[proxy] Add support for script blocking/allowing rules

1 files changed, 24 insertions, 8 deletions

diff --git a/src/hydrilla/url_patterns.py b/src/hydrilla/url_patterns.py
index 278827a..1b5fa10 100644
--- a/src/hydrilla/url_patterns.py
+++ b/src/hydrilla/url_patterns.py
@@ -57,7 +57,7 @@ class ParsedUrl:
     path_segments:      tuple[str, ...] = dc.field(hash=False, compare=False)
     query:              str             = dc.field(hash=False, compare=False)
     has_trailing_slash: bool            = dc.field(hash=False, compare=False)
-    port:               int             = dc.field(hash=False, compare=False)
+    port:               t.Optional[int] = dc.field(hash=False, compare=False)
 
     @property
     def url_without_path(self) -> str:
@@ -67,12 +67,12 @@ class ParsedUrl:
         netloc = '.'.join(reversed(self.domain_labels))
 
         if self.port is not None and \
-           default_ports[scheme] != self.port:
+           default_ports.get(scheme) != self.port:
             netloc += f':{self.port}'
 
         return f'{scheme}://{netloc}'
 
-    def _reconstruct_url(self) -> str:
+    def reconstruct_url(self) -> str:
         """...."""
         path = '/'.join(('', *self.path_segments))
         if self.has_trailing_slash:
@@ -82,7 +82,7 @@ class ParsedUrl:
 
     def path_append(self: ParsedUrlType, *new_segments: str) -> ParsedUrlType:
         """...."""
-        new_url = self._reconstruct_url()
+        new_url = self.reconstruct_url()
         if not self.has_trailing_slash:
             new_url += '/'
 
@@ -114,8 +114,11 @@ ParsedPattern = t.NewType('ParsedPattern', ParsedUrl)
 # actually supported by Hydrilla server and Haketilo proxy.
 supported_schemes = 'http', 'https', 'ftp', 'file'
 
-def _parse_pattern_or_url(url: str, orig_url: str, is_pattern: bool = False) \
-    -> ParsedUrl:
+def _parse_pattern_or_url(
+        url:               str,
+        orig_url:          str,
+        is_pattern:        bool = False
+) -> ParsedUrl:
     """...."""
     if not is_pattern:
         assert orig_url == url
@@ -164,7 +167,7 @@ def _parse_pattern_or_url(url: str, orig_url: str, is_pattern: bool = False) \
         else:
             raise HaketiloException(_('err.url_{}.bad_port').format(url))
 
-    port = t.cast(int, explicit_port or default_ports.get(parse_result.scheme))
+    port = explicit_port or default_ports.get(parse_result.scheme)
 
     # Make URL's hostname into a list of labels in reverse order. E.g.
     #     'https://a.bc..de.fg.com/h/i/' -> ['com', 'fg', 'de', 'bc', 'a']
@@ -215,8 +218,21 @@ def parse_pattern(url_pattern: str) -> t.Iterator[ParsedPattern]:
         patterns = [url_pattern]
 
     for pat in patterns:
-        yield ParsedPattern(_parse_pattern_or_url(pat, url_pattern, True))
+        yield ParsedPattern(
+            _parse_pattern_or_url(pat, url_pattern, True)
+        )
 
 def parse_url(url: str) -> ParsedUrl:
     """...."""
     return _parse_pattern_or_url(url, url)
+
+
+def normalize_pattern(url_pattern: str) -> str:
+    parsed = next(parse_pattern(url_pattern))
+
+    reconstructed = parsed.reconstruct_url()
+
+    if url_pattern.startswith('http*'):
+        reconstructed = replace_scheme_regex.sub('http*', reconstructed)
+
+    return reconstructed