From c45bbc61575fe33f21152b93e490cd77e5d07078 Mon Sep 17 00:00:00 2001
From: Jon Michael Aanes <jon.michael.aanes@secata.com>
Date: Tue, 29 Aug 2023 09:20:14 +0200
Subject: [PATCH] Lots of small fixes

---
 wikidata_ext.py | 28 +++++++++++++++++-----------
 1 file changed, 17 insertions(+), 11 deletions(-)

diff --git a/wikidata_ext.py b/wikidata_ext.py
index 9e80502..1d9f4aa 100644
--- a/wikidata_ext.py
+++ b/wikidata_ext.py
@@ -6,6 +6,8 @@ import requests
 import json
 import logging
 
+REQUEST_SESSION = requests.Session()
+
 def concept_uri(obj):
     assert isinstance(obj, wikidata.entity.Entity), obj
     if obj.id.startswith('P'):
@@ -24,19 +26,19 @@ def fmt_triple_value(obj, prefer_obj = False):
         return obj.geturl() if prefer_obj else fmt_triple_value(obj.geturl())
     elif isinstance(obj, wikidata.entity.Entity):
         uri = concept_uri(obj)
-        return fmt_triple_value(uri, prefer_obj)
+        return fmt_triple_value(uri, True)
     else:
         assert False, type(obj)
 
 @ratelimit.sleep_and_retry
 @ratelimit.limits(calls=10, period=60)
 def fetch_by_url(url, headers):
-    logging.debug('Fetching: %s', url)
-    result = requests.get(url, headers = headers)
-    if result.status_code != 200:
-        logging.error('Got %s error message: %s', result.status_code, repr((subject, predicate, object)))
+    logging.warning('Fetching: %s', url)
+    response = REQUEST_SESSION.get(url, headers = headers)
+    if response.status_code != 200:
+        logging.error('Got %s error message: %s', response.status_code, url)
         return None
-    return request
+    return response
 
 ITEMS_PER_PAGE = "http://www.w3.org/ns/hydra/core#itemsPerPage"
 TOTAL_ITEMS = "http://www.w3.org/ns/hydra/core#totalItems"
@@ -58,8 +60,8 @@ def get_triples_count(subject = None, predicate = None, object = None):
     '''
     params = fmt_params(subject, predicate, object)
     url = requests.Request(url = 'https://query.wikidata.org/bigdata/ldf', params = params).prepare().url
-    result = fetch_by_url(url, headers = {'accept': 'application/ld+json'})
-    json_data = json.loads(result)
+    response = fetch_by_url(url, headers = {'accept': 'application/ld+json'})
+    json_data = response.json()
     for item in json_data['@graph']:
         if TOTAL_ITEMS in item:
             return {
@@ -75,8 +77,8 @@ def get_triples_internal(subject, predicate, object):
     for current_page in range(1, pagination_data['num_pages']+1):
         params['page'] = current_page
         url = requests.Request(url = 'https://query.wikidata.org/bigdata/ldf', params = params).prepare().url
-        result = fetch_by_url(url, headers = {'accept': 'application/ld+json'})
-        json_data = json.loads(result)
+        response = fetch_by_url(url, headers = {'accept': 'application/ld+json'})
+        json_data = response.json()
 
         for item in json_data['@graph']:
             if item['@id'].startswith('_:b'):
@@ -88,7 +90,7 @@ def get_triples_internal(subject, predicate, object):
             yield item
 
         # Bookkeeping
-        del url, result, json_data
+        del url, response, json_data
 
 SCHEMA_ABOUT = urllib.parse.urlparse('http://schema.org/about')
 
@@ -112,3 +114,7 @@ def get_triples(client, subject = None, predicate = None, object = None):
             yield (s, predicate, o)
         del item, is_looking_for
 
+def get_backlinks(client, predicate, object):
+    for subject, _, _ in get_triples(client, subject = None, predicate = predicate, object = object):
+        yield subject
+