(original) (raw)

changeset: 75641:c06b94c5c609 parent: 75628:7cbc48324938 user: Andrew Svetlov andrew.svetlov@gmail.com date: Wed Mar 14 13:22:12 2012 -0700 files: Lib/idlelib/PyShell.py Lib/idlelib/rpc.py Lib/idlelib/run.py Misc/NEWS description: 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 -r 7cbc48324938 -r c06b94c5c609 Lib/idlelib/PyShell.py --- a/Lib/idlelib/PyShell.py Wed Mar 14 09:39:36 2012 -0700 +++ b/Lib/idlelib/PyShell.py Wed Mar 14 13:22:12 2012 -0700 @@ -1221,6 +1221,16 @@ 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 -r 7cbc48324938 -r c06b94c5c609 Lib/idlelib/rpc.py --- a/Lib/idlelib/rpc.py Wed Mar 14 09:39:36 2012 -0700 +++ b/Lib/idlelib/rpc.py Wed Mar 14 13:22:12 2012 -0700 @@ -196,8 +196,12 @@ 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 @@ 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 -r 7cbc48324938 -r c06b94c5c609 Lib/idlelib/run.py --- a/Lib/idlelib/run.py Wed Mar 14 09:39:36 2012 -0700 +++ b/Lib/idlelib/run.py Wed Mar 14 13:22:12 2012 -0700 @@ -6,6 +6,7 @@ import _thread as thread import threading import queue +import builtins from idlelib import CallTips from idlelib import AutoComplete @@ -261,6 +262,25 @@ 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 @@ 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 -r 7cbc48324938 -r c06b94c5c609 Misc/NEWS --- a/Misc/NEWS Wed Mar 14 09:39:36 2012 -0700 +++ b/Misc/NEWS Wed Mar 14 13:22:12 2012 -0700 @@ -24,6 +24,8 @@ 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. /andrew.svetlov@gmail.com