Issue #14200: Idle shell crash on printing non-BMP unicode character.

UnicodeEncodeError is raised for strings contains non-BMP characters.
For eval results unicode escaping is used, print() calls display
exception with traceback as usual.

diff --git a/Lib/idlelib/PyShell.py b/Lib/idlelib/PyShell.py
index f884b28d673600f54542e19ea8cc074d5b9be41b..6b75a8d4d0c414088df5a087425d867fed975368 100644 (file)
--- a/Lib/idlelib/PyShell.py
+++ b/Lib/idlelib/PyShell.py
@@ -1221,6 +1221,16 @@ class PyShell(OutputWindow):
         self.set_line_and_column()
 
     def write(self, s, tags=()):
+        if isinstance(s, str) and len(s) and max(s) > '\uffff':
+            # Tk doesn't support outputting non-BMP characters
+            # Let's assume what printed string is not very long,
+            # find first non-BMP character and construct informative
+            # UnicodeEncodeError exception.
+            for start, char in enumerate(s):
+                if char > '\uffff':
+                    break
+            raise UnicodeEncodeError("UCS-2", char, start, start+1,
+                                     'Non-BMP character not supported in Tk')
         try:
             self.text.mark_gravity("iomark", "right")
             OutputWindow.write(self, s, tags, "iomark")

diff --git a/Lib/idlelib/rpc.py b/Lib/idlelib/rpc.py
index def43945aee438aea98c3130e7ec98062c26b78d..301305ee90865d40574179c36fcd3345cf23dd20 100644 (file)
--- a/Lib/idlelib/rpc.py
+++ b/Lib/idlelib/rpc.py
@@ -196,8 +196,12 @@ class SocketIO(object):
                 return ("ERROR", "Unsupported message type: %s" % how)
         except SystemExit:
             raise
+        except KeyboardInterrupt:
+            raise
         except socket.error:
             raise
+        except Exception as ex:
+            return ("CALLEXC", ex)
         except:
             msg = "*** Internal Error: rpc.py:SocketIO.localcall()\n\n"\
                   " Object: %s \n Method: %s \n Args: %s\n"
@@ -257,6 +261,9 @@ class SocketIO(object):
         if how == "ERROR":
             self.debug("decoderesponse: Internal ERROR:", what)
             raise RuntimeError(what)
+        if how == "CALLEXC":
+            self.debug("decoderesponse: Call Exception:", what)
+            raise what
         raise SystemError(how, what)
 
     def decode_interrupthook(self):

diff --git a/Lib/idlelib/run.py b/Lib/idlelib/run.py
index 962c6c0a08d0aef66dfe631348498d76ffe66249..a161a93c4dad1d49c244710d163fffd15f4949cf 100644 (file)
--- a/Lib/idlelib/run.py
+++ b/Lib/idlelib/run.py
@@ -6,6 +6,7 @@ import traceback
 import _thread as thread
 import threading
 import queue
+import builtins
 
 from idlelib import CallTips
 from idlelib import AutoComplete
@@ -261,6 +262,25 @@ class MyRPCServer(rpc.RPCServer):
             thread.interrupt_main()
 
 
+def displayhook(value):
+    """Override standard display hook to use non-locale encoding"""
+    if value is None:
+        return
+    # Set '_' to None to avoid recursion
+    builtins._ = None
+    text = repr(value)
+    try:
+        sys.stdout.write(text)
+    except UnicodeEncodeError:
+        # let's use ascii while utf8-bmp codec doesn't present
+        encoding = 'ascii'
+        bytes = text.encode(encoding, 'backslashreplace')
+        text = bytes.decode(encoding, 'strict')
+        sys.stdout.write(text)
+    sys.stdout.write("\n")
+    builtins._ = value
+
+
 class MyHandler(rpc.RPCHandler):
 
     def handle(self):
@@ -270,6 +290,7 @@ class MyHandler(rpc.RPCHandler):
         sys.stdin = self.console = self.get_remote_proxy("stdin")
         sys.stdout = self.get_remote_proxy("stdout")
         sys.stderr = self.get_remote_proxy("stderr")
+        sys.displayhook = displayhook
         # page help() text to shell.
         import pydoc # import must be done here to capture i/o binding
         pydoc.pager = pydoc.plainpager

diff --git a/Misc/NEWS b/Misc/NEWS
index 0d6cdbb13d0dee54b6423c32d6e4df0fc2034fca..2e4f15d47a0e4e87f19d33c9c0a9598b75986c8c 100644 (file)
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -24,6 +24,8 @@ Core and Builtins
 Library
 -------
 
+- Issue #14200: Idle shell crash on printing non-BMP unicode character.
+
 - Issue #14291: Email now defaults to utf-8 for non-ASCII unicode headers
   instead of raising an error.  This fixes a regression relative to 2.7.