feat:支持解压Gzip Content-Encoding格式的数据包 - #142
Open
zhicheng233 wants to merge 1 commit into
Open
Conversation
Reviewer's Guide添加一个可选的模组配置，通过挂钩 NetHttpClient.Decompress 来支持 gzip 和更健壮的 zlib/deflate 处理，在无法解压时回退为纯文本。 针对 gzip/zlib/纯文本处理的 NetHttpClient.Decompress 补丁序列图sequenceDiagram
participant NetHttpClient
participant MoreContentEncoding
participant GZipStream
participant DeflateStream
NetHttpClient->>MoreContentEncoding: PreDecompress(__instance)
MoreContentEncoding->>MoreContentEncoding: Traverse.Field _temporaryStream/_memoryStream/_buffer
MoreContentEncoding->>MoreContentEncoding: memoryStream.SetLength(0)
alt [temporaryStream.Length == 0]
MoreContentEncoding-->>NetHttpClient: return false
else [raw starts with gzip header]
MoreContentEncoding->>GZipStream: new GZipStream(MemoryStream(raw), Decompress)
MoreContentEncoding->>MoreContentEncoding: CopyTo(gz, memoryStream, buffer)
else [raw starts with zlib header and TryInflateZlib succeeds]
MoreContentEncoding->>DeflateStream: new DeflateStream(MemoryStream(raw,2,...), Decompress)
MoreContentEncoding->>MoreContentEncoding: CopyTo(deflate, memoryStream, buffer)
else [no compression detected]
MoreContentEncoding->>MoreContentEncoding: memoryStream.Write(raw)
end
MoreContentEncoding->>MoreContentEncoding: memoryStream.Seek(0)
MoreContentEncoding->>MoreContentEncoding: temporaryStream.SetLength(0)
MoreContentEncoding-->>NetHttpClient: return false
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your Experience访问你的 dashboard 来：
Getting HelpOriginal review guide in EnglishReviewer's GuideAdds an optional mod configuration that hooks NetHttpClient.Decompress to support gzip and robust zlib/deflate handling, falling back to plaintext when decompression is not possible. Sequence diagram for patched NetHttpClient.Decompress with gzip/zlib/plaintext handlingsequenceDiagram
participant NetHttpClient
participant MoreContentEncoding
participant GZipStream
participant DeflateStream
NetHttpClient->>MoreContentEncoding: PreDecompress(__instance)
MoreContentEncoding->>MoreContentEncoding: Traverse.Field _temporaryStream/_memoryStream/_buffer
MoreContentEncoding->>MoreContentEncoding: memoryStream.SetLength(0)
alt [temporaryStream.Length == 0]
MoreContentEncoding-->>NetHttpClient: return false
else [raw starts with gzip header]
MoreContentEncoding->>GZipStream: new GZipStream(MemoryStream(raw), Decompress)
MoreContentEncoding->>MoreContentEncoding: CopyTo(gz, memoryStream, buffer)
else [raw starts with zlib header and TryInflateZlib succeeds]
MoreContentEncoding->>DeflateStream: new DeflateStream(MemoryStream(raw,2,...), Decompress)
MoreContentEncoding->>MoreContentEncoding: CopyTo(deflate, memoryStream, buffer)
else [no compression detected]
MoreContentEncoding->>MoreContentEncoding: memoryStream.Write(raw)
end
MoreContentEncoding->>MoreContentEncoding: memoryStream.Seek(0)
MoreContentEncoding->>MoreContentEncoding: temporaryStream.SetLength(0)
MoreContentEncoding-->>NetHttpClient: return false
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
There was a problem hiding this comment.
Pull request overview
This PR adds a new GameSystem mod to extend the game client’s HTTP payload decompression beyond the stock zlib/deflate path, enabling compatibility with servers/CDNs that deliver gzip-encoded responses (e.g., Cloudflare setups that don’t support deflate).
Changes:
- Introduces
GameSystem.MoreContentEncoding, a Harmony prefix patch overNetHttpClient.Decompressthat detects gzip vs zlib vs plaintext. - Registers the new config section key in
configSort.yamlso it appears in the intended ordering.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.
| File | Description |
|---|---|
| AquaMai/configSort.yaml | Adds GameSystem.MoreContentEncoding to config key ordering. |
| AquaMai.Mods/GameSystem/MoreContentEncoding.cs | New decompression patch supporting gzip + zlib + plaintext fallback. |
Suppressed comments (2)
AquaMai.Mods/GameSystem/MoreContentEncoding.cs:53
- The comment looks unfinished and contains an unmatched "(" which reads like a typo and may confuse future readers.
else if (raw.Length >= 6 && raw[0] == 0x78 && TryInflateZlib(raw, memoryStream, buffer))
{
// zlib 成功解压(
}
AquaMai.Mods/GameSystem/MoreContentEncoding.cs:89
- CopyTo reads with a hard-coded BufferSize (1024) but uses the instance-provided
buffer. If_bufferis ever smaller than 1024,Stream.Readwill throw becausecountexceedsbuffer.Length. It’s safer to cap the read size bybuffer.Length.
{
var count = from.Read(buffer, 0, BufferSize);
if (count <= 0) break;
to.Write(buffer, 0, count);
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
Comment on lines
+45
to
+49
| if (raw.Length >= 2 && raw[0] == 0x1F && raw[1] == 0x8B) | ||
| { | ||
| using var gz = new GZipStream(new MemoryStream(raw, writable: false), CompressionMode.Decompress); | ||
| CopyTo(gz, memoryStream, buffer); | ||
| } |
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
There was a problem hiding this comment.
Hey - 我发现了一个问题，并给出了一些总体反馈：
- 在
CopyTo中，你在Read调用里使用了BufferSize常量，但传入的缓冲区长度可能不同；建议要么移除BufferSize并始终使用buffer.Length，要么验证buffer.Length == BufferSize，以避免越界或低效的读取。 - Harmony 的前缀假设私有字段
_temporaryStream、_memoryStream和_buffer总是非空；添加空值检查或回退逻辑会让这个补丁在面对 NetHttpClient 内部实现的未来变更时更加健壮。
Prompt for AI Agents
Please address the comments from this code review:
## Overall Comments
- 在 `CopyTo` 中，你在 `Read` 调用里使用了 `BufferSize` 常量，但传入的缓冲区长度可能不同；建议要么移除 `BufferSize` 并始终使用 `buffer.Length`，要么验证 `buffer.Length == BufferSize`，以避免越界或低效的读取。
- Harmony 的前缀假设私有字段 `_temporaryStream`、`_memoryStream` 和 `_buffer` 总是非空；添加空值检查或回退逻辑会让这个补丁在面对 NetHttpClient 内部实现的未来变更时更加健壮。
## Individual Comments
### Comment 1
<location path="AquaMai.Mods/GameSystem/MoreContentEncoding.cs" line_range="23-32" />
<code_context>
+ private const int BufferSize = 1024;
</code_context>
<issue_to_address>
**issue (bug_risk):** CopyTo 使用固定的 BufferSize 而不是实际的缓冲区长度，如果底层缓冲区大小发生变化，就可能产生不匹配。
`CopyTo` 接受一个 `buffer` 参数，但在读取时总是使用 `BufferSize` 常量。如果 NetHttpClient 中的 `_buffer` 不是恰好 1024 字节，或者将来发生变化，就可能导致过度读取和运行时错误。考虑改为使用 `buffer.Length`，或者移除缓冲区参数并明确强制使用固定大小缓冲区的约定。
</issue_to_address>帮我变得更有用！请在每条评论上点击 👍 或 👎，我会根据你的反馈改进后续的审查。
Original comment in English
Hey - I've found 1 issue, and left some high level feedback:
- In CopyTo you use the BufferSize constant for the Read call but pass in a buffer whose length may differ; consider either removing BufferSize and always using buffer.Length or validating that buffer.Length == BufferSize to avoid overruns or inefficient reads.
- The Harmony prefix assumes the private fields _temporaryStream, _memoryStream, and _buffer are always non-null; adding null checks or fallback behavior would make the patch more robust against future changes in NetHttpClient’s internals.
Prompt for AI Agents
Please address the comments from this code review:
## Overall Comments
- In CopyTo you use the BufferSize constant for the Read call but pass in a buffer whose length may differ; consider either removing BufferSize and always using buffer.Length or validating that buffer.Length == BufferSize to avoid overruns or inefficient reads.
- The Harmony prefix assumes the private fields _temporaryStream, _memoryStream, and _buffer are always non-null; adding null checks or fallback behavior would make the patch more robust against future changes in NetHttpClient’s internals.
## Individual Comments
### Comment 1
<location path="AquaMai.Mods/GameSystem/MoreContentEncoding.cs" line_range="23-32" />
<code_context>
+ private const int BufferSize = 1024;
</code_context>
<issue_to_address>
**issue (bug_risk):** CopyTo uses a fixed BufferSize instead of the actual buffer length, which can cause mismatches if the underlying buffer size changes.
`CopyTo` takes a `buffer` argument but always uses the `BufferSize` constant when reading. If `_buffer` in `NetHttpClient` is not exactly 1024 bytes or changes in the future, this can lead to over-reads and runtime errors. Consider using `buffer.Length` instead, or remove the buffer parameter and enforce a fixed-size buffer contract explicitly.
</issue_to_address>Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.
Comment on lines
+23
to
+32
| private const int BufferSize = 1024; | ||
|
|
||
| [HarmonyPrefix] | ||
| [HarmonyPatch(typeof(NetHttpClient), "Decompress")] | ||
| public static bool PreDecompress(NetHttpClient __instance) | ||
| { | ||
| var traverse = Traverse.Create(__instance); | ||
| var temporaryStream = traverse.Field<MemoryStream>("_temporaryStream").Value; | ||
| var memoryStream = traverse.Field<MemoryStream>("_memoryStream").Value; | ||
| var buffer = traverse.Field<byte[]>("_buffer").Value; |
There was a problem hiding this comment.
issue (bug_risk): CopyTo 使用固定的 BufferSize 而不是实际的缓冲区长度，如果底层缓冲区大小发生变化，就可能产生不匹配。
CopyTo 接受一个 buffer 参数，但在读取时总是使用 BufferSize 常量。如果 NetHttpClient 中的 _buffer 不是恰好 1024 字节，或者将来发生变化，就可能导致过度读取和运行时错误。考虑改为使用 buffer.Length，或者移除缓冲区参数并明确强制使用固定大小缓冲区的约定。
Original comment in English
issue (bug_risk): CopyTo uses a fixed BufferSize instead of the actual buffer length, which can cause mismatches if the underlying buffer size changes.
CopyTo takes a buffer argument but always uses the BufferSize constant when reading. If _buffer in NetHttpClient is not exactly 1024 bytes or changes in the future, this can lead to over-reads and runtime errors. Consider using buffer.Length instead, or remove the buffer parameter and enforce a fixed-size buffer contract explicitly.
