feat: Added manually tested sync methods for PGVector to AlloyDB migration #209
Conversation
product-auto-label
bot
added
the
api: alloydb
twishabansal
changed the title
twishabansal
force-pushed
the
migration_methods
branch
6 times, most recently
from
502d2b8
to
e5ae00a
Compare
twishabansal
force-pushed
the
migration_methods
branch
from
e5ae00a
to
46fe6f7
Compare
| params = {} | ||
|
|
||
| if use_json_metadata: | ||
| insert_query = f"INSERT INTO {destination_table} (LANGCHAIN_ID, CONTENT, EMBEDDING, LANGCHAIN_METADATA) VALUES" |
There was a problem hiding this comment.
We get these default values from the file
There was a problem hiding this comment.
I see the default values hardcoded here:
Which file can we extract them from?
| The uuid corresponding to the collection. | ||
| """ | ||
| try: | ||
| collection_info = self._fetch( |
There was a problem hiding this comment.
The upcoming refactor we will be deleting fetch/execute. Any where that you use fetch or execute use the async pool directly like async with self.engine.connect() as conn:.... Then we will wrap the async methods with sync wrappers. For example extract_pgvector_collection could be _aextract_pgvector_collection. Then in extract_pgvector_collection we will use run_as_sync to run the _aextract_pgvector_collection methods. See https://github.com/googleapis/langchain-google-cloud-sql-pg-python/blob/c3a0f12276eef07a64ea40b20abe496496b5132c/src/langchain_google_cloud_sql_pg/engine.py#L364 for an example
There was a problem hiding this comment.
I'll refactor this PR once the async/sync refactor is pushed.
twishabansal
force-pushed
the
migration_methods
branch
2 times, most recently
from
630dd4f
to
c679a14
Compare
twishabansal
force-pushed
the
migration_methods
branch
from
c679a14
to
f50e52e
Compare
twishabansal
force-pushed
the
migration_methods
branch
from
e36078f
to
3f2584e
Compare
|
@twishabansal This PR can now be closed in favor of #235, correct? |
No description provided.