Garmelon/PFERD
1
0
Fork 0
mirror of https://github.com/Garmelon/PFERD.git synced 2026-04-12 15:35:05 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Download all TGI files and not just lectures

Browse source
This commit is contained in:
I-Al-Istannen 2019-10-17 22:14:32 +02:00
parent 1973c931bd
commit 1407c6d264
1 changed files with 2 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

5
PFERD/tgi.py
View file

@ -18,7 +18,6 @@ pretty = PrettyLogger(logger)
class TGI:
CRAWL_URL = "https://i11www.iti.kit.edu/teaching/{year}/tgi/index"
BASE_URL = "https://i11www.iti.kit.edu"
LINK_RE = re.compile(r"^/_media/teaching/.*?/(tgi-\d+-\d+-)([^/]*\.pdf)$")
def __init__(self, base_path, year="winter2019"):
self.base_path = base_path
@ -49,11 +48,11 @@ class TGI:
files = []
for found in soup.find_all("a", href=self.LINK_RE):
for found in soup.select("a.mediafile.mf_pdf"):
url = found["href"]
full_url = self.BASE_URL + url
filename = re.search(self.LINK_RE, url).group(2)
filename = re.search(r"\d+(/tgi)?/(.+.pdf)", url).group(2)
path = pathlib.PurePath(filename)
logger.debug(f"Found file {filename} at {full_url}")