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
4 issues found across 2 files
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="AquaMai.Mods/GameSystem/MoreContentEncoding.cs">
<violation number="1" location="AquaMai.Mods/GameSystem/MoreContentEncoding.cs:23">
P2: `BufferSize` 固定为 1024，但读取目标 `_buffer` 是游戏内部字段，其长度未知。若 `_buffer.Length < 1024`，`CopyTo` 中的 `Read(buffer, 0, 1024)` 会抛 `ArgumentOutOfRangeException`。建议按 `_buffer.Length` 读取（如 `Math.Min(BufferSize, buffer.Length)`），或直接使用 `Math.Min` 保证不超过缓冲区长度。</violation>
<violation number="2" location="AquaMai.Mods/GameSystem/MoreContentEncoding.cs:45">
P2: gzip 解压分支没有 try/catch，而 zlib 分支有。由于该 prefix 完全替换了 `Decompress`，损坏/被截断的 gzip 或碰巧以 0x1F 0x8B 开头的明文都会让 `GZipStream` 抛出未捕获的 `InvalidDataException`，传播出 prefix 后可能中断网络包处理。为 gzip 分支加上与 `TryInflateZlib` 一致的 try/catch，失败时回退到明文写入。</violation>
<violation number="3" location="AquaMai.Mods/GameSystem/MoreContentEncoding.cs:50">
P2: 使用非默认窗口大小的合法 zlib 响应不会被解压。不要只匹配 `raw[0] == 0x78`，应校验完整 zlib 头部（CM、CINFO 和 FCHECK）后再调用 `TryInflateZlib`。</violation>
<violation number="4" location="AquaMai.Mods/GameSystem/MoreContentEncoding.cs:71">
P2: 该分支会接受 `Adler32` 校验失败的 zlib 响应，可能把损坏数据当成正常网络数据。解压后校验 zlib 尾部的 `Adler32`，校验失败时应返回失败并清理输出。</violation>
</file>
Reply with feedback, questions, or to request a fix.
Re-trigger cubic
| try | ||
| { | ||
| // 跳过 2 字节的 zlib 头部，忽略末尾的 4 字节 Adler32 校验和。 | ||
| using var input = new MemoryStream(raw, 2, raw.Length - 6, writable: false); |
There was a problem hiding this comment.
P2: 该分支会接受 Adler32 校验失败的 zlib 响应，可能把损坏数据当成正常网络数据。解压后校验 zlib 尾部的 Adler32，校验失败时应返回失败并清理输出。
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At AquaMai.Mods/GameSystem/MoreContentEncoding.cs, line 71:
<comment>该分支会接受 `Adler32` 校验失败的 zlib 响应，可能把损坏数据当成正常网络数据。解压后校验 zlib 尾部的 `Adler32`，校验失败时应返回失败并清理输出。</comment>
<file context>
@@ -0,0 +1,92 @@
+ try
+ {
+ // 跳过 2 字节的 zlib 头部，忽略末尾的 4 字节 Adler32 校验和。
+ using var input = new MemoryStream(raw, 2, raw.Length - 6, writable: false);
+ using var deflate = new DeflateStream(input, CompressionMode.Decompress);
+ CopyTo(deflate, output, buffer);
</file context>
| using var gz = new GZipStream(new MemoryStream(raw, writable: false), CompressionMode.Decompress); | ||
| CopyTo(gz, memoryStream, buffer); | ||
| } | ||
| else if (raw.Length >= 6 && raw[0] == 0x78 && TryInflateZlib(raw, memoryStream, buffer)) |
There was a problem hiding this comment.
P2: 使用非默认窗口大小的合法 zlib 响应不会被解压。不要只匹配 raw[0] == 0x78，应校验完整 zlib 头部（CM、CINFO 和 FCHECK）后再调用 TryInflateZlib。
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At AquaMai.Mods/GameSystem/MoreContentEncoding.cs, line 50:
<comment>使用非默认窗口大小的合法 zlib 响应不会被解压。不要只匹配 `raw[0] == 0x78`，应校验完整 zlib 头部（CM、CINFO 和 FCHECK）后再调用 `TryInflateZlib`。</comment>
<file context>
@@ -0,0 +1,92 @@
+ using var gz = new GZipStream(new MemoryStream(raw, writable: false), CompressionMode.Decompress);
+ CopyTo(gz, memoryStream, buffer);
+ }
+ else if (raw.Length >= 6 && raw[0] == 0x78 && TryInflateZlib(raw, memoryStream, buffer))
+ {
+ // zlib 成功解压(
</file context>
|
|
||
| public class MoreContentEncoding | ||
| { | ||
| private const int BufferSize = 1024; |
There was a problem hiding this comment.
