Use proper HTML parsing for page title retrieval.

[plomlombot-irc.git] / plomlombot.py

diff --git a/plomlombot.py b/plomlombot.py
index 81be31fdb5f438349469e7d63afac9b4b20b1a7e..f34d97a821378c5e092cf064a78a2137d82350c7 100755 (executable)
--- a/plomlombot.py
+++ b/plomlombot.py
@@ -9,6 +9,7 @@ import re
 import urllib.request
 import http.client
 import html
+import html.parser
 
 # Defaults, may be overwritten by command line arguments.
 SERVER = "irc.freenode.net"
@@ -18,6 +19,22 @@ USERNAME = "plomlombot"
 NICKNAME = USERNAME
 
 
+class HTMLParser(html.parser.HTMLParser):
+    def __init__(self, html, tag):
+        super().__init__()
+        self._tag = ""
+        self.data = ""
+        self.feed(html)
+    def handle_starttag(self, tag, attrs):
+        if self.data == "":
+            self._tag = tag
+    def handle_endtag(self, tag):
+        self._tag = ""
+    def handle_data(self, data):
+        if self._tag != "":
+            self.data = data
+
+
 class ExceptionForRestart(Exception):
     pass
 
@@ -67,7 +84,11 @@ class IO:
                 self._pingtest(send_ping)
                 return None
             self.last_pong = time.time()
-            received_runes = self.socket.recv(1024).decode("UTF-8")
+            received_bytes = self.socket.recv(1024)
+            try:
+                received_runes = received_bytes.decode("UTF-8")
+            except UnicodeDecodeError:
+                received_runes = received_bytes.decode("latin1")
             if len(received_runes) == 0:
                 print("SOCKET CONNECTION BROKEN")
                 raise ExceptionForRestart
@@ -118,7 +139,8 @@ def lineparser_loop(io, nickname):
                     continue
                 charset = webpage.info().get_content_charset()
                 if not charset:
-                    charset = "utf-8"
+                    notice("TROUBLE READING PAGE TITLE: no charset in header")
+                    continue
                 content_type = webpage.info().get_content_type()
                 if content_type not in ('text/html', 'text/xml',
                                         'application/xhtml+xml'):
@@ -126,7 +148,7 @@ def lineparser_loop(io, nickname):
                            + content_type)
                     continue
                 content = webpage.read().decode(charset)
-                title = str(content).split('<title>')[1].split('</title>')[0]
+                title = HTMLParser(content, "title").data
                 title = html.unescape(title)
                 notice("PAGE TITLE FOR URL: " + title)