feat(protobuf): support any for protobuf message source #12291

xzhseh · 2023-09-13T22:44:41Z

I hereby agree to the terms of the RisingWave Labs, Inc. Contributor License Agreement.

What's changed and what's your intention?

resolve #12246

Checklist

I have written necessary rustdoc comments
I have added necessary unit tests and integration tests
I have added fuzzing tests or opened an issue to track them. (Optional, recommended for new SQL features Sqlsmith: Sql feature generation #7934).
My PR contains breaking changes. (If it deprecates some features, please create a tracking issue to remove them in the future).
All checks passed in ./risedev check (or alias, ./risedev c)
My PR changes performance-critical code. (Please run macro/micro-benchmarks and show the results.)

My PR contains critical fixes that are necessary to be merged into the latest release. (Please check out the details)

Documentation

My PR needs documentation updates. (Please use the Release note section below to summarize the impact on users)

Release note

Now we support any type for protobuf source, the relevant documentation maybe updated.

fuyufjh

Rest LGTM

src/connector/src/parser/protobuf/parser.rs

tabVersion · 2023-09-17T12:20:42Z

I'm curious where's the encoding part of the message

RisingWave doesn't wrap messages in Any. The producer may have such logic.

and which semantic of decoding we should follow when dealing with payload

I don't get the question

xzhseh · 2023-09-20T22:08:51Z

Now any type is supported, the returned type would be a ScalarImpl::Struct(...).
PTAL if this is expected, cc @fuyufjh @tabVersion.

albertoforcato · 2023-09-21T13:33:20Z

@BugenZhao When will this PR be merged ?

BugenZhao · 2023-09-22T04:06:15Z

@BugenZhao When will this PR be merged ?

We are currently awaiting approval from the module owners for this PR. We appreciate your patience, and you can subscribe to the notifications of this PR to get informed.

codecov · 2023-09-22T06:03:41Z

Codecov Report

Merging #12291 (ffb9e87) into main (c97e08c) will decrease coverage by 0.11%.
Report is 2 commits behind head on main.
The diff coverage is 49.90%.

@@            Coverage Diff             @@
##             main   #12291      +/-   ##
==========================================
- Coverage   68.26%   68.16%   -0.11%     
==========================================
  Files        1505     1506       +1     
  Lines      254901   255327     +426     
==========================================
+ Hits       174015   174034      +19     
- Misses      80886    81293     +407

Flag	Coverage Δ
rust	`68.16% <49.90%> (-0.11%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
src/common/src/metrics.rs	`13.15% <ø> (ø)`
src/connector/src/parser/unified/protobuf.rs	`86.36% <100.00%> (+3.03%)`	⬆️
src/connector/src/source/monitor/metrics.rs	`95.23% <ø> (-0.87%)`	⬇️
src/frontend/src/expr/pure.rs	`87.69% <ø> (ø)`
src/jni_core/src/macros.rs	`100.00% <ø> (ø)`
src/stream/src/executor/monitor/streaming_stats.rs	`95.69% <ø> (-0.12%)`	⬇️
src/frontend/src/binder/expr/binary_op.rs	`70.40% <0.00%> (ø)`
src/expr/impl/src/scalar/jsonb_access.rs	`0.00% <0.00%> (ø)`
src/stream/src/executor/actor.rs	`56.81% <0.00%> (+1.26%)`	⬆️
src/stream/src/executor/source/source_executor.rs	`78.22% <0.00%> (+0.58%)`	⬆️
... and 6 more

... and 20 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

src/connector/src/parser/protobuf/parser.rs

tabVersion · 2023-09-25T07:01:18Z

Now any type is supported, the returned type would be a ScalarImpl::Struct(...). PTAL if this is expected, cc @fuyufjh @tabVersion.

I think it is not expected.
When literally creating a source, RisingWave does not get any actual payload. You can only get Any in the schema but know nothing inside.

src/connector/src/parser/protobuf/parser.rs

xzhseh · 2023-10-26T05:16:17Z

In addition, any idea on the gitguardian error?

tabVersion · 2023-10-26T05:24:12Z

In addition, any idea on the gitguardian error?

Just ignore it, it is not related to the topic

tabVersion · 2023-10-26T05:43:17Z

There is a potential chance that naming conflict could happen, let's say user names the int64 field of B C as the same, i.e.,
message A {
  any v1 = 1;
  int64 v2 = 2;
}
message B {
  any v3 = 1;
  int64 v4 = 2;
}
message C {
  int64 v4 = 1;
  string v5 = 2;
  bytes v6 = 3;
}
In the above case, there could potentially exist two kinds of ingest messages: 1. A.v1 -> B.v4 ... 2. A.v1 -> C.v4 ... If we let the user extract the field only based on field name, in this case, select ((msg).A -> 'v1' -> 'v4')::bigint, the result is ambiguous and indeterministic. But this is the matter of how to restrict / control users' behavior. Do you prefer to let users themselves prevent naming conflict like this at the very beginning on their own, and add a hidden _type_ field for name hint, or we help them handle these cases by specifying full name, which would turn out to be, select ((msg).A -> 'A.v1' -> 'B.v4' or 'C.v4')::bigint.

Yes, you are right, it can be ambiguous if the message names collide. But we should do as little change to the original message content as possible, giving a prefix to each field is opposite to the idea.
we encourage users to match the message name and do the analysis, it matches the practice I've seen from my last company.

select case (msg).A -> '_type_'
  when 'messageB' then ... 
  when 'messageC' then ...
end

I think it is not confusing.

Rw gets both messageA and messageB wrapped in messageC. Rw will extract the inner fields without telling it is messageA or messageB

Yes, the customer told us that they will ensure only one message type in the any field. So I did not realize a name collide problem. After a second thought, I think adding a _type_ field is reasonable and enough to fix the problem.

fuyufjh · 2023-10-26T06:19:57Z

It sounds like the conclusion will be this, right?

{
  "__type": B.full_name(),
  "v3": {
    "__type": C.full_name(),
    "v5": 2,
    "v6": "some string",
    "v7": NULL // <-- json cannon handle bytes type, I think it is ok to leave NULL here and give a warning
  },
  "v4": 1,
}

LGTM 👍

A little suggestion: a single underscore (_type) is my favorite style, considering our hidden column _row_id follows this style, too. (Another common style is rw_, but it seems a bit weird here)

xzhseh · 2023-10-26T19:32:21Z

The current message will be displayed like this.

The example use case is as below.

The schema of the messages is as below.

message TestAny {
  int32 id = 1;
  google.protobuf.Any any_value = 2;
}
message StringStringInt32Value {
  string first = 1;
  StringInt32Value second = 2;
  Float32StringValue third = 3;
}

One thing to note is that, we now support _type hint and name mapping for nested any message (e.g., A.v1 -> B.v3 -> C from the discussion above), since the decoding part in from_protobuf_value will recursively handle any type resolution from bottom to top. See test_any_recursive for details.

match fields[1].clone() {
    Some(ScalarImpl::Jsonb(jv)) => {
        assert_eq!(
            jv,
            JsonbVal::from(json!({
                "_type": "test.AnyValue",
                "any_value_1": {
                    "_type": "test.StringValue",
                    "value": "114514",
                },
                "any_value_2": {
                    "_type": "test.Int32Value",
                    "value": 114514,
                }
            }))
        );
    }
    _ => panic!("Expected ScalarImpl::Jsonb"),
}

This PR could be merged after final reviews, cc @tabVersion @fuyufjh @Rossil2012.

…exist from upstream sink

Dismiss stale reviews.

tabVersion

LGTM! Thanks for finishing the pr🥹

tabVersion · 2023-10-31T06:48:55Z

src/connector/src/parser/protobuf/parser.rs

+                if key.is_empty() {
+                    key = "Int16".to_string();
+                }


I think we always get a struct here, when will we encounter the missing key scene?

When encountering any type holding a "nested" struct, we could not get the actual field name inside the struct.
e.g., For the proto below, we could only get the type_url and value for the second and third field in StringStringInt32Value, rather than the actual field name in those structs. In this case, the current workaround is use anonymous type name as the key, that's when the full_name_vec turns `None.

message TestAny { int32 id = 1; google.protobuf.Any any_value = 2; } message StringStringInt32Value { string first = 1; StringInt32Value second = 2; Float32StringValue third = 3; }

src/connector/src/parser/protobuf/parser.rs

xiangjinwu · 2024-09-03T07:03:42Z

This PR and its linked issue has no documentation of the expected behavior (but scattered in discussions), and the official document in https://github.com/risingwavelabs/risingwave-docs/pull/1506 has been wrong since it is published.

I am going to consider this as unreleased feature and do whatever (potentially breaking) changes to

better organize the code
keep behavior consistent with other parts of the system

By the way this feature itself is not a high priority, but the relevant code snippet is blocking me from a refactor that fixes a data type mismatch panic in the protobuf parser.

xzhseh added 2 commits September 12, 2023 16:26

feat(protobuf): support Any as message source

2b973e4

update

dbaa746

xzhseh requested review from neverchanje and tabVersion September 13, 2023 22:44

github-actions bot added the type/feature label Sep 13, 2023

This comment was marked as outdated.

Sign in to view

xzhseh requested a review from fuyufjh September 13, 2023 22:50

xzhseh self-assigned this Sep 13, 2023

fuyufjh reviewed Sep 16, 2023

View reviewed changes

src/connector/src/parser/protobuf/parser.rs Outdated Show resolved Hide resolved

src/connector/src/parser/protobuf/parser.rs Show resolved Hide resolved

This comment was marked as outdated.

Sign in to view

xzhseh added 2 commits September 20, 2023 18:02

support any protobuf

a23b50b

Merge branch 'main' into xzhseh/feat-protobuf-any

bb1cd88

fix test

3ab91ad

Merge branch 'main' into xzhseh/feat-protobuf-any

31bd84b

xzhseh requested a review from fuyufjh September 21, 2023 15:41

Merge branch 'main' into xzhseh/feat-protobuf-any

b38bfc5

tabVersion reviewed Sep 25, 2023

View reviewed changes

src/connector/src/parser/protobuf/parser.rs Outdated Show resolved Hide resolved

src/connector/src/parser/protobuf/parser.rs Show resolved Hide resolved

update

e90f067

tabVersion reviewed Sep 27, 2023

View reviewed changes

src/connector/src/parser/protobuf/parser.rs Outdated Show resolved Hide resolved

src/connector/src/parser/protobuf/parser.rs Outdated Show resolved Hide resolved

src/connector/src/parser/protobuf/parser.rs Outdated Show resolved Hide resolved

xzhseh added 2 commits September 27, 2023 18:52

support nested messages & address issues

eb85fcd

update the output format

ed7c424

xzhseh requested a review from a team as a code owner October 18, 2023 01:32

xzhseh requested a review from tabVersion October 18, 2023 01:33

Rossil2012 self-requested a review October 19, 2023 09:33

add log for e2e_test debug

2e5570f

xzhseh added 9 commits October 26, 2023 15:39

add type hint for any type

a32c64e

Merge branch 'main' into xzhseh/feat-protobuf-any

f7f4557

fix test

2130071

fix test

b143a7a

fix test

a93fcf4

add default value for any type when the corresponding value does not …

9e02c3d

…exist from upstream sink

fix format

232d2e7

Merge branch 'main' into xzhseh/feat-protobuf-any

643a8c4

Merge branch 'main' into xzhseh/feat-protobuf-any

707f601

xzhseh added the user-facing-changes Contains changes that are visible to users label Oct 27, 2023

tabVersion approved these changes Oct 31, 2023

View reviewed changes

xzhseh added 3 commits October 31, 2023 14:22

fix err_msg

b4155cf

Merge branch 'main' into xzhseh/feat-protobuf-any

828b46d

fix format

ffb9e87

xzhseh added this pull request to the merge queue Oct 31, 2023

Merged via the queue into main with commit 99fe0eb Oct 31, 2023

xzhseh deleted the xzhseh/feat-protobuf-any branch October 31, 2023 20:21

tabVersion mentioned this pull request Nov 8, 2023

tracking: support for data types in protobuf #10479

Closed

5 tasks

xxchan mentioned this pull request Nov 10, 2023

risingwave 1.4.0 risingwavelabs/homebrew-risingwave#17

Closed

xiangjinwu mentioned this pull request Sep 3, 2024

fix(source): Protobuf Any as canonical JSON #18380

Merged

9 tasks

fuyufjh mentioned this pull request Dec 24, 2024

support circular reference of protobuf #19915

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(protobuf): support any for protobuf message source #12291

feat(protobuf): support any for protobuf message source #12291

xzhseh commented Sep 13, 2023 •

edited

Loading

This comment was marked as outdated.

fuyufjh left a comment

tabVersion commented Sep 17, 2023

This comment was marked as outdated.

xzhseh commented Sep 20, 2023

albertoforcato commented Sep 21, 2023

BugenZhao commented Sep 22, 2023

codecov bot commented Sep 22, 2023 •

edited

Loading

tabVersion commented Sep 25, 2023

xzhseh commented Oct 26, 2023

tabVersion commented Oct 26, 2023

tabVersion commented Oct 26, 2023 •

edited

Loading

fuyufjh commented Oct 26, 2023 •

edited

Loading

xzhseh commented Oct 26, 2023 •

edited

Loading

tabVersion left a comment

tabVersion Oct 31, 2023

xzhseh Oct 31, 2023

xiangjinwu commented Sep 3, 2024 •

edited

Loading

feat(protobuf): support any for protobuf message source #12291

feat(protobuf): support any for protobuf message source #12291

Conversation

xzhseh commented Sep 13, 2023 • edited Loading

What's changed and what's your intention?

Checklist

Documentation

Release note

This comment was marked as outdated.

fuyufjh left a comment

Choose a reason for hiding this comment

tabVersion commented Sep 17, 2023

This comment was marked as outdated.

xzhseh commented Sep 20, 2023

albertoforcato commented Sep 21, 2023

BugenZhao commented Sep 22, 2023

codecov bot commented Sep 22, 2023 • edited Loading

Codecov Report

tabVersion commented Sep 25, 2023

xzhseh commented Oct 26, 2023

tabVersion commented Oct 26, 2023

tabVersion commented Oct 26, 2023 • edited Loading

fuyufjh commented Oct 26, 2023 • edited Loading

xzhseh commented Oct 26, 2023 • edited Loading

tabVersion left a comment

Choose a reason for hiding this comment

tabVersion Oct 31, 2023

Choose a reason for hiding this comment

xzhseh Oct 31, 2023

Choose a reason for hiding this comment

xiangjinwu commented Sep 3, 2024 • edited Loading

xzhseh commented Sep 13, 2023 •

edited

Loading

codecov bot commented Sep 22, 2023 •

edited

Loading

tabVersion commented Oct 26, 2023 •

edited

Loading

fuyufjh commented Oct 26, 2023 •

edited

Loading

xzhseh commented Oct 26, 2023 •

edited

Loading

xiangjinwu commented Sep 3, 2024 •

edited

Loading