQ", _EP, _CX)                     // MOVQ BX, CX
    self.Emit("MOVQ", _ET, _BX)                     // MOVQ AX, BX
    self.Emit("MOVQ", _IC, _AX)                     // MOVQ IC, AX
-    self.Emit("MOVQ", jit.Imm(0), _ARG_sp)        // MOVQ $0, sv.p<>+48(FP)
-    self.Emit("MOVQ", jit.Imm(0), _ARG_vp)        // MOVQ $0, sv.p<>+48(FP)
+    self.Emit("MOVQ", jit.Imm(0), _ARG_sp)          // MOVQ $0, sv.p<>+48(FP)
+    self.Emit("MOVQ", jit.Imm(0), _ARG_vp)          // MOVQ $0, sv.p<>+48(FP)
    self.Emit("MOVQ", jit.Imm(0), _ARG_sv_p)        // MOVQ $0, sv.p<>+48(FP)
    self.Emit("MOVQ", jit.Imm(0), _ARG_vk)          // MOVQ $0, vk<>+64(FP)
    self.Emit("MOVQ", jit.Ptr(_SP, _FP_offs), _BP)  // MOVQ _FP_offs(SP), BP
@ -480,6 +481,7 @@ func (self *_Assembler) call_vf(fn obj.Addr) {
 var (
    _V_stackOverflow              = jit.Imm(int64(uintptr(unsafe.Pointer(&stackOverflow))))
    _I_json_UnsupportedValueError = jit.Itab(_T_error, reflect.TypeOf(new(json.UnsupportedValueError)))
+    _I_json_MismatchTypeError     = jit.Itab(_T_error, reflect.TypeOf(new(MismatchTypeError)))
 )

 func (self *_Assembler) type_error() {
@ -489,7 +491,12 @@ func (self *_Assembler) type_error() {
 }

 func (self *_Assembler) mismatch_error() {
-    self.Link(_LB_mismatch_error)               // _type_error:
+    self.Link(_LB_mismatch_error)                     // _type_error:
+    self.Emit("MOVQ", _VAR_et, _ET)                   // MOVQ _VAR_et, ET
+    self.Emit("MOVQ", _VAR_ic, _EP)                   // MOVQ _VAR_ic, EP
+    self.Emit("MOVQ", _I_json_MismatchTypeError, _CX) // MOVQ _I_json_MismatchType, CX
+    self.Emit("CMPQ", _ET, _CX)                       // CMPQ ET, CX
+    self.Sjmp("JE"  , _LB_error)                      // JE _LB_error
    self.Emit("MOVQ", _ARG_sp, _AX)
    self.Emit("MOVQ", _ARG_sl, _BX)
    self.Emit("MOVQ", _VAR_ic, _CX)
@ -1119,7 +1126,7 @@ func (self *_Assembler) decode_dynamic(vt obj.Addr, vp obj.Addr) {
    self.Emit("MOVQ", _ARG_sp, _AX)            // MOVQ    sp, AX
    self.Emit("MOVQ", _ARG_sl, _BX)            // MOVQ    sp, BX
    self.Emit("MOVQ" , _IC, _CX)                // MOVQ    IC, CX
-    self.Emit("MOVQ" , _ST, _R8)                // MOVQ    ST, R8
+    self.Emit("MOVQ" , _ST, _R8)                // MOVQ    ST, R8 
    self.Emit("MOVQ" , _ARG_fv, _R9)            // MOVQ    fv, R9
    self.save(_REG_rt...)
    self.Emit("MOVQ", _F_decodeTypedPointer, _IL)  // MOVQ ${fn}, R11
@ -1129,7 +1136,13 @@ func (self *_Assembler) decode_dynamic(vt obj.Addr, vp obj.Addr) {
    self.Emit("MOVQ" , _BX, _ET)                // MOVQ    BX, ET
    self.Emit("MOVQ" , _CX, _EP)                // MOVQ    CX, EP
    self.Emit("TESTQ", _ET, _ET)                // TESTQ   ET, ET
-    self.Sjmp("JNZ"  , _LB_error)               // JNZ     _error
+    self.Sjmp("JE", "_decode_dynamic_end_{n}")  // JE, _decode_dynamic_end_{n}
+    self.Emit("MOVQ", _I_json_MismatchTypeError, _CX) // MOVQ _I_json_MismatchTypeError, CX
+    self.Emit("CMPQ", _ET, _CX)                 // CMPQ ET, CX
+    self.Sjmp("JNE",  _LB_error)                // JNE  LB_error
+    self.Emit("MOVQ", _EP, _VAR_ic)             // MOVQ EP, VAR_ic
+    self.Emit("MOVQ", _ET, _VAR_et)             // MOVQ ET, VAR_et
+    self.Link("_decode_dynamic_end_{n}")
 }

 /** OpCode Assembler Functions **/
@ -1155,6 +1168,8 @@ func (self *_Assembler) decode_dynamic(vt obj.Addr, vp obj.Addr) {

 var (
    _F_FieldMap_GetCaseInsensitive obj.Addr
+    _Empty_Slice = make([]byte, 0)
+    _Zero_Base = int64(uintptr(((*rt.GoSlice)(unsafe.Pointer(&_Empty_Slice))).Ptr))
 )

 const (
@ -1482,18 +1497,21 @@ func (self *_Assembler) _asm_OP_map_init(_ *_Instr) {
 func (self *_Assembler) _asm_OP_map_key_i8(p *_Instr) {
    self.parse_signed(int8Type, "", p.vi())                                                 // PARSE     int8
    self.range_signed_CX(_I_int8, _T_int8, math.MinInt8, math.MaxInt8)     // RANGE     int8
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                              // MAPASSIGN int8, mapassign, st.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_i16(p *_Instr) {
    self.parse_signed(int16Type, "", p.vi())                                                     // PARSE     int16
    self.range_signed_CX(_I_int16, _T_int16, math.MinInt16, math.MaxInt16)     // RANGE     int16
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                                  // MAPASSIGN int16, mapassign, st.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_i32(p *_Instr) {
    self.parse_signed(int32Type, "", p.vi())                                                     // PARSE     int32
    self.range_signed_CX(_I_int32, _T_int32, math.MinInt32, math.MaxInt32)     // RANGE     int32
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)                                  // MAPASSIGN int32, mapassign, st.Iv
    } else {
@ -1504,6 +1522,7 @@ func (self *_Assembler) _asm_OP_map_key_i32(p *_Instr) {

 func (self *_Assembler) _asm_OP_map_key_i64(p *_Instr) {
    self.parse_signed(int64Type, "", p.vi())                                 // PARSE     int64
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)              // MAPASSIGN int64, mapassign, st.Iv
    } else {
@ -1515,18 +1534,21 @@ func (self *_Assembler) _asm_OP_map_key_i64(p *_Instr) {
 func (self *_Assembler) _asm_OP_map_key_u8(p *_Instr) {
    self.parse_unsigned(uint8Type, "", p.vi())                                   // PARSE     uint8
    self.range_unsigned_CX(_I_uint8, _T_uint8, math.MaxUint8)  // RANGE     uint8
-    self.mapassign_std(p.vt(), _VAR_st_Iv)                  // MAPASSIGN uint8, vt.Iv
+    self.match_char('"')
+    self.mapassign_std(p.vt(), _VAR_st_Iv)                    // MAPASSIGN uint8, vt.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_u16(p *_Instr) {
    self.parse_unsigned(uint16Type, "", p.vi())                                       // PARSE     uint16
    self.range_unsigned_CX(_I_uint16, _T_uint16, math.MaxUint16)   // RANGE     uint16
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Iv)                      // MAPASSIGN uint16, vt.Iv
 }

 func (self *_Assembler) _asm_OP_map_key_u32(p *_Instr) {
    self.parse_unsigned(uint32Type, "", p.vi())                                       // PARSE     uint32
    self.range_unsigned_CX(_I_uint32, _T_uint32, math.MaxUint32)   // RANGE     uint32
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)                      // MAPASSIGN uint32, vt.Iv
    } else {
@ -1537,6 +1559,7 @@ func (self *_Assembler) _asm_OP_map_key_u32(p *_Instr) {

 func (self *_Assembler) _asm_OP_map_key_u64(p *_Instr) {
    self.parse_unsigned(uint64Type, "", p.vi())                                       // PARSE     uint64
+    self.match_char('"')
    if vt := p.vt(); !mapfast(vt) {
        self.mapassign_std(vt, _VAR_st_Iv)                      // MAPASSIGN uint64, vt.Iv
    } else {
@ -1549,11 +1572,13 @@ func (self *_Assembler) _asm_OP_map_key_f32(p *_Instr) {
    self.parse_number(float32Type, "", p.vi())                     // PARSE     NUMBER
    self.range_single_X0()                     // RANGE     float32
    self.Emit("MOVSS", _X0, _VAR_st_Dv)     // MOVSS     X0, st.Dv
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Dv)  // MAPASSIGN ${p.vt()}, mapassign, st.Dv
 }

 func (self *_Assembler) _asm_OP_map_key_f64(p *_Instr) {
    self.parse_number(float64Type, "", p.vi())                     // PARSE     NUMBER
+    self.match_char('"')
    self.mapassign_std(p.vt(), _VAR_st_Dv)  // MAPASSIGN ${p.vt()}, mapassign, st.Dv
 }

@ -1612,6 +1637,24 @@ func (self *_Assembler) _asm_OP_slice_init(p *_Instr) {
    self.Link("_done_{n}")                          // _done_{n}
 }

+func (self *_Assembler) _asm_OP_check_empty(p *_Instr) {
+    rbracket := p.vb()
+    if rbracket == ']' {
+        self.check_eof(1)
+        self.Emit("LEAQ", jit.Ptr(_IC, 1), _AX)                              // LEAQ    1(IC), AX
+        self.Emit("CMPB", jit.Sib(_IP, _IC, 1, 0), jit.Imm(int64(rbracket))) // CMPB    (IP)(IC), ']'
+        self.Sjmp("JNE" , "_not_empty_array_{n}")                            // JNE     _not_empty_array_{n}
+        self.Emit("MOVQ", _AX, _IC)                                          // MOVQ    AX, IC
+        self.StorePtr(_Zero_Base, jit.Ptr(_VP, 0), _AX)                      // MOVQ    $zerobase, (VP)
+        self.Emit("PXOR", _X0, _X0)                                          // PXOR    X0, X0
+        self.Emit("MOVOU", _X0, jit.Ptr(_VP, 8))                             // MOVOU   X0, 8(VP)
+        self.Xjmp("JMP" , p.vi())                                            // JMP     {p.vi()}
+        self.Link("_not_empty_array_{n}")
+    } else {
+        panic("only implement check empty array here!")
+    }
+}
+
 func (self *_Assembler) _asm_OP_slice_append(p *_Instr) {
    self.Emit("MOVQ" , jit.Ptr(_VP, 8), _AX)            // MOVQ    8(VP), AX
    self.Emit("CMPQ" , _AX, jit.Ptr(_VP, 16))           // CMPQ    AX, 16(VP)
@ -1626,6 +1669,28 @@ func (self *_Assembler) _asm_OP_slice_append(p *_Instr) {
    self.WritePtrAX(8, jit.Ptr(_VP, 0), false)          // MOVQ    AX, (VP)
    self.Emit("MOVQ" , _BX, jit.Ptr(_VP, 8))            // MOVQ    BX, 8(VP)
    self.Emit("MOVQ" , _CX, jit.Ptr(_VP, 16))           // MOVQ    CX, 16(VP)
+
+    // because growslice not zero memory {oldcap, newlen} when append et not has ptrdata.
+    // but we should zero it, avoid decode it as random values.
+    if rt.UnpackType(p.vt()).PtrData == 0 {
+        self.Emit("MOVQ" , _CX, _DI)                        // MOVQ    CX, DI
+        self.Emit("SUBQ" , _BX, _DI)                        // MOVQ    BX, DI
+    
+        self.Emit("ADDQ" , jit.Imm(1), jit.Ptr(_VP, 8))     // ADDQ    $1, 8(VP)
+        self.Emit("MOVQ" , _AX, _VP)                        // MOVQ    AX, VP
+        self.Emit("MOVQ" , jit.Imm(int64(p.vlen())), _CX)   // MOVQ    ${p.vlen()}, CX
+        self.Emit("MOVQ" , _BX, _AX)                        // MOVQ    BX, AX 
+        self.From("MULQ" , _CX)                             // MULQ    CX
+        self.Emit("ADDQ" , _AX, _VP)                        // ADDQ    AX, VP
+
+        self.Emit("MOVQ" , _DI, _AX)                        // MOVQ    SI, AX
+        self.From("MULQ" , _CX)                             // MULQ    BX
+        self.Emit("MOVQ" , _AX, _BX)                        // ADDQ    AX, BX
+        self.Emit("MOVQ" , _VP, _AX)                        // MOVQ    VP, AX
+        self.mem_clear_fn(true)                             // CALL_GO memclr{Has,NoHeap}
+        self.Sjmp("JMP", "_append_slice_end_{n}")
+    }
+
    self.Emit("MOVQ" , _BX, _AX)                        // MOVQ    BX, AX
    self.Link("_index_{n}")                             // _index_{n}:
    self.Emit("ADDQ" , jit.Imm(1), jit.Ptr(_VP, 8))     // ADDQ    $1, 8(VP)
@ -1633,6 +1698,7 @@ func (self *_Assembler) _asm_OP_slice_append(p *_Instr) {
    self.Emit("MOVQ" , jit.Imm(int64(p.vlen())), _CX)   // MOVQ    ${p.vlen()}, CX
    self.From("MULQ" , _CX)                             // MULQ    CX
    self.Emit("ADDQ" , _AX, _VP)                        // ADDQ    AX, VP
+    self.Link("_append_slice_end_{n}")
 }

 func (self *_Assembler) _asm_OP_object_skip(_ *_Instr) {
@ -1767,8 +1833,12 @@ func (self *_Assembler) lspace(subfix string) {
 }

 func (self *_Assembler) _asm_OP_match_char(p *_Instr) {
+    self.match_char(p.vb())
+}
+
+func (self *_Assembler) match_char(char byte) {
    self.check_eof(1)
-    self.Emit("CMPB", jit.Sib(_IP, _IC, 1, 0), jit.Imm(int64(p.vb())))  // CMPB (IP)(IC), ${p.vb()}
+    self.Emit("CMPB", jit.Sib(_IP, _IC, 1, 0), jit.Imm(int64(char)))  // CMPB (IP)(IC), ${p.vb()}
    self.Sjmp("JNE" , _LB_char_0_error)                                 // JNE  _char_0_error
    self.Emit("ADDQ", jit.Imm(1), _IC)                                  // ADDQ $1, IC
 }
--- a/vendor/github.com/bytedance/sonic/internal/decoder/compiler.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/compiler.go
@ -98,6 +98,7 @@
    _OP_dismatch_err
    _OP_go_skip
    _OP_add
+    _OP_check_empty
    _OP_debug
 )

@ -174,6 +175,9 @@
    _OP_check_char_0     : "check_char_0",
    _OP_dismatch_err     : "dismatch_err",
    _OP_add              : "add",
+    _OP_go_skip          : "go_skip",
+    _OP_check_empty      : "check_empty",
+    _OP_debug            : "debug",
 }

 func (self _Op) String() string {
@ -639,11 +643,6 @@ func (self *_Compiler) compileMapOp(p *_Program, sp int, vt reflect.Type, op _Op
    skip2 := p.pc()
    p.rtt(op, vt)

-    /* match the closing quote if needed */
-    if op != _OP_map_key_str && op != _OP_map_key_utext && op != _OP_map_key_utext_p {
-        p.chr(_OP_match_char, '"')
-    }
-
    /* match the value separator */
    p.add(_OP_lspace)
    p.chr(_OP_match_char, ':')
@ -660,11 +659,6 @@ func (self *_Compiler) compileMapOp(p *_Program, sp int, vt reflect.Type, op _Op
    skip3 := p.pc()
    p.rtt(op, vt)

-    /* match the closing quote if needed */
-    if op != _OP_map_key_str && op != _OP_map_key_utext && op != _OP_map_key_utext_p {
-        p.chr(_OP_match_char, '"')
-    }
-
    /* match the value separator */
    p.add(_OP_lspace)
    p.chr(_OP_match_char, ':')
@ -689,12 +683,37 @@ func (self *_Compiler) compilePtr(p *_Program, sp int, et reflect.Type) {

    /* dereference all the way down */
    for et.Kind() == reflect.Ptr {
+        if et.Implements(jsonUnmarshalerType) {
+            p.rtt(_OP_unmarshal_p, et)
+            return
+        }
+
+        if et.Implements(encodingTextUnmarshalerType) {
+            p.add(_OP_lspace)
+            self.compileUnmarshalTextPtr(p, et)
+            return
+        }
+
        et = et.Elem()
        p.rtt(_OP_deref, et)
    }

-    /* compile the element type */
-    self.compileOne(p, sp + 1, et)
+    /* check for recursive nesting */
+    ok := self.tab[et]
+    if ok {
+        p.rtt(_OP_recurse, et)
+    } else {
+        /* enter the recursion */
+        p.add(_OP_lspace)
+        self.tab[et] = true
+        
+        /* not inline the pointer type
+        * recursing the defined pointer type's elem will casue issue379.
+        */
+        self.compileOps(p, sp, et)
+    }
+    delete(self.tab, et)
+
    j := p.pc()
    p.add(_OP_goto)
    p.pin(i)
@ -791,11 +810,11 @@ func (self *_Compiler) compileSliceList(p *_Program, sp int, vt reflect.Type) {
 }

 func (self *_Compiler) compileSliceBody(p *_Program, sp int, et reflect.Type) {
-    p.rtt(_OP_slice_init, et)
-    p.add(_OP_save)
    p.add(_OP_lspace)
    j := p.pc()
-    p.chr(_OP_check_char, ']')
+    p.chr(_OP_check_empty, ']')
+    p.rtt(_OP_slice_init, et)
+    p.add(_OP_save)
    p.rtt(_OP_slice_append, et)
    self.compileOne(p, sp + 1, et)
    p.add(_OP_load)
@ -808,9 +827,9 @@ func (self *_Compiler) compileSliceBody(p *_Program, sp int, et reflect.Type) {
    self.compileOne(p, sp + 1, et)
    p.add(_OP_load)
    p.int(_OP_goto, k0)
-    p.pin(j)
    p.pin(k1)
    p.add(_OP_drop)
+    p.pin(j)
 }

 func (self *_Compiler) compileString(p *_Program, vt reflect.Type) {
--- a/vendor/github.com/bytedance/sonic/internal/decoder/debug.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/debug.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/decoder.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/decoder.go
@ -17,6 +17,7 @@
 package decoder

 import (
+    `unsafe`
    `encoding/json`
    `reflect`
    `runtime`
@ -127,8 +128,17 @@ func (self *Decoder) Decode(val interface{}) error {
        return &json.InvalidUnmarshalError{Type: vv.Type.Pack()}
    }

+    etp := rt.PtrElem(vv.Type)
+
+    /* check the defined pointer type for issue 379 */
+    if vv.Type.IsNamed() {
+        newp := vp
+        etp  = vv.Type
+        vp   = unsafe.Pointer(&newp)
+    }
+
    /* create a new stack, and call the decoder */
-    sb, etp := newStack(), rt.PtrElem(vv.Type)
+    sb := newStack()
    nb, err := decodeTypedPointer(self.s, self.i, etp, vp, sb, self.f)
    /* return the stack back */
    self.i = nb
--- a/vendor/github.com/bytedance/sonic/internal/decoder/errors.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/errors.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_go116.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_go116.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_go117.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_go117.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_go117_test.s
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_go117_test.s
--- a/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_test.s
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/generic_amd64_test.s
--- a/vendor/github.com/bytedance/sonic/internal/decoder/pools.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/pools.go
@ -26,7 +26,7 @@
 )

 const (
-    _MinSlice = 16
+    _MinSlice = 2
    _MaxStack = 4096 // 4k slots
    _MaxStackBytes = _MaxStack * _PtrBytes
    _MaxDigitNums = 800  // used in atof fallback algorithm
--- a/vendor/github.com/bytedance/sonic/internal/decoder/primitives.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/primitives.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/stream.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/stream.go
@ -21,15 +21,15 @@
    `io`
    `sync`

+    `github.com/bytedance/sonic/option`
    `github.com/bytedance/sonic/internal/native/types`
 )

 var (
-    defaultBufferSize    uint = 4096
-    growSliceFactorShift uint = 1
-    minLeftBufferShift   uint = 2
+    minLeftBufferShift   uint = 1
 )

+// StreamDecoder is the decoder context object for streaming input.
 type StreamDecoder struct {
    r       io.Reader
    buf     []byte
@ -41,7 +41,7 @@ type StreamDecoder struct {

 var bufPool = sync.Pool{
    New: func () interface{} {
-        return make([]byte, 0, defaultBufferSize)
+        return make([]byte, 0, option.DefaultDecoderBufferSize)
    },
 }

@ -206,8 +206,8 @@ func realloc(buf *[]byte) {
    c := uint(cap(*buf))
    if c - l <= c >> minLeftBufferShift {
        e := l+(l>>minLeftBufferShift)
-        if e < defaultBufferSize {
-            e = defaultBufferSize
+        if e < option.DefaultDecoderBufferSize {
+            e = option.DefaultDecoderBufferSize
        }
        tmp := make([]byte, l, e)
        copy(tmp, *buf)
--- a/vendor/github.com/bytedance/sonic/internal/decoder/stubs_go115.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/stubs_go115.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/stubs_go120.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/stubs_go120.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/types.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/types.go
--- a/vendor/github.com/bytedance/sonic/internal/decoder/utils.go
+++ b/vendor/github.com/bytedance/sonic/internal/decoder/utils.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/asm.s
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/asm.s
--- a/vendor/github.com/bytedance/sonic/internal/encoder/assembler_amd64_go116.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/assembler_amd64_go116.go
@ -514,7 +514,8 @@ func (self *_Assembler) call_marshaler(fn obj.Addr, it *rt.GoType, vt reflect.Ty
    switch vt.Kind() {
        case reflect.Interface       : self.call_marshaler_i(fn, it)
        case reflect.Ptr, reflect.Map: self.call_marshaler_v(fn, it, vt, true)
-        default                      : self.call_marshaler_v(fn, it, vt, false)
+        // struct/array of 1 direct iface type can be direct
+        default                      : self.call_marshaler_v(fn, it, vt, !rt.UnpackType(vt).Indirect())
    }
 }

--- a/vendor/github.com/bytedance/sonic/internal/encoder/assembler_amd64_go117.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/assembler_amd64_go117.go
@ -539,7 +539,8 @@ func (self *_Assembler) call_marshaler(fn obj.Addr, it *rt.GoType, vt reflect.Ty
    switch vt.Kind() {
        case reflect.Interface        : self.call_marshaler_i(fn, it)
        case reflect.Ptr, reflect.Map : self.call_marshaler_v(fn, it, vt, true)
-        default                       : self.call_marshaler_v(fn, it, vt, false)
+        // struct/array of 1 direct iface type can be direct
+        default                       : self.call_marshaler_v(fn, it, vt, !rt.UnpackType(vt).Indirect())
    }
 }

--- a/vendor/github.com/bytedance/sonic/internal/encoder/compiler.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/compiler.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/debug_go116.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/debug_go116.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/debug_go117.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/debug_go117.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/encoder.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/encoder.go
@ -21,6 +21,7 @@
    `encoding/json`
    `reflect`
    `runtime`
+    `unsafe`

    `github.com/bytedance/sonic/internal/native`
    `github.com/bytedance/sonic/internal/native/types`
@ -161,8 +162,10 @@ func Quote(s string) string {

 // Encode returns the JSON encoding of val, encoded with opts.
 func Encode(val interface{}, opts Options) ([]byte, error) {
+    var ret []byte
+
    buf := newBytes()
-    err := EncodeInto(&buf, val, opts)
+    err := encodeInto(&buf, val, opts)

    /* check for errors */
    if err != nil {
@ -170,12 +173,20 @@ func Encode(val interface{}, opts Options) ([]byte, error) {
        return nil, err
    }

-    if opts & EscapeHTML != 0 || opts & ValidateString != 0  {
+    /* htmlescape or correct UTF-8 if opts enable */
+    old := buf
+    buf = encodeFinish(old, opts)
+    pbuf := ((*rt.GoSlice)(unsafe.Pointer(&buf))).Ptr
+    pold := ((*rt.GoSlice)(unsafe.Pointer(&old))).Ptr
+
+    /* return when allocated a new buffer */
+    if pbuf != pold {
+        freeBytes(old)
        return buf, nil
    }

    /* make a copy of the result */
-    ret := make([]byte, len(buf))
+    ret = make([]byte, len(buf))
    copy(ret, buf)

    freeBytes(buf)
@ -186,6 +197,15 @@ func Encode(val interface{}, opts Options) ([]byte, error) {
 // EncodeInto is like Encode but uses a user-supplied buffer instead of allocating
 // a new one.
 func EncodeInto(buf *[]byte, val interface{}, opts Options) error {
+    err := encodeInto(buf, val, opts)
+    if err != nil {
+        return err
+    }
+    *buf = encodeFinish(*buf, opts)
+    return err
+}
+
+func encodeInto(buf *[]byte, val interface{}, opts Options) error {
    stk := newStack()
    efv := rt.UnpackEface(val)
    err := encodeTypedPointer(buf, efv.Type, &efv.Value, stk, uint64(opts))
@ -196,25 +216,22 @@ func EncodeInto(buf *[]byte, val interface{}, opts Options) error {
    }
    freeStack(stk)

-    /* EscapeHTML needs to allocate a new buffer*/
-    if opts & EscapeHTML != 0 {
-        dest := HTMLEscape(nil, *buf)
-        freeBytes(*buf) // free origin used buffer
-        *buf = dest
-    }
-
-    if opts & ValidateString != 0 && !utf8.Validate(*buf) {
-        dest := utf8.CorrectWith(nil, *buf, `\ufffd`)
-        freeBytes(*buf) // free origin used buffer
-        *buf = dest
-    }
-
    /* avoid GC ahead */
    runtime.KeepAlive(buf)
    runtime.KeepAlive(efv)
    return err
 }

+func encodeFinish(buf []byte, opts Options) []byte {
+    if opts & EscapeHTML != 0 {
+        buf = HTMLEscape(nil, buf)
+    }
+    if opts & ValidateString != 0 && !utf8.Validate(buf) {
+        buf = utf8.CorrectWith(nil, buf, `\ufffd`)
+    }
+    return buf
+}
+
 var typeByte = rt.UnpackType(reflect.TypeOf(byte(0)))

