Fix buggy HTMLParser.

[plomlombot-irc.git] / plomlombot.py

diff --git a/plomlombot.py b/plomlombot.py
index c1c23089ba01450ef5f2cd686f000580e44398cb..9e9d0a44ce21bc1d07204dcb9780c54ac1b31d0e 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 +19,23 @@ USERNAME = "plomlombot"
 NICKNAME = USERNAME
 
 
+class HTMLParser(html.parser.HTMLParser):
+    def __init__(self, html, tag):
+        super().__init__()
+        self._tag_to_check = tag
+        self._tag = ""
+        self.data = ""
+        self.feed(html)
+    def handle_starttag(self, tag, attrs):
+        if self.data == "" and tag == self._tag_to_check:
+            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 +85,11 @@ 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
@@ -100,29 +122,36 @@ def lineparser_loop(io, nickname):
     def act_on_privmsg(tokens):
 
         def url_check(msg):
-            matches = re.findall("(https?://[^\s]+)", msg)
+
+            def notice(msg):
+                io.send_line("NOTICE " + target + " :" + msg)
+
+            matches = re.findall("(https?://[^\s>]+)", msg)
             for i in range(len(matches)):
                 url = matches[i]
+                request = urllib.request.Request(url, headers={
+                    "User-Agent": "plomlombot"
+                })
                 try:
-                    webpage = urllib.request.urlopen(url, timeout=15)
+                    webpage = urllib.request.urlopen(request, timeout=15)
                 except (urllib.error.HTTPError, urllib.error.URLError,
                         UnicodeError, http.client.BadStatusLine) as error:
-                    print("TROUBLE FOLLOWING URL: " + str(error))
+                    notice("TROUBLE FOLLOWING URL: " + str(error))
                     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'):
-                    print("TROUBLE INTERPRETING URL: bad content type "
-                          + content_type)
+                    notice("TROUBLE READING PAGE TITLE: bad content type "
+                           + 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)
-                io.send_line("PRIVMSG " + target + " :page title for url: "
-                             + title)
+                notice("PAGE TITLE FOR URL: " + title)
 
         sender = ""
         for rune in tokens[0]: