fundor333/fundor333.com
1
0
Fork 0
Code Issues Pull Requests Actions 5 Packages Projects Releases Wiki Activity

Fix file syndication

Browse Source
This commit is contained in:
Fundor333
2025-06-16 03:25:25 +02:00
parent a1f3f16e21
commit aa7387154f
1 changed files with 11 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
action_script/syndication-collector.py
View File

@@ -84,7 +84,7 @@ class FeedFinder:
return entry.get("content") return entry.get("content")
def run(self, rss_url: str, domain: str, output: dict): def run(self, rss_url: str, domain: str, output: dict, data: dict):
feed = feedparser.parse(rss_url) feed = feedparser.parse(rss_url)
if feed.status == 200: if feed.status == 200:
@@ -93,7 +93,9 @@ class FeedFinder:
for e in self.find_urls(self.get_label(feed.version, entry)): for e in self.find_urls(self.get_label(feed.version, entry)):
if domain in e: if domain in e:
temp = e
e = clean_slug(e) e = clean_slug(e)
data[e] = temp
if output.get(e, False): if output.get(e, False):
output[e].append(link.strip()) output[e].append(link.strip())
else: else:
@@ -118,14 +120,16 @@ class IndieWebOrg:
return entry.get("content") return entry.get("content")
def run(self, rss_url: str, site: str, domain: str, output: dict): def run(self, rss_url: str, site: str, domain: str, output: dict, data: dict):
feed = feedparser.parse(rss_url) feed = feedparser.parse(rss_url)
if feed.status == 200: if feed.status == 200:
for entry in feed.entries: for entry in feed.entries:
for e in self.find_urls(self.get_label(feed.version, entry)): for e in self.find_urls(self.get_label(feed.version, entry)):
if domain in e: if domain in e:
temp = e
e = clean_slug(e) e = clean_slug(e)
data[e] = temp
if output.get(e, False): if output.get(e, False):
output[e].append(site) output[e].append(site)
else: else:
@@ -143,6 +147,7 @@ class WriterSyndication:
domain: str, domain: str,
): ):
self.output = {} self.output = {}
self.data = {}
self.rss = rss self.rss = rss
self.indiweb = indiweb self.indiweb = indiweb
self.domain = domain self.domain = domain
@@ -151,25 +156,26 @@ class WriterSyndication:
def data_gathering(self): def data_gathering(self):
ff = FeedFinder() ff = FeedFinder()
for i in self.rss: for i in self.rss:
ff.run(i, self.domain, self.output) ff.run(i, self.domain, self.output, self.data)
ii = IndieWebOrg() ii = IndieWebOrg()
for a, b in self.indiweb: for a, b in self.indiweb:
ii.run(a, b, self.domain, output=self.output) ii.run(a, b, self.domain, output=self.output, data=self.data)
hn = HackerNewsFinder(self.hacker_news_username, self.output) hn = HackerNewsFinder(self.hacker_news_username, self.output)
hn.run() hn.run()
print(self.data)
def write(self): def write(self):
for key in self.output.keys(): for key in self.output.keys():
path_folder = os.path.join("data", "syndication") path_folder = os.path.join("data", "syndication")
Path(path_folder).mkdir(parents=True, exist_ok=True) Path(path_folder).mkdir(parents=True, exist_ok=True)
path_file = os.path.join(path_folder, key) path_file = os.path.join(path_folder, key)
if os.path.exists(path_file + ".json"): if os.path.exists(path_file + ".json"):
with open(path_file + ".json") as fp: with open(path_file + ".json") as fp:
data = json.load(fp) data = json.load(fp)
data["source"] = self.data.get(key, None)
if data.get("syndication", False): if data.get("syndication", False):
data["syndication"] = sorted( data["syndication"] = sorted(
set(data["syndication"] + self.output[key]) set(data["syndication"] + self.output[key])