mirror of
https://github.com/boykopovar/AnyPS5.git
synced 2026-11-05 13:20:11 +00:00
MarcZX10
22d76a1f79
GitHub returns some open pull request bodies with an escape that is not valid JSON, for example a backslash followed by u00XX. gh itself rejects the page, and pr_overlap.py stops with "Invalid \uXXXX escape" on the whole open-PR list. Add a test for a page containing such a body, and one that checks valid escapes are still decoded to the same text. The first test fails on upstream main with that error. The second passes.
149 lines
6.6 KiB
Python
149 lines
6.6 KiB
Python
import json
|
|
import sys
|
|
import threading
|
|
import unittest
|
|
from pathlib import Path
|
|
from unittest.mock import patch
|
|
|
|
sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
|
|
import pr_overlap
|
|
|
|
URL = "repos/owner/project/pulls?state=open&base=main&per_page=100"
|
|
|
|
|
|
def pull(number):
|
|
return {"number": number, "user": {"login": f"author{number}"}, "head": {"sha": f"sha{number}"}, "body": None}
|
|
|
|
|
|
def response(numbers, last=None, newline="\r\n"):
|
|
headers = ["HTTP/2.0 200 OK", "Content-Type: application/json; charset=utf-8"]
|
|
if last:
|
|
headers.append(f'Link: <https://api.github.com/repositories/1/pulls?state=open&base=main&per_page=100&page=2>; rel="next", '
|
|
f'<https://api.github.com/repositories/1/pulls?state=open&base=main&per_page=100&page={last}>; rel="last"')
|
|
return newline.join(headers) + newline + newline + json.dumps([pull(n) for n in numbers])
|
|
|
|
|
|
class ServedPages:
|
|
def __init__(self, pages, last):
|
|
self.pages = pages
|
|
self.last = last
|
|
self.requested = []
|
|
self.lock = threading.Lock()
|
|
|
|
def __call__(self, *args):
|
|
self.assertion(args)
|
|
number = int(args[2].rsplit("&page=", 1)[1])
|
|
with self.lock:
|
|
self.requested.append(number)
|
|
return response(self.pages.get(number, []), self.last if number == 1 else None)
|
|
|
|
@staticmethod
|
|
def assertion(args):
|
|
if args[:2] != ("api", "--include") or not args[2].startswith(URL + "&page="):
|
|
raise AssertionError(f"unexpected gh call {args}")
|
|
|
|
|
|
class ParallelPageTests(unittest.TestCase):
|
|
def open_prs(self, served):
|
|
with patch.object(pr_overlap, "REPO", "owner/project", create=True), \
|
|
patch.object(pr_overlap, "gh", side_effect=served), \
|
|
patch.object(pr_overlap, "progress"):
|
|
return pr_overlap.open_prs("main")
|
|
|
|
def test_every_page_named_by_the_last_link_is_fetched(self):
|
|
pages = {1: range(1, 101), 2: range(101, 201), 3: range(201, 251)}
|
|
served = ServedPages(pages, last=3)
|
|
found = self.open_prs(served)
|
|
self.assertEqual(set(found), set(range(1, 251)))
|
|
self.assertCountEqual(served.requested, [1, 2, 3])
|
|
self.assertEqual(found[250], {"ref": "refs/pr/250", "sha": "sha250", "author": "author250", "depends": set()})
|
|
|
|
def test_a_single_page_needs_one_request(self):
|
|
served = ServedPages({1: range(1, 43)}, last=None)
|
|
self.assertEqual(set(self.open_prs(served)), set(range(1, 43)))
|
|
self.assertEqual(served.requested, [1])
|
|
|
|
def test_next_links_without_a_last_link_are_followed_even_on_short_pages(self):
|
|
pages = {1: range(1, 3), 2: range(3, 5), 3: range(5, 6)}
|
|
served = ServedPages(pages, last=None)
|
|
|
|
def next_only(*args):
|
|
result = served(*args)
|
|
number = int(args[2].rsplit("&page=", 1)[1])
|
|
if number < 3:
|
|
link = f'Link: <https://api.github.com/repositories/1/pulls?page={number + 1}&per_page=100>; rel="next"'
|
|
result = result.replace("\r\n\r\n", "\r\n" + link + "\r\n\r\n", 1)
|
|
return result
|
|
|
|
found = self.open_prs(next_only)
|
|
self.assertEqual(set(found), set(range(1, 6)))
|
|
self.assertEqual(served.requested, [1, 2, 3])
|
|
|
|
def test_next_only_full_final_page_checks_for_a_new_following_page(self):
|
|
pages = {1: range(1, 101), 2: range(101, 201), 3: range(201, 202)}
|
|
served = ServedPages(pages, last=None)
|
|
|
|
def next_only(*args):
|
|
result = served(*args)
|
|
number = int(args[2].rsplit("&page=", 1)[1])
|
|
if number == 1:
|
|
link = 'Link: <https://api.github.com/repositories/1/pulls?page=2&per_page=100>; rel="next"'
|
|
result = result.replace("\r\n\r\n", "\r\n" + link + "\r\n\r\n", 1)
|
|
return result
|
|
|
|
self.assertEqual(set(self.open_prs(next_only)), set(range(1, 202)))
|
|
self.assertEqual(served.requested, [1, 2, 3])
|
|
|
|
def test_invalid_or_repeated_next_pages_are_rejected(self):
|
|
for query in ("per_page=100", "page=one", "page=0", "page=1", "page=2&page=3", "page=&page=2"):
|
|
with self.subTest(query=query):
|
|
link = f'Link: <https://api.github.com/repositories/1/pulls?{query}>; rel="next"'
|
|
output = response(range(1, 3)).replace("\r\n\r\n", "\r\n" + link + "\r\n\r\n", 1)
|
|
with patch.object(pr_overlap, "REPO", "owner/project", create=True), \
|
|
patch.object(pr_overlap, "gh", return_value=output):
|
|
with self.assertRaisesRegex(ValueError, "invalid next page link"):
|
|
pr_overlap.open_prs("main")
|
|
|
|
def test_a_full_last_page_is_followed_by_the_next_one(self):
|
|
pages = {1: range(1, 101), 2: range(101, 201), 3: range(201, 202)}
|
|
served = ServedPages(pages, last=2)
|
|
self.assertEqual(set(self.open_prs(served)), set(range(1, 202)))
|
|
self.assertEqual(sorted(served.requested), [1, 2, 3])
|
|
|
|
def test_a_full_last_page_followed_by_an_empty_one_stops(self):
|
|
served = ServedPages({1: range(1, 101), 2: range(101, 201)}, last=2)
|
|
self.assertEqual(set(self.open_prs(served)), set(range(1, 201)))
|
|
self.assertEqual(sorted(served.requested), [1, 2, 3])
|
|
|
|
def test_a_pull_request_shifted_onto_two_pages_is_kept_once(self):
|
|
served = ServedPages({1: range(1, 101), 2: range(100, 150)}, last=2)
|
|
found = self.open_prs(served)
|
|
self.assertEqual(len(found), 149)
|
|
|
|
def test_headers_with_plain_newlines_are_split_from_the_body(self):
|
|
with patch.object(pr_overlap, "gh", return_value=response(range(1, 3), newline="\n")):
|
|
head, body = pr_overlap.page(URL + "&page=1")
|
|
self.assertTrue(head.startswith("HTTP/2.0 200 OK"))
|
|
self.assertEqual(body, json.dumps([pull(1), pull(2)]))
|
|
|
|
def test_a_non_array_page_is_rejected(self):
|
|
with self.assertRaisesRegex(ValueError, "non-array page"):
|
|
pr_overlap.json_pages('{"message": "Not Found"}')
|
|
|
|
def test_a_malformed_unicode_escape_in_a_pull_request_body_is_read(self):
|
|
page = r'[{"number": 14172, "body": "lowercase `\\\u00XX`, unpaired"}]'
|
|
try:
|
|
pages = pr_overlap.json_pages(page)
|
|
except ValueError as error:
|
|
self.fail(f"a page with a malformed escape was rejected: {error}")
|
|
self.assertEqual([[pull["number"] for pull in found] for found in pages], [[14172]])
|
|
|
|
def test_valid_escapes_are_decoded_unchanged(self):
|
|
page = r'[{"title": "café \\ \"quoted\" \\u0041", "number": 1}]'
|
|
pages = pr_overlap.json_pages(page)
|
|
self.assertEqual(pages[0][0]["title"], 'café \\ "quoted" \\u0041')
|
|
|
|
|
|
if __name__ == "__main__":
|
|
unittest.main()
|