Integration of crawl module in histories

crawl/crawl.py

@@ -1,4 +1,5 @@
 from threading import Thread
+from queue import Queue
 from urllib.robotparser import RobotFileParser
 from urllib.error import URLError
 from urllib.parse import urlparse
@@ -26,10 +27,10 @@ MAX_PER_PAGE = 10
 FOOTER_URL = re.compile(".*footer.*")
 
 class Settings:
-    USER_AGENT = 'Blah'
+    USER_AGENT = 'Default User'
 
 settings = Settings()
-startup_time = datetime.now()
+startup_time = datetime.min
 
 
 def url_getter(html, current_page, root_url):
@@ -161,16 +162,24 @@ class CrawlingThread(Thread):
     """ A separate thread for the crawling task. This is needed to use asyncio,
     since the thread will need its own event loop. """
 
-    def __init__(self):
+    def __init__(self, user, url, queue):
+        global settings
+        self.queue = queue
         super(CrawlingThread, self).__init__()
+        if user:
+            settings.USER_AGENT = user.serialize_headers()
+        self.url = url
 
     def run(self):
+        global startup_time
         tasks = []
         #tasks.append(async_crawler("http://plus.google.com/+Python"))
-        tasks.append(async_crawler('https://python.org/'))
+        #tasks.append(async_crawler('https://python.org/'))
+        tasks.append(async_crawler(self.url, self.queue))
 
         loop = asyncio.new_event_loop()
         asyncio.set_event_loop(loop)
+        startup_time = datetime.now()
         loop.run_until_complete(asyncio.wait(tasks))
         loop.close()
 
@@ -211,13 +220,13 @@ async def async_print(url):
             url,
             datetime.now() - startup_time))
 
-async def async_crawler(url):
-    queue = [url]
+async def async_crawler(url, queue):
+    queued = [url]
     crawled = []
-    while queue and (len(crawled) < HARD_LIMIT):
+    while queued and (len(crawled) < HARD_LIMIT):
         async with aiohttp.ClientSession() as session:
             try:
-                url = queue.pop(0)
+                url = queued.pop(0)
             except IndexError:
                 print("Error queue is empty")
                 return crawled
@@ -236,13 +245,14 @@ async def async_crawler(url):
                         new_urls,
                         randrange(min(MAX_PER_PAGE, len(new_urls)))
                     )
-                    queue += [sample_url for sample_url in sampled if
-                              sample_url not in queue and sample_url not in
+                    queued += [sample_url for sample_url in sampled if
+                              sample_url not in queued and sample_url not in
                               crawled]
     print(crawled)
-    return crawled
+    queue.put(crawled)
 
 if __name__ == '__main__':
-    crawl = CrawlingThread()
+    queue = Queue()
+    crawl = CrawlingThread(None, "https://python.org/", queue)
     crawl.start()
     crawl.join()

histories/models.py

@@ -5,9 +5,10 @@ interests, keywords...
 
 import random
 from math import floor
+from queue import Queue
 from django.db import models
 import profiles.models as profiles
-#from crawl import crawl
+from crawl import crawl
 from pinocchio.settings import HISTORY_MIN
 
 
@@ -63,7 +64,7 @@ class History(models.Model):
         self.save()
 
 
-def generate_partial_history(user, t_start, history):
+def generate_partial_history(user, t_start):
     """ Generate the part of the history resulting from the crawl starting at
     the given url.
     """
@@ -71,7 +72,14 @@ def generate_partial_history(user, t_start, history):
     basis = generate_first_url(user)
     result.append((basis, t_start))
     t_start += 5* random.weibullvariate(1, 1.5)
-    #crawler = crawl.CrawlingThread()
+    queue = Queue()
+    crawler = crawl.CrawlingThread(user, basis, queue)
+    crawler.start()
+    crawler.join()
+    urls = queue.get()
+    for url in urls:
+        t_start += 5* random.weibullvariate(1, 1.5)
+        result.append((url, t_start)
     return result
 
 def generate_first_url(user):
@@ -104,7 +112,7 @@ def generate_history(user, ts_start):
 
     while history_line < length:
         ts_start += 5 * random.weibullvariate(1, 2.8)
-        history_list = generate_partial_history(user, ts_start, history)
+        history_list = generate_partial_history(user, ts_start)
         ts_start = history_list[-1].timestamp + 5 * random.weibullvariate(1, 5)
         for (url, timestamp) in history_list:
             new_line = HistoryEntry(