summaryrefslogtreecommitdiff
path: root/tests/coveragetest.py
blob: e1f42bb435a8fadea28f592d97dde4fbe1486ff8 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
"""Base test case class for coverage testing."""

import glob, os, random, re, shlex, shutil, sys

import coverage
from coverage.backunittest import TestCase
from coverage.backward import StringIO, import_local_file
from coverage.control import _TEST_NAME_FILE
from coverage.test_helpers import (
    EnvironmentAwareMixin, StdStreamCapturingMixin, TempDirMixin,
)

from nose.plugins.skip import SkipTest

from tests.backtest import run_command


# Status returns for the command line.
OK, ERR = 0, 1

class CoverageTest(
    EnvironmentAwareMixin,
    StdStreamCapturingMixin,
    TempDirMixin,
    TestCase
):
    """A base class for Coverage test cases."""

    # Standard unittest setting: show me diffs even if they are very long.
    maxDiff = None

    # Tell newer unittest implementations to print long helpful messages.
    longMessage = True

    def setUp(self):
        super(CoverageTest, self).setUp()

        if _TEST_NAME_FILE:                                 # pragma: debugging
            with open(_TEST_NAME_FILE, "w") as f:
                f.write("%s_%s" % (
                    self.__class__.__name__, self._testMethodName,
                ))

    def skip(self, reason):
        self.class_behavior().skipped += 1
        raise SkipTest(reason)

    def clean_local_file_imports(self):
        """Clean up the results of calls to `import_local_file`.

        Use this if you need to `import_local_file` the same file twice in
        one test.

        """
        # So that we can re-import files, clean them out first.
        self.cleanup_modules()
        # Also have to clean out the .pyc file, since the timestamp
        # resolution is only one second, a changed file might not be
        # picked up.
        for pyc in glob.glob('*.pyc'):
            os.remove(pyc)
        if os.path.exists("__pycache__"):
            shutil.rmtree("__pycache__")

    def import_local_file(self, modname):
        """Import a local file as a module.

        Opens a file in the current directory named `modname`.py, imports it
        as `modname`, and returns the module object.

        """
        return import_local_file(modname)

    def start_import_stop(self, cov, modname):
        """Start coverage, import a file, then stop coverage.

        `cov` is started and stopped, with an `import_local_file` of
        `modname` in the middle.

        The imported module is returned.

        """
        cov.start()
        try:                                    # pragma: nested
            # Import the Python file, executing it.
            mod = self.import_local_file(modname)
        finally:                                # pragma: nested
            # Stop Coverage.
            cov.stop()
        return mod

    def get_module_name(self):
        """Return the module name to use for this test run."""
        return 'coverage_test_' + str(random.random())[2:]

    # Map chars to numbers for arcz_to_arcs
    _arcz_map = {'.': -1}
    _arcz_map.update(dict((c, ord(c)-ord('0')) for c in '123456789'))
    _arcz_map.update(dict(
        (c, 10+ord(c)-ord('A')) for c in 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
    ))

    def arcz_to_arcs(self, arcz):
        """Convert a compact textual representation of arcs to a list of pairs.

        The text has space-separated pairs of letters.  Period is -1, 1-9 are
        1-9, A-Z are 10 through 36.  The resulting list is sorted regardless of
        the order of the input pairs.

        ".1 12 2." --> [(-1,1), (1,2), (2,-1)]

        Minus signs can be included in the pairs:

        "-11, 12, 2-5" --> [(-1,1), (1,2), (2,-5)]

        """
        arcs = []
        for pair in arcz.split():
            asgn = bsgn = 1
            if len(pair) == 2:
                a,b = pair
            else:
                assert len(pair) == 3
                if pair[0] == '-':
                    _,a,b = pair
                    asgn = -1
                else:
                    assert pair[1] == '-'
                    a,_,b = pair
                    bsgn = -1
            arcs.append((asgn*self._arcz_map[a], bsgn*self._arcz_map[b]))
        return sorted(arcs)

    def assert_equal_args(self, a1, a2, msg=None):
        """Assert that the arc lists `a1` and `a2` are equal."""
        # Make them into multi-line strings so we can see what's going wrong.
        s1 = "\n".join(repr(a) for a in a1) + "\n"
        s2 = "\n".join(repr(a) for a in a2) + "\n"
        self.assertMultiLineEqual(s1, s2, msg)

    def check_coverage(self, text, lines=None, missing="", report="",
            excludes=None, partials="",
            arcz=None, arcz_missing="", arcz_unpredicted=""):
        """Check the coverage measurement of `text`.

        The source `text` is run and measured.  `lines` are the line numbers
        that are executable, or a list of possible line numbers, any of which
        could match. `missing` are the lines not executed, `excludes` are
        regexes to match against for excluding lines, and `report` is the text
        of the measurement report.

        For arc measurement, `arcz` is a string that can be decoded into arcs
        in the code (see `arcz_to_arcs` for the encoding scheme),
        `arcz_missing` are the arcs that are not executed, and
        `arcs_unpredicted` are the arcs executed in the code, but not deducible
        from the code.