 // HTMLEscape appends to dst the JSON-encoded src with <, >, &, U+2028 and U+2029
--- a/vendor/github.com/bytedance/sonic/internal/encoder/errors.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/errors.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/mapiter.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/mapiter.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/pools.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/pools.go
@ -30,7 +30,6 @@

 const (
    _MaxStack  = 4096      // 4k states
-    _MaxBuffer = 1048576    // 1MB buffer size

    _StackSize = unsafe.Sizeof(_Stack{})
 )
@ -92,7 +91,7 @@ func newBytes() []byte {
    if ret := bytesPool.Get(); ret != nil {
        return ret.([]byte)
    } else {
-        return make([]byte, 0, _MaxBuffer)
+        return make([]byte, 0, option.DefaultEncoderBufferSize)
    }
 }

@ -112,7 +111,7 @@ func newBuffer() *bytes.Buffer {
    if ret := bufferPool.Get(); ret != nil {
        return ret.(*bytes.Buffer)
    } else {
-        return bytes.NewBuffer(make([]byte, 0, _MaxBuffer))
+        return bytes.NewBuffer(make([]byte, 0, option.DefaultEncoderBufferSize))
    }
 }

--- a/vendor/github.com/bytedance/sonic/internal/encoder/primitives.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/primitives.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/sort.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/sort.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/stream.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/stream.go
@ -21,7 +21,7 @@
    `io`
 )

-// StreamEncoder uses io.Writer as 
+// StreamEncoder uses io.Writer as input.
 type StreamEncoder struct {
    w io.Writer
    Encoder
--- a/vendor/github.com/bytedance/sonic/internal/encoder/stubs_go116.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/stubs_go116.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/stubs_go117.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/stubs_go117.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/stubs_go120.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/stubs_go120.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/types.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/types.go
--- a/vendor/github.com/bytedance/sonic/internal/encoder/utils.go
+++ b/vendor/github.com/bytedance/sonic/internal/encoder/utils.go
--- a/vendor/github.com/bytedance/sonic/internal/jit/assembler_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/jit/assembler_amd64.go
@ -72,6 +72,18 @@ func (self *BaseAssembler) NOPn(n int) {
    }
 }

+func (self *BaseAssembler) StorePtr(ptr int64, to obj.Addr, tmp obj.Addr) {
+    if (to.Type != obj.TYPE_MEM) || (tmp.Type != obj.TYPE_REG) {
+        panic("must store imm to memory, tmp must be register")
+    }
+    if (ptr >> 32) != 0 {
+        self.Emit("MOVQ", Imm(ptr), tmp)
+        self.Emit("MOVQ", tmp, to)
+    } else {
+        self.Emit("MOVQ", Imm(ptr), to);
+    }
+}
+
 func (self *BaseAssembler) Byte(v ...byte) {
    for ; len(v) >= 8; v = v[8:] { self.From("QUAD", Imm(rt.Get64(v))) }
    for ; len(v) >= 4; v = v[4:] { self.From("LONG", Imm(int64(rt.Get32(v)))) }
--- a/vendor/github.com/bytedance/sonic/internal/native/avx/native_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/avx/native_amd64.go
@ -122,7 +122,7 @@ func __validate_one(s *string, p *int, m *types.StateMachine) (ret int)
 //go:nosplit
 //go:noescape
 //goland:noinspection GoUnusedParameter
-func __get_by_path(s *string, p *int, path *[]interface{}) (ret int)
+func __get_by_path(s *string, p *int, path *[]interface{}, m *types.StateMachine) (ret int)

 //go:nosplit
 //go:noescape
--- a/vendor/github.com/bytedance/sonic/internal/native/avx/native_amd64.s
+++ b/vendor/github.com/bytedance/sonic/internal/native/avx/native_amd64.s
--- a/vendor/github.com/bytedance/sonic/internal/native/avx/native_subr_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/avx/native_subr_amd64.go
@ -9,53 +9,53 @@
 func __native_entry__() uintptr

 var (
-    _subr__f32toa             = __native_entry__() + 29744
-    _subr__f64toa             = __native_entry__() + 496
-    _subr__get_by_path        = __native_entry__() + 27424
-    _subr__html_escape        = __native_entry__() + 9968
-    _subr__i64toa             = __native_entry__() + 4112
-    _subr__lspace             = __native_entry__() + 80
-    _subr__quote              = __native_entry__() + 5792
-    _subr__skip_array         = __native_entry__() + 20576
-    _subr__skip_number        = __native_entry__() + 23920
-    _subr__skip_object        = __native_entry__() + 22496
-    _subr__skip_one           = __native_entry__() + 24080
-    _subr__skip_one_fast      = __native_entry__() + 24320
-    _subr__u64toa             = __native_entry__() + 4384
-    _subr__unquote            = __native_entry__() + 7488
-    _subr__validate_one       = __native_entry__() + 24144
-    _subr__validate_utf8      = __native_entry__() + 28464
-    _subr__validate_utf8_fast = __native_entry__() + 29136
-    _subr__value              = __native_entry__() + 14672
-    _subr__vnumber            = __native_entry__() + 18320
-    _subr__vsigned            = __native_entry__() + 19856
-    _subr__vstring            = __native_entry__() + 16864
-    _subr__vunsigned          = __native_entry__() + 20208
+    _subr__f32toa             = __native_entry__() + 31264
+    _subr__f64toa             = __native_entry__() + 192
+    _subr__get_by_path        = __native_entry__() + 25856
+    _subr__html_escape        = __native_entry__() + 9040
+    _subr__i64toa             = __native_entry__() + 3488
+    _subr__lspace             = __native_entry__() + 16
+    _subr__quote              = __native_entry__() + 4880
+    _subr__skip_array         = __native_entry__() + 17952
+    _subr__skip_number        = __native_entry__() + 21952
+    _subr__skip_object        = __native_entry__() + 20368
+    _subr__skip_one           = __native_entry__() + 22112
+    _subr__skip_one_fast      = __native_entry__() + 22352
+    _subr__u64toa             = __native_entry__() + 3600
+    _subr__unquote            = __native_entry__() + 6672
+    _subr__validate_one       = __native_entry__() + 22176
+    _subr__validate_utf8      = __native_entry__() + 30000
+    _subr__validate_utf8_fast = __native_entry__() + 30672
+    _subr__value              = __native_entry__() + 12224
+    _subr__vnumber            = __native_entry__() + 15616
+    _subr__vsigned            = __native_entry__() + 17232
+    _subr__vstring            = __native_entry__() + 14064
+    _subr__vunsigned          = __native_entry__() + 17600
 )

 const (
-    _stack__f32toa = 56
+    _stack__f32toa = 48
    _stack__f64toa = 80
-    _stack__get_by_path = 312
+    _stack__get_by_path = 304
    _stack__html_escape = 64
    _stack__i64toa = 16
    _stack__lspace = 8
-    _stack__quote = 80
+    _stack__quote = 56
    _stack__skip_array = 128
    _stack__skip_number = 72
    _stack__skip_object = 128
    _stack__skip_one = 128
-    _stack__skip_one_fast = 208
+    _stack__skip_one_fast = 200
    _stack__u64toa = 8
-    _stack__unquote = 128
+    _stack__unquote = 88
    _stack__validate_one = 128
    _stack__validate_utf8 = 48
    _stack__validate_utf8_fast = 24
-    _stack__value = 368
-    _stack__vnumber = 280
+    _stack__value = 328
+    _stack__vnumber = 240
    _stack__vsigned = 16
-    _stack__vstring = 128
-    _stack__vunsigned = 24
+    _stack__vstring = 136
+    _stack__vunsigned = 16
 )

 var (
--- a/vendor/github.com/bytedance/sonic/internal/native/avx2/native_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/avx2/native_amd64.go
@ -122,7 +122,7 @@ func __validate_one(s *string, p *int, m *types.StateMachine) (ret int)
 //go:nosplit
 //go:noescape
 //goland:noinspection GoUnusedParameter
-func __get_by_path(s *string, p *int, path *[]interface{}) (ret int)
+func __get_by_path(s *string, p *int, path *[]interface{}, m *types.StateMachine) (ret int)

 //go:nosplit
 //go:noescape
--- a/vendor/github.com/bytedance/sonic/internal/native/avx2/native_amd64.s
+++ b/vendor/github.com/bytedance/sonic/internal/native/avx2/native_amd64.s
--- a/vendor/github.com/bytedance/sonic/internal/native/avx2/native_subr_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/avx2/native_subr_amd64.go
@ -9,53 +9,53 @@
 func __native_entry__() uintptr

 var (
-    _subr__f32toa             = __native_entry__() + 35216
-    _subr__f64toa             = __native_entry__() + 752
-    _subr__get_by_path        = __native_entry__() + 30384
-    _subr__html_escape        = __native_entry__() + 11712
-    _subr__i64toa             = __native_entry__() + 4368
-    _subr__lspace             = __native_entry__() + 224
-    _subr__quote              = __native_entry__() + 6160
-    _subr__skip_array         = __native_entry__() + 22864
-    _subr__skip_number        = __native_entry__() + 26928
-    _subr__skip_object        = __native_entry__() + 24864
-    _subr__skip_one           = __native_entry__() + 27088
-    _subr__skip_one_fast      = __native_entry__() + 27504
-    _subr__u64toa             = __native_entry__() + 4640
-    _subr__unquote            = __native_entry__() + 8960
-    _subr__validate_one       = __native_entry__() + 27152
-    _subr__validate_utf8      = __native_entry__() + 31552
-    _subr__validate_utf8_fast = __native_entry__() + 32496
-    _subr__value              = __native_entry__() + 16816
-    _subr__vnumber            = __native_entry__() + 20608
-    _subr__vsigned            = __native_entry__() + 22144
-    _subr__vstring            = __native_entry__() + 19312
-    _subr__vunsigned          = __native_entry__() + 22496
+    _subr__f32toa             = __native_entry__() + 33888
+    _subr__f64toa             = __native_entry__() + 288
+    _subr__get_by_path        = __native_entry__() + 28336
+    _subr__html_escape        = __native_entry__() + 10496
+    _subr__i64toa             = __native_entry__() + 3584
+    _subr__lspace             = __native_entry__() + 64
+    _subr__quote              = __native_entry__() + 5072
+    _subr__skip_array         = __native_entry__() + 20688
+    _subr__skip_number        = __native_entry__() + 24912
+    _subr__skip_object        = __native_entry__() + 22736
+    _subr__skip_one           = __native_entry__() + 25072
+    _subr__skip_one_fast      = __native_entry__() + 25488
+    _subr__u64toa             = __native_entry__() + 3696
+    _subr__unquote            = __native_entry__() + 7888
+    _subr__validate_one       = __native_entry__() + 25136
+    _subr__validate_utf8      = __native_entry__() + 30320
+    _subr__validate_utf8_fast = __native_entry__() + 31280
+    _subr__value              = __native_entry__() + 15024
+    _subr__vnumber            = __native_entry__() + 18352
+    _subr__vsigned            = __native_entry__() + 19968
+    _subr__vstring            = __native_entry__() + 17024
+    _subr__vunsigned          = __native_entry__() + 20336
 )

 const (
-    _stack__f32toa = 56
+    _stack__f32toa = 48
    _stack__f64toa = 80
-    _stack__get_by_path = 320
+    _stack__get_by_path = 296
    _stack__html_escape = 72
    _stack__i64toa = 16
    _stack__lspace = 8
-    _stack__quote = 72
-    _stack__skip_array = 120
-    _stack__skip_number = 80
-    _stack__skip_object = 120
-    _stack__skip_one = 120
-    _stack__skip_one_fast = 216
+    _stack__quote = 56
+    _stack__skip_array = 128
+    _stack__skip_number = 72
+    _stack__skip_object = 128
+    _stack__skip_one = 128
+    _stack__skip_one_fast = 208
    _stack__u64toa = 8
-    _stack__unquote = 128
-    _stack__validate_one = 120
+    _stack__unquote = 72
+    _stack__validate_one = 128
    _stack__validate_utf8 = 48
-    _stack__validate_utf8_fast = 200
-    _stack__value = 368
-    _stack__vnumber = 280
+    _stack__validate_utf8_fast = 176
+    _stack__value = 328
+    _stack__vnumber = 240
    _stack__vsigned = 16
-    _stack__vstring = 104
-    _stack__vunsigned = 24
+    _stack__vstring = 112
+    _stack__vunsigned = 16
 )

 var (
--- a/vendor/github.com/bytedance/sonic/internal/native/dispatch_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/dispatch_amd64.go
@ -94,7 +94,7 @@ func SkipOneFast(s *string, p *int) int
 //go:nosplit
 //go:noescape
 //goland:noinspection GoUnusedParameter
-func GetByPath(s *string, p *int, path *[]interface{}) int
+func GetByPath(s *string, p *int, path *[]interface{}, m *types.StateMachine) int

 //go:nosplit
 //go:noescape
--- a/vendor/github.com/bytedance/sonic/internal/native/dispatch_amd64.s
+++ b/vendor/github.com/bytedance/sonic/internal/native/dispatch_amd64.s
@ -72,7 +72,7 @@ TEXT ·SkipOneFast(SB), NOSPLIT, $0 - 24
    JMP  github·com∕bytedance∕sonic∕internal∕native∕avx·__skip_one_fast(SB)
    JMP  github·com∕bytedance∕sonic∕internal∕native∕sse·__skip_one_fast(SB)

