|
|
# encoding: utf-8
|
|
|
"""Tests for IPython.utils.text"""
|
|
|
|
|
|
#-----------------------------------------------------------------------------
|
|
|
# Copyright (C) 2011 The IPython Development Team
|
|
|
#
|
|
|
# Distributed under the terms of the BSD License. The full license is in
|
|
|
# the file COPYING, distributed as part of this software.
|
|
|
#-----------------------------------------------------------------------------
|
|
|
|
|
|
#-----------------------------------------------------------------------------
|
|
|
# Imports
|
|
|
#-----------------------------------------------------------------------------
|
|
|
|
|
|
import os
|
|
|
import math
|
|
|
import random
|
|
|
|
|
|
import nose.tools as nt
|
|
|
|
|
|
from nose import with_setup
|
|
|
|
|
|
from IPython.testing import decorators as dec
|
|
|
from IPython.utils import text
|
|
|
|
|
|
#-----------------------------------------------------------------------------
|
|
|
# Globals
|
|
|
#-----------------------------------------------------------------------------
|
|
|
|
|
|
def test_columnize():
|
|
|
"""Basic columnize tests."""
|
|
|
size = 5
|
|
|
items = [l*size for l in 'abc']
|
|
|
out = text.columnize(items, displaywidth=80)
|
|
|
nt.assert_equal(out, 'aaaaa bbbbb ccccc\n')
|
|
|
out = text.columnize(items, displaywidth=12)
|
|
|
nt.assert_equal(out, 'aaaaa ccccc\nbbbbb\n')
|
|
|
out = text.columnize(items, displaywidth=10)
|
|
|
nt.assert_equal(out, 'aaaaa\nbbbbb\nccccc\n')
|
|
|
|
|
|
def test_columnize_random():
|
|
|
"""Test with random input to hopfully catch edge case """
|
|
|
for nitems in [random.randint(2,70) for i in range(2,20)]:
|
|
|
displaywidth = random.randint(20,200)
|
|
|
rand_len = [random.randint(2,displaywidth) for i in range(nitems)]
|
|
|
items = ['x'*l for l in rand_len]
|
|
|
out = text.columnize(items, displaywidth=displaywidth)
|
|
|
longer_line = max([len(x) for x in out.split('\n')])
|
|
|
longer_element = max(rand_len)
|
|
|
if longer_line > displaywidth:
|
|
|
print "Columnize displayed something lager than displaywidth : %s " % longer_line
|
|
|
print "longer element : %s " % longer_element
|
|
|
print "displaywidth : %s " % displaywidth
|
|
|
print "number of element : %s " % nitems
|
|
|
print "size of each element :\n %s" % rand_len
|
|
|
assert False
|
|
|
|
|
|
def test_columnize_medium():
|
|
|
"""Test with inputs than shouldn't be wider tahn 80 """
|
|
|
size = 40
|
|
|
items = [l*size for l in 'abc']
|
|
|
out = text.columnize(items, displaywidth=80)
|
|
|
nt.assert_equal(out, '\n'.join(items+['']))
|
|
|
|
|
|
def test_columnize_long():
|
|
|
"""Test columnize with inputs longer than the display window"""
|
|
|
size = 11
|
|
|
items = [l*size for l in 'abc']
|
|
|
out = text.columnize(items, displaywidth=size-1)
|
|
|
nt.assert_equal(out, '\n'.join(items+['']))
|
|
|
|
|
|
def eval_formatter_check(f):
|
|
|
ns = dict(n=12, pi=math.pi, stuff='hello there', os=os, u=u"café", b="café")
|
|
|
s = f.format("{n} {n//4} {stuff.split()[0]}", **ns)
|
|
|
nt.assert_equal(s, "12 3 hello")
|
|
|
