Improve gfonts spider start working on start_urls from search result spider

ger_gfonts/ger_gfonts/spiders/gfonts_spider.py

@@ -1,14 +1,17 @@
 import scrapy
+from scrapy.spiders import Spider,CrawlSpider, Rule
 import socket
 import re
 import json
 from urllib.request import urlopen
 from urllib.parse import urlparse
+from utility.countries import *
 
-class GFontsSpider(scrapy.Spider):
+class GFontsSpider(Spider):
     name = "gfonts"
     #start_urls = self.getUrls()
     checked_domains = []
+    eu_domains = []
     custom_settings = {'CLOSESPIDER_PAGECOUNT': 2000}
 
     def __init__(self):
@@ -18,12 +21,14 @@ class GFontsSpider(scrapy.Spider):
         # check current url - german or eu (todo)
         # check if api.gfonts.com is found
         # @todo: check if google analytics is found
-        parsed = urlparse(response.url)
-        self.checked_domains.append(parsed.hostname)
-        self.logNewDomain(response.url)
-        if self.isGerman(response.url):
-            self.findGFonts(response)
-            self.parseOn(response)
+        if self.isEuropean(response.url):
+            self.writeTrackers(response)
+            parsed = urlparse(response.url)
+            self.eu_domains.append(parsed.hostname)
+            self.logNewDomain(response.url)
+        else:
+            print("NOT EUROPEAN: " + response.url)
+        self.parseOn(response)
 
     def getCountryOfUrl(self, url):
         ip = socket.gethostbyname(url)
@@ -44,24 +49,60 @@ class GFontsSpider(scrapy.Spider):
         if not self.isGermanTLD(url):
             return self.isCountryGerman(url)
         return True
+    def isEuropean(self, url):
+        eu_tlds = self.getEuTlds()
+        parts = urlparse(url)
+        tld = parts.hostname[-3:]
+        if tld in eu_tlds:
+            return eu_tlds[tld]
+        country = self.getCountryOfUrl(url)
+        if country in eu_tlds.values():
+            return country
+        return False
 
     def findGFonts(self, response):
         for links in response.css('head link'):
-            if 'fonts.googleapis.com' in links.attrib['href']:
-                yield {
-                    'url': response.url,
-                    'gfonts': True,
+            return 'fonts.googleapis.com' in links.attrib['href']
+    def findGTrackers(self, response):
+        trackers = { 'ga' : 'www.google-analytics.com',
+                'gt' : 'www.googletagmanager.com'}
+        result = {'ga':0, 'gt':0}
+        for script in response.css('script::text').getall():
+            if script.find(trackers['ga']) > 0:
+                result['ga'] = 1
+            if script.find(trackers['gt']) > 0:
+                result['gt'] = 1
+        return result
+    def findMetaPixel(self, response):
+        for img in response.css('img'):
+            if img.attrib['src'].find('www.facebook.com/tr?id='):
+                return TRUE
+        return FALSE
+    def writeTrackers(self,response):
+        gtrackers = self.findGTrackers(response)
+        yield {
+            'domain': urlparse(response.url).netloc,
+            'country': self.isEuropean(response.url),
+            'gf': self.findGFonts(response),
+            'ga': gtrackers['ga'],
+            'gt': gtrackers['gm'],
+            'mp': self.findMetaPixel(response)
                 }
 
+
     def parseOn(self, response):
-        for links in response.css('a'):
-            url = links.attrib['href']
+        links = response.css('a');
+        print('FOUND: ' + str(len(links)) + ' LINKS')
+        for link in links:
+            url = link.attrib['href']
             # parse valid urls
-            if validators.url(url) and bool(urlparse(url).netloc):
+            found = urlparse(url)
+            if validators.url(url) and bool(found.netloc):
                 current = urlparse(response.url)
-                found = urlparse(url)
-                if current.hostname != found.hostname and found.hostname not in self.checked_domains:
+                if current.hostname != found.hostname:
                     yield response.follow(url, callback=self.parse)
+                else:
+                    print("NOT FOLLOWING: " + url)
 
     def getUrls(self):
         with open('sites.txt') as sites_file:
@@ -72,3 +113,11 @@ class GFontsSpider(scrapy.Spider):
         print('############################################')
         print('######   ' + url + '   #######')
         print('############################################')
+
+    def getEuTlds(self):
+        return { '.ad': 'AD', '.at': 'AT', '.be':'BE', '.ch':'CH', '.cz':'CZ', 
+        '.de':'DE', '.dk':'DK', '.ee':'EE', '.es':'ES', '.eu':'EU', '.fi':'FI',
+        '.fr':'FR', '.gr':'GR', '.hr':'HR', '.hu':'HU', '.ie':'IE', '.it':'IT',
+        '.li':'LI', '.lt':'LT', '.lu':'LU', '.lv':'LV', '.nl':'NL', '.no':'NO',
+        '.pl':'PL', '.pt':'PT', '.ro':'RO', '.se':'SE', '.si':'SI', '.sk':'SK',
+        '.bg':'BG', '.cy':'CY', '.mt':'MT'}

ger_gfonts/ger_gfonts/spiders/sites.txt

@@ -75,7 +75,7 @@ kik.de
 wetterzentrale.de
 service.bund.de
 katholisch.de
-homboldt-foundation.de
+humboldt-foundation.de
 deginvest.de
 comdirect.de
 standaard.be

ger_gfonts/ger_gfonts/spiders/start_urls_spider.py

@@ -0,0 +1,17 @@
+import scrapy
+#import pandas
+from scrapy.linkextractors import LinkExtractor
+#from utility.countries import getEuTlds
+from utility import countries
+
+class firstSpider(scrapy.Spider):
+    name = "start_urls"
+
+    def __init__(self):
+        eu_tlds = countries.getEuTlds()
+        self.start_urls = map(lambda t: 'https://www.google.com/search?q=inurl%3A' + t, eu_tlds.keys())
+
+    def parse(self, response):
+        xlink = LinkExtractor()
+        for link in xlink.extract_links(response):
+            print(link)

ger_gfonts/utility/countries.py

@@ -0,0 +1,47 @@
+# utility functions for countries
+# import pycountry
+import socket
+from urllib.request import urlopen
+import json
+from urllib.parse import urlparse
+
+def getEuTlds():
+    # map tld to alpha_2
+    return { '.ad': 'AD', '.at': 'AT', '.be':'BE', '.ch':'CH', '.cz':'CZ', 
+        '.de':'DE', '.dk':'DK', '.ee':'EE', '.es':'ES', '.eu':'EU', '.fi':'FI',
+        '.fr':'FR', '.gr':'GR', '.hr':'HR', '.hu':'HU', '.ie':'IE', '.it':'IT',
+        '.li':'LI', '.lt':'LT', '.lu':'LU', '.lv':'LV', '.nl':'NL', '.no':'NO',
+        '.pl':'PL', '.pt':'PT', '.ro':'RO', '.se':'SE', '.si':'SI', '.sk':'SK',
+        '.bg':'BG', '.cy':'CY', '.mt':'MT'}
+
+def getCountryOfUrl(url):
+    ip = socket.gethostbyname(url)
+    api_url = 'https://ipinfo.io/' + ip + '/json'
+    response = urlopen(api_url)
+    data = json.load(response)
+    return data['country']
+
+def isCountryGerman(self, url):
+    return 'DE' == getCountryOfUrl(url)
+
+def isGermanTLD(url):
+    parts = urlparse(url)
+    tld = parts.hostname[-3:]
+    return tld == '.de'
+
+def isGerman(self, url):
+    if not isGermanTLD(url):
+        return isCountryGerman(url)
+    return True
+
+def isEuropean(url):
+    eu_tlds = getEuTlds()
+    parts = urlparse(url)
+    tld = parts.hostname[-3:]
+    if tld in eu_tlds:
+        return eu_tlds[tld]
+    country = getCountryOfUrl(url)
+    if country in eu_tlds.values():
+        return country
+    return False
+