anond.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import feedparser
import pymongo
import datetime

anond_feed = feedparser.parse("http://anond.hatelabo.jp/rss")
client = pymongo.MongoClient()
db = client["anond"]
posts = db.posts

for entry in anond_feed["entries"]:
    entry_id = entry["id"][25:]
    entry_title = entry["title"]
    entry_text = entry["content"][0]["value"]
    entry_updated = entry["updated"]
    entry_link = entry["link"]

    if posts.find_one({"id":entry_id}):
#        print str(entry_id) + " is already inserted"
        pass
    else:
        posts.insert({"id":entry_id,"title":entry_title,"text":entry_text,"updated":entry_updated,"link":entry_link})
        print str(entry_id) + " inserted"


print str(datetime.datetime.now()) + " / " + str(posts.count()) + " posts archived"
print "========="

load_anond.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

# usage: python load_anond.py <id>

import pymongo
import sys

client = pymongo.MongoClient()
db = client["anond"]
posts = db.posts

for e in posts.find({"id":sys.argv[1]}):
    print "title: " + e["title"]
    print e["text"].replace("<p>","").replace("</p>","")
    print "Updated: " + e["updated"]
    print "link: " + e["link"]
    print ""