s = f.format(' '.join(['{n//%i}'%i for i in range(1,8)]), **ns)
|
|
|
nt.assert_equal(s, "12 6 4 3 2 2 1")
|
|
|
s = f.format('{[n//i for i in range(1,8)]}', **ns)
|
|
|
nt.assert_equal(s, "[12, 6, 4, 3, 2, 2, 1]")
|
|
|
s = f.format("{stuff!s}", **ns)
|
|
|
nt.assert_equal(s, ns['stuff'])
|
|
|
s = f.format("{stuff!r}", **ns)
|
|
|
nt.assert_equal(s, repr(ns['stuff']))
|
|
|
|
|
|
# Check with unicode:
|
|
|
s = f.format("{u}", **ns)
|
|
|
nt.assert_equal(s, ns['u'])
|
|
|
# This decodes in a platform dependent manner, but it shouldn't error out
|
|
|
s = f.format("{b}", **ns)
|
|
|
|
|
|
nt.assert_raises(NameError, f.format, '{dne}', **ns)
|
|
|
|
|
|
def eval_formatter_slicing_check(f):
|
|
|
ns = dict(n=12, pi=math.pi, stuff='hello there', os=os)
|
|
|
s = f.format(" {stuff.split()[:]} ", **ns)
|
|
|
nt.assert_equal(s, " ['hello', 'there'] ")
|
|
|
s = f.format(" {stuff.split()[::-1]} ", **ns)
|
|
|
nt.assert_equal(s, " ['there', 'hello'] ")
|
|
|
s = f.format("{stuff[::2]}", **ns)
|
|
|
nt.assert_equal(s, ns['stuff'][::2])
|
|
|
|
|
|
nt.assert_raises(SyntaxError, f.format, "{n:x}", **ns)
|
|
|
|
|
|
def eval_formatter_no_slicing_check(f):
|
|
|
ns = dict(n=12, pi=math.pi, stuff='hello there', os=os)
|
|
|
|
|
|
s = f.format('{n:x} {pi**2:+f}', **ns)
|
|
|
nt.assert_equal(s, "c +9.869604")
|
|
|
|
|
|
s = f.format('{stuff[slice(1,4)]}', **ns)
|
|
|
nt.assert_equal(s, 'ell')
|
|
|
|
|
|
nt.assert_raises(SyntaxError, f.format, "{a[:]}")
|
|
|
|
|
|
def test_eval_formatter():
|
|
|
f = text.EvalFormatter()
|
|
|
eval_formatter_check(f)
|
|
|
eval_formatter_no_slicing_check(f)
|
|
|
|
|
|
def test_full_eval_formatter():
|
|
|
f = text.FullEvalFormatter()
|
|
|
eval_formatter_check(f)
|
|
|
eval_formatter_slicing_check(f)
|
|
|
|
|
|
def test_dollar_formatter():
|
|
|
f = text.DollarFormatter()
|
|
|
eval_formatter_check(f)
|
|
|
eval_formatter_slicing_check(f)
|
|
|
|
|
|
ns = dict(n=12, pi=math.pi, stuff='hello there', os=os)
|
|
|
s = f.format("$n", **ns)
|
|
|
nt.assert_equal(s, "12")
|
|
|
s = f.format("$n.real", **ns)
|
|
|
nt.assert_equal(s, "12")
|
|
|
s = f.format("$n/{stuff[:5]}", **ns)
|
|
|
nt.assert_equal(s, "12/hello")
|
|
|
s = f.format("$n $$HOME", **ns)
|
|
|
nt.assert_equal(s, "12 $HOME")
|
|
|
s = f.format("${foo}", foo="HOME")
|
|
|
nt.assert_equal(s, "$HOME")
|
|
|
|
|
|
|
|
|
def test_long_substr():
|
|
|
data = ['hi']
|
|
|
nt.assert_equal(text.long_substr(data), 'hi')
|
|
|
|
|
|
|
|
|
def test_long_substr2():
|
|
|
data = ['abc', 'abd', 'abf', 'ab']
|
|
|
nt.assert_equal(text.long_substr(data), 'ab')
|
|
|
|
|
|
def test_long_substr_empty():
|
|
|
data = []
|
|
|
nt.assert_equal(text.long_substr(data), '')
|
|
|
|
|
|
def test_strip_email():
|
|
|
src = """\
|
|
|
>> >>> def f(x):
|
|
|
>> ... return x+1
|
|
|
>> ...
|
|
|
>> >>> zz = f(2.5)"""
|
|
|
cln = """\
|
|
|
>>> def f(x):
|
|
|
... return x+1
|
|
|
...
|
|
|
>>> zz = f(2.5)"""
|
|
|
nt.assert_equal(text.strip_email_quotes(src), cln)
|
|
|
|
|
|
|
|
|
def test_strip_email2():
|
|
|
src = '> > > list()'
|
|
|
cln = 'list()'
|
|
|
nt.assert_equal(text.strip_email_quotes(src), cln)
|
|
|
|