-TEXT ·GetByPath(SB), NOSPLIT, $0 - 32
+TEXT ·GetByPath(SB), NOSPLIT, $0 - 40
    CMPB github·com∕bytedance∕sonic∕internal∕cpu·HasAVX2(SB), $0
    JE   2(PC)
    JMP  github·com∕bytedance∕sonic∕internal∕native∕avx2·__get_by_path(SB)
--- a/vendor/github.com/bytedance/sonic/internal/native/native_amd64.tmpl
+++ b/vendor/github.com/bytedance/sonic/internal/native/native_amd64.tmpl
@ -120,7 +120,7 @@ func __validate_one(s *string, p *int, m *types.StateMachine) (ret int)
 //go:nosplit
 //go:noescape
 //goland:noinspection GoUnusedParameter
-func __get_by_path(s *string, p *int, path *[]interface{}) (ret int)
+func __get_by_path(s *string, p *int, path *[]interface{}, m *types.StateMachine) (ret int)

 //go:nosplit
 //go:noescape
--- a/vendor/github.com/bytedance/sonic/internal/native/sse/native_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/sse/native_amd64.go
@ -122,7 +122,7 @@ func __validate_one(s *string, p *int, m *types.StateMachine) (ret int)
 //go:nosplit
 //go:noescape
 //goland:noinspection GoUnusedParameter
-func __get_by_path(s *string, p *int, path *[]interface{}) (ret int)
+func __get_by_path(s *string, p *int, path *[]interface{}, m *types.StateMachine) (ret int)

 //go:nosplit
 //go:noescape
--- a/vendor/github.com/bytedance/sonic/internal/native/sse/native_amd64.s
+++ b/vendor/github.com/bytedance/sonic/internal/native/sse/native_amd64.s
--- a/vendor/github.com/bytedance/sonic/internal/native/sse/native_subr_amd64.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/sse/native_subr_amd64.go
@ -9,53 +9,53 @@
 func __native_entry__() uintptr

 var (
-    _subr__f32toa             = __native_entry__() + 28688
-    _subr__f64toa             = __native_entry__() + 464
-    _subr__get_by_path        = __native_entry__() + 26432
-    _subr__html_escape        = __native_entry__() + 9584
-    _subr__i64toa             = __native_entry__() + 3744
-    _subr__lspace             = __native_entry__() + 80
-    _subr__quote              = __native_entry__() + 5472
-    _subr__skip_array         = __native_entry__() + 19184
-    _subr__skip_number        = __native_entry__() + 22528
-    _subr__skip_object        = __native_entry__() + 21088
-    _subr__skip_one           = __native_entry__() + 22688
-    _subr__skip_one_fast      = __native_entry__() + 22912
-    _subr__u64toa             = __native_entry__() + 4016
-    _subr__unquote            = __native_entry__() + 7184
-    _subr__validate_one       = __native_entry__() + 22736
-    _subr__validate_utf8      = __native_entry__() + 27456
-    _subr__validate_utf8_fast = __native_entry__() + 28128
-    _subr__value              = __native_entry__() + 13216
-    _subr__vnumber            = __native_entry__() + 16928
-    _subr__vsigned            = __native_entry__() + 18464
-    _subr__vstring            = __native_entry__() + 15408
-    _subr__vunsigned          = __native_entry__() + 18816
+    _subr__f32toa             = __native_entry__() + 31760
+    _subr__f64toa             = __native_entry__() + 160
+    _subr__get_by_path        = __native_entry__() + 26384
+    _subr__html_escape        = __native_entry__() + 9072
+    _subr__i64toa             = __native_entry__() + 3424
+    _subr__lspace             = __native_entry__() + 16
+    _subr__quote              = __native_entry__() + 4864
+    _subr__skip_array         = __native_entry__() + 18112
+    _subr__skip_number        = __native_entry__() + 22128
+    _subr__skip_object        = __native_entry__() + 20512
+    _subr__skip_one           = __native_entry__() + 22288
+    _subr__skip_one_fast      = __native_entry__() + 22512
+    _subr__u64toa             = __native_entry__() + 3552
+    _subr__unquote            = __native_entry__() + 6704
+    _subr__validate_one       = __native_entry__() + 22336
+    _subr__validate_utf8      = __native_entry__() + 30528
+    _subr__validate_utf8_fast = __native_entry__() + 31200
+    _subr__value              = __native_entry__() + 12272
+    _subr__vnumber            = __native_entry__() + 15728
+    _subr__vsigned            = __native_entry__() + 17376
+    _subr__vstring            = __native_entry__() + 14112
+    _subr__vunsigned          = __native_entry__() + 17760
 )

 const (
-    _stack__f32toa = 56
+    _stack__f32toa = 48
    _stack__f64toa = 80
-    _stack__get_by_path = 264
+    _stack__get_by_path = 240
    _stack__html_escape = 64
    _stack__i64toa = 16
    _stack__lspace = 8
-    _stack__quote = 80
+    _stack__quote = 64
    _stack__skip_array = 128
    _stack__skip_number = 72
    _stack__skip_object = 128
    _stack__skip_one = 128
-    _stack__skip_one_fast = 160
+    _stack__skip_one_fast = 136
    _stack__u64toa = 8
-    _stack__unquote = 128
+    _stack__unquote = 88
    _stack__validate_one = 128
    _stack__validate_utf8 = 48
    _stack__validate_utf8_fast = 24
-    _stack__value = 368
-    _stack__vnumber = 280
+    _stack__value = 328
+    _stack__vnumber = 240
    _stack__vsigned = 16
-    _stack__vstring = 128
-    _stack__vunsigned = 24
+    _stack__vstring = 136
+    _stack__vunsigned = 16
 )

 var (
--- a/vendor/github.com/bytedance/sonic/internal/native/types/types.go
+++ b/vendor/github.com/bytedance/sonic/internal/native/types/types.go
@ -76,6 +76,10 @@
    ERR_FLOAT_INFINITY     ParsingError = 8
    ERR_MISMATCH           ParsingError = 9
    ERR_INVALID_UTF8       ParsingError = 10
+
+    // error code used in ast
+    ERR_NOT_FOUND          ParsingError = 33
+    ERR_UNSUPPORT_TYPE     ParsingError = 34
 )

 var _ParsingErrors = []string{
--- a/vendor/github.com/bytedance/sonic/internal/resolver/resolver.go
+++ b/vendor/github.com/bytedance/sonic/internal/resolver/resolver.go
@ -152,8 +152,8 @@ func resolveFields(vt reflect.Type) []FieldMeta {
        }

        /* get the index to the last offset */
-        fvt := fv.typ
        idx := len(path) - 1
+        fvt := path[idx].Type

        /* do not dereference into fields */
        if path[idx].Kind == F_deref {
--- a/vendor/github.com/bytedance/sonic/internal/rt/fastmem.go
+++ b/vendor/github.com/bytedance/sonic/internal/rt/fastmem.go
@ -87,7 +87,6 @@ func GuardSlice(buf *[]byte, n int) {
 		copy(tmp, *buf)
 		*buf = tmp
 	}
-	return
 }

 //go:nosplit
--- a/vendor/github.com/bytedance/sonic/internal/rt/fastvalue.go
+++ b/vendor/github.com/bytedance/sonic/internal/rt/fastvalue.go
@ -22,14 +22,23 @@
 )

 var (
-	reflectRtypeItab = findReflectRtypeItab()
+    reflectRtypeItab = findReflectRtypeItab()
 )

+// GoType.KindFlags const
 const (
    F_direct    = 1 << 5
    F_kind_mask = (1 << 5) - 1
 )

+// GoType.Flags const
+const (
+    tflagUncommon      uint8 = 1 << 0
+    tflagExtraStar     uint8 = 1 << 1
+    tflagNamed         uint8 = 1 << 2
+    tflagRegularMemory uint8 = 1 << 3
+)
+
 type GoType struct {
    Size       uintptr
    PtrData    uintptr
@ -44,6 +53,10 @@ type GoType struct {
    PtrToSelf  int32
 }

+func (self *GoType) IsNamed() bool {
+    return (self.Flags & tflagNamed) != 0
+}
+
 func (self *GoType) Kind() reflect.Kind {
    return reflect.Kind(self.KindFlags & F_kind_mask)
 }
--- a/vendor/github.com/bytedance/sonic/internal/rt/stackmap.go
+++ b/vendor/github.com/bytedance/sonic/internal/rt/stackmap.go
@ -152,6 +152,7 @@ type StackMapBuilder struct {
    b Bitmap
 }

+//go:nocheckptr
 func (self *StackMapBuilder) Build() (p *StackMap) {
    nb := len(self.b.B)
    bm := mallocgc(_StackMapSize + uintptr(nb) - 1, byteType, false)
--- a/vendor/github.com/bytedance/sonic/introduction-1.png
+++ b/vendor/github.com/bytedance/sonic/introduction-1.png
--- a/vendor/github.com/bytedance/sonic/introduction-2.png
+++ b/vendor/github.com/bytedance/sonic/introduction-2.png
--- a/vendor/github.com/bytedance/sonic/loader/funcdata_go115.go
+++ b/vendor/github.com/bytedance/sonic/loader/funcdata_go115.go
@ -1,5 +1,5 @@
-// go:build go1.15 && !go1.18
-// +build go1.15,!go1.18
+//go:build go1.15 && !go1.16
+// +build go1.15,!go1.16

 /*
 * Copyright 2021 ByteDance Inc.
@ -28,7 +28,7 @@
 )

 const (
-    _Magic uint32 = 0xfffffff0
+    _Magic uint32 = 0xfffffffa
 )

 type pcHeader struct {
@ -38,7 +38,6 @@ type pcHeader struct {
    ptrSize        uint8   // size of a ptr in bytes
    nfunc          int     // number of functions in the module
    nfiles         uint    // number of entries in the file tab
-    textStart      uintptr // base for function entry PC offsets in this module, equal to moduledata.text
    funcnameOffset uintptr // offset to the funcnametab variable from pcHeader
    cuOffset       uintptr // offset to the cutab variable from pcHeader
    filetabOffset  uintptr // offset to the filetab variable from pcHeader
@ -64,9 +63,7 @@ funcnametab  []byte
    noptrbss, enoptrbss   uintptr
    end, gcdata, gcbss    uintptr
    types, etypes         uintptr
-    rodata                uintptr
-    gofunc                uintptr // go.func.* is actual funcinfo object in image
-
+    
    textsectmap []textSection // see runtime/symtab.go: textAddr()
    typelinks   []int32 // offsets from types
    itablinks   []*rt.GoItab
@ -91,7 +88,7 @@ funcnametab  []byte
 }

 type _func struct {
-    entryOff uint32 // start pc, as offset from moduledata.text/pcHeader.textStart
+    entry    uintptr // start pc, as offset from moduledata.text/pcHeader.textStart
    nameOff  int32  // function name, as index into moduledata.funcnametab.

    args        int32  // in/out args size
@ -103,8 +100,7 @@ type _func struct {
    npcdata   uint32
    cuOffset  uint32 // runtime.cutab offset of this function's CU
    funcID    uint8  // set for certain special runtime functions
-    flag      uint8
-    _         [1]byte // pad
+    _         [2]byte // pad
    nfuncdata uint8   // 
    
    // The end of the struct is followed immediately by two variable-length
@ -131,8 +127,8 @@ funcID    uint8  // set for certain special runtime functions
 }

 type funcTab struct {
-    entry   uint32
-    funcoff uint32
+    entry   uintptr
+    funcoff uintptr
 }

 type bitVector struct {
@ -170,6 +166,11 @@ type findfuncbucket struct {
    _SUBBUCKETS [16]byte
 }

+
+type compilationUnit struct {
+    fileNames []string
+}
+
 // func name table format: 
 //   nameOff[0] -> namePartA namePartB namePartC \x00 
 //   nameOff[1] -> namePartA namePartB namePartC \x00
@ -192,10 +193,6 @@ func makeFuncnameTab(funcs []Func) (tab []byte, offs []int32) {
    return
 }

-type compilationUnit struct {
-    fileNames []string
-}
-
 // CU table format:
 //  cuOffsets[0] -> filetabOffset[0] filetabOffset[1] ... filetabOffset[len(CUs[0].fileNames)-1]
 //  cuOffsets[1] -> filetabOffset[len(CUs[0].fileNames)] ... filetabOffset[len(CUs[0].fileNames) + len(CUs[1].fileNames)-1]
@ -266,68 +263,76 @@ funcdataOffs[i] = append(funcdataOffs[i], _INVALID_FUNCDATA_OFFSET)
    return 
 }

