Compare commits

..

No commits in common. "7cc3d1b7e46778df43ca7193cd2b348286b281e0" and "07f8b1ef311330fa4f708424e10bb3e799ec7886" have entirely different histories.

View file

@ -1,4 +1,4 @@
#! python3 #! python
import locale import locale
import logging import logging
import os import os
@ -74,15 +74,12 @@ def download(id: int):
author = None author = None
try: try:
category = pC.find("span", {"class": "categoryInfo"}).find_all() category = pC.find("span", {"class": "categoryInfo"}).find().text
category = [c.text for c in category]
category = ";".join(category)
except AttributeError: except AttributeError:
category = None category = None
try: try:
tags = [x.text for x in pC.find("div", {"class": "tagsInfo"}).find_all("a")] tags = [x.text for x in pC.find("div", {"class": "tagsInfo"}).find_all("a")]
tags = ";".join(tags)
except AttributeError: except AttributeError:
tags = None tags = None
@ -130,6 +127,7 @@ def run_downloads(min_id: int, max_id: int, num_threads: int = 8):
# sqlite can't handle lists so let's convert them to a single row csv # sqlite can't handle lists so let's convert them to a single row csv
# TODO: make sure our database is properly normalized # TODO: make sure our database is properly normalized
df = pd.DataFrame(res) df = pd.DataFrame(res)
df.tags = df.tags.apply(lambda x: "; ".join(x) if x is not None else None)
return df return df