Refactor adblock parsing

(cherry picked from commit 4c3c86081fedef7b5e9192f55cb29e8774c965cf)

1 files changed, 50 insertions, 35 deletions

diff --git a/qutebrowser/browser/adblock.py b/qutebrowser/browser/adblock.py
index 6e2adf923..53e3165ca 100644
--- a/qutebrowser/browser/adblock.py
+++ b/qutebrowser/browser/adblock.py
@@ -205,6 +205,54 @@ class HostBlocker:
                 download.finished.connect(
                     functools.partial(self.on_download_finished, download))
 
+    def _parse_line(self, line):
+        """Parse a line from a host file.
+
+        Args:
+            line: The bytes object to parse.
+
+        Returns:
+            True if parsing succeeded, False otherwise.
+        """
+        if line.startswith(b'#'):
+            # Ignoring comments early so we don't have to care about
+            # encoding errors in them.
+            return True
+
+        try:
+            line = line.decode('utf-8')
+        except UnicodeDecodeError:
+            log.misc.error("Failed to decode: {!r}".format(line))
+            return False
+
+        # Remove comments
+        try:
+            hash_idx = line.index('#')
+            line = line[:hash_idx]
+        except ValueError:
+            pass
+
+        line = line.strip()
+        # Skip empty lines
+        if not line:
+            return True
+
+        parts = line.split()
+        if len(parts) == 1:
+            # "one host per line" format
+            host = parts[0]
+        elif len(parts) == 2:
+            # /etc/hosts format
+            host = parts[1]
+        else:
+            log.misc.error("Failed to parse: {!r}".format(line))
+            return False
+
+        if host not in self.WHITELISTED:
+            self._blocked_hosts.add(host)
+
+        return True
+
     def _merge_file(self, byte_io):
         """Read and merge host files.
 
@@ -225,43 +273,10 @@ class HostBlocker:
             return
 
         for line in f:
-            if line.startswith(b'#'):
-                # Ignoring comments early so we don't have to care about
-                # encoding errors in them.
-                continue
-
             line_count += 1
-
-            try:
-                line = line.decode('utf-8')
-            except UnicodeDecodeError:
-                log.misc.error("Failed to decode: {!r}".format(line))
-                error_count += 1
-                continue
-
-            # Remove comments
-            try:
-                hash_idx = line.index('#')
-                line = line[:hash_idx]
-            except ValueError:
-                pass
-            line = line.strip()
-            # Skip empty lines
-            if not line:
-                continue
-            parts = line.split()
-            if len(parts) == 1:
-                # "one host per line" format
-                host = parts[0]
-            elif len(parts) == 2:
-                # /etc/hosts format
-                host = parts[1]
-            else:
+            ok = self._parse_line(line)
+            if not ok:
                 error_count += 1
-                log.misc.error("Failed to parse: {!r}".format(line))
-                continue
-            if host not in self.WHITELISTED:
-                self._blocked_hosts.add(host)
 
         log.misc.debug("{}: read {} lines".format(byte_io.name, line_count))
         if error_count > 0: