api: native crud module support

[GRISHNOV]

Adds native api support for crud module to use it from a connection object.

Below there are examples of using the api.

>>> import tarantool
>>> from tarantool.error import CrudModuleError, CrudModuleManyError, DatabaseError
>>> conn = tarantool.Connection(host='localhost',port=3301)

>>> conn.crud_
    conn.crud_count(                conn.crud_insert(               conn.crud_insert_object_many(   
    conn.crud_min(                  conn.crud_replace_object(       conn.crud_stats(                
    conn.crud_unflatten_rows(       conn.crud_upsert_many(          conn.crud_delete(               
    conn.crud_insert_many(          conn.crud_len(                  conn.crud_replace(              
    conn.crud_replace_object_many(  conn.crud_storage_info(         conn.crud_update(               
    conn.crud_upsert_object(        conn.crud_get(                  conn.crud_insert_object(        
    conn.crud_max(                  conn.crud_replace_many(         conn.crud_select(               
    conn.crud_truncate(             conn.crud_upsert(               conn.crud_upsert_object_many(

Inserting data via crud:

# Insert without exception:
>>> res = conn.crud_insert('tester', (3500,300,'Rob'))
>>> res
<tarantool.crud.CrudResult object at 0x11a56e320>
>>> res.
res.metadata  res.rows
>>> res.rows
[[3500, 300, 'Rob']]
>>> res.metadata
[{'name': 'id', 'type': 'unsigned'}, {'name': 'bucket_id', 'type': 'unsigned'}, {'name': 'name', 'type': 'string'}]

# Insert with exception (duplicate key exists):
>>> try:
...     res = conn.crud_insert('tester', (3500,300,'Rob'))
... except CrudModuleError as e:
...     exc_crud = e
... 
>>> exc_crud
CrudModuleError(0, 'Failed to insert: Duplicate key exists in unique index "primary_index" in space "tester" with old tuple - [3500, 300, "Rob"] and new tuple - [3500, 300, "Rob"]')
>>> exc_crud.extra_info_error
<tarantool.crud.CrudError object at 0x10a276950>
>>> exc_crud.extra_info_error.
exc_crud.extra_info_error.class_name  exc_crud.extra_info_error.err         exc_crud.extra_info_error.file        exc_crud.extra_info_error.line        exc_crud.extra_info_error.str       
>>> exc_crud.extra_info_error.class_name
'InsertError'
>>> exc_crud.extra_info_error.str
'InsertError: Failed to insert: Duplicate key exists in unique index "primary_index" in space "tester" with old tuple - [3500, 300, "Rob"] and new tuple - [3500, 300, "Rob"]'

# In case of batch operation (*_many), CrudModuleManyError exception contains both result and errors (if there is a problem with at least one row).
>>> try:
...     res = conn.crud_insert_object_many('tester', ({'id':3,'bucket_id':100,'name':'Ann'}, {'id':4,'bucket_id':100,'name':'Sam'}), {'timeout':100, 'rollback_on_error':False})
... except CrudModuleManyError as e:
...     exc_crud = e
... 
>>> exc_crud
CrudModuleManyError(0, 'Got multiple errors, see errors_list')
>>> exc_crud.success_list # some of the rows were inserted.
<tarantool.crud.CrudResult object at 0x11a56f310>
>>> exc_crud.success_list.rows
[[1, 100, 'Bob'], [2, 100, 'Rob']]
>>> exc_crud.errors_list # some of the rows were not inserted.
[<tarantool.crud.CrudError object at 0x11a56e9e0>, <tarantool.crud.CrudError object at 0x11a56f490>]
>>> exc_crud.errors_list[0].str
'CallError: Failed for 037adb3a-b9e3-4f78-a6d1-9f0cdb6cbefc: Function returned an error: Duplicate key exists in unique index "primary_index" in space "tester" with old tuple - [3500, 300, "Rob"] and new tuple - [3500, 100, "Mike"]'
>>> exc_crud.errors_list[1].str
'InsertManyError: Failed to flatten object: FlattenError: Object is specified in bad format: FlattenError: Unknown field "second_name" is specified'

