feat: Cache HTTP responses (#84)

* Implement HTTP caching

* Respect cache in test workflow

* Expire cache after 1 day

* Save cache regardless of hit during restore

.github/workflows/test.yml

@@ -34,6 +34,13 @@ jobs:
       run: |
         poetry env use ${{ matrix.python-version }}
         poetry install
+    - name: Cache HTTP responses
+      uses: actions/cache@v4
+      with:
+        path: ~/.cache/tap-f1.sqlite
+        key: http-responses-${{ github.run_id }}
+        restore-keys: |
+          http-responses-
     - name: Test with pytest
       run: |
         poetry run pytest

pyproject.toml

@@ -15,6 +15,7 @@ python = "^3.8"
 singer-sdk = { version="~=0.42.1" }
 fs-s3fs = { version = "~=1.1.1", optional = true }
 requests = "~=2.32.3"
+requests-cache = "^1.2.1"
 
 [tool.poetry.group.dev.dependencies]
 pytest = ">=7.4.0"
@@ -30,6 +31,8 @@ warn_unused_configs = true
 [tool.ruff]
 ignore = [
     "ANN001",  # missing-type-function-argument
+    "ANN002",  # missing-type-args
+    "ANN003",  # missing-type-kwargs
     "ANN101",  # missing-type-self
     "ANN102",  # missing-type-cls
     "ANN201",  # missing-return-type-undocumented-public-function

tap_f1/client.py

@@ -1,5 +1,8 @@
 """REST client handling, including F1Stream base class."""
 
+from datetime import timedelta
+
+from requests_cache import CachedSession
 from singer_sdk.streams import RESTStream
 from typing_extensions import override
 
@@ -12,6 +15,15 @@ class F1Stream(RESTStream):
     url_base = "https://ergast.com/api/f1"
     _limit = 1000
 
+    def __init__(self, *args, **kwargs) -> None:
+        """Initialise the F1 stream."""
+        super().__init__(*args, **kwargs)
+        self._requests_session = CachedSession(
+            self.tap_name,
+            use_cache_dir=True,
+            expire_after=timedelta(days=1),
+        )
+
     @override
     def get_new_paginator(self):
         return F1Paginator(0, self._limit)