[SAC-28693] - Python upgrade

[mukeshbhatt18gl]

Description of change

Ticket: SAC-28693

Manual QA steps

• Discovery: Running
• Sync: Running
• Unit Tests: Running
• Integration Tests: Running

Rollback steps

• revert this branch

[Copilot]

Pull request overview

Upgrades the project’s CI environment to run on Python 3.12 and adds a suite of unit + integration tests to validate core tap behaviors (sync, discovery, pagination, bookmarking) under the upgraded runtime.

Changes:

• Update CircleCI to create the virtualenv with Python 3.12 and run unit + integration test suites.
• Add focused unit tests for client pagination/query params, discovery helpers, incremental bookmark logic, and sync_stream.
• Add integration-style tests validating schema discovery metadata, selected-stream sync behavior, bookmark/start-date handling, and pagination behavior.

Reviewed changes

Copilot reviewed 13 out of 13 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
.circleci/config.yml	Switch CI venv to Python 3.12; validate schemas from repo path; add unit & integration test execution steps.
tests/unittests/test_sync.py	Unit coverage for sync_stream emitting records and state for incremental streams.
tests/unittests/test_incremental_sync.py	Unit coverage for bookmark comparison and update helpers on incremental streams.
tests/unittests/test_discovery.py	Unit coverage for schema translation/type inference, merge behavior, and discovery stream enumeration.
tests/unittests/test_client.py	Unit coverage for pagination params/URL construction in the client get() method.
tests/base.py	Shared integration test utilities (schema loading, record generation, expected metadata).
tests/test_sync.py	Integration test ensuring only selected streams are synced and state is written once.
tests/test_start_date.py	Integration tests for start-date fallback bookmark behavior and client call argument.
tests/test_pagination.py	Integration tests for multi-page reads and stopping on short page.
tests/test_discovery.py	Integration test validating expected streams and key replication metadata from discovery output.
tests/test_bookmark.py	Integration tests for bookmark reading and incremental bookmark advancement.
tests/test_automatic_fields.py	Integration test validating “automatic” inclusion is set for key properties where present.
tests/test_all_fields.py	Integration test sanity-checking schema-based record generation across all streams.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[RushiT0122]

We have chargify test account available please write standard integration tests.

[Copilot]

Pull request overview

Copilot reviewed 19 out of 20 changed files in this pull request and generated 2 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Chargify.get() determines pagination via len(response.json()) < per_page, which only works when the JSON root is a list. This change makes invoices() expect a dict root (i["invoices"]), so len(response.json()) will be the number of dict keys (likely < 100) and pagination will stop after the first page even when more invoices exist. Consider updating get() to compute page size from the actual records list (e.g., accept a results_key/extractor, or detect dict-vs-list and use the list under the appropriate key) so invoice pagination remains correct.

Suggested change

	for j in i["invoices"]:
	yield j
	for j in i:
	yield j["invoice"]

[mittal-tushar]

The suggested change breaks at runtime. for j in i on a dict iterates its keys (strings "invoices", "meta"), so j["invoice"] raises TypeError: string indices must be integers.

The current code is correct, unlike other streams where each element is {"customer": {...}} requiring j["customer"], the invoices endpoint returns records directly as plain objects inside the "invoices" array, so yield j (no unwrapping) is right.

[mittal-tushar]

Pages correction made in: 04b543c

[Copilot]

The mocked invoice record uses a full timestamp (2025-04-20T10:30:00Z) for due_date, but the tap code/comments and the invoices schema define due_date as a date-only field (format: date, typically YYYY-MM-DD). Aligning the mock data with the actual field shape will make this test more representative and reduce the chance of missing date-only parsing/comparison issues.

Suggested change

	record = {"id": 100, "due_date": "2025-04-20T10:30:00Z"}
	mock_client.invoices.return_value = iter([record])
	results = list(stream.sync(state))
	self.assertEqual(len(results), 1)
	self.assertEqual(results[0][1]["due_date"], "2025-04-20T10:30:00Z")
	record = {"id": 100, "due_date": "2025-04-20"}
	mock_client.invoices.return_value = iter([record])
	results = list(stream.sync(state))
	self.assertEqual(len(results), 1)
	self.assertEqual(results[0][1]["due_date"], "2025-04-20")

[mittal-tushar]

Updated due_date with format date-time in invoices.json schema.
Here: 04b543c