feat(consume): add support for ethereum/{tests,legacytests} tarball urls (ethereum#1306)

* feat(consume): add support for ethereum/{tests,legacytests} tarball urls

* docs: update changelog

* fix(consume): don't extract to hard-coded fixtures directory

Detect single directory within archive instead of a hard-coded "fixtures" directory.

* fix(consume): fix tag matching when fetching releases

* docs: add advisory to changelog about cache dir changes

* fix(consume): revert incorrect "fix" from 39b0f54

Author: danceratopz

docs/CHANGELOG.md

@@ -2,16 +2,31 @@
 
 Test fixtures for use by clients are available for each release on the [Github releases page](https://github.com/ethereum/execution-spec-tests/releases).
 
-**Key:** ✨ = New, 🐞 = Fixed, 🔀 = Changed.
+**Key:** ✨ = New, 🐞 = Fixed, 🔀 = Changed. 💥 = Breaking
 
 ## 🔜 [Unreleased]
 
+**Note**: Although not a breaking change, `consume` users should delete the cache directory (typically located at `~/.cache/ethereum-execution-spec-tests`) used to store downloaded fixture release tarballs. This release adds support for [ethereum/tests](https://github.com/ethereum/tests) and [ethereum/legacytests](https://github.com/ethereum/legacytests) fixture release downloads and the structure of the cache directory has been updated to accommodate this change.
+
+To try this feature:
+
+```shell
+consume direct --input=https://github.com/ethereum/tests/releases/download/v17.0/fixtures_blockchain_tests.tgz
+```
+
+To determine the cache directory location, see the `--cache-folder` entry from the command:
+
+```shell
+consume cache --help
+```
+
 ### 💥 Breaking Change
 
 ### 🛠️ Framework
 
 #### `consume`
 
+- ✨ Add support for [ethereum/tests](https://github.com/ethereum/tests) and [ethereum/legacytests](https://github.com/ethereum/legacytests) release tarball download via URL to the `--input` flag of `consume` commands ([#1306](https://github.com/ethereum/execution-spec-tests/pull/1306)).
 - ✨ Add support for Nethermind's `nethtest` command to `consume direct` ([#1250](https://github.com/ethereum/execution-spec-tests/pull/1250)).
 - ✨ Allow filtering of test cases by fork via pytest marks (e.g., `-m "Cancun or Prague"`) ([#1304](https://github.com/ethereum/execution-spec-tests/pull/1304), [#1318](https://github.com/ethereum/execution-spec-tests/pull/1318)).
 - ✨ Allow filtering of test cases by fixture format via pytest marks (e.g., `-m blockchain_test`) ([#1314](https://github.com/ethereum/execution-spec-tests/pull/1314)).

src/pytest_plugins/consume/consume.py

@@ -43,6 +43,58 @@ def default_html_report_file_path() -> str:
     return ".meta/report_consume.html"
 
 
+class FixtureDownloader:
+    """Handles downloading and extracting fixture archives."""
+
+    def __init__(self, url: str, base_directory: Path):  # noqa: D107
+        self.url = url
+        self.base_directory = base_directory
+        self.parsed_url = urlparse(url)
+        self.org_repo = self.extract_github_repo()
+        self.version = Path(self.parsed_url.path).parts[-2]
+        self.archive_name = self.strip_archive_extension(Path(self.parsed_url.path).name)
+        self.extract_to = base_directory / self.org_repo / self.version / self.archive_name
+
+    def download_and_extract(self) -> Tuple[bool, Path]:
+        """Download the URL and extract it locally if it hasn't already been downloaded."""
+        if self.extract_to.exists():
+            return True, self.detect_extracted_directory()
+
+        return False, self.fetch_and_extract()
+
+    def extract_github_repo(self) -> str:
+        """Extract <username>/<repo> from GitHub URLs, otherwise return 'other'."""
+        parts = self.parsed_url.path.strip("/").split("/")
+        return (
+            f"{parts[0]}/{parts[1]}"
+            if self.parsed_url.netloc == "github.com" and len(parts) >= 2
+            else "other"
+        )
+
+    @staticmethod
+    def strip_archive_extension(filename: str) -> str:
+        """Remove .tar.gz or .tgz extensions from filename."""
+        return filename.removesuffix(".tar.gz").removesuffix(".tgz")
+
+    def fetch_and_extract(self) -> Path:
+        """Download and extract an archive from the given URL."""
+        self.extract_to.mkdir(parents=True, exist_ok=False)
+        response = requests.get(self.url)
+        response.raise_for_status()
+
+        with tarfile.open(fileobj=BytesIO(response.content), mode="r:gz") as tar:
+            tar.extractall(path=self.extract_to)
+
+        return self.detect_extracted_directory()
+
+    def detect_extracted_directory(self) -> Path:
+        """
+        Detect a single top-level dir within the extracted archive, otherwise return extract_to.
+        """  # noqa: D200
+        extracted_dirs = [d for d in self.extract_to.iterdir() if d.is_dir()]
+        return extracted_dirs[0] if len(extracted_dirs) == 1 else self.extract_to
+
+
 @dataclass
 class FixturesSource:
     """Represents the source of test fixtures."""
@@ -70,7 +122,8 @@ def from_input(cls, input_source: str) -> "FixturesSource":
     def from_url(cls, url: str) -> "FixturesSource":
         """Create a fixture source from a direct URL."""
         release_page = get_release_page_url(url)
-        was_cached, path = download_and_extract(url, CACHED_DOWNLOADS_DIRECTORY)
+        downloader = FixtureDownloader(url, CACHED_DOWNLOADS_DIRECTORY)
+        was_cached, path = downloader.download_and_extract()
         return cls(
             input_option=url,
             path=path,
@@ -85,7 +138,8 @@ def from_release_spec(cls, spec: str) -> "FixturesSource":
         """Create a fixture source from a release spec (e.g., develop@latest)."""
         url = get_release_url(spec)
         release_page = get_release_page_url(url)
-        was_cached, path = download_and_extract(url, CACHED_DOWNLOADS_DIRECTORY)
+        downloader = FixtureDownloader(url, CACHED_DOWNLOADS_DIRECTORY)
+        was_cached, path = downloader.download_and_extract()
         return cls(
             input_option=spec,
             path=path,
@@ -111,25 +165,6 @@ def is_url(string: str) -> bool:
     return all([result.scheme, result.netloc])
 
 
-def download_and_extract(url: str, base_directory: Path) -> Tuple[bool, Path]:
-    """Download the URL and extract it locally if it hasn't already been downloaded."""
-    parsed_url = urlparse(url)
-    filename = Path(parsed_url.path).name
-    version = Path(parsed_url.path).parts[-2]
-    extract_to = base_directory / version / filename.removesuffix(".tar.gz")
-    already_cached = extract_to.exists()
-    if already_cached:
-        return already_cached, extract_to / "fixtures"
-
-    extract_to.mkdir(parents=True, exist_ok=False)
-    response = requests.get(url)
-    response.raise_for_status()
-
-    with tarfile.open(fileobj=BytesIO(response.content), mode="r:gz") as tar:
-        tar.extractall(path=extract_to)
-    return already_cached, extract_to / "fixtures"
-
-
 class SimLimitBehavior:
     """Represents options derived from the `--sim.limit` argument."""
 

src/pytest_plugins/consume/releases.py

@@ -2,6 +2,7 @@
 
 import json
 import os
+import re
 from dataclasses import dataclass
 from datetime import datetime
 from pathlib import Path
@@ -11,13 +12,12 @@
 import requests
 from pydantic import BaseModel, Field, RootModel
 
-RELEASE_INFORMATION_URL = "https://api.github.com/repos/ethereum/execution-spec-tests/releases"
-
-
 CACHED_RELEASE_INFORMATION_FILE = (
     Path(platformdirs.user_cache_dir("ethereum-execution-spec-tests")) / "release_information.json"
 )
 
+SUPPORTED_REPOS = ["ethereum/execution-spec-tests", "ethereum/tests", "ethereum/legacytests"]
+
 
 class NoSuchReleaseError(Exception):
     """Raised when a release does not exist."""
@@ -154,19 +154,20 @@ def download_release_information(destination_file: Path | None) -> List[ReleaseI
     crucial for finding older version or latest releases.
     """
     all_releases = []
-    current_url: str | None = RELEASE_INFORMATION_URL
-    max_pages = 2
-    while current_url and max_pages > 0:
-        max_pages -= 1
-        response = requests.get(current_url)
-        response.raise_for_status()
-        all_releases.extend(response.json())
-        current_url = None
-        if "link" in response.headers:
-            for link in requests.utils.parse_header_links(response.headers["link"]):
-                if link["rel"] == "next":
-                    current_url = link["url"]
-                    break
+    for repo in SUPPORTED_REPOS:
+        current_url: str | None = f"https://api.github.com/repos/{repo}/releases"
+        max_pages = 2
+        while current_url and max_pages > 0:
+            max_pages -= 1
+            response = requests.get(current_url)
+            response.raise_for_status()
+            all_releases.extend(response.json())
+            current_url = None
+            if "link" in response.headers:
+                for link in requests.utils.parse_header_links(response.headers["link"]):
+                    if link["rel"] == "next":
+                        current_url = link["url"]
+                        break
 
     if destination_file:
         destination_file.parent.mkdir(parents=True, exist_ok=True)
@@ -200,17 +201,17 @@ def get_release_page_url(release_string: str) -> str:
     Return the GitHub Release page URL for a specific release descriptor.
 
     This function can handle:
-    - A standard release string (e.g., "eip7692@latest").
+    - A standard release string (e.g., "eip7692@latest") - from execution-spec-tests only.
     - A direct asset download link (e.g.,
         "https://github.com/ethereum/execution-spec-tests/releases/download/v4.0.0/fixtures_eip7692.tar.gz").
     """
     release_information = get_release_information()
 
     # Case 1: If it's a direct GitHub Releases download link,
     #         find which release in `release_information` has an asset with this exact URL.
-    if release_string.startswith(
-        "https://github.com/ethereum/execution-spec-tests/releases/download/"
-    ):
+    repo_pattern = "|".join(re.escape(repo) for repo in SUPPORTED_REPOS)
+    regex_pattern = rf"https://github\.com/({repo_pattern})/releases/download/"
+    if re.match(regex_pattern, release_string):
         for release in release_information:
             for asset in release.assets.root:
                 if asset.url == release_string:

whitelist.txt

@@ -857,3 +857,4 @@ ize
 nectos
 fibonacci
 CPython
+legacytests