-func makeFtab(funcs []_func, lastFuncSize uint32) (ftab []funcTab) {
+func makeFtab(funcs []_func, lastFuncSize uint32) (ftab []funcTab, pclntabSize int64, startLocations []uint32) {
    // Allocate space for the pc->func table. This structure consists of a pc offset
    // and an offset to the func structure. After that, we have a single pc
    // value that marks the end of the last function in the binary.
-    var size int64 = int64(len(funcs)*2*4 + 4)
-    var startLocations = make([]uint32, len(funcs))
+    pclntabSize = int64(len(funcs)*2*int(_PtrSize) + int(_PtrSize))
+    startLocations = make([]uint32, len(funcs))
    for i, f := range funcs {
-        size = rnd(size, int64(_PtrSize))
+        pclntabSize = rnd(pclntabSize, int64(_PtrSize))
        //writePCToFunc
-        startLocations[i] = uint32(size)
-        size += int64(uint8(_FUNC_SIZE)+f.nfuncdata*4+uint8(f.npcdata)*4)
+        startLocations[i] = uint32(pclntabSize)
+        pclntabSize += int64(uint8(_FUNC_SIZE) + f.nfuncdata*_PtrSize + uint8(f.npcdata)*4)
    }
-
    ftab = make([]funcTab, 0, len(funcs)+1)

    // write a map of pc->func info offsets 
    for i, f := range funcs {
-        ftab = append(ftab, funcTab{uint32(f.entryOff), uint32(startLocations[i])})
+        ftab = append(ftab, funcTab{uintptr(f.entry), uintptr(startLocations[i])})
    }

    // Final entry of table is just end pc offset.
    lastFunc := funcs[len(funcs)-1]
-    ftab = append(ftab, funcTab{uint32(lastFunc.entryOff + lastFuncSize), 0})
+    ftab = append(ftab, funcTab{lastFunc.entry + uintptr(lastFuncSize), 0})

    return
 }

 // Pcln table format: [...]funcTab + [...]_Func
-func makePclntable(funcs []_func, lastFuncSize uint32, pcdataOffs [][]uint32, funcdataOffs [][]uint32) (pclntab []byte) {
-    // Allocate space for the pc->func table. This structure consists of a pc offset
-    // and an offset to the func structure. After that, we have a single pc
-    // value that marks the end of the last function in the binary.
-    var size int64 = int64(len(funcs)*2*4 + 4)
-    var startLocations = make([]uint32, len(funcs))
-    for i := range funcs {
-        size = rnd(size, int64(_PtrSize))
-        //writePCToFunc
-        startLocations[i] = uint32(size)
-        size += int64(int(_FUNC_SIZE)+len(funcdataOffs[i])*4+len(pcdataOffs[i])*4)
-    }
-
+func makePclntable(size int64, startLocations []uint32, funcs []_func, lastFuncSize uint32, pcdataOffs [][]uint32, funcdataAddr uintptr, funcdataOffs [][]uint32) (pclntab []byte) {
    pclntab = make([]byte, size, size)

    // write a map of pc->func info offsets 
    offs := 0
    for i, f := range funcs {
-        byteOrder.PutUint32(pclntab[offs:offs+4], uint32(f.entryOff))
-        byteOrder.PutUint32(pclntab[offs+4:offs+8], uint32(startLocations[i]))
-        offs += 8
+        byteOrder.PutUint64(pclntab[offs:offs+8], uint64(f.entry))
+        byteOrder.PutUint64(pclntab[offs+8:offs+16], uint64(startLocations[i]))
+        offs += 16
    }
    // Final entry of table is just end pc offset.
    lastFunc := funcs[len(funcs)-1]
-    byteOrder.PutUint32(pclntab[offs:offs+4], uint32(lastFunc.entryOff+lastFuncSize))
+    byteOrder.PutUint64(pclntab[offs:offs+8], uint64(lastFunc.entry)+uint64(lastFuncSize))
+    offs += 8

    // write func info table
    for i, f := range funcs {
        off := startLocations[i]

        // write _func structure to pclntab
-        fb := rt.BytesFrom(unsafe.Pointer(&f), int(_FUNC_SIZE), int(_FUNC_SIZE))
-        copy(pclntab[off:off+uint32(_FUNC_SIZE)], fb)
-        off += uint32(_FUNC_SIZE)
+        byteOrder.PutUint64(pclntab[off:off+8], uint64(f.entry))
+        off += 8
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.nameOff))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.args))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.deferreturn))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.pcsp))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.pcfile))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.pcln))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.npcdata))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.cuOffset))
+        off += 4
+        pclntab[off] = f.funcID
+        // NOTICE: _[2]byte alignment
+        off += 3
+        pclntab[off] = f.nfuncdata
+        off += 1

        // NOTICE: _func.pcdata always starts from PcUnsafePoint, which is index 3
        for j := 3; j < len(pcdataOffs[i]); j++ {
@ -335,12 +340,17 @@ func makePclntable(funcs []_func, lastFuncSize uint32, pcdataOffs [][]uint32, fu
            off += 4
        }

+        off = uint32(rnd(int64(off), int64(_PtrSize)))
+
        // funcdata refs as offsets from gofunc
        for _, funcdata := range funcdataOffs[i] {
-            byteOrder.PutUint32(pclntab[off:off+4], uint32(funcdata))
-            off += 4
+            if funcdata == _INVALID_FUNCDATA_OFFSET {
+                byteOrder.PutUint64(pclntab[off:off+8], 0)
+            } else {
+                byteOrder.PutUint64(pclntab[off:off+8], uint64(funcdataAddr)+uint64(funcdata))
+            }
+            off += 8
        }
-
    }

    return
@ -364,14 +374,14 @@ func writeFindfunctab(out *[]byte, ftab []funcTab) (start int) {
    tab := make([]findfuncbucket, 0, nbuckets)
    var s, e = 0, 0
    for i := 0; i<int(nbuckets); i++ {
-        var pc = min + uint32((i+1)*_BUCKETSIZE)
+        var pc = min + uintptr((i+1)*_BUCKETSIZE)
        // find the end func of the bucket
        for ; e < len(ftab)-1 && ftab[e+1].entry <= pc; e++ {}
        // store the start func of the bucket
        var fb = findfuncbucket{idx: uint32(s)}

        for j := 0; j<_SUBBUCKETS && (i*_SUBBUCKETS+j)<int(n); j++ {
-            pc = min + uint32(i*_BUCKETSIZE) + uint32((j+1)*_SUB_BUCKETSIZE)
+            pc = min + uintptr(i*_BUCKETSIZE) + uintptr((j+1)*_SUB_BUCKETSIZE)
            var ss = s
            // find the end func of the subbucket
            for ; ss < len(ftab)-1 && ftab[ss+1].entry <= pc; ss++ {}
@ -408,29 +418,6 @@ func makeModuledata(name string, filenames []string, funcs []Func, text []byte)
    funcnametab, nameOffs := makeFuncnameTab(funcs)
    mod.funcnametab = funcnametab

-    // make pcdata table
-    // NOTICE: _func only use offset to index pcdata, thus no need mmap() pcdata 
-    pctab, pcdataOffs, _funcs := makePctab(funcs, cuOffs, nameOffs)
-    mod.pctab = pctab
-
-    // write func data
-    // NOTICE: _func use mod.gofunc+offset to directly point funcdata, thus need cache funcdata
-    // TODO: estimate accurate capacity
-    cache := make([]byte, 0, len(funcs)*int(_PtrSize)) 
-    fstart, funcdataOffs := writeFuncdata(&cache, funcs)
-
-    // make pc->func (binary search) func table
-    lastFuncsize := funcs[len(funcs)-1].TextSize
-    ftab := makeFtab(_funcs, lastFuncsize)
-    mod.ftab = ftab
-
-    // write pc->func (modmap) findfunc table
-    ffstart := writeFindfunctab(&cache, ftab)
-
-    // make pclnt table
-    pclntab := makePclntable(_funcs, lastFuncsize, pcdataOffs, funcdataOffs)
-    mod.pclntable = pclntab
-
    // mmap() text and funcdata segements
    p := os.Getpagesize()
    size := int(rnd(int64(len(text)), int64(p)))
@ -441,18 +428,41 @@ funcnametab, nameOffs := makeFuncnameTab(funcs)
    // make it executable
    mprotect(addr, size)

-    // assign addresses
-    mod.text = addr
-    mod.etext = addr + uintptr(size)
-    mod.minpc = addr
-    mod.maxpc = addr + uintptr(len(text))
+    // make pcdata table
+    // NOTICE: _func only use offset to index pcdata, thus no need mmap() pcdata 
+    pctab, pcdataOffs, _funcs := makePctab(funcs, addr, cuOffs, nameOffs)
+    mod.pctab = pctab
+
+    // write func data
+    // NOTICE: _func use mod.gofunc+offset to directly point funcdata, thus need cache funcdata
+    // TODO: estimate accurate capacity
+    cache := make([]byte, 0, len(funcs)*int(_PtrSize)) 
+    fstart, funcdataOffs := writeFuncdata(&cache, funcs)
+
+    // make pc->func (binary search) func table
+    lastFuncsize := funcs[len(funcs)-1].TextSize
+    ftab, pclntSize, startLocations := makeFtab(_funcs, lastFuncsize)
+    mod.ftab = ftab
+
+    // write pc->func (modmap) findfunc table
+    ffstart := writeFindfunctab(&cache, ftab)

    // cache funcdata and findfuncbucket
    moduleCache.Lock()
    moduleCache.m[mod] = cache
    moduleCache.Unlock()
-    mod.gofunc = uintptr(unsafe.Pointer(&cache[fstart]))
-    mod.findfunctab = uintptr(unsafe.Pointer(&cache[ffstart]))
+    mod.findfunctab = uintptr(rt.IndexByte(cache, ffstart))
+    funcdataAddr := uintptr(rt.IndexByte(cache, fstart))
+
+    // make pclnt table
+    pclntab := makePclntable(pclntSize, startLocations, _funcs, lastFuncsize, pcdataOffs, funcdataAddr, funcdataOffs)
+    mod.pclntable = pclntab
+
+    // assign addresses
+    mod.text = addr
+    mod.etext = addr + uintptr(size)
+    mod.minpc = addr
+    mod.maxpc = addr + uintptr(len(text))

    // make pc header
    mod.pcHeader = &pcHeader {
@ -461,7 +471,6 @@ funcnametab, nameOffs := makeFuncnameTab(funcs)
        ptrSize : _PtrSize,
        nfunc   : len(funcs),
        nfiles: uint(len(cu)),
-        textStart: mod.text,
        funcnameOffset: getOffsetOf(moduledata{}, "funcnametab"),
        cuOffset: getOffsetOf(moduledata{}, "cutab"),
        filetabOffset: getOffsetOf(moduledata{}, "filetab"),
@ -478,7 +487,7 @@ funcnameOffset: getOffsetOf(moduledata{}, "funcnametab"),

 // makePctab generates pcdelta->valuedelta tables for functions,
 // and returns the table and the entry offset of every kind pcdata in the table.
-func makePctab(funcs []Func, cuOffset []uint32, nameOffset []int32) (pctab []byte, pcdataOffs [][]uint32, _funcs []_func) {
+func makePctab(funcs []Func, addr uintptr, cuOffset []uint32, nameOffset []int32) (pctab []byte, pcdataOffs [][]uint32, _funcs []_func) {
    _funcs = make([]_func, len(funcs))

    // Pctab offsets of 0 are considered invalid in the runtime. We respect
@ -523,7 +532,7 @@ func makePctab(funcs []Func, cuOffset []uint32, nameOffset []int32) (pctab []byt
        writer(f.PcInlTreeIndex)
        writer(f.PcArgLiveIndex)
        
-        _f.entryOff = f.EntryOff
+        _f.entry = addr + uintptr(f.EntryOff)
        _f.nameOff = nameOffset[i]
        _f.args = f.ArgsSize
        _f.deferreturn = f.DeferReturn
@ -531,7 +540,6 @@ func makePctab(funcs []Func, cuOffset []uint32, nameOffset []int32) (pctab []byt
        _f.npcdata = uint32(_N_PCDATA)
        _f.cuOffset = cuOffset[i]
        _f.funcID = f.ID
-        _f.flag = f.Flag
        _f.nfuncdata = uint8(_N_FUNCDATA)
    }

--- a/vendor/github.com/bytedance/sonic/loader/funcdata_go116.go
+++ b/vendor/github.com/bytedance/sonic/loader/funcdata_go116.go
@ -0,0 +1,549 @@
+//go:build go1.16 && !go1.18
+// +build go1.16,!go1.18
+
+/*
+ * Copyright 2021 ByteDance Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package loader
+
+import (
+   `encoding`
+   `os`
+   `unsafe`
+
+   `github.com/bytedance/sonic/internal/rt`
+)
+
+const (
+    _Magic uint32 = 0xfffffffa
+)
+
+type pcHeader struct {
+    magic          uint32  // 0xFFFFFFF0
+    pad1, pad2     uint8   // 0,0
+    minLC          uint8   // min instruction size
+    ptrSize        uint8   // size of a ptr in bytes
+    nfunc          int     // number of functions in the module
+    nfiles         uint    // number of entries in the file tab
+    funcnameOffset uintptr // offset to the funcnametab variable from pcHeader
+    cuOffset       uintptr // offset to the cutab variable from pcHeader
+    filetabOffset  uintptr // offset to the filetab variable from pcHeader
+    pctabOffset    uintptr // offset to the pctab variable from pcHeader
+    pclnOffset     uintptr // offset to the pclntab variable from pcHeader
+}
+
+type moduledata struct {
+    pcHeader     *pcHeader
+    funcnametab  []byte
+    cutab        []uint32
+    filetab      []byte
+    pctab        []byte
+    pclntable    []byte
+    ftab         []funcTab
+    findfunctab  uintptr
+    minpc, maxpc uintptr // first func address, last func address + last func size
+
+    text, etext           uintptr // start/end of text, (etext-text) must be greater than MIN_FUNC
+    noptrdata, enoptrdata uintptr
+    data, edata           uintptr
+    bss, ebss             uintptr
+    noptrbss, enoptrbss   uintptr
+    end, gcdata, gcbss    uintptr
+    types, etypes         uintptr
+    
+    textsectmap []textSection // see runtime/symtab.go: textAddr()
+    typelinks   []int32 // offsets from types
+    itablinks   []*rt.GoItab
+
+    ptab []ptabEntry
+
+    pluginpath string
+    pkghashes  []modulehash
+
+    modulename   string
+    modulehashes []modulehash
+
+    hasmain uint8 // 1 if module contains the main function, 0 otherwise
+
+    gcdatamask, gcbssmask bitVector
+
+    typemap map[int32]*rt.GoType // offset to *_rtype in previous module
+
+    bad bool // module failed to load and should be ignored
+
+    next *moduledata
+}
+
+type _func struct {
+    entry    uintptr // start pc, as offset from moduledata.text/pcHeader.textStart
+    nameOff  int32  // function name, as index into moduledata.funcnametab.
+
+    args        int32  // in/out args size
+    deferreturn uint32 // offset of start of a deferreturn call instruction from entry, if any.
+
+    pcsp      uint32 
+    pcfile    uint32
+    pcln      uint32
+    npcdata   uint32
+    cuOffset  uint32 // runtime.cutab offset of this function's CU
+    funcID    uint8  // set for certain special runtime functions
+    _         [2]byte // pad
+    nfuncdata uint8   // 
+    
+    // The end of the struct is followed immediately by two variable-length
+    // arrays that reference the pcdata and funcdata locations for this
+    // function.
+
+    // pcdata contains the offset into moduledata.pctab for the start of
+    // that index's table. e.g.,
+    // &moduledata.pctab[_func.pcdata[_PCDATA_UnsafePoint]] is the start of
+    // the unsafe point table.
+    //
+    // An offset of 0 indicates that there is no table.
+    //
+    // pcdata [npcdata]uint32
+
+    // funcdata contains the offset past moduledata.gofunc which contains a
+    // pointer to that index's funcdata. e.g.,
+    // *(moduledata.gofunc +  _func.funcdata[_FUNCDATA_ArgsPointerMaps]) is
+    // the argument pointer map.
+    //
+    // An offset of ^uint32(0) indicates that there is no entry.
+    //
+    // funcdata [nfuncdata]uint32
+}
+
+type funcTab struct {
+    entry   uintptr
+    funcoff uintptr
+}
+
+type bitVector struct {
+    n        int32 // # of bits
+    bytedata *uint8
+}
+
+type ptabEntry struct {
+    name int32
+    typ  int32
+}
+
+type textSection struct {
+    vaddr    uintptr // prelinked section vaddr
+    end      uintptr // vaddr + section length
+    baseaddr uintptr // relocated section address
+}
+
+type modulehash struct {
+    modulename   string
+    linktimehash string
+    runtimehash  *string
+}
+
+// findfuncbucket is an array of these structures.
+// Each bucket represents 4096 bytes of the text segment.
+// Each subbucket represents 256 bytes of the text segment.
+// To find a function given a pc, locate the bucket and subbucket for
+// that pc. Add together the idx and subbucket value to obtain a
+// function index. Then scan the functab array starting at that
+// index to find the target function.
+// This table uses 20 bytes for every 4096 bytes of code, or ~0.5% overhead.
+type findfuncbucket struct {
+    idx        uint32
+    _SUBBUCKETS [16]byte
+}
+
+
+type compilationUnit struct {
+    fileNames []string
+}
+
+// func name table format: 
+//   nameOff[0] -> namePartA namePartB namePartC \x00 
+//   nameOff[1] -> namePartA namePartB namePartC \x00
+//  ...
+func makeFuncnameTab(funcs []Func) (tab []byte, offs []int32) {
+    offs = make([]int32, len(funcs))
+    offset := 0
+
+    for i, f := range funcs {
+        offs[i] = int32(offset)
+
+        a, b, c := funcNameParts(f.Name)
+        tab = append(tab, a...)
+        tab = append(tab, b...)
+        tab = append(tab, c...)
+        tab = append(tab, 0)
+        offset += len(a) + len(b) + len(c) + 1
+    }
+
+    return
+}
+
+// CU table format:
+//  cuOffsets[0] -> filetabOffset[0] filetabOffset[1] ... filetabOffset[len(CUs[0].fileNames)-1]
+//  cuOffsets[1] -> filetabOffset[len(CUs[0].fileNames)] ... filetabOffset[len(CUs[0].fileNames) + len(CUs[1].fileNames)-1]
+//  ...
+//
+// file name table format:
+//  filetabOffset[0] -> CUs[0].fileNames[0] \x00
+//  ...
+//  filetabOffset[len(CUs[0]-1)] -> CUs[0].fileNames[len(CUs[0].fileNames)-1] \x00
+//  ...
+//  filetabOffset[SUM(CUs,fileNames)-1] -> CUs[len(CU)-1].fileNames[len(CUs[len(CU)-1].fileNames)-1] \x00
+func makeFilenametab(cus []compilationUnit) (cutab []uint32, filetab []byte, cuOffsets []uint32) {
+    cuOffsets = make([]uint32, len(cus))
+    cuOffset := 0
+    fileOffset := 0
+
+    for i, cu := range cus {
+        cuOffsets[i] = uint32(cuOffset)
+
+        for _, name := range cu.fileNames {
+            cutab = append(cutab, uint32(fileOffset))
+
+            fileOffset += len(name) + 1
+            filetab = append(filetab, name...)
+            filetab = append(filetab, 0)
+        }
+
+        cuOffset += len(cu.fileNames)
+    }
+
+    return
+}
+
+func writeFuncdata(out *[]byte, funcs []Func) (fstart int, funcdataOffs [][]uint32) {
+    fstart = len(*out)
+    *out = append(*out, byte(0))
+    offs := uint32(1)
+
+    funcdataOffs = make([][]uint32, len(funcs))
+    for i, f := range funcs {
+
+        var writer = func(fd encoding.BinaryMarshaler) {
+            var ab []byte
+            var err error
+            if fd != nil {
+                ab, err = fd.MarshalBinary()
+                if err != nil {
+                    panic(err)
+                }
+                funcdataOffs[i] = append(funcdataOffs[i], offs)
+            } else {
+                ab = []byte{0}
+                funcdataOffs[i] = append(funcdataOffs[i], _INVALID_FUNCDATA_OFFSET)
+            }
+            *out = append(*out, ab...)
+            offs += uint32(len(ab))
+        }
+
+        writer(f.ArgsPointerMaps)
+        writer(f.LocalsPointerMaps)
+        writer(f.StackObjects)
+        writer(f.InlTree)
+        writer(f.OpenCodedDeferInfo)
+        writer(f.ArgInfo)
+        writer(f.ArgLiveInfo)
+        writer(f.WrapInfo)
+    }
+    return 
+}
+
+func makeFtab(funcs []_func, lastFuncSize uint32) (ftab []funcTab, pclntabSize int64, startLocations []uint32) {
+    // Allocate space for the pc->func table. This structure consists of a pc offset
+    // and an offset to the func structure. After that, we have a single pc
+    // value that marks the end of the last function in the binary.
+    pclntabSize = int64(len(funcs)*2*int(_PtrSize) + int(_PtrSize))
+    startLocations = make([]uint32, len(funcs))
+    for i, f := range funcs {
+        pclntabSize = rnd(pclntabSize, int64(_PtrSize))
+        //writePCToFunc
+        startLocations[i] = uint32(pclntabSize)
+        pclntabSize += int64(uint8(_FUNC_SIZE) + f.nfuncdata*_PtrSize + uint8(f.npcdata)*4)
+    }
+    ftab = make([]funcTab, 0, len(funcs)+1)
+
+    // write a map of pc->func info offsets 
+    for i, f := range funcs {
+        ftab = append(ftab, funcTab{uintptr(f.entry), uintptr(startLocations[i])})
+    }
+
+    // Final entry of table is just end pc offset.
+    lastFunc := funcs[len(funcs)-1]
+    ftab = append(ftab, funcTab{lastFunc.entry + uintptr(lastFuncSize), 0})
+
+    return
+}
+
+// Pcln table format: [...]funcTab + [...]_Func
+func makePclntable(size int64, startLocations []uint32, funcs []_func, lastFuncSize uint32, pcdataOffs [][]uint32, funcdataAddr uintptr, funcdataOffs [][]uint32) (pclntab []byte) {
+    pclntab = make([]byte, size, size)
+
+    // write a map of pc->func info offsets 
+    offs := 0
+    for i, f := range funcs {
+        byteOrder.PutUint64(pclntab[offs:offs+8], uint64(f.entry))
+        byteOrder.PutUint64(pclntab[offs+8:offs+16], uint64(startLocations[i]))
+        offs += 16
+    }
+    // Final entry of table is just end pc offset.
+    lastFunc := funcs[len(funcs)-1]
+    byteOrder.PutUint64(pclntab[offs:offs+8], uint64(lastFunc.entry)+uint64(lastFuncSize))
+    offs += 8
+
+    // write func info table
+    for i, f := range funcs {
+        off := startLocations[i]
+
+        // write _func structure to pclntab
+        byteOrder.PutUint64(pclntab[off:off+8], uint64(f.entry))
+        off += 8
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.nameOff))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.args))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.deferreturn))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.pcsp))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.pcfile))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.pcln))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.npcdata))
+        off += 4
+        byteOrder.PutUint32(pclntab[off:off+4], uint32(f.cuOffset))
+        off += 4
+        pclntab[off] = f.funcID
+        // NOTICE: _[2]byte alignment
+        off += 3
+        pclntab[off] = f.nfuncdata
+        off += 1
+
+        // NOTICE: _func.pcdata always starts from PcUnsafePoint, which is index 3
+        for j := 3; j < len(pcdataOffs[i]); j++ {
+            byteOrder.PutUint32(pclntab[off:off+4], uint32(pcdataOffs[i][j]))
+            off += 4
+        }
+
+        off = uint32(rnd(int64(off), int64(_PtrSize)))
+
+        // funcdata refs as offsets from gofunc
+        for _, funcdata := range funcdataOffs[i] {
+            if funcdata == _INVALID_FUNCDATA_OFFSET {
+                byteOrder.PutUint64(pclntab[off:off+8], 0)
+            } else {
+                byteOrder.PutUint64(pclntab[off:off+8], uint64(funcdataAddr)+uint64(funcdata))
+            }
+            off += 8
+        }
+    }
+
+    return
+}
+
+// findfunc table used to map pc to belonging func, 
+// returns the index in the func table.
+//
+// All text section are divided into buckets sized _BUCKETSIZE(4K):
+//   every bucket is divided into _SUBBUCKETS sized _SUB_BUCKETSIZE(64),
+//   and it has a base idx to plus the offset stored in jth subbucket.
+// see findfunc() in runtime/symtab.go
+func writeFindfunctab(out *[]byte, ftab []funcTab) (start int) {
+    start = len(*out)
+
+    max := ftab[len(ftab)-1].entry
+    min := ftab[0].entry
+    nbuckets := (max - min + _BUCKETSIZE - 1) / _BUCKETSIZE
+    n := (max - min + _SUB_BUCKETSIZE - 1) / _SUB_BUCKETSIZE
+
+    tab := make([]findfuncbucket, 0, nbuckets)
+    var s, e = 0, 0
+    for i := 0; i<int(nbuckets); i++ {
+        var pc = min + uintptr((i+1)*_BUCKETSIZE)
+        // find the end func of the bucket
+        for ; e < len(ftab)-1 && ftab[e+1].entry <= pc; e++ {}
+        // store the start func of the bucket
+        var fb = findfuncbucket{idx: uint32(s)}
+
+        for j := 0; j<_SUBBUCKETS && (i*_SUBBUCKETS+j)<int(n); j++ {
+            pc = min + uintptr(i*_BUCKETSIZE) + uintptr((j+1)*_SUB_BUCKETSIZE)
+            var ss = s
+            // find the end func of the subbucket
+            for ; ss < len(ftab)-1 && ftab[ss+1].entry <= pc; ss++ {}
+            // store the start func of the subbucket
+            fb._SUBBUCKETS[j] = byte(uint32(s) - fb.idx)
+            s = ss
+        }
+        s = e
+        tab = append(tab, fb)
+    }
+
+    // write findfuncbucket
+    if len(tab) > 0 {
+        size := int(unsafe.Sizeof(findfuncbucket{}))*len(tab)
+        *out = append(*out, rt.BytesFrom(unsafe.Pointer(&tab[0]), size, size)...)
+    }
+    return 
+}
+
+func makeModuledata(name string, filenames []string, funcs []Func, text []byte) (mod *moduledata) {
+    mod = new(moduledata)
+    mod.modulename = name
+
+    // make filename table
+    cu := make([]string, 0, len(filenames))
+    for _, f := range filenames {
+        cu = append(cu, f)
+    }
+    cutab, filetab, cuOffs := makeFilenametab([]compilationUnit{{cu}})
+    mod.cutab = cutab
+    mod.filetab = filetab
+
+    // make funcname table
+    funcnametab, nameOffs := makeFuncnameTab(funcs)
+    mod.funcnametab = funcnametab
+
+    // mmap() text and funcdata segements
+    p := os.Getpagesize()
+    size := int(rnd(int64(len(text)), int64(p)))
+    addr := mmap(size)
+    // copy the machine code
+    s := rt.BytesFrom(unsafe.Pointer(addr), len(text), size)
+    copy(s, text)
+    // make it executable
+    mprotect(addr, size)
+
+    // make pcdata table
+    // NOTICE: _func only use offset to index pcdata, thus no need mmap() pcdata 
+    pctab, pcdataOffs, _funcs := makePctab(funcs, addr, cuOffs, nameOffs)
+    mod.pctab = pctab
+
+    // write func data
+    // NOTICE: _func use mod.gofunc+offset to directly point funcdata, thus need cache funcdata
+    // TODO: estimate accurate capacity
+    cache := make([]byte, 0, len(funcs)*int(_PtrSize)) 
+    fstart, funcdataOffs := writeFuncdata(&cache, funcs)
+
+    // make pc->func (binary search) func table
+    lastFuncsize := funcs[len(funcs)-1].TextSize
+    ftab, pclntSize, startLocations := makeFtab(_funcs, lastFuncsize)
+    mod.ftab = ftab
+
+    // write pc->func (modmap) findfunc table
+    ffstart := writeFindfunctab(&cache, ftab)
+
+    // cache funcdata and findfuncbucket
+    moduleCache.Lock()
+    moduleCache.m[mod] = cache
+    moduleCache.Unlock()
+    mod.findfunctab = uintptr(rt.IndexByte(cache, ffstart))
+    funcdataAddr := uintptr(rt.IndexByte(cache, fstart))
+
+    // make pclnt table
+    pclntab := makePclntable(pclntSize, startLocations, _funcs, lastFuncsize, pcdataOffs, funcdataAddr, funcdataOffs)
+    mod.pclntable = pclntab
+
+    // assign addresses
+    mod.text = addr
+    mod.etext = addr + uintptr(size)
+    mod.minpc = addr
+    mod.maxpc = addr + uintptr(len(text))
+
+    // make pc header
+    mod.pcHeader = &pcHeader {
+        magic   : _Magic,
+        minLC   : _MinLC,
+        ptrSize : _PtrSize,
+        nfunc   : len(funcs),
+        nfiles: uint(len(cu)),
+        funcnameOffset: getOffsetOf(moduledata{}, "funcnametab"),
+        cuOffset: getOffsetOf(moduledata{}, "cutab"),
+        filetabOffset: getOffsetOf(moduledata{}, "filetab"),
+        pctabOffset: getOffsetOf(moduledata{}, "pctab"),
+        pclnOffset: getOffsetOf(moduledata{}, "pclntable"),
+    }
+
+    // sepecial case: gcdata and gcbss must by non-empty
+    mod.gcdata = uintptr(unsafe.Pointer(&emptyByte))
+    mod.gcbss = uintptr(unsafe.Pointer(&emptyByte))
+
+    return
+}
+
+// makePctab generates pcdelta->valuedelta tables for functions,
+// and returns the table and the entry offset of every kind pcdata in the table.
+func makePctab(funcs []Func, addr uintptr, cuOffset []uint32, nameOffset []int32) (pctab []byte, pcdataOffs [][]uint32, _funcs []_func) {
+    _funcs = make([]_func, len(funcs))
+
+    // Pctab offsets of 0 are considered invalid in the runtime. We respect
+    // that by just padding a single byte at the beginning of runtime.pctab,
+    // that way no real offsets can be zero.
+    pctab = make([]byte, 1, 12*len(funcs)+1)
+    pcdataOffs = make([][]uint32, len(funcs))
+
+    for i, f := range funcs {
+        _f := &_funcs[i]
+
+        var writer = func(pc *Pcdata) {
+            var ab []byte
+            var err error
+            if pc != nil {
+                ab, err = pc.MarshalBinary()
+                if err != nil {
+                    panic(err)
+                }
+                pcdataOffs[i] = append(pcdataOffs[i], uint32(len(pctab)))
+            } else {
+                ab = []byte{0}
+                pcdataOffs[i] = append(pcdataOffs[i], _PCDATA_INVALID_OFFSET)
+            }
+            pctab = append(pctab, ab...)
+        }
+
+        if f.Pcsp != nil {
+            _f.pcsp = uint32(len(pctab))
+        }
+        writer(f.Pcsp)
+        if f.Pcfile != nil {
+            _f.pcfile = uint32(len(pctab))
+        }
+        writer(f.Pcfile)
+        if f.Pcline != nil {
+            _f.pcln = uint32(len(pctab))
+        }
+        writer(f.Pcline)
+        writer(f.PcUnsafePoint)
+        writer(f.PcStackMapIndex)
+        writer(f.PcInlTreeIndex)
+        writer(f.PcArgLiveIndex)
+        
+        _f.entry = addr + uintptr(f.EntryOff)
+        _f.nameOff = nameOffset[i]
+        _f.args = f.ArgsSize
+        _f.deferreturn = f.DeferReturn
+        // NOTICE: _func.pcdata is always as [PCDATA_UnsafePoint(0) : PCDATA_ArgLiveIndex(3)]
+        _f.npcdata = uint32(_N_PCDATA)
+        _f.cuOffset = cuOffset[i]
+        _f.funcID = f.ID
+        _f.nfuncdata = uint8(_N_FUNCDATA)
+    }
+
+    return
+}
+
+func registerFunction(name string, pc uintptr, textSize uintptr, fp int, args int, size uintptr, argptrs uintptr, localptrs uintptr) {} 
--- a/vendor/github.com/bytedance/sonic/loader/loader_go115.go
+++ b/vendor/github.com/bytedance/sonic/loader/loader_go115.go
@ -1,5 +1,5 @@
-//go:build go1.15 && !go1.18
-// +build go1.15,!go1.18
+//go:build go1.15 && !go1.16
+// +build go1.15,!go1.16

