All Downloads are FREE. Search and download functionalities are using the official Maven repository.

lib-python.2.7.test.test_difflib.py Maven / Gradle / Ivy

Go to download

Jython is an implementation of the high-level, dynamic, object-oriented language Python written in 100% Pure Java, and seamlessly integrated with the Java platform. It thus allows you to run Python on any Java platform.

There is a newer version: 2.7.4
Show newest version
import difflib
from test.test_support import run_unittest, findfile
import unittest
import doctest
import sys


class TestWithAscii(unittest.TestCase):
    def test_one_insert(self):
        sm = difflib.SequenceMatcher(None, 'b' * 100, 'a' + 'b' * 100)
        self.assertAlmostEqual(sm.ratio(), 0.995, places=3)
        self.assertEqual(list(sm.get_opcodes()),
            [   ('insert', 0, 0, 0, 1),
                ('equal', 0, 100, 1, 101)])
        sm = difflib.SequenceMatcher(None, 'b' * 100, 'b' * 50 + 'a' + 'b' * 50)
        self.assertAlmostEqual(sm.ratio(), 0.995, places=3)
        self.assertEqual(list(sm.get_opcodes()),
            [   ('equal', 0, 50, 0, 50),
                ('insert', 50, 50, 50, 51),
                ('equal', 50, 100, 51, 101)])

    def test_one_delete(self):
        sm = difflib.SequenceMatcher(None, 'a' * 40 + 'c' + 'b' * 40, 'a' * 40 + 'b' * 40)
        self.assertAlmostEqual(sm.ratio(), 0.994, places=3)
        self.assertEqual(list(sm.get_opcodes()),
            [   ('equal', 0, 40, 0, 40),
                ('delete', 40, 41, 40, 40),
                ('equal', 41, 81, 40, 80)])


class TestAutojunk(unittest.TestCase):
    """Tests for the autojunk parameter added in 2.7"""
    def test_one_insert_homogenous_sequence(self):
        # By default autojunk=True and the heuristic kicks in for a sequence
        # of length 200+
        seq1 = 'b' * 200
        seq2 = 'a' + 'b' * 200

        sm = difflib.SequenceMatcher(None, seq1, seq2)
        self.assertAlmostEqual(sm.ratio(), 0, places=3)

        # Now turn the heuristic off
        sm = difflib.SequenceMatcher(None, seq1, seq2, autojunk=False)
        self.assertAlmostEqual(sm.ratio(), 0.9975, places=3)


class TestSFbugs(unittest.TestCase):
    def test_ratio_for_null_seqn(self):
        # Check clearing of SF bug 763023
        s = difflib.SequenceMatcher(None, [], [])
        self.assertEqual(s.ratio(), 1)
        self.assertEqual(s.quick_ratio(), 1)
        self.assertEqual(s.real_quick_ratio(), 1)

    def test_comparing_empty_lists(self):
        # Check fix for bug #979794
        group_gen = difflib.SequenceMatcher(None, [], []).get_grouped_opcodes()
        self.assertRaises(StopIteration, group_gen.next)
        diff_gen = difflib.unified_diff([], [])
        self.assertRaises(StopIteration, diff_gen.next)

    def test_added_tab_hint(self):
        # Check fix for bug #1488943
        diff = list(difflib.Differ().compare(["\tI am a buggy"],["\t\tI am a bug"]))
        self.assertEqual("- \tI am a buggy", diff[0])
        self.assertEqual("?            --\n", diff[1])
        self.assertEqual("+ \t\tI am a bug", diff[2])
        self.assertEqual("? +\n", diff[3])

patch914575_from1 = """
   1. Beautiful is beTTer than ugly.
   2. Explicit is better than implicit.
   3. Simple is better than complex.
   4. Complex is better than complicated.
"""

patch914575_to1 = """
   1. Beautiful is better than ugly.
   3.   Simple is better than complex.
   4. Complicated is better than complex.
   5. Flat is better than nested.
"""

patch914575_from2 = """
\t\tLine 1: preceeded by from:[tt] to:[ssss]
  \t\tLine 2: preceeded by from:[sstt] to:[sssst]
  \t \tLine 3: preceeded by from:[sstst] to:[ssssss]
Line 4:  \thas from:[sst] to:[sss] after :
Line 5: has from:[t] to:[ss] at end\t
"""

patch914575_to2 = """
    Line 1: preceeded by from:[tt] to:[ssss]
    \tLine 2: preceeded by from:[sstt] to:[sssst]
      Line 3: preceeded by from:[sstst] to:[ssssss]
Line 4:   has from:[sst] to:[sss] after :
Line 5: has from:[t] to:[ss] at end
"""

patch914575_from3 = """line 0
1234567890123456789012345689012345
line 1
line 2
line 3
line 4   changed
line 5   changed
line 6   changed
line 7
line 8  subtracted
line 9
1234567890123456789012345689012345
short line
just fits in!!
just fits in two lines yup!!
the end"""

patch914575_to3 = """line 0
1234567890123456789012345689012345
line 1
line 2    added
line 3
line 4   chanGEd
line 5a  chanGed
line 6a  changEd
line 7
line 8
line 9
1234567890
another long line that needs to be wrapped
just fitS in!!
just fits in two lineS yup!!
the end"""

class TestSFpatches(unittest.TestCase):

    def test_html_diff(self):
        # Check SF patch 914575 for generating HTML differences
        f1a = ((patch914575_from1 + '123\n'*10)*3)
        t1a = (patch914575_to1 + '123\n'*10)*3
        f1b = '456\n'*10 + f1a
        t1b = '456\n'*10 + t1a
        f1a = f1a.splitlines()
        t1a = t1a.splitlines()
        f1b = f1b.splitlines()
        t1b = t1b.splitlines()
        f2 = patch914575_from2.splitlines()
        t2 = patch914575_to2.splitlines()
        f3 = patch914575_from3
        t3 = patch914575_to3
        i = difflib.HtmlDiff()
        j = difflib.HtmlDiff(tabsize=2)
        k = difflib.HtmlDiff(wrapcolumn=14)

        full = i.make_file(f1a,t1a,'from','to',context=False,numlines=5)
        tables = '\n'.join(
            [
             '

