qt5base-lts/tests/auto/testlib/selftests/generate_expected_output.py

211 lines
8.9 KiB
Python
Raw Normal View History

#!/usr/bin/env python3
#############################################################################
##
## Copyright (C) 2016 The Qt Company Ltd.
## Contact: https://www.qt.io/licensing/
##
## This file is part of the release tools of the Qt Toolkit.
##
## $QT_BEGIN_LICENSE:GPL-EXCEPT$
## Commercial License Usage
## Licensees holding valid commercial Qt licenses may use this file in
## accordance with the commercial license agreement provided with the
## Software or, alternatively, in accordance with the terms contained in
## a written agreement between you and The Qt Company. For licensing terms
## and conditions see https://www.qt.io/terms-conditions. For further
## information use the contact form at https://www.qt.io/contact-us.
##
## GNU General Public License Usage
## Alternatively, this file may be used under the terms of the GNU
## General Public License version 3 as published by the Free Software
## Foundation with exceptions as appearing in the file LICENSE.GPL3-EXCEPT
## included in the packaging of this file. Please review the following
## information to ensure the GNU General Public License requirements will
## be met: https://www.gnu.org/licenses/gpl-3.0.html.
##
## $QT_END_LICENSE$
##
#############################################################################
Major re-write of generate_expected_output.py Restructured, separated the canonicalising of output lines out (into an object that prepares the necessary regexes and replacements) suppress more, changed the path-stripping to strip qtbase's parent rather than os.getcwd() and took account of shadow builds (so both source tree and build tree provide prefixes we want to strip from paths). Also cope with $PWD potentially having symlinks in it, where os.getcwd() is canonical. It's possible some output might name files elsewhere in the source tree; these won't be filtered by the prior cwd prefix removal; and, in any case, the problem with cwd is only that the ancestry of qtbase is apt to vary; paths relative to there should be consistent between test runs. This change shall lead to a one-off rewrite of all expected_* files; but it should now catch all paths. By stripping both build root and source root (when different) it also avoids differences for those doing out-of-source ("shadow") builds. In our XML formats, any hyphens in root paths (e.g. I had Qt-5.6 in my build root's path) got represented by a character entity, confounding the replacement; so also do replacement that catches this. We may discover other character entity subsitutions needed along with this. Now filtering line numbers and timing information, including benchmark results; these numbers all get replaced with 0 to avoid noisy diffs. Also purging dangling hspace, to placate sanity-bot. The module can now be imported - the code it runs is packaged as a main() function that a __name__ == '__main__' stanza runs - and all data is localised to where it's needed, rather than held in globals. Tidied up and organized the existing regexes. There are doc-strings; there is a short usage comment. Data is localised rather than global and modern pythonic idioms get used where apt. Regexes are compiled once instead of repeatedly. An object looks after the list of patterns to apply and its construction handles all anticipated problems. Failures are mediated by an exception. The output file now gets written once, instead of twice (once before editing, then over-write to edit), and Popen uses text mode, so that write can do the same. Its command is delivered as an array, avoiding the need to invoke a shell. Instead of relying on qmake being in our path (which might give us a bogus QT_VERSION if the one in path doesn't match our build tree), use the relative path to qmake - we rely on being run in a specific directory in the build tree, after all. Escape dots in the version properly, so that 51730 doesn't get mistaken for 5.7.0 (for example), and moved this check later in the sequence (matching a smaller target makes it more likely to falsely match). Overtly check we are in the right directory and tell the user what we actually need, if run from the wrong place. Simplify handling of the unsupported use-case for MS-Windows (but note what would be needed for it). Change-Id: Ibdff8f8cae173f6c31492648148cc345ae29022b Reviewed-by: Mitch Curtis <mitch.curtis@qt.io> Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
2016-07-25 13:55:47 +00:00
# Regenerate all tests' output.
#
# Usage: cd to the build directory corresponding to this script's
# location; invoke this script; optionally pass the names of sub-dirs
# to limit which tests to regenerate expected_* files for.
import os
import subprocess
Major re-write of generate_expected_output.py Restructured, separated the canonicalising of output lines out (into an object that prepares the necessary regexes and replacements) suppress more, changed the path-stripping to strip qtbase's parent rather than os.getcwd() and took account of shadow builds (so both source tree and build tree provide prefixes we want to strip from paths). Also cope with $PWD potentially having symlinks in it, where os.getcwd() is canonical. It's possible some output might name files elsewhere in the source tree; these won't be filtered by the prior cwd prefix removal; and, in any case, the problem with cwd is only that the ancestry of qtbase is apt to vary; paths relative to there should be consistent between test runs. This change shall lead to a one-off rewrite of all expected_* files; but it should now catch all paths. By stripping both build root and source root (when different) it also avoids differences for those doing out-of-source ("shadow") builds. In our XML formats, any hyphens in root paths (e.g. I had Qt-5.6 in my build root's path) got represented by a character entity, confounding the replacement; so also do replacement that catches this. We may discover other character entity subsitutions needed along with this. Now filtering line numbers and timing information, including benchmark results; these numbers all get replaced with 0 to avoid noisy diffs. Also purging dangling hspace, to placate sanity-bot. The module can now be imported - the code it runs is packaged as a main() function that a __name__ == '__main__' stanza runs - and all data is localised to where it's needed, rather than held in globals. Tidied up and organized the existing regexes. There are doc-strings; there is a short usage comment. Data is localised rather than global and modern pythonic idioms get used where apt. Regexes are compiled once instead of repeatedly. An object looks after the list of patterns to apply and its construction handles all anticipated problems. Failures are mediated by an exception. The output file now gets written once, instead of twice (once before editing, then over-write to edit), and Popen uses text mode, so that write can do the same. Its command is delivered as an array, avoiding the need to invoke a shell. Instead of relying on qmake being in our path (which might give us a bogus QT_VERSION if the one in path doesn't match our build tree), use the relative path to qmake - we rely on being run in a specific directory in the build tree, after all. Escape dots in the version properly, so that 51730 doesn't get mistaken for 5.7.0 (for example), and moved this check later in the sequence (matching a smaller target makes it more likely to falsely match). Overtly check we are in the right directory and tell the user what we actually need, if run from the wrong place. Simplify handling of the unsupported use-case for MS-Windows (but note what would be needed for it). Change-Id: Ibdff8f8cae173f6c31492648148cc345ae29022b Reviewed-by: Mitch Curtis <mitch.curtis@qt.io> Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
2016-07-25 13:55:47 +00:00
class Fail (Exception): pass
class Cleaner (object):
"""Tool to clean up test output to make diff-ing runs useful.
We care about whether tests pass or fail - if that changes,
something that matters has happened - and we care about some
changes to what they say when they do fail; but we don't care
exactly what line of what file the failing line of code now
occupies, nor do we care how many milliseconds each test took to
run; and changes to the Qt version number mean nothing to us.
Create one singleton instance; it'll do mildly expensive things
once and you can use its .clean() method to tidy up your test
output."""
def __init__(self, here, command):
"""Set up the details we need for later cleaning.
Takes two parameters: here is $PWD and command is how this
script was invoked, from which we'll work out where it is; in
a shadow build, the former is the build tree's location
corresponding to this last. Checks $PWD does look as expected
in a build tree - raising Fail() if not - then invokes qmake
to discover Qt version (saved as .version for the benefit of
clients) and prepares the sequence of (regex, replace) pairs
that .clean() needs to do its job."""
self.version, self.__replace = self.__getPatterns(here, command)
import re
@staticmethod
def __getPatterns(here, command,
patterns = (
# Timings:
(r'( *<Duration msecs=)"[\d\.]+"/>', r'\1"0"/>'), # xml, lightxml
(r'(Totals:.*,) *[0-9.]+ms', r'\1 0ms'), # txt
# Benchmarks:
(r'[0-9,.]+( (?:CPU ticks|msecs) per iteration \(total:) [0-9,.]+ ', r'0\1 0, '), # txt
(r'(<BenchmarkResult metric="(?:CPUTicks|WalltimeMilliseconds)".*\bvalue=)"[^"]+"', r'\1"0"'), # xml, lightxml
# Build details:
(r'(Config: Using QtTest library).*', r'\1'), # txt
(r'( *<QtBuild)>[^<]+</QtBuild>', r'\1/>'), # xml, lightxml
(r'(<property value=")[^"]+(" name="QtBuild"/>)', r'\1\2'), # xunitxml
# Line numbers in source files:
(r'(Loc: \[[^[\]()]+)\(\d+\)', r'\1(0)'), # txt
(r'(\[Loc: [^[\]()]+)\(\d+\)', r'\1(0)'), # teamcity
Major re-write of generate_expected_output.py Restructured, separated the canonicalising of output lines out (into an object that prepares the necessary regexes and replacements) suppress more, changed the path-stripping to strip qtbase's parent rather than os.getcwd() and took account of shadow builds (so both source tree and build tree provide prefixes we want to strip from paths). Also cope with $PWD potentially having symlinks in it, where os.getcwd() is canonical. It's possible some output might name files elsewhere in the source tree; these won't be filtered by the prior cwd prefix removal; and, in any case, the problem with cwd is only that the ancestry of qtbase is apt to vary; paths relative to there should be consistent between test runs. This change shall lead to a one-off rewrite of all expected_* files; but it should now catch all paths. By stripping both build root and source root (when different) it also avoids differences for those doing out-of-source ("shadow") builds. In our XML formats, any hyphens in root paths (e.g. I had Qt-5.6 in my build root's path) got represented by a character entity, confounding the replacement; so also do replacement that catches this. We may discover other character entity subsitutions needed along with this. Now filtering line numbers and timing information, including benchmark results; these numbers all get replaced with 0 to avoid noisy diffs. Also purging dangling hspace, to placate sanity-bot. The module can now be imported - the code it runs is packaged as a main() function that a __name__ == '__main__' stanza runs - and all data is localised to where it's needed, rather than held in globals. Tidied up and organized the existing regexes. There are doc-strings; there is a short usage comment. Data is localised rather than global and modern pythonic idioms get used where apt. Regexes are compiled once instead of repeatedly. An object looks after the list of patterns to apply and its construction handles all anticipated problems. Failures are mediated by an exception. The output file now gets written once, instead of twice (once before editing, then over-write to edit), and Popen uses text mode, so that write can do the same. Its command is delivered as an array, avoiding the need to invoke a shell. Instead of relying on qmake being in our path (which might give us a bogus QT_VERSION if the one in path doesn't match our build tree), use the relative path to qmake - we rely on being run in a specific directory in the build tree, after all. Escape dots in the version properly, so that 51730 doesn't get mistaken for 5.7.0 (for example), and moved this check later in the sequence (matching a smaller target makes it more likely to falsely match). Overtly check we are in the right directory and tell the user what we actually need, if run from the wrong place. Simplify handling of the unsupported use-case for MS-Windows (but note what would be needed for it). Change-Id: Ibdff8f8cae173f6c31492648148cc345ae29022b Reviewed-by: Mitch Curtis <mitch.curtis@qt.io> Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
2016-07-25 13:55:47 +00:00
(r'(<Incident.*\bfile=.*\bline=)"\d+"', r'\1"0"'), # lightxml, xml
),
precook = re.compile):
"""Private implementation details of __init__()."""
qmake = ('..',) * 4 + ('bin', 'qmake')
qmake = os.path.join(*qmake)
if os.path.sep in command:
scriptPath = os.path.abspath(command)
elif os.path.exists(command):
# e.g. if you typed "python3 generate_expected_output.py"
scriptPath = os.path.join(here, command)
else:
# From py 3.2: could use os.get_exec_path() here.
for d in os.environ.get('PATH', '').split(os.pathsep):
scriptPath = os.path.join(d, command)
if os.path.isfile(scriptPath):
break
else: # didn't break
raise Fail('Unable to find', command, 'in $PATH')
# Are we being run from the right place ?
myNames = scriptPath.split(os.path.sep)
if not (here.split(os.path.sep)[-5:] == myNames[-6:-1]
and os.path.isfile(qmake)):
raise Fail('Run', myNames[-1], 'in its directory of a completed build')
try:
qtver = subprocess.check_output([qmake, '-query', 'QT_VERSION'])
except OSError as what:
raise Fail(what.strerror)
qtver = qtver.strip().decode('utf-8')
scriptPath = os.path.dirname(scriptPath) # ditch leaf file-name
sentinel = os.path.sep + 'qtbase' + os.path.sep # '/qtbase/'