        """
        # We write the code into a file so that we can import it.
        # Coverage wants to deal with things as modules with file names.
        modname = self.get_module_name()

        self.make_file(modname+".py", text)

        arcs = arcs_missing = arcs_unpredicted = None
        if arcz is not None:
            arcs = self.arcz_to_arcs(arcz)
            arcs_missing = self.arcz_to_arcs(arcz_missing or "")
            arcs_unpredicted = self.arcz_to_arcs(arcz_unpredicted or "")

        # Start up Coverage.
        cov = coverage.coverage(branch=(arcs_missing is not None))
        cov.erase()
        for exc in excludes or []:
            cov.exclude(exc)
        for par in partials or []:
            cov.exclude(par, which='partial')

        mod = self.start_import_stop(cov, modname)

        # Clean up our side effects
        del sys.modules[modname]

        # Get the analysis results, and check that they are right.
        analysis = cov._analyze(mod)
        statements = sorted(analysis.statements)
        if lines is not None:
            if type(lines[0]) == type(1):
                # lines is just a list of numbers, it must match the statements
                # found in the code.
                self.assertEqual(statements, lines)
            else:
                # lines is a list of possible line number lists, one of them
                # must match.
                for line_list in lines:
                    if statements == line_list:
                        break
                else:
                    self.fail(
                        "None of the lines choices matched %r" % statements
                        )

            missing_formatted = analysis.missing_formatted()
            if type(missing) == type(""):
                self.assertEqual(missing_formatted, missing)
            else:
                for missing_list in missing:
                    if missing_formatted == missing_list:
                        break
                else:
                    self.fail(
                        "None of the missing choices matched %r" %
                        missing_formatted
                        )

        if arcs is not None:
            self.assert_equal_args(
                analysis.arc_possibilities(), arcs, "Possible arcs differ"
                )

            if arcs_missing is not None:
                self.assert_equal_args(
                    analysis.arcs_missing(), arcs_missing,
                    "Missing arcs differ"
                    )

            if arcs_unpredicted is not None:
                self.assert_equal_args(
                    analysis.arcs_unpredicted(), arcs_unpredicted,
                    "Unpredicted arcs differ"
                    )

        if report:
            frep = StringIO()
            cov.report(mod, file=frep)
            rep = " ".join(frep.getvalue().split("\n")[2].split()[1:])
            self.assertEqual(report, rep)

    def nice_file(self, *fparts):
        """Canonicalize the filename composed of the parts in `fparts`."""
        fname = os.path.join(*fparts)
        return os.path.normcase(os.path.abspath(os.path.realpath(fname)))

    def assert_same_files(self, flist1, flist2):
        """Assert that `flist1` and `flist2` are the same set of file names."""
        flist1_nice = [self.nice_file(f) for f in flist1]
        flist2_nice = [self.nice_file(f) for f in flist2]
        self.assertCountEqual(flist1_nice, flist2_nice)

    def assert_exists(self, fname):
        """Assert that `fname` is a file that exists."""
        msg = "File %r should exist" % fname
        self.assertTrue(os.path.exists(fname), msg)

    def assert_doesnt_exist(self, fname):
        """Assert that `fname` is a file that doesn't exist."""
        msg = "File %r shouldn't exist" % fname
        self.assertTrue(not os.path.exists(fname), msg)

    def assert_starts_with(self, s, prefix, msg=None):
        """Assert that `s` starts with `prefix`."""
        if not s.startswith(prefix):
            self.fail(msg or ("%r doesn't start with %r" % (s, prefix)))

    def command_line(self, args, ret=OK, _covpkg=None):
        """Run `args` through the command line.

        Use this when you want to run the full coverage machinery, but in the
        current process.  Exceptions may be thrown from deep in the code.
        Asserts that `ret` is returned by `CoverageScript.command_line`.

        Compare with `run_command`.

        Returns None.

        """
        script = coverage.CoverageScript(_covpkg=_covpkg)
        ret_actual = script.command_line(shlex.split(args))
        self.assertEqual(ret_actual, ret)

    def run_command(self, cmd):
        """Run the command-line `cmd` in a sub-process, and print its output.

        Use this when you need to test the process behavior of coverage.

        Compare with `command_line`.

        Returns the process' stdout text.

        """
        # Running Python sub-processes can be tricky.  Use the real name of our
        # own executable.  So "python foo.py" might get executed as
        # "python3.3 foo.py".  This is important because Python 3.x doesn't
        # install as "python", so you might get a Python 2 executable instead
        # if you don't use the executable's basename.
        if cmd.startswith("python "):
            cmd = os.path.basename(sys.executable) + cmd[6:]

        _, output = self.run_command_status(cmd)
        return output

    def run_command_status(self, cmd):
        """Run the command-line `cmd` in a sub-process, and print its output.

        Use this when you need to test the process behavior of coverage.

        Compare with `command_line`.

        Returns a pair: the process' exit status and stdout text.

        """
        # Add our test modules directory to PYTHONPATH.  I'm sure there's too
        # much path munging here, but...
        here = os.path.dirname(self.nice_file(coverage.__file__, ".."))
        testmods = self.nice_file(here, 'tests/modules')
        zipfile = self.nice_file(here, 'tests/zipmods.zip')
        pypath = os.getenv('PYTHONPATH', '')
        if pypath:
            pypath += os.pathsep
        pypath += testmods + os.pathsep + zipfile
        self.set_environ('PYTHONPATH', pypath)

        status, output = run_command(cmd)
        print(output)
        return status, output

    def report_from_command(self, cmd):
        """Return the report from the `cmd`, with some convenience added."""
        report = self.run_command(cmd).replace('\\', '/')
        self.assertNotIn("error", report.lower())
        return report

    def report_lines(self, report):
        """Return the lines of the report, as a list."""
        lines = report.split('\n')
        self.assertEqual(lines[-1], "")
        return lines[:-1]

    def line_count(self, report):
        """How many lines are in `report`?"""
        return len(self.report_lines(report))

    def squeezed_lines(self, report):
        """Return a list of the lines in report, with the spaces squeezed."""
        lines = self.report_lines(report)
        return [re.sub(r"\s+", " ", l.strip()) for l in lines]

    def last_line_squeezed(self, report):
        """Return the last line of `report` with the spaces squeezed down."""
        return self.squeezed_lines(report)[-1]