 /*
 * Copyright 2021 ByteDance Inc.
@ -20,14 +20,9 @@
 package loader

 import (
-
    `github.com/bytedance/sonic/internal/loader`
 )

 func (self Loader) LoadOne(text []byte, funcName string, frameSize int, argSize int, argStackmap []bool, localStackmap []bool) Function {
    return Function(loader.Loader(text).Load(funcName, frameSize, argSize, argStackmap, localStackmap))
-}
-
-func Load(modulename string, filenames []string, funcs []Func, text []byte) (out []Function) {
-    panic("not implemented") 
 }
--- a/vendor/github.com/bytedance/sonic/loader/loader_go116.go
+++ b/vendor/github.com/bytedance/sonic/loader/loader_go116.go
@ -1,5 +1,5 @@
-//go:build go1.18 && !go1.21
-// +build go1.18,!go1.21
+//go:build go1.16 && !go1.21
+// +build go1.16,!go1.21

 /*
 * Copyright 2021 ByteDance Inc.
--- a/vendor/github.com/bytedance/sonic/loader/stubs.go
+++ b/vendor/github.com/bytedance/sonic/loader/stubs.go
@ -1,12 +1,12 @@
 /**
 * Copyright 2023 ByteDance Inc.
- * 
+ *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
- * 
+ *
 *     http://www.apache.org/licenses/LICENSE-2.0
- * 
+ *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
@ -17,16 +17,21 @@
 package loader

 import (
-	_ `unsafe`
+    `sync`
+    _ `unsafe`
 )

 //go:linkname lastmoduledatap runtime.lastmoduledatap
 //goland:noinspection GoUnusedGlobalVariable
 var lastmoduledatap *moduledata

+var moduledataMux sync.Mutex
+
 func registerModule(mod *moduledata) {
+    moduledataMux.Lock()
    lastmoduledatap.next = mod
    lastmoduledatap = mod
+    moduledataMux.Unlock()
 }

 //go:linkname moduledataverify1 runtime.moduledataverify1
--- a/vendor/github.com/bytedance/sonic/option/option.go
+++ b/vendor/github.com/bytedance/sonic/option/option.go
@ -16,6 +16,14 @@

 package option

+var (
+    // DefaultDecoderBufferSize is the initial buffer size of StreamDecoder
+    DefaultDecoderBufferSize  uint = 128 * 1024
+
+    // DefaultEncoderBufferSize is the initial buffer size of Encoder
+    DefaultEncoderBufferSize  uint = 128 * 1024
+)
+
 // CompileOptions includes all options for encoder or decoder compiler.
 type CompileOptions struct {
    // the maximum depth for compilation inline
--- a/vendor/github.com/bytedance/sonic/other-langs.png
+++ b/vendor/github.com/bytedance/sonic/other-langs.png
--- a/vendor/github.com/bytedance/sonic/unquote/unquote.go
+++ b/vendor/github.com/bytedance/sonic/unquote/unquote.go
@ -18,6 +18,7 @@

 import (
    `unsafe`
+    `runtime`

    `github.com/bytedance/sonic/internal/native`
    `github.com/bytedance/sonic/internal/native/types`
@ -43,7 +44,8 @@ func intoBytesUnsafe(s string, m *[]byte) types.ParsingError {
    pos := -1
    slv := (*rt.GoSlice)(unsafe.Pointer(m))
    str := (*rt.GoString)(unsafe.Pointer(&s))
-    ret := native.Unquote(str.Ptr, str.Len, slv.Ptr, &pos, 0)
+    /* unquote as the default configuration, replace invalid unicode with \ufffd */
+    ret := native.Unquote(str.Ptr, str.Len, slv.Ptr, &pos, types.F_UNICODE_REPLACE)

    /* check for errors */
    if ret < 0 {
@ -52,5 +54,6 @@ func intoBytesUnsafe(s string, m *[]byte) types.ParsingError {

    /* update the length */
    slv.Len = ret
+    runtime.KeepAlive(s)
    return 0
 }
--- a/vendor/github.com/gin-gonic/gin/.golangci.yml
+++ b/vendor/github.com/gin-gonic/gin/.golangci.yml
@ -19,6 +19,22 @@ linters:
    - nolintlint
    - revive
    - wastedassign
+
+linters-settings:
+  gosec:
+    # To select a subset of rules to run.
+    # Available rules: https://github.com/securego/gosec#available-rules
+    # Default: [] - means include all rules
+    includes:
+      - G102
+      - G106
+      - G108
+      - G109
+      - G111
+      - G112
+      - G201
+      - G203
+
 issues:
  exclude-rules:
    - linters:
@ -37,3 +53,6 @@ issues:
    - path: _test\.go
      linters:
        - gosec # security is not make sense in tests
+    - linters:
+        - revive
+      path: _test\.go
--- a/vendor/github.com/gin-gonic/gin/CHANGELOG.md
+++ b/vendor/github.com/gin-gonic/gin/CHANGELOG.md
@ -1,5 +1,26 @@
 # Gin ChangeLog

+## Gin v1.9.1
+
+### BUG FIXES 
+
+* fix Request.Context() checks [#3512](https://github.com/gin-gonic/gin/pull/3512)
+
+### SECURITY
+
+* fix lack of escaping of filename in Content-Disposition [#3556](https://github.com/gin-gonic/gin/pull/3556) 
+
+### ENHANCEMENTS
+
+* refactor: use bytes.ReplaceAll directly [#3455](https://github.com/gin-gonic/gin/pull/3455)
+* convert strings and slices using the officially recommended way [#3344](https://github.com/gin-gonic/gin/pull/3344)
+* improve render code coverage [#3525](https://github.com/gin-gonic/gin/pull/3525)
+
+### DOCS
+
+* docs: changed documentation link for trusted proxies [#3575](https://github.com/gin-gonic/gin/pull/3575)
+* chore: improve linting, testing, and GitHub Actions setup [#3583](https://github.com/gin-gonic/gin/pull/3583)
+
 ## Gin v1.9.0

 ### BREAK CHANGES
--- a/vendor/github.com/gin-gonic/gin/README.md
+++ b/vendor/github.com/gin-gonic/gin/README.md
@ -6,7 +6,6 @@
 [![codecov](https://codecov.io/gh/gin-gonic/gin/branch/master/graph/badge.svg)](https://codecov.io/gh/gin-gonic/gin)
 [![Go Report Card](https://goreportcard.com/badge/github.com/gin-gonic/gin)](https://goreportcard.com/report/github.com/gin-gonic/gin)
 [![GoDoc](https://pkg.go.dev/badge/github.com/gin-gonic/gin?status.svg)](https://pkg.go.dev/github.com/gin-gonic/gin?tab=doc)
-[![Join the chat at https://gitter.im/gin-gonic/gin](https://badges.gitter.im/Join%20Chat.svg)](https://gitter.im/gin-gonic/gin?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
 [![Sourcegraph](https://sourcegraph.com/github.com/gin-gonic/gin/-/badge.svg)](https://sourcegraph.com/github.com/gin-gonic/gin?badge)
 [![Open Source Helpers](https://www.codetriage.com/gin-gonic/gin/badges/users.svg)](https://www.codetriage.com/gin-gonic/gin)
 [![Release](https://img.shields.io/github/release/gin-gonic/gin.svg?style=flat-square)](https://github.com/gin-gonic/gin/releases)
@ -31,7 +30,7 @@ Gin is a web framework written in [Go](https://go.dev/). It features a martini-l

 ### Prerequisites

- **[Go](https://go.dev/)**: ~~any one of the **three latest major** [releases](https://go.dev/doc/devel/release)~~ (now version **1.16+** is required).
+- **[Go](https://go.dev/)**: any one of the **three latest major** [releases](https://go.dev/doc/devel/release) (we test it with these).

 ### Getting Gin

@ -176,4 +175,4 @@ Awesome project lists using [Gin](https://github.com/gin-gonic/gin) web framewor

 Gin is the work of hundreds of contributors. We appreciate your help!

-Please see [CONTRIBUTING](CONTRIBUTING.md) for details on submitting patches and the contribution workflow.
+Please see [CONTRIBUTING](CONTRIBUTING.md) for details on submitting patches and the contribution workflow.
--- a/vendor/github.com/gin-gonic/gin/any.go
+++ b/vendor/github.com/gin-gonic/gin/any.go
@ -1,10 +0,0 @@
-// Copyright 2022 Gin Core Team. All rights reserved.
-// Use of this source code is governed by a MIT style
-// license that can be found in the LICENSE file.
-
-//go:build !go1.18
-// +build !go1.18
-
-package gin
-
-type any = interface{}
--- a/vendor/github.com/gin-gonic/gin/binding/any.go
+++ b/vendor/github.com/gin-gonic/gin/binding/any.go
@ -1,10 +0,0 @@
-// Copyright 2022 Gin Core Team. All rights reserved.
-// Use of this source code is governed by a MIT style
-// license that can be found in the LICENSE file.
-
-//go:build !go1.18
-// +build !go1.18
-
-package binding
-
-type any = interface{}
--- a/vendor/github.com/gin-gonic/gin/binding/binding.go
+++ b/vendor/github.com/gin-gonic/gin/binding/binding.go
@ -3,7 +3,6 @@
 // license that can be found in the LICENSE file.

 //go:build !nomsgpack
-// +build !nomsgpack

 package binding

--- a/vendor/github.com/gin-gonic/gin/binding/binding_nomsgpack.go
+++ b/vendor/github.com/gin-gonic/gin/binding/binding_nomsgpack.go
@ -3,7 +3,6 @@
 // license that can be found in the LICENSE file.

 //go:build nomsgpack
-// +build nomsgpack

 package binding

--- a/vendor/github.com/gin-gonic/gin/binding/json.go
+++ b/vendor/github.com/gin-gonic/gin/binding/json.go
@ -15,7 +15,7 @@

 // EnableDecoderUseNumber is used to call the UseNumber method on the JSON
 // Decoder instance. UseNumber causes the Decoder to unmarshal a number into an
-// interface{} as a Number instead of as a float64.
+// any as a Number instead of as a float64.
 var EnableDecoderUseNumber = false

 // EnableDecoderDisallowUnknownFields is used to call the DisallowUnknownFields method
--- a/vendor/github.com/gin-gonic/gin/binding/msgpack.go
+++ b/vendor/github.com/gin-gonic/gin/binding/msgpack.go
@ -3,7 +3,6 @@
 // license that can be found in the LICENSE file.

 //go:build !nomsgpack
-// +build !nomsgpack

 package binding

--- a/vendor/github.com/gin-gonic/gin/codecov.yml
+++ b/vendor/github.com/gin-gonic/gin/codecov.yml
@ -1,5 +0,0 @@
-coverage:
-  notify:
-    gitter:
-      default:
-        url: https://webhooks.gitter.im/e/d90dcdeeab2f1e357165
--- a/vendor/github.com/gin-gonic/gin/context.go
+++ b/vendor/github.com/gin-gonic/gin/context.go
@ -652,7 +652,7 @@ func (c *Context) BindYAML(obj any) error {
 }

 // BindTOML is a shortcut for c.MustBindWith(obj, binding.TOML).
-func (c *Context) BindTOML(obj interface{}) error {
+func (c *Context) BindTOML(obj any) error {
 	return c.MustBindWith(obj, binding.TOML)
 }

@ -717,7 +717,7 @@ func (c *Context) ShouldBindYAML(obj any) error {
 }

 // ShouldBindTOML is a shortcut for c.ShouldBindWith(obj, binding.TOML).
-func (c *Context) ShouldBindTOML(obj interface{}) error {
+func (c *Context) ShouldBindTOML(obj any) error {
 	return c.ShouldBindWith(obj, binding.TOML)
 }

@ -995,7 +995,7 @@ func (c *Context) YAML(code int, obj any) {
 }

 // TOML serializes the given struct as TOML into the response body.
-func (c *Context) TOML(code int, obj interface{}) {
+func (c *Context) TOML(code int, obj any) {
 	c.Render(code, render.TOML{Data: obj})
 }

@ -1052,11 +1052,17 @@ func (c *Context) FileFromFS(filepath string, fs http.FileSystem) {
 	http.FileServer(fs).ServeHTTP(c.Writer, c.Request)
 }

+var quoteEscaper = strings.NewReplacer("\\", "\\\\", `"`, "\\\"")
+
+func escapeQuotes(s string) string {
+	return quoteEscaper.Replace(s)
+}
+
 // FileAttachment writes the specified file into the body stream in an efficient way
 // On the client side, the file will typically be downloaded with the given filename
 func (c *Context) FileAttachment(filepath, filename string) {
 	if isASCII(filename) {
-		c.Writer.Header().Set("Content-Disposition", `attachment; filename="`+filename+`"`)
+		c.Writer.Header().Set("Content-Disposition", `attachment; filename="`+escapeQuotes(filename)+`"`)
 	} else {
 		c.Writer.Header().Set("Content-Disposition", `attachment; filename*=UTF-8''`+url.QueryEscape(filename))
 	}
@ -1174,9 +1180,16 @@ func (c *Context) SetAccepted(formats ...string) {
 /***** GOLANG.ORG/X/NET/CONTEXT *****/
 /************************************/

+// hasRequestContext returns whether c.Request has Context and fallback.
+func (c *Context) hasRequestContext() bool {
+	hasFallback := c.engine != nil && c.engine.ContextWithFallback
+	hasRequestContext := c.Request != nil && c.Request.Context() != nil
+	return hasFallback && hasRequestContext
+}
+
 // Deadline returns that there is no deadline (ok==false) when c.Request has no Context.
 func (c *Context) Deadline() (deadline time.Time, ok bool) {
-	if !c.engine.ContextWithFallback || c.Request == nil || c.Request.Context() == nil {
+	if !c.hasRequestContext() {
 		return
 	}
 	return c.Request.Context().Deadline()
@ -1184,7 +1197,7 @@ func (c *Context) Deadline() (deadline time.Time, ok bool) {

 // Done returns nil (chan which will wait forever) when c.Request has no Context.
 func (c *Context) Done() <-chan struct{} {
-	if !c.engine.ContextWithFallback || c.Request == nil || c.Request.Context() == nil {
+	if !c.hasRequestContext() {
 		return nil
 	}
 	return c.Request.Context().Done()
@ -1192,7 +1205,7 @@ func (c *Context) Done() <-chan struct{} {

 // Err returns nil when c.Request has no Context.
 func (c *Context) Err() error {
-	if !c.engine.ContextWithFallback || c.Request == nil || c.Request.Context() == nil {
+	if !c.hasRequestContext() {
 		return nil
 	}
 	return c.Request.Context().Err()
@ -1213,7 +1226,7 @@ func (c *Context) Value(key any) any {
 			return val
 		}
 	}
-	if !c.engine.ContextWithFallback || c.Request == nil || c.Request.Context() == nil {
+	if !c.hasRequestContext() {
 		return nil
 	}
 	return c.Request.Context().Value(key)
--- a/vendor/github.com/gin-gonic/gin/context_appengine.go
+++ b/vendor/github.com/gin-gonic/gin/context_appengine.go
@ -3,7 +3,6 @@
 // license that can be found in the LICENSE file.

 //go:build appengine
-// +build appengine

 package gin

--- a/vendor/github.com/gin-gonic/gin/debug.go
+++ b/vendor/github.com/gin-gonic/gin/debug.go
@ -12,7 +12,7 @@
 	"strings"
 )

-const ginSupportMinGoVer = 16
+const ginSupportMinGoVer = 18

 // IsDebugging returns true if the framework is running in debug mode.
 // Use SetMode(gin.ReleaseMode) to disable debug mode.
@ -67,7 +67,7 @@ func getMinVer(v string) (uint64, error) {

 func debugPrintWARNINGDefault() {
 	if v, e := getMinVer(runtime.Version()); e == nil && v < ginSupportMinGoVer {
-		debugPrint(`[WARNING] Now Gin requires Go 1.16+.
+		debugPrint(`[WARNING] Now Gin requires Go 1.18+.

 `)
 	}
--- a/vendor/github.com/gin-gonic/gin/deprecated.go
+++ b/vendor/github.com/gin-gonic/gin/deprecated.go
@ -13,7 +13,7 @@
 // BindWith binds the passed struct pointer using the specified binding engine.
 // See the binding package.
 func (c *Context) BindWith(obj any, b binding.Binding) error {
-	log.Println(`BindWith(\"interface{}, binding.Binding\") error is going to
+	log.Println(`BindWith(\"any, binding.Binding\") error is going to
 	be deprecated, please check issue #662 and either use MustBindWith() if you
 	want HTTP 400 to be automatically returned if any error occur, or use
 	ShouldBindWith() if you need to manage the error.`)
--- a/vendor/github.com/gin-gonic/gin/fs.go
+++ b/vendor/github.com/gin-gonic/gin/fs.go
@ -39,7 +39,7 @@ func (fs onlyFilesFS) Open(name string) (http.File, error) {
 }

 // Readdir overrides the http.File default implementation.
-func (f neuteredReaddirFile) Readdir(count int) ([]os.FileInfo, error) {
+func (f neuteredReaddirFile) Readdir(_ int) ([]os.FileInfo, error) {
 	// this disables directory listing
 	return nil, nil
 }
--- a/vendor/github.com/gin-gonic/gin/gin.go
+++ b/vendor/github.com/gin-gonic/gin/gin.go
@ -515,7 +515,7 @@ func (engine *Engine) RunUnix(file string) (err error) {

 	if engine.isUnsafeTrustedProxies() {
 		debugPrint("[WARNING] You trusted all proxies, this is NOT safe. We recommend you to set a value.\n" +
-			"Please check https://pkg.go.dev/github.com/gin-gonic/gin#readme-don-t-trust-all-proxies for details.")
+			"Please check https://github.com/gin-gonic/gin/blob/master/docs/doc.md#dont-trust-all-proxies for details.")
 	}

 	listener, err := net.Listen("unix", file)
@ -538,7 +538,7 @@ func (engine *Engine) RunFd(fd int) (err error) {

 	if engine.isUnsafeTrustedProxies() {
 		debugPrint("[WARNING] You trusted all proxies, this is NOT safe. We recommend you to set a value.\n" +
-			"Please check https://pkg.go.dev/github.com/gin-gonic/gin#readme-don-t-trust-all-proxies for details.")
+			"Please check https://github.com/gin-gonic/gin/blob/master/docs/doc.md#dont-trust-all-proxies for details.")
 	}

 	f := os.NewFile(uintptr(fd), fmt.Sprintf("fd@%d", fd))
@ -559,7 +559,7 @@ func (engine *Engine) RunListener(listener net.Listener) (err error) {

 	if engine.isUnsafeTrustedProxies() {
 		debugPrint("[WARNING] You trusted all proxies, this is NOT safe. We recommend you to set a value.\n" +
-			"Please check https://pkg.go.dev/github.com/gin-gonic/gin#readme-don-t-trust-all-proxies for details.")
+			"Please check https://github.com/gin-gonic/gin/blob/master/docs/doc.md#dont-trust-all-proxies for details.")
 	}

 	err = http.Serve(listener, engine.Handler())
--- a/vendor/github.com/gin-gonic/gin/internal/bytesconv/bytesconv_1.19.go
+++ b/vendor/github.com/gin-gonic/gin/internal/bytesconv/bytesconv_1.19.go
@ -2,6 +2,8 @@
 // Use of this source code is governed by a MIT style
 // license that can be found in the LICENSE file.

+//go:build !go1.20
+
 package bytesconv

 import (
--- a/vendor/github.com/gin-gonic/gin/internal/bytesconv/bytesconv_1.20.go
+++ b/vendor/github.com/gin-gonic/gin/internal/bytesconv/bytesconv_1.20.go
@ -0,0 +1,23 @@
+// Copyright 2023 Gin Core Team. All rights reserved.
+// Use of this source code is governed by a MIT style
+// license that can be found in the LICENSE file.
+
+//go:build go1.20
+
+package bytesconv
+
+import (
+	"unsafe"
+)
+
+// StringToBytes converts string to byte slice without a memory allocation.
+// For more details, see https://github.com/golang/go/issues/53003#issuecomment-1140276077.
+func StringToBytes(s string) []byte {
+	return unsafe.Slice(unsafe.StringData(s), len(s))
+}
+
+// BytesToString converts byte slice to string without a memory allocation.
+// For more details, see https://github.com/golang/go/issues/53003#issuecomment-1140276077.
+func BytesToString(b []byte) string {
+	return unsafe.String(unsafe.SliceData(b), len(b))
+}
--- a/vendor/github.com/gin-gonic/gin/internal/json/go_json.go
+++ b/vendor/github.com/gin-gonic/gin/internal/json/go_json.go
@ -3,7 +3,6 @@
 // license that can be found in the LICENSE file.

 //go:build go_json
-// +build go_json

 package json

--- a/Show More
+++ b/Show More