Edit: This only happens when I run the code from inside iPython notebook. It works fine from a regular .py file
I'm just getting started with learning myHDL and I'm getting compilation errors using @instance or @always_comb generators like so:
TypeError: compile() expected string without null bytes
For example: 3.2 Signals, ports, and concurrency:
from myhdl import Signal, delay, always, instance, now, Simulation
def ClkDriver(clk, period=20):
lowTime = int(period/2)
highTime = period - lowTime
@instance
def driveClk():
while True:
yield delay(lowTime)
clk.next = 1
yield delay(highTime)
clk.next = 0
return driveClk
clk = Signal(0)
clkdriver_inst = ClkDriver(clk)
Gives the stack trace:
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-47-4252bd2f72e0> in <module>()
16
17 clk = Signal(0)
---> 18 clkdriver_inst = ClkDriver(clk)
<ipython-input-47-4252bd2f72e0> in ClkDriver(clk, period)
5 highTime = period - lowTime
6
----> 7 @instance
8 def driveClk():
9 while True:
C:\Python27\lib\site-packages\myhdl\_instance.pyc in instance(genFunc)
40 if genFunc.func_code.co_argcount > 0:
41 raise InstanceError(_error.NrOfArgs)
---> 42 return _Instantiator(genFunc)
43
44 class _Instantiator(object):
C:\Python27\lib\site-packages\myhdl\_instance.pyc in __init__(self, genFunc)
47 self.genfunc = genFunc
48 self.gen = genFunc()
---> 49 self.waiter = _inferWaiter(self.gen)
50
C:\Python27\lib\site-packages\myhdl\_Waiter.pyc in _inferWaiter(gen)
209 s = inspect.getsource(f)
210 s = _dedent(s)
--> 211 root = ast.parse(s)
212 root.symdict = f.f_globals.copy()
213 root.symdict.update(f.f_locals)
C:\Python27\lib\ast.pyc in parse(source, filename, mode)
35 Equivalent to compile(source, filename, mode, PyCF_ONLY_AST).
36 """
---> 37 return compile(source, filename, mode, PyCF_ONLY_AST)
38
39
TypeError: compile() expected string without null bytes
Any clues on what I'm doing wrong?
This looks like an issue with unicode support in myhdl._util._dedent()
Here is snippet that illustrates the problem:
For a quick fix I added the following code:
def _dedent(s):
"""Dedent python code string."""
# RL convert to ascii
s = s.encode('ascii','ignore')
result = [t[:2] for t in generate_tokens(StringIO(s).readline)]
# set initial indent to 0 if any
if result[0][0] == INDENT:
result[0] = (INDENT, '')
return untokenize(result)