# If there are no problems with any rows, the entire response will be contained in the res variable.
>>> res = conn.crud_insert_object_many('tester', ({'id':3,'bucket_id':100,'name':'Ann'}, {'id':4,'bucket_id':100,'name':'Sam'}), {'timeout':100, 'rollback_on_error':False})
>>> res.rows
[[3, 100, 'Ann'], [4, 100, 'Sam']]

Crud not found on the router:

>>> try:
...     res = conn.crud_insert('tester', (22221,300,'Rob'))
... except DatabaseError as e:
...     exc_db = e
... 
 >>> exc_db
DatabaseError(33, "Ensure that you're calling crud.router and user has sufficient grants")
>>> exc_db.extra_info
BoxError(type='ClientError', file='/tmp/tarantool-20221003-6335-edruh3/tarantool-2.10.3/src/box/lua/call.c', line=112, message="Procedure 'crud.insert' is not defined", errno=0, errcode=33, fields=None, prev=None)

Select and unflatten_rows via crud:

>>> try:
...     res = conn.crud_select('tester', {}, {'first':2})
... except CrudModuleError as e:
...     exc_crud = e
... 
>>> res
<tarantool.crud.CrudResult object at 0x10a276d10>
>>> res.
res.metadata  res.rows      
>>> res.rows
[[1, 100, 'Mike'], [2, 100, 'Mike']]
>>> res.metadata
[{'name': 'id', 'type': 'unsigned'}, {'name': 'bucket_id', 'type': 'unsigned'}, {'name': 'name', 'type': 'string'}]
>>> r = conn.crud_unflatten_rows(res.rows, res.metadata)
>>> r
[{'id': 1, 'bucket_id': 100, 'name': 'Mike'}, {'id': 2, 'bucket_id': 100, 'name': 'Mike'}]

Truncate and len via crud:

>>> try:
...     res = conn.crud_len('tester')
... except CrudModuleError as e:
...     exc_crud = e
>>> res
26
>>> try:
...     res = conn.crud_truncate('tester')
... except CrudModuleError as e:
...     exc_crud = e
>>> res
True

Encoding=None in Connection:

>>> conn = tarantool.Connection(host='localhost',port=3301,encoding=None)

>>> try:
...     res = conn.crud_insert('tester', (3500,300,'Rob'))
... except CrudModuleError as e:
...     exc_crud = e
... 
>>> res
<tarantool.crud.CrudResult object at 0x1188a7fd0>
>>> res.
res.metadata  res.rows      
>>> res.metadata
[{b'name': b'id', b'type': b'unsigned'}, {b'name': b'bucket_id', b'type': b'unsigned'}, {b'name': b'name', b'type': b'string'}]
>>> res.rows
[[3500, 300, b'Rob']]

Closes #205

[GRISHNOV]

Thank you for your feedback!

I have made most of the edits (with the exception of the pairs implementation). If the current API is suitable, I'll move on to writing documentation and tests

[DifferentialOrange]

>>> exc_crud.res is None # use only in case of *_many

I don't get this comment.

[GRISHNOV]

>>> exc_crud.res is None # use only in case of *_many

I don't get this comment.

I mean, in exc_crud.res, the result can only appear when the *_many() is called (when part of the batch is successful and part is not), in other cases exc_crud.res will always be None

[GRISHNOV]

Thank you! At the moment, I have implemented tests and documentation, and removed the draft status from the PR

[DifferentialOrange]

For unknown reasons, your PR doesn't trigger packing pipeline. I'll check it out.

[DifferentialOrange]

For unknown reasons, your PR doesn't trigger packing pipeline. I'll check it out.

Doc:

Note: Workflows will not run on pull_request activity if the pull request has a merge conflict.
The merge conflict must be resolved first.

It would be nice if you've resolved them first so CI runs would be full (packing has test runs too).

[GRISHNOV]

Thank you for your feedback! I have updated the PR description and tried to answer all the comments on the code review