zz

mongo/mongocwarler.py

@@ -9,10 +9,14 @@ import courlan
 import urllib
 from datetime import datetime
 import click
+import logging as LOGGER
+import os
 
-LANGUAGE="sk"
-DOMAIN = "sk"
-BATCHSIZE=10
+LANGUAGE= os.getenv("SUCKER_LANGUAGE","sk")
+DOMAIN = os.getenv("SUCKER_DOMAIN","sk")
+BATCHSIZE=os.getenv("SUCKER_BATCHSIZE",10)
+CONNECTION=os.getenv("SUCKER_CONNECTION","mongodb://root:example@localhost:27017/")
+DBNAME=os.getenv("SUCKER_DBNAME","crawler")
 MINFILESIZE=300
 MAXFILESIZE=10000000
 MINTEXTSIZE=200
@@ -48,7 +52,7 @@ def calculate_checksums(text):
 
 def is_robot_good(link,rules):
     # check robots.txt rules
-    if rules is not None and not rules.can_fetch("*", llink):
+    if rules is not None and not rules.can_fetch("*", link):
         return False
     return True
 
@@ -61,10 +65,10 @@ def is_link_good(link):
     print(llink,ldomain)
     # domain rules
     if not ldomain.endswith(DOMAIN):
-        print("bad domain")
+        LOGGER.debug("bad domain")
         return None
     if courlan.is_not_crawlable(llink):
-        print("not crawlable")
+        LOGGER.debug("not crawlable")
         return None
     return llink
 
@@ -112,14 +116,14 @@ def fetch_pages(link_batch):
             good = True
             if response.status != 200:
                 good = False
-                #LOGGER.error('not a 200 response: %s for URL %s', response.status, url)
+                LOGGER.error('not a 200 response: %s for URL %s', response.status, url)
             elif response.data is None or len(response.data) < MINFILESIZE:
-                #LOGGER.error('too small/incorrect for URL %s', url)
+                LOGGER.error('too small/incorrect for URL %s', url)
                 good = False
             # raise error instead?
             elif len(response.data) > MAXFILESIZE:
                 good = False
-            #LOGGER.error('too large: length %s for URL %s', len(response.data), url)
+                LOGGER.error('too large: length %s for URL %s', len(response.data), url)
             if good:
                 html = trafilatura.utils.decode_response(response) 
                 final_link = response.url
@@ -133,12 +137,12 @@ def fetch_pages(link_batch):
 
 def fetch_robot(base_url):
     rules = urllib.robotparser.RobotFileParser()
-    rules.set_url(base_url + '/robots.txt')
+    rules.set_url("https://" + base_url + '/robots.txt')
     # exceptions happening here
     try:
         rules.read()
     except Exception as exc:
-        #LOGGER.error('cannot read robots.txt: %s', exc)
+        LOGGER.error('cannot read robots.txt: %s', exc)
         rules = None
     return rules
 
@@ -166,25 +170,31 @@ def index_pages(db,domain,extracted_pages):
     links = []
     for original_link,final_link,html,doc in extracted_pages:
         state = "good"
+        link = original_link
+        if original_link != final_link:
+            linkcol.insert_one(get_link_doc(original_link,"redirect"))
+            link = final_link
         if html is None:
             state = "html_error"
         elif doc is None:
             state = "content_error"
-        if original_link != final_link:
-            linkcol.insert_one(get_link_doc(final_link,state))
-            state = "redirect"
-        linkcol.update_one({"url":original_link},{"$set":{"status":state}})
         if doc is not None:
-            if html is not None:
-                htmlcol.insert_one({"url":final_link,"html":html,"html_size":len(html),"created_at":datetime.utcnow()})
             checksums,sizes = calculate_checksums(doc["text"])
-            doc["created_at"] = datetime.utcnow()
             doc["text_size"] = len(doc["text"])
             doc["paragraph_checksums"] = checksums
             doc["paragraph_sizes"] = sizes
+            if len(checksums) < 1:
+                state = "trash"
+        if state == "good":
+            htdoc = get_link_doc(link,state)
+            htdoc["html"] = html
+            htdoc["html_size"] = len(html)
+            htmlcol.insert_one(htdoc)
+            doc.update(get_link_doc(link,"good"))
             # todo extract links
             print(doc)
             contentcol.insert_one(doc)
+        linkcol.update_one({"url":original_link},{"$set":{"status":state}})
 
 
 def extract_links(link_batch,responses,domain,rules,default_status="frontlink"):
@@ -244,15 +254,32 @@ def process_links(db,domain,status,links=[],rules=None,batch_size=BATCHSIZE):
 def link_summary(db,domain):
     linkcol = db["links"]
     #res = linkcol.distinct("domain",{"hostname":domain})
-
+    
+    # count links
     res = linkcol.aggregate([
         {"$match":{"host":domain}},
         {"$group":{"_id":"$status","count":{"$sum":1}}},
     ])
     for item in res:
         print(item)
+    contentcol = db["content"]
+    res = contentcol.aggregate([
+        {"$match":{"hostname":domain}},
+        {"$group":{"_id":None,"text_size_sum":{"$sum":"text_size"}}},
+    ])
+    for item in res:
+        print(item)
 
-def create_indices(db):
+global DB
+
+@click.group()
+def cli():
+    pass
+
+@cli.command()
+def dropdb():
+    myclient = pymongo.MongoClient(CONNECTION)
+    db=myclient[DBNAME]
     linkcol = db["links"]
     linkcol.create_index({"url":1},{"name":"url"})
     linkcol.create_index({"host":1,"status":1},{"name":"hostname_status"})
@@ -263,17 +290,14 @@ def create_indices(db):
     htmlcol = db["html"]
     htmlcol.create_index({"url":1})
 
-@click.group()
-def cli():
-    pass
-
-@click.command()
+@cli.command()
 @click.argument("start_link")
-def simple_visit(start_link):
+def visit(start_link):
+    myclient = pymongo.MongoClient(CONNECTION)
+    db=myclient[DBNAME]
     start_link,domain = courlan.check_url(start_link)
-    myclient = pymongo.MongoClient("mongodb://root:example@localhost:27017/")
-    db=myclient["crawler"]
     rules = fetch_robot(domain)
+    print(rules)
     batch_size = BATCHSIZE
     navigation_links = get_links(db,domain,"navigation",batch_size)
     if start_link is not None: