test(fts): add integration tests for FTS data plane

[jhamon]

Summary

This PR adds comprehensive integration tests for the new FTS (Full-Text Search) data plane functionality:

• search_documents() with text queries: Tests text search using text_query(), phrase matching with quoted strings, required terms (+term), boost and slop parameters, and $text_match filter operator
• search_documents() with vector queries: Tests dense vector search using vector_query(), combined with metadata filtering and include_fields
• upsert_documents(): Tests single/batch document upserts, document updates, and response validation
• Document and DocumentSearchResponse access patterns: Tests attribute access, dict-style access, get() with defaults, iteration methods, and to_dict()
• Backwards compatibility: Verifies that existing query(), upsert(), fetch(), and delete() methods continue to work with schema-based indexes

Also updates the __init__.pyi stub file to export new FTS types for proper mypy support:

• TextQuery, VectorQuery, Document, DocumentSearchResponse
• text_query, vector_query helper functions
• Schema field types (TextField, IntegerField, etc.) and deployment classes

Test Plan

• Tests pass in CI (pending SDK-116 to re-enable linting/unit tests)
• Manual verification of test collection
• All test files pass Python syntax validation

Related Issues

• Linear: SDK-114
• Depends on: SDK-113 (Integration tests - control plane)

Made with Cursor

---

Note

Medium Risk
Adds a large suite of REST-only integration tests that create/delete real indexes and upsert/search data, which can be flaky or slow depending on environment and backend availability. Production library logic changes are limited to type-stub exports, so functional risk to SDK runtime is low.

Overview
Adds a new REST-sync integration test suite for FTS/schema-based indexes, including fixtures that create an index with a metadata schema, seed documents, and clean up via tagged deletion.

Covers upsert_documents() and search_documents() for both text (TextQuery/text_query) and vector (VectorQuery/vector_query) scoring, plus filters (including $text_match) and include_fields; also validates Document/DocumentSearchResponse access patterns and verifies backwards compatibility of existing query/upsert/fetch/delete/describe_index_stats on schema-enabled indexes.

Updates pinecone/__init__.pyi to export the new FTS query/document types and helper functions, and re-export schema field and deployment types for typing/mypy support.

^{Written by Cursor Bugbot for commit bdcf685. This will update automatically on new commits. Configure here.}