Mirrors/ArchiveBox
2
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox synced 2024-11-24 21:23:22 +00:00
Code Issues Projects Releases Packages Wiki Activity

Merge with latest dev

Browse source
This commit is contained in:
jim winstead 2024-03-01 12:05:43 -08:00
commit 0f402df42f
1 changed files with 19 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

19
tests/test_add.py
View file

@ -153,6 +153,21 @@ def test_generic_rss(tmp_path, process, disable_extractors_dict):
with open('../../mock_server/templates/example.rss', 'r', encoding='utf-8') as f:
arg_process = subprocess.run(
["archivebox", "add", "--index-only", "--parser=rss"],
stdin=f,
capture_output=True,
env=disable_extractors_dict,
)
conn = sqlite3.connect("index.sqlite3")
c = conn.cursor()
urls = c.execute("SELECT url from core_snapshot").fetchall()
tags = c.execute("SELECT name from core_tag").fetchall()
conn.commit()
conn.close()
urls = list(map(lambda x: x[0], urls))
assert "http://127.0.0.1:8080/static/example.com.html" in urls
# if the following URL appears, we must have fallen back to another parser
assert not "http://purl.org/dc/elements/1.1/" in urls
tags = list(map(lambda x: x[0], tags))
@ -197,3 +212,7 @@ def test_atom(tmp_path, process, disable_extractors_dict):
assert "http://127.0.0.1:8080/static/example.com.html" in urls
# if the following URL appears, we must have fallen back to another parser
assert not "http://www.w3.org/2005/Atom" in urls
tags = list(map(lambda x: x[0], tags))
assert "Tag1" in tags
assert "Tag2" in tags