OPN-415: updates to url_database.py, bi-weekly generation

bin/resource_management/data/.gitignore

@@ -0,0 +1 @@
+*.csv

bin/resource_management/email_message.txt

@@ -0,0 +1 @@
+Automated email of the OGP Broken_Resource_Links report.

bin/resource_management/recipients.config

@@ -0,0 +1,2 @@
+recipients=''
+venv=''

bin/resource_management/reports/.gitignore

@@ -0,0 +1 @@
+*.csv

bin/resource_management/url_database.py

@@ -5,8 +5,9 @@
 Outputs status to 'url_database.csv'
 
 Arguments:
-fileinput - metadata file to be read ('od-do-canada.jsonl')
-batch_size - maximum number of URL's to test in parallel
+fileinput - metadata file to be read ('od-do-canada.jsonl') or retest 'N/A' responses from previous url_database.csv
+batch_size - INT maximum number of URL's to test in parallel
+filename - name of file to export
 """
 import sys
 import grequests
@@ -20,19 +21,21 @@
 
 requests.packages.urllib3.disable_warnings(InsecureRequestWarning)
 
-file = sys.argv[1]
-batch_size = int(sys.argv[2])
+# use a generic user agent to detect services that render information pages instead of the actual data when a web browser user visits
+GENERIC_WEB_CLIENT_AGENT = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.77 Safari/537.36"}
 
-print(file, batch_size)
+def create_database(filename):
+    with open(filename, "w") as f:
+        writer = csv.writer(f)
+        writer.writerow(("url", "date", "response",
+                         "content-type", "content-length"))
+    f.close()
 
-prev_i = 1
-urls = set()
-batch_urls = []
-url_match = []
-responses = []
-date = []
-ftp_urls = []
-headers = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.77 Safari/537.36"}
+def write_to_database(filename, rows):
+    with open(filename, "a") as f:
+        writer = csv.writer(f)
+        writer.writerows(rows)
+    f.close()
 
 def check_for_connection():
     url = "https://www.google.ca"
@@ -45,110 +48,191 @@ def check_for_connection():
         except (requests.ConnectionError, requests.Timeout) as exception:
             print("No internet connection.")
 
-print("Starting...")
-print("Reading and testing URL's")
-
-for i, dataset in enumerate(open(file), 1):
-    line = json.loads(dataset)
-    resources = line["resources"]
-    for l in range(len(resources)):
-        url = resources[l]["url"].encode('utf-8')
-        if url in urls:
+def get_batch_response(batch_urls, i, prev_i):
+    sys.stderr.write("\r")
+    sys.stderr.write("Testing Datasets {0} - {1}".format(prev_i, i))
+    rs = (grequests.head(u, timeout=60, headers=GENERIC_WEB_CLIENT_AGENT, verify=False, allow_redirects=True, stream=False) for u in
+          batch_urls)
+    batch_response = grequests.map(rs)
+    for j, r in enumerate(batch_response):
+        if not r is None:
+            r.close()
+        else:
+            batch_response[j] = "N/A"
+    return batch_response
+
+def get_batch_response_ftp(batch_urls):
+    ftp_responses = []
+    ftp_dates = []
+    requests_ftp.monkeypatch_session()
+    for i, url in enumerate(batch_urls):
+        sys.stderr.write("\r")
+        sys.stderr.write("Testing FTP {0} of {1}".format(i, len(batch_urls)))
+        s = requests.Session()
+        now = datetime.now()
+        dt_string = now.strftime("%Y-%m-%d %H:%M:%S")
+        ftp_dates.append(dt_string.encode('utf-8'))
+        try:
+            resp = s.head(url, timeout=120, headers=GENERIC_WEB_CLIENT_AGENT, verify=False, allow_redirects=True, stream=False)
+
+            if not resp is None:
+                s.close()
+            else:
+                resp = "N/A"
+            ftp_responses.append(resp)
+        except requests.exceptions.RequestException as e:
+            ftp_responses.append("N/A")
             continue
-        elif 'ftp://' in url:
-            ftp_urls.append(url)
+        except UnicodeEncodeError as e:
+            ftp_responses.append("N/A")
+            continue
+    return ftp_responses, ftp_dates
+
+def get_batch_content(batch_response):
+    content_lengths = []
+    content_types = []
+    for r in batch_response:
+        if "N/A" in r:
+            content_lengths.append('N/A')
+            content_types.append('N/A')
         else:
-            urls.add(url)
-            now = datetime.now()
-            dt_string = now.strftime("%Y-%m-%d %H:%M:%S")
-            date.append(dt_string.encode('utf-8'))
-            batch_urls.append(url)
-
+            cl = r.headers.get("Content-Length")
+            ct = r.headers.get("Content-Type")
+            if cl is None:
+                cl = 'N/A'
+            if ct is None:
+                ct = 'N/A'
+            content_lengths.append(cl.encode('utf-8'))
+            content_types.append(ct.encode('utf-8'))
+    return content_lengths, content_types
+
+def main(import_file, batch_size, filename):
+    # set local vars
+    prev_i = 1
+    all_urls = set()
+    ftp_urls = []
+    # batch lists
+    batch_urls = []
+    batch_dates = []
+    # create url_database file
+    create_database(filename)
+    # Open JSONL and retrieve urls
+    for i, dataset in enumerate(open(import_file), 1):
+        line = json.loads(dataset, "utf-8")
+        resources = line["resources"]
+        for l in range(len(resources)):
+            # append urls to temp list
+            url = resources[l]["url"]
+            url = str(url.encode('utf-8'))
+            if url in all_urls:
+                continue
+            elif 'ftp://' in url:
+                all_urls.add(url)
+                ftp_urls.append(url)
+            else:
+                all_urls.add(url)
+                batch_urls.append(url)
+                now = datetime.now()
+                dt_string = now.strftime("%Y-%m-%d %H:%M:%S")
+                batch_dates.append(dt_string.encode('utf-8'))
+            # when len(list) == batch_size
             if len(batch_urls) == batch_size:
-                url_match.append(batch_urls)
-                sys.stderr.write("\r")
-                sys.stderr.write("Testing Datasets {0} - {1}"
-                                 .format(prev_i, i))
+                check_for_connection()
+                # get batch_responses as list
+                batch_response = get_batch_response(batch_urls, i, prev_i)
                 prev_i = i
-                rs = (grequests.head(u, timeout=120, headers=headers, verify=False, allow_redirects=True, stream=False) for u in batch_urls)
-                batch_response = grequests.map(rs)
-                responses.append(batch_response)
-                for r in batch_response:
-                    if not r is None:
-                        r.close()
+                # get content-length and content-type from responses list
+                batch_cl , batch_ct = get_batch_content(batch_response)
+                # zip urls, dates, responses, content-type, content-length
+                rows = zip(batch_urls, batch_dates, batch_response, batch_ct, batch_cl)
+                # write rows to csv
+                write_to_database(filename, rows)
+                # clear lists and repeat
                 batch_urls = []
-                check_for_connection()
-
-# Check last urls not covered in loop
-url_match.append(batch_urls)
-sys.stderr.write("\r")
-sys.stderr.write("Testing Datasets {0} - {1}".format(prev_i, i))
-rs = (grequests.head(u, timeout=120, headers=headers, verify=False, allow_redirects=True, stream=False) for u in batch_urls)
-batch_response = grequests.map(rs)
-responses.append(batch_response)
-for r in batch_response:
-    if not r is None:
-        r.close()
-
-#Testing FTP urls
-ftp_batch = []
-ftp_response = []
+                batch_dates = []
+
+    # Get response info for last URLS not included in final batch
+    check_for_connection()
+    batch_response = get_batch_response(batch_urls, i, prev_i)
+    batch_cl, batch_ct = get_batch_content(batch_response)
+    rows = zip(batch_urls, batch_dates, batch_response, batch_ct, batch_cl)
+    write_to_database(filename, rows)
+    batch_urls = []
+    batch_dates = []
+
+    # get response info for FTP links
+    check_for_connection()
+    ftp_batch_response, ftp_dates  = get_batch_response_ftp(ftp_urls)
+    batch_cl, batch_ct = get_batch_content(ftp_batch_response)
+    rows = zip(ftp_urls, ftp_dates, ftp_batch_response, batch_ct, batch_cl)
+    write_to_database(filename, rows)
+
+def retest(import_file, batch_size, filename):
+    create_database(filename)
+    with open(import_file, "r") as csvfile:
+        reader = csv.reader(csvfile)
+        next(reader)
+        prev_i = 1
+        ftp_urls = []
+        # batch lists
+        batch_urls = []
+        batch_dates = []
+        for i, row in enumerate(reader):
+            url = row[0]
+            date = row[1]
+            response = row[2]
+            content_length = row[3]
+            content_type = row[4]
+            if 'N/A' in response:
+                if 'ftp://' in url:
+                    ftp_urls.append(url)
+                else:
+                    batch_urls.append(url)
+                    now = datetime.now()
+                    dt_string = now.strftime("%Y-%m-%d %H:%M:%S")
+                    batch_dates.append(dt_string.encode('utf-8'))
+                    # when len(list) == batch_size
+                if len(batch_urls) == batch_size:
+                    check_for_connection()
+                    # get batch_responses as list
+                    batch_response = get_batch_response(batch_urls, i, prev_i)
+                    prev_i = i
+                    # get content-length and content-type from responses list
+                    batch_cl, batch_ct = get_batch_content(batch_response)
+                    # zip urls, dates, responses, content-type, content-length
+                    rows = zip(batch_urls, batch_dates, batch_response, batch_ct, batch_cl)
+                    # write rows to csv
+                    write_to_database(filename, rows)
+                    # clear lists and repeat
+                    batch_urls = []
+                    batch_dates = []
+            else:
+                write_to_database(filename,[row])
+    csvfile.close()
+    # Get response info for last URLS not included in final batch
+    check_for_connection()
+    batch_response = get_batch_response(batch_urls, i, prev_i)
+    batch_cl, batch_ct = get_batch_content(batch_response)
+    rows = zip(batch_urls, batch_dates, batch_response, batch_ct, batch_cl)
+    write_to_database(filename, rows)
+    batch_urls = []
+    batch_dates = []
+
+    # get response info for FTP links
+    check_for_connection()
+    ftp_batch_response, ftp_dates = get_batch_response_ftp(ftp_urls)
+    batch_cl, batch_ct = get_batch_content(ftp_batch_response)
+    rows = zip(ftp_urls, ftp_dates, ftp_batch_response, batch_ct, batch_cl)
+    write_to_database(filename, rows)
+
+if __name__ == '__main__':
+    import_file = sys.argv[1]
+    batch_size = int(sys.argv[2])
+    filename = sys.argv[3]
+    print(import_file, batch_size, filename)
+    if 'jsonl' in import_file or '.jl' in import_file:
+        main(import_file, batch_size, filename)
+    elif '.csv' in import_file:
+        retest(import_file, batch_size, filename)
 