# Identify the path prefix of our qtbase ancestor directory
# (source, build and $PWD, when different); trim such prefixes
# off all paths we see.
roots = tuple(r[:r.find(sentinel) + 1].encode('unicode-escape').decode('utf-8')
for r in set((here, scriptPath, os.environ.get('PWD', '')))
if sentinel in r)
patterns += tuple((root, r'') for root in roots) + (
(r'\.'.join(qtver.split('.')), r'@INSERT_QT_VERSION_HERE@'),)
if any('-' in r for r in roots):
# Our xml formats replace hyphens with a character entity:
patterns += tuple((root.replace('-', '&#x0*2D;'), r'')
for root in roots if '-' in root)
return qtver, tuple((precook(p), r) for p, r in patterns)
del re
def clean(self, data):
"""Remove volatile details from test output.
Takes the full test output as a single (possibly huge)
multi-line string; iterates over cleaned lines of output."""
for line in data.split('\n'):
# Replace all occurrences of each regex:
for searchRe, replaceExp in self.__replace:
line = searchRe.sub(replaceExp, line)
yield line
def generateTestData(testname, clean,
formats = ('xml', 'txt', 'xunitxml', 'lightxml', 'teamcity'),
Major re-write of generate_expected_output.py Restructured, separated the canonicalising of output lines out (into an object that prepares the necessary regexes and replacements) suppress more, changed the path-stripping to strip qtbase's parent rather than os.getcwd() and took account of shadow builds (so both source tree and build tree provide prefixes we want to strip from paths). Also cope with $PWD potentially having symlinks in it, where os.getcwd() is canonical. It's possible some output might name files elsewhere in the source tree; these won't be filtered by the prior cwd prefix removal; and, in any case, the problem with cwd is only that the ancestry of qtbase is apt to vary; paths relative to there should be consistent between test runs. This change shall lead to a one-off rewrite of all expected_* files; but it should now catch all paths. By stripping both build root and source root (when different) it also avoids differences for those doing out-of-source ("shadow") builds. In our XML formats, any hyphens in root paths (e.g. I had Qt-5.6 in my build root's path) got represented by a character entity, confounding the replacement; so also do replacement that catches this. We may discover other character entity subsitutions needed along with this. Now filtering line numbers and timing information, including benchmark results; these numbers all get replaced with 0 to avoid noisy diffs. Also purging dangling hspace, to placate sanity-bot. The module can now be imported - the code it runs is packaged as a main() function that a __name__ == '__main__' stanza runs - and all data is localised to where it's needed, rather than held in globals. Tidied up and organized the existing regexes. There are doc-strings; there is a short usage comment. Data is localised rather than global and modern pythonic idioms get used where apt. Regexes are compiled once instead of repeatedly. An object looks after the list of patterns to apply and its construction handles all anticipated problems. Failures are mediated by an exception. The output file now gets written once, instead of twice (once before editing, then over-write to edit), and Popen uses text mode, so that write can do the same. Its command is delivered as an array, avoiding the need to invoke a shell. Instead of relying on qmake being in our path (which might give us a bogus QT_VERSION if the one in path doesn't match our build tree), use the relative path to qmake - we rely on being run in a specific directory in the build tree, after all. Escape dots in the version properly, so that 51730 doesn't get mistaken for 5.7.0 (for example), and moved this check later in the sequence (matching a smaller target makes it more likely to falsely match). Overtly check we are in the right directory and tell the user what we actually need, if run from the wrong place. Simplify handling of the unsupported use-case for MS-Windows (but note what would be needed for it). Change-Id: Ibdff8f8cae173f6c31492648148cc345ae29022b Reviewed-by: Mitch Curtis <mitch.curtis@qt.io> Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
2016-07-25 13:55:47 +00:00
extraArgs = {
"commandlinedata": "fiveTablePasses fiveTablePasses:fiveTablePasses_data1 -v2",
"benchlibcallgrind": "-callgrind",
"benchlibeventcounter": "-eventcounter",
"benchliboptions": "-eventcounter",
"benchlibtickcounter": "-tickcounter",
"badxml": "-eventcounter",
"benchlibcounting": "-eventcounter",
"printdatatags": "-datatags",
"printdatatagswithglobaltags": "-datatags",
"silent": "-silent",
"verbose1": "-v1",
"verbose2": "-v2",
}):
"""Run one test and save its cleaned results.