P2: BufferSize 固定为 1024，但读取目标 _buffer 是游戏内部字段，其长度未知。若 _buffer.Length < 1024，CopyTo 中的 Read(buffer, 0, 1024) 会抛 ArgumentOutOfRangeException。建议按 _buffer.Length 读取（如 Math.Min(BufferSize, buffer.Length)），或直接使用 Math.Min 保证不超过缓冲区长度。
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At AquaMai.Mods/GameSystem/MoreContentEncoding.cs, line 23:
<comment>`BufferSize` 固定为 1024，但读取目标 `_buffer` 是游戏内部字段，其长度未知。若 `_buffer.Length < 1024`，`CopyTo` 中的 `Read(buffer, 0, 1024)` 会抛 `ArgumentOutOfRangeException`。建议按 `_buffer.Length` 读取（如 `Math.Min(BufferSize, buffer.Length)`），或直接使用 `Math.Min` 保证不超过缓冲区长度。</comment>
<file context>
@@ -0,0 +1,92 @@
+
+public class MoreContentEncoding
+{
+ private const int BufferSize = 1024;
+
+ [HarmonyPrefix]
</file context>
| // - 0x1F 0x8B -> gzip | ||
| // - 0x78 ?? + valid zlib -> zlib (the stock format: zlib header + raw deflate + adler32) | ||
| // - otherwise -> treat as plaintext | ||
| if (raw.Length >= 2 && raw[0] == 0x1F && raw[1] == 0x8B) |
There was a problem hiding this comment.
P2: gzip 解压分支没有 try/catch，而 zlib 分支有。由于该 prefix 完全替换了 Decompress，损坏/被截断的 gzip 或碰巧以 0x1F 0x8B 开头的明文都会让 GZipStream 抛出未捕获的 InvalidDataException，传播出 prefix 后可能中断网络包处理。为 gzip 分支加上与 TryInflateZlib 一致的 try/catch，失败时回退到明文写入。
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At AquaMai.Mods/GameSystem/MoreContentEncoding.cs, line 45:
<comment>gzip 解压分支没有 try/catch，而 zlib 分支有。由于该 prefix 完全替换了 `Decompress`，损坏/被截断的 gzip 或碰巧以 0x1F 0x8B 开头的明文都会让 `GZipStream` 抛出未捕获的 `InvalidDataException`，传播出 prefix 后可能中断网络包处理。为 gzip 分支加上与 `TryInflateZlib` 一致的 try/catch，失败时回退到明文写入。</comment>
<file context>
@@ -0,0 +1,92 @@
+ // - 0x1F 0x8B -> gzip
+ // - 0x78 ?? + valid zlib -> zlib (the stock format: zlib header + raw deflate + adler32)
+ // - otherwise -> treat as plaintext
+ if (raw.Length >= 2 && raw[0] == 0x1F && raw[1] == 0x8B)
+ {
+ using var gz = new GZipStream(new MemoryStream(raw, writable: false), CompressionMode.Decompress);
</file context>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
开启后游戏将支持解压非deflate格式的数据包，比如Gzip;
用于服务器使用Cloudflare等不支持deflate格式的CDN;
Summary by Sourcery
添加一个可选的游戏系统设置，在解压网络响应时支持额外的 HTTP
content-encoding格式。新功能：
MoreContentEncoding游戏系统模块，将解压支持从仅deflate扩展到包括gzip和zlib编码的负载，并在必要时回退到纯文本。增强：
MoreContentEncoding配置选项，以便可以在设置中启用或禁用扩展的content-encoding支持。Original summary in English
Summary by Sourcery
Add an optional game system setting to support additional HTTP content-encoding formats when decompressing network responses.
New Features:
Enhancements: