feat: Support NebulaGraph

[GG2002]

Which issue does this PR close?

Closes #4553.

Rationale for this change

What changes are included in this PR?

Support read, write, list, delete ops on NebulaGraph.

To access NebulaGraph, users must provide the username and password for NebulaGraph, as well as specify the space, tag, and the key field and value field of the tag.

Since each vertex in NebulaGraph must have a unique ID(called VID), but NebulaGraph lacks a self-incrementing ID, the unique IDs are generated using the Snowflake algorithm. Additionally, the set operation involves first delete existing original vertex, followed by inserting a new vertex.

NebulaGraph does not support the Blob type, hence Base64+string type is employed for file access. If it's supported in the future, just remove Base64.

Are there any user-facing changes?

No(?)

[GG2002]

Test results are as follows:

$ OPENDAL_TEST=nebulagraph cargo test behavior --features tests -- --nocapture --test-threads=20
    Finished `test` profile [unoptimized + debuginfo] target(s) in 0.32s
     Running unittests src/lib.rs (target/debug/deps/opendal-20c211e669cef355)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 128 filtered out; finished in 0.00s

     Running tests/behavior/main.rs (target/debug/deps/behavior-0bda2893f3ecdc95)

running 90 tests
test behavior::test_delete_with_version                       ... ok
test behavior::test_delete_with_not_existing_version          ... ok
test behavior::test_delete_not_existing                       ... ok
test behavior::test_list_with_start_after                     ... ok
test behavior::test_check                                     ... ok
test behavior::test_list_non_exist_dir                        ... ok
test behavior::test_delete_empty_dir                          ... ok
test behavior::test_create_dir                                ... ok
test behavior::test_create_dir_existing                       ... ok
test behavior::test_list_files_with_version                   ... ok
test behavior::test_list_with_version_and_limit               ... ok
test behavior::test_list_with_version_and_start_after         ... ok
test behavior::test_list_root_with_recursive                  ... ok
test behavior::test_list_sub_dir                              ... ok
test behavior::test_list_file_with_recursive                  ... ok
test behavior::test_list_empty_dir                            ... ok
test behavior::test_read_with_if_match                        ... ok
test behavior::test_read_with_if_none_match                   ... ok
test behavior::test_list_dir                                  ... ok
test behavior::test_read_not_exist                            ... ok
test behavior::test_read_with_override_cache_control          ... ok
test behavior::test_read_with_override_content_disposition    ... ok
test behavior::test_read_with_override_content_type           ... ok
test behavior::test_read_with_version                         ... ok
test behavior::test_read_with_not_existing_version            ... ok
test behavior::test_read_with_dir_path                        ... ok
test behavior::test_stat_dir                                  ... ok
test behavior::test_list_nested_dir                           ... ok
test behavior::test_list_dir_with_recursive_no_trailing_slash ... ok
test behavior::test_list_dir_with_recursive                   ... ok
test behavior::test_stat_not_exist                            ... ok
test behavior::test_stat_with_if_match                        ... ok
test behavior::test_stat_with_if_none_match                   ... ok
test behavior::test_stat_with_override_cache_control          ... ok
test behavior::test_stat_with_override_content_disposition    ... ok
test behavior::test_stat_with_override_content_type           ... ok
test behavior::test_stat_root                                 ... ok
test behavior::test_stat_with_version                         ... ok
test behavior::stat_with_not_existing_version                 ... ok
test behavior::test_remove_all                                ... ok
test behavior::test_write_only                                ... ok
test behavior::test_write_with_dir_path                       ... ok
test behavior::test_write_with_empty_content                  ... ok
test behavior::test_write_with_cache_control                  ... ok
test behavior::test_write_with_content_type                   ... ok
test behavior::test_write_with_content_disposition            ... ok
test behavior::test_write_with_user_metadata                  ... ok
test behavior::test_writer_write                              ... ok
test behavior::test_stat_nested_parent_dir                    ... ok
test behavior::test_writer_write_with_concurrent              ... ok
test behavior::test_writer_sink                               ... ok
test behavior::test_writer_sink_with_concurrent               ... ok
test behavior::test_list_rich_dir                             ... ok
test behavior::test_writer_abort                              ... ok
test behavior::test_writer_futures_copy                       ... ok
test behavior::test_writer_futures_copy_with_concurrent       ... ok
test behavior::test_blocking_create_dir                       ... ok
test behavior::test_write_with_special_chars                  ... ok
test behavior::test_blocking_create_dir_existing              ... ok
test behavior::test_blocking_delete_file                      ... ok
test behavior::test_reader                                    ... ok
test behavior::test_writer_abort_with_concurrent              ... ok
test behavior::test_read_range                                ... ok
test behavior::test_blocking_read_not_exist                   ... ok
test behavior::test_read_with_special_chars                   ... ok
test behavior::test_blocking_stat_dir                         ... ok
test behavior::test_list_dir_with_file_path                   ... ok
test behavior::test_blocking_stat_not_exist                   ... ok
test behavior::test_blocking_stat_with_special_chars          ... ok
test behavior::test_list_prefix                               ... ok
test behavior::test_blocking_write_with_dir_path              ... ok
test behavior::test_blocking_stat_file                        ... ok
test behavior::test_stat_file                                 ... ok
test behavior::test_remove_one_file                           ... ok
test behavior::test_blocking_write_file                       ... ok
test behavior::test_blocking_remove_all_basic                 ... ok
test behavior::test_stat_not_cleaned_path                     ... ok
test behavior::test_delete_with_special_chars                 ... ok
test behavior::test_delete_file                               ... ok
test behavior::test_read_full                                 ... ok
test behavior::test_list_dir_with_metakey                     ... ok
test behavior::test_stat_with_special_chars                   ... ok
test behavior::test_blocking_read_full                        ... ok
test behavior::test_blocking_remove_one_file                  ... ok
test behavior::test_blocking_read_range                       ... ok
test behavior::test_list_dir_with_metakey_complete            ... ok
test behavior::test_delete_stream                             ... ok
test behavior::test_blocking_write_with_special_chars         ... ok
test behavior::test_remove_all_basic                          ... ok
test behavior::test_writer_write_with_overwrite               ... ok

test result: ok. 90 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 1.56s

[Xuanwo]

Who is depending on futures 0.1?

[GG2002]

It's futures-util, fbthrift-* used in rust-nebula and bb8 depends on futures and futures depends on futures-util. futures-utils/dependencies displayed that futures=0.1.25.

cargo tree is as follows:

├── bb8 v0.8.5
│   ├── async-trait v0.1.82 (proc-macro) (*)
│   ├── futures-util v0.3.30
│   │   ├── futures v0.1.31
│   │   ├── futures-channel v0.3.30
│   │   │   ├── futures-core v0.3.30
│   │   │   └── futures-sink v0.3.30

Does it need to be removed? I just don't know how to do it:(

[Xuanwo]

Seems someone has enabled futures-util's compat feature in wrong.

---

Ok, I got it: https://github.com/bk-rs/fbthrift-git-rs/blob/36ddca22a4aca8bbf02482ca20ff3fe525a24a87/Cargo.toml#L17

[GG2002]

vkill has contributed almost all of fbThrift's packages, can we remove this without communicating with him😇?

[Xuanwo]

I'm not sure. Can you contact @vkill? Or can we obtain write permission for the bk-rs organization? Maybe @wey-gu can help a bit.

[GG2002]

Perhaps it's possible, but I don't have high hopes. If I can establish a stable connection with vkill, I won't have to release a new rust-nebula crate. I will make modifications directly on his nebula-rust crate:(

[GG2002]

bk-rs/nebula-rs#12

Hi @GG2002 The feature compat is also enabled in the latest official version. I don't think there is any problem with it.

https://github.com/facebook/fbthrift/blob/v2024.09.23.00/thrift/lib/rust/Cargo.toml#L18

@vkill says that.

What should I do now?

[GG2002]

I briefly reviewed all the fbthrift related code that rust-nebula depends on and found that there is nothing that requires the use of compat. Therefore, I temporarily moved all these code to rust-nebula, removed compat, and made the necessary changes.

Now rust-nebula does not rely on futures v0.1, and it has also passed all tests (as above, I edited it).

I will also continue to communicate with vkill and Facebook, and if they are willing to make changes, I will update my code. If they don't want to, I may release these fbthrift package separately after rust-nebula becomes more mature (because there are 8 crates to be changed, so I don't really want to publish them directly to occupy so many names now), but this has no impact on OpenDAL.

[Xuanwo]

Hi, @GG2002. It might take us a while to complete the communication. Would you like to split this PR into parts so we can merge some of them first? I believe we can add the builder, config first without introduce the client impl in.

[GG2002]

Hi, @GG2002. It might take us a while to complete the communication. Would you like to split this PR into parts so we can merge some of them first? I believe we can add the builder, config first without introduce the client impl in.

It's ok.