Required arguments are the name of the test directory (the binary
it contains is expected to have the same name) and a function
that'll clean a test-run's output; see Cleaner.clean().
"""
# MS-Win: shall need to add .exe to this
path = os.path.join(testname, testname)
if not os.path.isfile(path):
print("Warning: directory", testname, "contains no test executable")
return
print(" running", testname)
for format in formats:
Major re-write of generate_expected_output.py Restructured, separated the canonicalising of output lines out (into an object that prepares the necessary regexes and replacements) suppress more, changed the path-stripping to strip qtbase's parent rather than os.getcwd() and took account of shadow builds (so both source tree and build tree provide prefixes we want to strip from paths). Also cope with $PWD potentially having symlinks in it, where os.getcwd() is canonical. It's possible some output might name files elsewhere in the source tree; these won't be filtered by the prior cwd prefix removal; and, in any case, the problem with cwd is only that the ancestry of qtbase is apt to vary; paths relative to there should be consistent between test runs. This change shall lead to a one-off rewrite of all expected_* files; but it should now catch all paths. By stripping both build root and source root (when different) it also avoids differences for those doing out-of-source ("shadow") builds. In our XML formats, any hyphens in root paths (e.g. I had Qt-5.6 in my build root's path) got represented by a character entity, confounding the replacement; so also do replacement that catches this. We may discover other character entity subsitutions needed along with this. Now filtering line numbers and timing information, including benchmark results; these numbers all get replaced with 0 to avoid noisy diffs. Also purging dangling hspace, to placate sanity-bot. The module can now be imported - the code it runs is packaged as a main() function that a __name__ == '__main__' stanza runs - and all data is localised to where it's needed, rather than held in globals. Tidied up and organized the existing regexes. There are doc-strings; there is a short usage comment. Data is localised rather than global and modern pythonic idioms get used where apt. Regexes are compiled once instead of repeatedly. An object looks after the list of patterns to apply and its construction handles all anticipated problems. Failures are mediated by an exception. The output file now gets written once, instead of twice (once before editing, then over-write to edit), and Popen uses text mode, so that write can do the same. Its command is delivered as an array, avoiding the need to invoke a shell. Instead of relying on qmake being in our path (which might give us a bogus QT_VERSION if the one in path doesn't match our build tree), use the relative path to qmake - we rely on being run in a specific directory in the build tree, after all. Escape dots in the version properly, so that 51730 doesn't get mistaken for 5.7.0 (for example), and moved this check later in the sequence (matching a smaller target makes it more likely to falsely match). Overtly check we are in the right directory and tell the user what we actually need, if run from the wrong place. Simplify handling of the unsupported use-case for MS-Windows (but note what would be needed for it). Change-Id: Ibdff8f8cae173f6c31492648148cc345ae29022b Reviewed-by: Mitch Curtis <mitch.curtis@qt.io> Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
2016-07-25 13:55:47 +00:00
cmd = [path, '-' + format]
if testname in extraArgs:
cmd += extraArgs[testname].split()
data = subprocess.Popen(cmd, stdout=subprocess.PIPE,
universal_newlines=True).communicate()[0]
with open('expected_' + testname + '.' + format, 'w') as out:
out.write('\n'.join(clean(data))) # write() appends a newline, too
def main(name, *args):
"""Minimal argument parsing and driver for the real work"""
os.environ['LC_ALL'] = 'C'
herePath = os.getcwd()
cleaner = Cleaner(herePath, name)
tests = args if args else [d for d in os.listdir('.') if os.path.isdir(d)]
print("Generating", len(tests), "test results for", cleaner.version, "in:", herePath)
for path in tests:
generateTestData(path, cleaner.clean)
if __name__ == '__main__':
# Executed when script is run, not when imported (e.g. to debug)
import sys
if sys.platform.startswith('win'):
print("This script does not work on Windows.")
exit()
try:
main(*sys.argv)
except Fail as what:
sys.stderr.write('Failed: ' + ' '.join(what.args) + '\n')
exit(1)