[document_store] Raise warning when labels are overwritten

[akkefa]

Proposed changes:

• Raise warning when labels are overwritten Give warning when labels are overwritten in DocumentStore.write_labels() #1140

Status (please check what you already did):

• First draft (up for discussions & feedback)
• Final code
• Added tests
• Updated documentation

[akkefa]

@brandenchan Please review this pull request. labels overwrite warning added for SQL based document store. If I am on right track I will add the same functionality for other stores.

[brandenchan]

The latest changes look great! Thanks for those. Excited to see this working for the other document stores too!

[akkefa]

The latest changes look great! Thanks for those. Excited to see this working for the other document stores too!

I will add this functionality to the base class as the get_duplicate_lables() method. All document stores will use that generic method.

[akkefa]

@brandenchan Please review my changes for all document stores.

[brandenchan]

Hey @akkefa, I just tried out these changes here are my notes.

I ran this script

from haystack.document_store import ElasticsearchDocumentStore, SQLDocumentStore
from haystack import Label
from haystack.utils import launch_es

launch_es()

labels = [Label(question="", answer="", is_correct_answer=False, is_correct_document=False, origin="asdf", id="1"),
          Label(question="", answer="", is_correct_answer=False, is_correct_document=False, origin="asdf", id="1")]

es_ds = ElasticsearchDocumentStore()
es_ds.write_labels(labels)

sql_ds = SQLDocumentStore()
sql_ds.write_labels(labels)

mem_ds = InMemoryDocumentStore()
mem_ds.write_labels(labels)

1. ES seems to work
2. SQL throws an error when there are duplicate ids. Let's leave it like this for now.
3. Could you add support for the InMemoryDocumentStore too?

Once you address the comments and add support for InMemoryDocumentStore, I think we should be just about ready to merge!

[akkefa]

@brandenchan

Once you address the comments and add support for InMemoryDocumentStore, I think we should be just about ready to merge!

@brandenchan InMemoryDocumentStore changes add. Pull request is ready for merge. 😴

[brandenchan]

Great thank you for your work on this! LGTM