Repository navigation
%magic% doesn't work in Emacs 25.1-2 ipython (run-python) buffer on Windows 7. #10211
Description
Programmer - normally I use ipython in emacs, in particular Ubuntu, but that machine is down for the count at the moment and so I'm, reluctantly, using windows and reconstructing the environment I like. Since there's a part not working thought I might have a try at the code.
IPython 5.1.0, Emacs 25.1-2 on Windows 7.
I run %magic% and get this trace. (same operation in ipython in a cmd window works fine)
(Pdb) c
BdbQuitTraceback (most recent call last)
<ipython-input-6-d4375e0cd73a> in <module>()
----> 1 get_ipython().magic(u'magic %')
c:\python27\lib\site-packages\IPython\core\interactiveshell.pyc in magic(self, arg_s)
2156 magic_name, _, magic_arg_s = arg_s.partition(' ')
2157 magic_name = magic_name.lstrip(prefilter.ESC_MAGIC)
-> 2158 return self.run_line_magic(magic_name, magic_arg_s)
2159
2160 #-------------------------------------------------------------------------
c:\python27\lib\site-packages\IPython\core\interactiveshell.pyc in run_line_magic(self, magic_name, line)
2077 kwargs['local_ns'] = sys._getframe(stack_depth).f_locals
2078 with self.builtin_trap:
-> 2079 result = fn(*args,**kwargs)
2080 return result
2081
<decorator-gen-32> in magic(self, parameter_s)
c:\python27\lib\site-packages\IPython\core\magic.pyc in <lambda>(f, *a, **k)
186 # but it's overkill for just that one bit of state.
187 def magic_deco(arg):
--> 188 call = lambda f, *a, **k: f(*a, **k)
189
190 if callable(arg):
c:\python27\lib\site-packages\IPython\core\magics\basic.pyc in magic(self, parameter_s)
257 str(self.lsmagic()),
258 ]
--> 259 page.page('\n'.join(out))
260
261
c:\python27\lib\site-packages\IPython\core\page.py in page(data, start, screen_lines, pager_cmd)
273
274 # fallback on default pager
--> 275 return pager_page(data, start, screen_lines, pager_cmd)
276
277
c:\python27\lib\site-packages\IPython\core\page.py in pager_page(strng, start, screen_lines, pager_cmd)
186 numlines = max(num_newlines,int(len_str/80)+1)
187 pdb.set_trace()
--> 188 screen_lines_def = get_terminal_size()[1]
189
190 # auto-determine screen size
c:\python27\lib\site-packages\IPython\core\page.py in pager_page(strng, start, screen_lines, pager_cmd)
186 numlines = max(num_newlines,int(len_str/80)+1)
187 pdb.set_trace()
--> 188 screen_lines_def = get_terminal_size()[1]
189
190 # auto-determine screen size
c:\python27\lib\bdb.pyc in trace_dispatch(self, frame, event, arg)
47 return # None
48 if event == 'line':
---> 49 return self.dispatch_line(frame)
50 if event == 'call':
51 return self.dispatch_call(frame, arg)
c:\python27\lib\bdb.pyc in dispatch_line(self, frame)
65 def dispatch_line(self, frame):
66 if self.stop_here(frame) or self.break_here(frame):
---> 67 self.user_line(frame)
68 if self.quitting: raise BdbQuit
69 return self.trace_dispatch
BdbQuit:
It's crashing basically because get_terminal_size isn't working - here:
def _get_terminal_size(fd):
pdb.set_trace()
columns = lines = 0
try:
handle = _handles[fd]
csbi = create_string_buffer(22)
res = windll.kernel32.GetConsoleScreenBufferInfo(handle, csbi)
if res:
res = struct.unpack("hhhhHhhhhhh", csbi.raw)
left, top, right, bottom = res[5:9]
columns = right - left + 1
lines = bottom - top + 1
except Exception:
pass
windll.kernel32.GetConsoleScreenBufferInfo(handle, csbi) fails and returns 0. I can, later on, in pdb plug in a reasonable value for LINES and it works.
What's needed to have a crack at this?
Activity
Carreau commentedon Jan 28, 2017
If it's from
_get_terminal_size, then it's a bug from or dependencies, so it will need to be reported here. Though the maintainer seem to have stopped releasing, so I would first try to install said package from master (see if it is fixed). And if necessary we can try to poke the maintainer, and if no response ask the PyPI folks in order to take over the package and publish an update.What's the raw excpetion returned by _get_terminal_size (if you don't use PDB) we can catch it and return a dummy value (like 80x40) as a workaround.
patfla commentedon Jan 30, 2017
Thanks Carreau. Actually when I first ran it was with python 2.7, when I now rerun with 3.3 some of the problems go away. %magic% and %quickref work fine. The problem at the moment, and I'm stepping through the code to figure it out, is %recall returns nothing. The previous command is not put onto a next input line.
takluyver commentedon Jan 30, 2017
If you're using emacs 'inferior shell' mode, IPython switches to a simple fallback interface, because the full interface needs a real terminal. I expect that
%recallis one of the things that doesn't work in the fallback.patfla commentedon Jan 30, 2017
Yes, it's inferior mode. What might be done to get %recall to work?