webtoolkit 0.0.185__tar.gz → 0.0.187__tar.gz

{webtoolkit-0.0.185 → webtoolkit-0.0.187}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.3
 Name: webtoolkit
-Version: 0.0.185
+Version: 0.0.187
 Summary: Web tools and interfaces for Internet data processing.
 License: GPL3
 Author: Iwan Grozny
@@ -12,6 +12,7 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Requires-Dist: beautifulsoup4 (>=4.13.5,<5.0.0)
 Requires-Dist: brutefeedparser (>=0.10.5,<0.11.0)
 Requires-Dist: lxml (>=5.4.0,<6.0.0)

{webtoolkit-0.0.185 → webtoolkit-0.0.187}/pyproject.toml

@@ -3,7 +3,7 @@
 
 [tool.poetry]
 name = "webtoolkit"
-version = "0.0.185"
+version = "0.0.187"
 description = "Web tools and interfaces for Internet data processing."
 authors = ["Iwan Grozny <renegat@renegat0x0.ddns.net>"]
 license = "GPL3"

{webtoolkit-0.0.185 → webtoolkit-0.0.187}/webtoolkit/domaincache.py

@@ -176,6 +176,9 @@ class DomainCache(object):
     respect_robots_txt = True
 
     def get_object(url, url_builder):
+        """
+        API
+        """
 
         if DomainCache.object is None:
             DomainCache.object = DomainCache(
@@ -193,7 +196,7 @@ class DomainCache(object):
         respect_robots_txt=True,
     ):
         """
-        @note Not public
+        @note Not API
         """
         self.cache_size = cache_size
         self.cache = {}
@@ -209,6 +212,18 @@ class DomainCache(object):
 
         return self.cache[input_url]["domain"]
 
+    def get_length(self):
+        """
+        Returns length of cache
+        """
+        return len(self.cache)
+
+    def get_max_length(self):
+        """
+        Returns length of cache
+        """
+        return self.cache_size)
+
     def read_info(self, domain_url):
         return DomainCacheInfo(
             domain_url,
@@ -216,7 +231,7 @@ class DomainCache(object):
         )
 
     def remove_from_cache(self):
-        if len(self.cache) < self.cache_size:
+        if self.get_length() < self.get_max_length():
             return
 
         thelist = []

{webtoolkit-0.0.185 → webtoolkit-0.0.187}/webtoolkit/urllocation.py

@@ -376,6 +376,8 @@ class UrlLocation(object):
             parsed_url = urlparse(url)
             query_params = parse_qs(parsed_url.query)
             param_value = query_params.get("q", [None])[0]
+            if not param_value:
+                return url
 
             param_value = unquote(param_value)
             param_value = UrlLocation.get_cleaned_link(param_value)