mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-14 23:24:30 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

enforce utf8 on literally all file operations because windows sucks

Browse source
This commit is contained in:
Nick Sweeting 2021-03-27 01:01:29 -04:00
parent 185d2f9f9b
commit bd6d9c165b
9 changed files with 29 additions and 28 deletions
Download patch file Download diff file Expand all files Collapse all files

8
tests/test_extractors.py
View file

@ -86,7 +86,7 @@ def test_headers_retrieved(tmp_path, process, disable_extractors_dict):
output_file = archived_item_path / "headers.json"
assert output_file.exists()
headers_file = archived_item_path / 'headers.json'
with open(headers_file) as f:
with open(headers_file, 'r', encoding='utf-8') as f:
headers = pyjson.load(f)
assert headers['Content-Language'] == 'en'
assert headers['Content-Script-Type'] == 'text/javascript'
@ -98,7 +98,7 @@ def test_headers_redirect_chain(tmp_path, process, disable_extractors_dict):
capture_output=True, env=disable_extractors_dict)
archived_item_path = list(tmp_path.glob("archive/**/*"))[0]
output_file = archived_item_path / "headers.json"
with open(output_file) as f:
with open(output_file, 'r', encoding='utf-8') as f:
headers = pyjson.load(f)
assert headers['Content-Language'] == 'en'
assert headers['Content-Script-Type'] == 'text/javascript'
@ -110,6 +110,6 @@ def test_headers_400_plus(tmp_path, process, disable_extractors_dict):
capture_output=True, env=disable_extractors_dict)
archived_item_path = list(tmp_path.glob("archive/**/*"))[0]
output_file = archived_item_path / "headers.json"
with open(output_file) as f:
with open(output_file, 'r', encoding='utf-8') as f:
headers = pyjson.load(f)
assert headers["Status-Code"] == "200"
assert headers["Status-Code"] == "200"