Anon/DanbooruBot
Archived
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Refactored to retry when overshooting pagination to fetch the last page and try again

Browse Source
This commit is contained in:
Anon 2023-10-19 21:22:07 -07:00
parent 88e11f58bc
commit 22c5a6e231
1 changed files with 125 additions and 106 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

229
src/konachan_backend.py
View File

@ -22,146 +22,165 @@
import requests import requests
import os import os
import random import random
import re
import time
def random_tag(*tags): def random_tag(*tags):
return len(tags) == 1 and tags[0].lower() == "random" return len(tags) == 1 and tags[0].lower() == "random"
def collect_tags(post): def collect_tags(post):
return post["tags"].strip().lower().split() return post["tags"].strip().lower().split()
def is_banned(post, profile): def is_banned(post, profile):
tag_response = collect_tags(post) tag_response = collect_tags(post)
tag_banned = profile["banned_tags"] tag_banned = profile["banned_tags"]
for tag in tag_banned: for tag in tag_banned:
if tag in tag_response: if tag in tag_response:
return tag return tag
return None return None
def get_nsfw(post): def get_nsfw(post):
return post["rating"] in ("q", "e") return post["rating"] in ("q", "e")
def select_from_response(response, profile, max_size=None): def select_from_response(response, profile, max_size=None):
for post in response: for post in response:
if is_banned(post, profile): if is_banned(post, profile):
continue continue
elif "file_url" not in post: elif "file_url" not in post:
continue continue
# Select only nsfw # Select only nsfw
elif ( profile["force_nsfw"] is not None and elif ( profile["force_nsfw"] is not None and
profile["force_nsfw"] != get_nsfw(post) profile["force_nsfw"] != get_nsfw(post)
): ):
continue continue
# Make sure serverside size is not larger than max_size # Make sure serverside size is not larger than max_size
elif ( max_size != None and elif ( max_size != None and
"file_size" in post and "file_size" in post and
post["file_size"] > max_size post["file_size"] > max_size
): ):
continue continue
return post return post
return None return None
class downloader: class downloader:
username = None def __init__(self, backend_credentials):
password = None self.api_endpoint = "{}/post/index.json?limit={}&page={}"
max_size = None self.html_endpoint = "{}/post/index?limit={}&page={}"
tmp = None self.tag_url = "&tags={}"
url = "" self.limit = 100
api_endpoint = "post.json?random=true&limit=100" self.retry_limit = 3
api_tags = "&tags={}"
api_limit = "&limit={}"
api_offset = "&page={}"
limit = 100
max_depth = 200
def __init__(self, backend_credentials): self.username = backend_credentials["username"]
self.username = backend_credentials["username"] self.password = backend_credentials["password"]
self.password = backend_credentials["password"] self.depth = backend_credentials["max_size"]
self.max_size = backend_credentials["max_size"] self.tmp = backend_credentials["tmp_dir"]
self.tmp = backend_credentials["tmp_dir"] self.url = backend_credentials["url"]
self.url = backend_credentials["url"] self.max_depth = backend_credentials["max_depth"]
self.max_depth = backend_credentials["max_depth"] random.seed(os.urandom(16))
random.seed(os.urandom(16))
def download_post(self, post): def download_post(self, post):
file_url = post["file_url"] file_url = post["file_url"]
full_path = post["full_path"] full_path = post["full_path"]
remote_image = requests.get(file_url) remote_image = requests.get(file_url)
if remote_image.status_code != 200: if remote_image.status_code != 200:
print("Remote image request returned:", remote_image.status_code) print("Remote image request returned:", remote_image.status_code)
return None return None
with open(full_path, "wb") as f: with open(full_path, "wb") as f:
f.write(remote_image.content) f.write(remote_image.content)
return post return post
def get_full_url(self, limit=100, offset=0, *tags): def search(self, search_url):
search_url = "/".join((self.url, self.api_endpoint)) search_request = None
search_url += self.api_limit.format(str(limit)) if self.username and self.password:
search_url += self.api_offset.format(str(offset)) search_request = requests.get(search_url,
if tags and not random_tag(*tags): auth=(self.username, self.password)
search_tags = "+".join(tags) )
search_url += self.api_tags.format(search_tags) else:
return search_url search_request = requests.get(search_url)
return search_request
# I suck at regex :(
def get_max_page(self, html):
match = re.findall('page=[0-9]*', html)
if match:
last_group = match[len(match) - 1]
last_page = last_group.rsplit("=", 1)[1]
return int(last_page)
else:
return None
def search(self, search_url): def fetch_post(self, profile):
search_request = None tags = profile["tags"]
if self.username and self.password: selected = dict()
search_request = requests.get(search_url, max_depth = self.max_depth
auth=(self.username, self.password) search_url_tags = "+".join(tags)
) search_url = ""
else:
search_request = requests.get(search_url)
return search_request
for _ in range(0, self.retry_limit):
page_offset = random.randint(0, max_depth)
search_url = self.api_endpoint.format(self.url, self.limit, page_offset)
search_url_html = self.html_endpoint.format(self.url, self.limit, page_offset)
if search_url_tags:
search_url += self.tag_url.format(search_url_tags)
search_url_html += self.tag_url.format(search_url_tags)
def fetch_post(self, profile): search_request = self.search(search_url)
# Search ratings: s=safe, e=nsfw
tags = profile["tags"]
page_offset = random.randint(0, self.max_depth) if search_request.status_code != 200:
search_url = self.get_full_url(self.limit, page_offset, *tags) print("Search request returned:", search_request.status_code)
search_request = self.search(search_url) continue
if search_request.status_code != 200: posts = search_request.json()
print("Search request returned:", search_request.status_code) random.shuffle(posts)
return None
posts = search_request.json() selected = select_from_response(posts, profile)
random.shuffle(posts)
selected = select_from_response(posts, profile, self.max_size) if selected is None:
print("Could not select image based on criteria")
time.sleep(2)
search_request = self.search(search_url_html)
if selected is None: if search_request.status_code == 200:
print("Could not select image based on criteria") new_max_depth = self.get_max_page(search_request.text)
return None if new_max_depth < max_depth:
max_depth = new_max_depth
else:
max_depth = max_depth // 2
else:
max_depth = max_depth // 2
continue
tag_response = collect_tags(selected) break
nsfw = get_nsfw(selected)
file_url = selected["file_url"]
basename = "{}.{}".format(selected["md5"], file_url.rsplit(".", 1)[1]) if not selected:
full_path = os.path.join(self.tmp, basename) return None
r = { tag_response = collect_tags(selected)
# Add profile to dictioanry nsfw = get_nsfw(selected)
"profile": profile, file_url = selected["file_url"]
# Query results basename = "{}.{}".format(selected["md5"], file_url.rsplit(".", 1)[1])
"search_url": search_url, full_path = os.path.join(self.tmp, basename)
"file_url": file_url,
"full_path": full_path,
"tag_response": tag_response,
"nsfw": nsfw
}
return r r = {
# Add profile to dictioanry
"profile": profile,
# Query results
"search_url": search_url,
"file_url": file_url,
"full_path": full_path,
"tag_response": tag_response,
"nsfw": nsfw
}
return r