refactor(tests): Use pytest collection to load JSON fixtures

.github/workflows/test.yaml

@@ -105,13 +105,35 @@ jobs:
       - uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955
         with:
           submodules: recursive
+          fetch-depth: 0  # Fetch full history for commit comparison
       - name: Setup Python
         uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065
         with:
           python-version: "3.11"
       - uses: ./.github/actions/setup-env
+      - name: Get changed files and save to disk
+        id: get-changed-files
+        run: |
+          if [ "${{ github.event_name }}" = "pull_request" ]; then
+            BASE_SHA="${{ github.event.pull_request.base.sha }}"
+            HEAD_SHA="${{ github.event.pull_request.head.sha }}"
+          else
+          # On push or force push to the feature branch
+            BASE_SHA="${{ github.event.before }}"
+            HEAD_SHA="${{ github.sha }}"
+          fi
+
+          echo "Diffing commits: $BASE_SHA..$HEAD_SHA"
+
+          # Get changed files and save to disk
+          FILE_LIST="changed_files.txt"
+          git diff --name-only "$BASE_SHA" "$HEAD_SHA" > "$FILE_LIST"
+          echo "Changed files saved to $FILE_LIST"
+          echo "file_list=$FILE_LIST" >> $GITHUB_OUTPUT
+          echo "List of files changed in the PR"
+          cat $FILE_LIST
       - name: Run json infra tests
-        run: tox -e json_infra
+        run: tox -e json_infra -- --file-list="${{ steps.get-changed-files.outputs.file_list }}"
       - name: Upload coverage reports to Codecov
         uses: codecov/codecov-action@5a1091511ad55cbe89839c7260b706298ca349f7
         with:

src/ethereum_spec_tools/evm_tools/statetest/__init__.py

@@ -9,7 +9,7 @@
 from copy import deepcopy
 from dataclasses import dataclass
 from io import StringIO
-from typing import Any, Dict, Iterable, List, Optional, TextIO
+from typing import Any, Dict, Generator, Iterable, List, Optional, TextIO
 
 from ethereum.utils.hexadecimal import hex_to_bytes
 
@@ -35,6 +35,41 @@ class TestCase:
     transaction: Dict
 
 
+def read_test_case(
+    test_file_path: str, key: str, test: Dict[str, Any]
+) -> Generator[TestCase, None, None]:
+    """
+    Given a key and a value, return a `TestCase` object.
+    """
+    env = test["env"]
+    if not isinstance(env, dict):
+        raise TypeError("env not dict")
+
+    pre = test["pre"]
+    if not isinstance(pre, dict):
+        raise TypeError("pre not dict")
+
+    transaction = test["transaction"]
+    if not isinstance(transaction, dict):
+        raise TypeError("transaction not dict")
+
+    for fork_name, content in test["post"].items():
+        for idx, post in enumerate(content):
+            if not isinstance(post, dict):
+                raise TypeError(f'post["{fork_name}"] not dict')
+
+            yield TestCase(
+                path=test_file_path,
+                key=key,
+                index=idx,
+                fork_name=fork_name,
+                post=post,
+                env=env,
+                pre=pre,
+                transaction=transaction,
+            )
+
+
 def read_test_cases(test_file_path: str) -> Iterable[TestCase]:
     """
     Given a path to a filled state test in JSON format, return all the
@@ -44,33 +79,7 @@ def read_test_cases(test_file_path: str) -> Iterable[TestCase]:
         tests = json.load(test_file)
 
     for key, test in tests.items():
-        env = test["env"]
-        if not isinstance(env, dict):
-            raise TypeError("env not dict")
-
-        pre = test["pre"]
-        if not isinstance(pre, dict):
-            raise TypeError("pre not dict")
-
-        transaction = test["transaction"]
-        if not isinstance(transaction, dict):
-            raise TypeError("transaction not dict")
-
-        for fork_name, content in test["post"].items():
-            for idx, post in enumerate(content):
-                if not isinstance(post, dict):
-                    raise TypeError(f'post["{fork_name}"] not dict')
-
-                yield TestCase(
-                    path=test_file_path,
-                    key=key,
-                    index=idx,
-                    fork_name=fork_name,
-                    post=post,
-                    env=env,
-                    pre=pre,
-                    transaction=transaction,
-                )
+        yield from read_test_case(test_file_path, key, test)
 
 
 def run_test_case(

tests/json_infra/__init__.py

@@ -1,9 +1,11 @@
 """Tests related to json infrastructure."""
 
-from typing import Dict, Optional, TypedDict
+from typing import Dict, TypedDict
 
 from typing_extensions import NotRequired
 
+from .hardfork import TestHardfork
+
 
 class _FixtureSource(TypedDict):
     url: str
@@ -31,113 +33,6 @@ class _FixtureSource(TypedDict):
 }
 
 
-def _get_fixture_path(key: str) -> str:
-    return TEST_FIXTURES[key]["fixture_path"]
-
-
-def _build_ethereum_test_paths(
-    base_path: str, legacy_fork: Optional[str] = None
-) -> tuple:
-    if legacy_fork:
-        bc_path = f"{base_path}/LegacyTests/{legacy_fork}/BlockchainTests/"
-        state_path = (
-            f"{base_path}/LegacyTests/{legacy_fork}/GeneralStateTests/"
-        )
-    else:
-        bc_path = f"{base_path}/BlockchainTests/"
-        state_path = f"{base_path}/GeneralStateTests/"
-    return bc_path, state_path
-
-
-def _build_eest_test_paths(base_path: str) -> tuple:
-    bc_path = f"{base_path}/fixtures/blockchain_tests/"
-    state_path = f"{base_path}/fixtures/state_tests/"
-    return bc_path, state_path
-
-
-# Base paths
-ETHEREUM_TESTS_BASE = _get_fixture_path("ethereum_tests")
-EEST_TESTS_BASE = _get_fixture_path("latest_fork_tests")
-
-# Ethereum test paths
-(
-    PRE_CONSTANTINOPLE_BC_ETHEREUM_TESTS,
-    PRE_CONSTANTINOPLE_STATE_ETHEREUM_TESTS,
-) = _build_ethereum_test_paths(ETHEREUM_TESTS_BASE, "Constantinople")
-(
-    PRE_CANCUN_BC_ETHEREUM_TESTS,
-    PRE_CANCUN_STATE_ETHEREUM_TESTS,
-) = _build_ethereum_test_paths(ETHEREUM_TESTS_BASE, "Cancun")
-BC_ETHEREUM_TESTS, STATE_ETHEREUM_TESTS = _build_ethereum_test_paths(
-    ETHEREUM_TESTS_BASE
-)
-
-# EEST test paths
-EEST_BC_TESTS, EEST_STATE_TESTS = _build_eest_test_paths(EEST_TESTS_BASE)
-
-ForkConfig = TypedDict(
-    "ForkConfig",
-    {
-        "eels_fork": str,
-        "blockchain_test_dirs": list[str],
-        "state_test_dirs": list[str],
-    },
-)
-
-
-def _create_fork_config(
-    eels_fork: str, bc_dirs: list, state_dirs: list
-) -> ForkConfig:
-    return {
-        "eels_fork": eels_fork,
-        "blockchain_test_dirs": bc_dirs,
-        "state_test_dirs": state_dirs,
-    }
-
-
-PRE_CONSTANTINOPLE_DIRS = (
-    [PRE_CONSTANTINOPLE_BC_ETHEREUM_TESTS, EEST_BC_TESTS],
-    [PRE_CONSTANTINOPLE_STATE_ETHEREUM_TESTS, EEST_STATE_TESTS],
-)
-
-PRE_CANCUN_DIRS = (
-    [PRE_CANCUN_BC_ETHEREUM_TESTS, EEST_BC_TESTS],
-    [PRE_CANCUN_STATE_ETHEREUM_TESTS, EEST_STATE_TESTS],
-)
-
-CURRENT_DIRS = (
-    [BC_ETHEREUM_TESTS, EEST_BC_TESTS],
-    [STATE_ETHEREUM_TESTS, EEST_STATE_TESTS],
-)
-
-FORKS: Dict[str, ForkConfig] = {
-    **{
-        json_fork: _create_fork_config(eels_fork, *PRE_CONSTANTINOPLE_DIRS)
-        for json_fork, eels_fork in [
-            ("Frontier", "frontier"),
-            ("Homestead", "homestead"),
-            ("EIP150", "tangerine_whistle"),
-            ("EIP158", "spurious_dragon"),
-            ("Byzantium", "byzantium"),
-            ("ConstantinopleFix", "constantinople"),
-        ]
-    },
-    **{
-        json_fork: _create_fork_config(eels_fork, *PRE_CANCUN_DIRS)
-        for json_fork, eels_fork in [
-            ("Istanbul", "istanbul"),
-            ("Berlin", "berlin"),
-            ("London", "london"),
-            ("Paris", "paris"),
-            ("Shanghai", "shanghai"),
-        ]
-    },
-    **{
-        json_fork: _create_fork_config(eels_fork, *CURRENT_DIRS)
-        for json_fork, eels_fork in [
-            ("Cancun", "cancun"),
-            ("Prague", "prague"),
-            ("Osaka", "osaka"),
-        ]
-    },
+FORKS: Dict[str, TestHardfork] = {
+    fork.json_test_name: fork for fork in TestHardfork.discover()
 }