Support parse jobs posts

Author: polyrabbit

hacker_news/llm/openai.py

@@ -117,7 +117,7 @@ def call_openai_family(content: str, sys_prompt: str) -> str:
     for line in answer.split('\n'):
         if not line.strip():
             continue
-        if 'summary' in line.lower() and len(line) <= 100:
+        if 'summary' in line.lower() and line.strip()[-1] == ':':
             continue
         answer = line
         break

hacker_news/news.py

@@ -80,6 +80,9 @@ def get_score(self) -> int:
         except:
             return 0
 
+    def is_hiring_job(self) -> bool:
+        return self.get_score() == 0 and not self.author and 'YC ' in self.title
+
     def slug(self):
         return slugify(self.title or 'no title')
 
@@ -130,7 +133,8 @@ def summarize_by_openai(self, content):
         if not openai.api_key:
             logger.info("OpenAI API key is not set")
             return ''
-        if self.get_score() < config.openai_score_threshold:  # Avoid expensive openai
+        if (self.get_score() < config.openai_score_threshold # Avoid expensive openai
+                and not self.is_hiring_job()):
             logger.info("Score %d is too small, ignore openai", self.get_score())
             return ''
 

test/test_hackernews_parser.py

@@ -2,6 +2,7 @@
 from datetime import datetime, timedelta
 
 from hacker_news.algolia_api import get_news
+from hacker_news.news import News
 from hacker_news.parser import HackerNewsParser
 
 
@@ -52,3 +53,7 @@ def test_algolia_api(self):
         date = news_list[0].submit_time.date()
         for news in news_list:
             self.assertEqual(date, news.submit_time.date())
+
+    def test_maybe_jobs_post(self):
+        news = News(title='MixRank (YC S11) Is Hiring Software Engineers and Founders Globally')
+        self.assertTrue(news.is_hiring_job())