bandaid unicode/str.encode-related crash bug

[z_archive/twitter.git] / twitter / util.py

diff --git a/twitter/util.py b/twitter/util.py
index db497f9d3c569d87a6c04eff63674aa52613d0e5..5a5f50710f4d8ed8d7d5a0c65f4bc97671f5f072 100644 (file)
--- a/twitter/util.py
+++ b/twitter/util.py
@@ -14,6 +14,8 @@ import textwrap
 import time
 import socket
 
+PY_3_OR_HIGHER = sys.version_info >= (3, 0)
+
 try:
     from html.entities import name2codepoint
     unichr = chr
@@ -49,14 +51,32 @@ def printNicely(string):
     if hasattr(sys.stdout, 'buffer'):
         sys.stdout.buffer.write(string.encode('utf8'))
         print()
+        sys.stdout.buffer.flush()
+        sys.stdout.flush()
     else:
         print(string.encode('utf8'))
 
-__all__ = ["htmlentitydecode", "smrt_input"]
+def actually_bytes(stringy):
+    if PY_3_OR_HIGHER:
+        if type(stringy) == bytes:
+            pass
+        elif type(stringy) != str:
+            stringy = str(stringy)
+        if type(stringy) == str:
+            stringy = stringy.encode("utf-8")
+    else:
+        if type(stringy) == str:
+            pass
+        elif type(stringy) != unicode:
+            stringy = str(stringy)
+        if type(stringy) == unicode:
+            stringy = stringy.encode("utf-8")
+    return stringy
 
 def err(msg=""):
     print(msg, file=sys.stderr)
 
+
 class Fail(object):
     """A class to count fails during a repetitive task.
 
@@ -120,15 +140,22 @@ def follow_redirects(link, sites= None):
     try:
         with contextlib.closing(opener.open(req,timeout=1)) as site:
             return site.url
-    except (urllib2.HTTPError, urllib2.URLError, socket.timeout):
+    except:
         return redirect_handler.last_url if redirect_handler.last_url else link
 
 def expand_line(line, sites):
     """Expand the links in the line for the given sites."""
-    l = line.strip()
-    msg_format, links = find_links(l)
-    args = tuple(follow_redirects(l, sites) for l in links)
-    return msg_format % args
+    try:
+        l = line.strip()
+        msg_format, links = find_links(l)
+        args = tuple(follow_redirects(l, sites) for l in links)
+        line = msg_format % args
+    except Exception as e:
+        try:
+            err("expanding line %s failed due to %s" % (line, unicode(e)))
+        except:
+            pass
+    return line
 
 def parse_host_list(list_of_hosts):
     """Parse the comma separated list of hosts."""
@@ -145,3 +172,6 @@ def align_text(text, left_margin=17, max_width=160):
         lines.append('\n'.join(temp_lines))
     ret = '\n'.join(lines)
     return ret.lstrip()
+
+
+__all__ = ["htmlentitydecode", "smrt_input"]