mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-06-03 16:18:59 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat: depth=0 crawls the current page only

Browse source
This commit is contained in:
Cristian 2020-07-07 09:49:28 -05:00
parent 8b22a2a7dd
commit 2db0324539
2 changed files with 21 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

12
tests/test_args.py
View file

@ -1,7 +1,15 @@
import subprocess
import json
from .fixtures import *
def test_depth_flag_is_accepted(tmp_path, process):
def test_depth_flag_is_accepted(process):
arg_process = subprocess.run(["archivebox", "add", "https://example.com", "--depth=0"], capture_output=True)
assert 'unrecognized arguments: --depth' not in arg_process.stderr.decode('utf-8')
assert 'unrecognized arguments: --depth' not in arg_process.stderr.decode('utf-8')
def test_depth_flag_0_crawls_only_the_arg_page(tmp_path, process):
arg_process = subprocess.run(["archivebox", "add", "https://example.com", "--depth=0"], capture_output=True)
archived_item_path = list(tmp_path.glob('archive/**/*'))[0]
with open(archived_item_path / "index.json", "r") as f:
output_json = json.load(f)
assert output_json["base_url"] == "example.com"