Context (first diff within numlines=5(default))

', i.make_table(f1a,t1a,'from','to',context=True), '

Context (first diff after numlines=5(default))

', i.make_table(f1b,t1b,'from','to',context=True), '

Context (numlines=6)

', i.make_table(f1a,t1a,'from','to',context=True,numlines=6), '

Context (numlines=0)

', i.make_table(f1a,t1a,'from','to',context=True,numlines=0), '

Same Context

', i.make_table(f1a,f1a,'from','to',context=True), '

Same Full

', i.make_table(f1a,f1a,'from','to',context=False), '

Empty Context

', i.make_table([],[],'from','to',context=True), '

Empty Full

', i.make_table([],[],'from','to',context=False), '

tabsize=2

', j.make_table(f2,t2), '

tabsize=default

', i.make_table(f2,t2), '

Context (wrapcolumn=14,numlines=0)

', k.make_table(f3.splitlines(),t3.splitlines(),context=True,numlines=0), '

wrapcolumn=14,splitlines()

', k.make_table(f3.splitlines(),t3.splitlines()), '

wrapcolumn=14,splitlines(True)

', k.make_table(f3.splitlines(True),t3.splitlines(True)), ]) actual = full.replace('','\n%s\n' % tables) # temporarily uncomment next two lines to baseline this test #with open('test_difflib_expect.html','w') as fp: # fp.write(actual) with open(findfile('test_difflib_expect.html')) as fp: self.assertEqual(actual, fp.read()) def test_recursion_limit(self): # Check if the problem described in patch #1413711 exists. limit = sys.getrecursionlimit() old = [(i%2 and "K:%d" or "V:A:%d") % i for i in range(limit*2)] new = [(i%2 and "K:%d" or "V:B:%d") % i for i in range(limit*2)] difflib.SequenceMatcher(None, old, new).get_opcodes() class TestOutputFormat(unittest.TestCase): def test_tab_delimiter(self): args = ['one', 'two', 'Original', 'Current', '2005-01-26 23:30:50', '2010-04-02 10:20:52'] ud = difflib.unified_diff(*args, lineterm='') self.assertEqual(list(ud)[0:2], [ "--- Original\t2005-01-26 23:30:50", "+++ Current\t2010-04-02 10:20:52"]) cd = difflib.context_diff(*args, lineterm='') self.assertEqual(list(cd)[0:2], [ "*** Original\t2005-01-26 23:30:50", "--- Current\t2010-04-02 10:20:52"]) def test_no_trailing_tab_on_empty_filedate(self): args = ['one', 'two', 'Original', 'Current'] ud = difflib.unified_diff(*args, lineterm='') self.assertEqual(list(ud)[0:2], ["--- Original", "+++ Current"]) cd = difflib.context_diff(*args, lineterm='') self.assertEqual(list(cd)[0:2], ["*** Original", "--- Current"]) def test_range_format_unified(self): # Per the diff spec at http://www.unix.org/single_unix_specification/ spec = '''\ Each field shall be of the form: %1d", if the range contains exactly one line, and: "%1d,%1d", , otherwise. If a range is empty, its beginning line number shall be the number of the line just before the range, or 0 if the empty range starts the file. ''' fmt = difflib._format_range_unified self.assertEqual(fmt(3,3), '3,0') self.assertEqual(fmt(3,4), '4') self.assertEqual(fmt(3,5), '4,2') self.assertEqual(fmt(3,6), '4,3') self.assertEqual(fmt(0,0), '0,0') def test_range_format_context(self): # Per the diff spec at http://www.unix.org/single_unix_specification/ spec = '''\ The range of lines in file1 shall be written in the following format if the range contains two or more lines: "*** %d,%d ****\n", , and the following format otherwise: "*** %d ****\n", The ending line number of an empty range shall be the number of the preceding line, or 0 if the range is at the start of the file. Next, the range of lines in file2 shall be written in the following format if the range contains two or more lines: "--- %d,%d ----\n", , and the following format otherwise: "--- %d ----\n", ''' fmt = difflib._format_range_context self.assertEqual(fmt(3,3), '3') self.assertEqual(fmt(3,4), '4') self.assertEqual(fmt(3,5), '4,5') self.assertEqual(fmt(3,6), '4,6') self.assertEqual(fmt(0,0), '0') def test_main(): difflib.HtmlDiff._default_prefix = 0 Doctests = doctest.DocTestSuite(difflib) run_unittest( TestWithAscii, TestAutojunk, TestSFpatches, TestSFbugs, TestOutputFormat, Doctests) if __name__ == '__main__': test_main()




© 2015 - 2024 Weber Informatics LLC | Privacy Policy