From 62d68da415704425b1b1c623d93acfbbe7326e16 Mon Sep 17 00:00:00 2001
From: Tanner Collin <git@tannercollin.com>
Date: Fri, 23 Aug 2019 08:23:48 +0000
Subject: [PATCH] Finish prototype api server

---
 apiserver/feeds/hackernews.py | 56 +++++++++++++++++++++
 apiserver/server.py           | 95 +++++++++++++++++++++++++++++++++++
 2 files changed, 151 insertions(+)
 create mode 100644 apiserver/feeds/hackernews.py
 create mode 100644 apiserver/server.py

diff --git a/apiserver/feeds/hackernews.py b/apiserver/feeds/hackernews.py
new file mode 100644
index 0000000..b8fc8a7
--- /dev/null
+++ b/apiserver/feeds/hackernews.py
@@ -0,0 +1,56 @@
+import requests
+
+API_TOPSTORIES = lambda x: 'https://hacker-news.firebaseio.com/v0/topstories.json'
+API_ITEM = lambda x : 'https://hn.algolia.com/api/v1/items/{}'.format(x)
+
+SITE_LINK = lambda x : 'https://news.ycombinator.com/item?id={}'.format(x)
+SITE_AUTHOR_LINK = lambda x : 'https://news.ycombinator.com/user?id={}'.format(x)
+
+def api(route, ref=None):
+    r = requests.get(route(ref), timeout=5)
+    return r.json()
+
+def feed():
+    return api(API_TOPSTORIES)[:30]
+
+def comment(i):
+    c = {}
+    c['author'] = i.get('author', '')
+    c['score'] = i.get('points', 0)
+    c['date'] = i.get('created_at_i', 0)
+    c['text'] = i.get('text', '')
+    c['link'] = SITE_LINK(i['id'])
+    c['comments'] = [comment(j) for j in i['children']]
+    return c
+
+def comment_count(i):
+    alive = 1 if i['author'] else 0
+    return sum([comment_count(c) for c in i['comments']]) + alive
+
+def story(ref):
+    r = api(API_ITEM, ref)
+
+    if 'deleted' in r:
+        return False
+    elif r.get('type', '') != 'story':
+        return False
+
+    s = {}
+    s['author'] = r.get('author', '')
+    s['author_link'] = SITE_AUTHOR_LINK(r.get('author', ''))
+    s['score'] = r.get('points', 0)
+    s['date'] = r.get('created_at_i', 0)
+    s['title'] = r.get('title', '')
+    s['link'] = SITE_LINK(ref)
+    s['source'] = 'hackernews'
+    s['url'] = r.get('url', '')
+    s['comments'] = [comment(i) for i in r['children']]
+    s['num_comments'] = comment_count(s) - 1
+
+    if 'text' in r and r['text']:
+        s['text'] = r['text']
+
+    return s
+
+if __name__ == '__main__':
+    print(story(20763961))
diff --git a/apiserver/server.py b/apiserver/server.py
new file mode 100644
index 0000000..1f43635
--- /dev/null
+++ b/apiserver/server.py
@@ -0,0 +1,95 @@
+import logging
+logging.basicConfig(
+        format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+        level=logging.INFO)
+
+import copy
+import threading
+import time
+import random
+import requests
+import string
+
+from feeds import hackernews
+from flask import abort, Flask, request
+from flask_cors import CORS
+
+CACHE_LENGTH = 300
+READ_API = 'http://127.0.0.1:33843'
+
+news_index = 0
+news_list = []
+news_ref_to_id = {}
+news_cache = {}
+
+flask_app = Flask(__name__)
+cors = CORS(flask_app)
+
+@flask_app.route('/')
+def index():
+    front_page = [news_cache[news_ref_to_id[ref]] for ref in news_list]
+    front_page = [copy.deepcopy(x) for x in front_page if 'title' in x]
+    for story in front_page:
+        if 'comments' in story: story.pop('comments')
+        if 'text' in story: story.pop('text')
+    return {'stories': front_page}
+
+@flask_app.route('/<id>')
+def comments(id):
+    if id in news_cache:
+        return {'story': news_cache[id]}
+    else:
+        abort(404)
+
+print('Starting Flask...')
+web_thread = threading.Thread(target=flask_app.run, kwargs={'port': 33842})
+web_thread.setDaemon(True)
+web_thread.start()
+
+def new_id():
+    return ''.join(random.choice(string.ascii_uppercase) for _ in range(4))
+
+def get_article(url):
+    try:
+        r = requests.post(READ_API, data=dict(url=url), timeout=10)
+
+        if r.status_code != 200:
+            raise
+
+        return r.text
+    except:
+        return '<p>Problem parsing article :(</p>'
+
+while True:
+    if news_index == 0:
+        feed = hackernews.feed()
+        new_refs = [ref for ref in feed if ref not in news_list]
+        for ref in new_refs:
+            news_list.insert(0, ref)
+            nid = new_id()
+            news_ref_to_id[ref] = nid
+            news_cache[nid] = dict(id=nid, ref=ref)
+
+        if len(new_refs):
+            logging.info('Added {} new refs.'.format(len(new_refs)))
+
+        while len(news_list) > CACHE_LENGTH:
+            old_ref = news_list.pop()
+            del news_cache[news_ref_to_id[old_ref]]
+            del news_ref_to_id[old_ref]
+            logging.info('Removed ref {}.'.format(old_ref))
+
+    if news_index < len(news_list):
+        update_ref = news_list[news_index]
+        update_id = news_ref_to_id[update_ref]
+        news_story = news_cache[update_id]
+        story = hackernews.story(update_ref)
+        if story:
+            news_story.update(story)
+        if news_story.get('url', '') and not news_story.get('text', ''):
+            news_story['text'] = get_article(news_story['url'])
+
+    time.sleep(1)
+
+    news_index += 1
+    if news_index == CACHE_LENGTH: news_index = 0