-requests_ftp.monkeypatch_session()
 
-for i, url in enumerate(ftp_urls):
-    sys.stderr.write("\r")
-    sys.stderr.write("Testing FTP {0} of {1}".format(i, len(ftp_urls)))
-    s = requests.Session()
-    try:
-        resp = s.head(url,timeout=120, headers=headers, verify=False, allow_redirects=True, stream=False)
-        now = datetime.now()
-        dt_string = now.strftime("%Y-%m-%d %H:%M:%S")
-        date.append(dt_string.encode('utf-8'))
-        ftp_batch.append(url)
-        ftp_response.append(resp)
-        if not resp is None:
-            s.close()
-        if i%batch_size == 0:
-            check_for_connection()
-    except requests.exceptions.RequestException as e:
-        print str(e)
-        ftp_batch.append(url)
-        ftp_response.append(None)
-        continue
-
-responses.append(ftp_response)
-url_match.append(ftp_batch)
-
-print("Fetching content data...")
-
-responses = sum(responses, [])
-url_match = sum(url_match, [])
-content_lengths = []
-content_types = []
-for z, r in enumerate(responses):
-    if r is None:
-        content_lengths.append('N/A')
-        content_types.append('N/A')
-        responses[z] = 'N/A'
-    else:
-        cl = r.headers.get("Content-Length")
-        ct = r.headers.get("Content-Type")
-        if cl is None:
-            cl = 'N/A'
-        if ct is None:
-            ct = 'N/A'
-        content_lengths.append(cl.encode('utf-8'))
-        content_types.append(ct.encode('utf-8'))
-
-print("Exporting to csv...")
-rows = zip(url_match, date, responses, content_types, content_lengths)
-
-with open('url_database.csv', "w") as f:
-    writer = csv.writer(f)
-    writer.writerow(("url", "date", "response", "content-type",
-                     "content-length"))
-    for row in rows:
-        writer.writerow(row)
-f.close()
-print("Done.")

bin/resource_management/url_database_script.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+dir=$1
+cd $dir
+. recipients.config
+wget http://open.canada.ca/static/od-do-canada.jl.gz
+gzip -d od-do-canada.jl.gz
+source $venv
+python url_database.py od-do-canada.jl 500 data/url_database_draft.csv
+python url_database.py data/url_database_draft.csv 200 data/url_database.csv
+rm data/url_database_draft.csv
+python url_metadata_match.py od-do-canada.jl data/url_database.csv
+rm od-do-canada.jl
+the_date=$(date +"%Y-%m-%d")
+mail -s "Broken_Resource_Links-$the_date.xlsx" -a reports/Broken_Resource_Links-$the_date.xlsx $recipients < email_message.txt