Mirrors/ArchiveBox
2
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox synced 2024-11-10 06:34:16 +00:00
Code Issues Projects Releases Packages Wiki Activity

tests: add test for mercury-parser

Browse source
This commit is contained in:
ttimasdf 2020-09-22 16:47:43 +08:00 committed by Cristian Vargas
parent 776fda947d
commit e3329be291
2 changed files with 9 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
tests/fixtures.py
View file

@ -16,6 +16,7 @@ def disable_extractors_dict():
"USE_WGET": "false",
"USE_SINGLEFILE": "false",
"USE_READABILITY": "false",
"USE_MERCURY": "false",
"SAVE_PDF": "false",
"SAVE_SCREENSHOT": "false",
"SAVE_DOM": "false",

8
tests/test_extractors.py
View file

@ -30,6 +30,14 @@ def test_readability_works(tmp_path, process, disable_extractors_dict):
output_file = archived_item_path / "readability" / "content.html"
assert output_file.exists()
def test_mercury_works(tmp_path, process, disable_extractors_dict):
disable_extractors_dict.update({"USE_MERCURY": "true"})
add_process = subprocess.run(['archivebox', 'add', 'http://127.0.0.1:8080/static/example.com.html'],
capture_output=True, env=disable_extractors_dict)
archived_item_path = list(tmp_path.glob("archive/**/*"))[0]
output_file = archived_item_path / "mercury" / "content.html"
assert output_file.exists()
def test_readability_works_with_wget(tmp_path, process, disable_extractors_dict):
disable_extractors_dict.update({"USE_READABILITY": "true", "USE_WGET": "true"})
add_process = subprocess.run(['archivebox', 'add', 'http://127.0.0.1:8080/static/example.com.html'],