Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
lucene/dev-tools/scripts/smokeTestRelease.py

1189 lines
45 KiB
Python
Executable File
Raw Blame History

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import argparse
import codecs
import datetime
import filecmp
import hashlib
import http.client
import os
import platform
import re
import shutil
import subprocess
import sys
import textwrap
import traceback
import urllib.error
import urllib.parse
import urllib.parse
import urllib.request
import xml.etree.ElementTree as ET
import zipfile
from collections import namedtuple
import scriptutil
# This tool expects to find /lucene off the base URL. You
# must have a working gpg, tar, unzip in your path. This has been
# tested on Linux and on Cygwin under Windows 7.
cygwin = platform.system().lower().startswith('cygwin')
cygwinWindowsRoot = os.popen('cygpath -w /').read().strip().replace('\\','/') if cygwin else ''
def unshortenURL(url):
parsed = urllib.parse.urlparse(url)
if parsed[0] in ('http', 'https'):
h = http.client.HTTPConnection(parsed.netloc)
h.request('HEAD', parsed.path)
response = h.getresponse()
if int(response.status/100) == 3 and response.getheader('Location'):
return response.getheader('Location')
return url
# TODO
# - make sure jars exist inside bin release
# - make sure docs exist
reHREF = re.compile('<a href="(.*?)">(.*?)</a>')
# Set to False to avoid re-downloading the packages...
FORCE_CLEAN = True
def getHREFs(urlString):
# Deref any redirects
while True:
url = urllib.parse.urlparse(urlString)
if url.scheme == "http":
h = http.client.HTTPConnection(url.netloc)
elif url.scheme == "https":
h = http.client.HTTPSConnection(url.netloc)
else:
raise RuntimeError("Unknown protocol: %s" % url.scheme)
h.request('HEAD', url.path)
r = h.getresponse()
newLoc = r.getheader('location')
if newLoc is not None:
urlString = newLoc
else:
break
links = []
try:
html = load(urlString)
except:
print('\nFAILED to open url %s' % urlString)
traceback.print_exc()
raise
for subUrl, text in reHREF.findall(html):
fullURL = urllib.parse.urljoin(urlString, subUrl)
links.append((text, fullURL))
return links
def load(urlString):
try:
content = urllib.request.urlopen(urlString).read().decode('utf-8')
except Exception as e:
print('Retrying download of url %s after exception: %s' % (urlString, e))
content = urllib.request.urlopen(urlString).read().decode('utf-8')
return content
def noJavaPackageClasses(desc, file):
with zipfile.ZipFile(file) as z2:
for name2 in z2.namelist():
if name2.endswith('.class') and (name2.startswith('java/') or name2.startswith('javax/')):
raise RuntimeError('%s contains sheisty class "%s"' % (desc, name2))
def decodeUTF8(bytes):
return codecs.getdecoder('UTF-8')(bytes)[0]
MANIFEST_FILE_NAME = 'META-INF/MANIFEST.MF'
NOTICE_FILE_NAME = 'META-INF/NOTICE.txt'
LICENSE_FILE_NAME = 'META-INF/LICENSE.txt'
def checkJARMetaData(desc, jarFile, gitRevision, version):
with zipfile.ZipFile(jarFile, 'r') as z:
for name in (MANIFEST_FILE_NAME, NOTICE_FILE_NAME, LICENSE_FILE_NAME):
try:
# The Python docs state a KeyError is raised ... so this None
# check is just defensive:
if z.getinfo(name) is None:
raise RuntimeError('%s is missing %s' % (desc, name))
except KeyError:
raise RuntimeError('%s is missing %s' % (desc, name))
s = decodeUTF8(z.read(MANIFEST_FILE_NAME))
for verify in (
'Specification-Vendor: The Apache Software Foundation',
'Implementation-Vendor: The Apache Software Foundation',
'Specification-Title: Lucene Search Engine:',
'Implementation-Title: org.apache.lucene',
'X-Compile-Source-JDK: 11',
'X-Compile-Target-JDK: 11',
'Specification-Version: %s' % version,
'X-Build-JDK: 11.',
'Extension-Name: org.apache.lucene'):
if type(verify) is not tuple:
verify = (verify,)
for x in verify:
if s.find(x) != -1:
break
else:
if len(verify) == 1:
raise RuntimeError('%s is missing "%s" inside its META-INF/MANIFEST.MF' % (desc, verify[0]))
else:
raise RuntimeError('%s is missing one of "%s" inside its META-INF/MANIFEST.MF' % (desc, verify))
if gitRevision != 'skip':
# Make sure this matches the version and git revision we think we are releasing:
match = re.search("Implementation-Version: (.+\r\n .+)", s, re.MULTILINE)
if match:
implLine = match.group(1).replace("\r\n ", "")
verifyRevision = '%s %s' % (version, gitRevision)
if implLine.find(verifyRevision) == -1:
raise RuntimeError('%s is missing "%s" inside its META-INF/MANIFEST.MF (wrong git revision?)' % \
(desc, verifyRevision))
else:
raise RuntimeError('%s is missing Implementation-Version inside its META-INF/MANIFEST.MF' % desc)
notice = decodeUTF8(z.read(NOTICE_FILE_NAME))
lucene_license = decodeUTF8(z.read(LICENSE_FILE_NAME))
if LUCENE_LICENSE is None:
raise RuntimeError('BUG in smokeTestRelease!')
if LUCENE_NOTICE is None:
raise RuntimeError('BUG in smokeTestRelease!')
if notice != LUCENE_NOTICE:
raise RuntimeError('%s: %s contents doesn\'t match main NOTICE.txt' % \
(desc, NOTICE_FILE_NAME))
if lucene_license != LUCENE_LICENSE:
raise RuntimeError('%s: %s contents doesn\'t match main LICENSE.txt' % \
(desc, LICENSE_FILE_NAME))
def normSlashes(path):
return path.replace(os.sep, '/')
def checkAllJARs(topDir, gitRevision, version):
print(' verify JAR metadata/identity/no javax.* or java.* classes...')
for root, dirs, files in os.walk(topDir):
normRoot = normSlashes(root)
for file in files:
if file.lower().endswith('.jar'):
if normRoot.endswith('/replicator/lib') and file.startswith('javax.servlet'):
continue
fullPath = '%s/%s' % (root, file)
noJavaPackageClasses('JAR file "%s"' % fullPath, fullPath)
if file.lower().find('lucene') != -1:
checkJARMetaData('JAR file "%s"' % fullPath, fullPath, gitRevision, version)
def checkSigs(urlString, version, tmpDir, isSigned, keysFile):
print(' test basics...')
ents = getDirEntries(urlString)
artifact = None
changesURL = None
mavenURL = None
artifactURL = None
expectedSigs = []
if isSigned:
expectedSigs.append('asc')
expectedSigs.extend(['sha512'])
sigs = []
artifacts = []
for text, subURL in ents:
if text == 'KEYS':
raise RuntimeError('lucene: release dir should not contain a KEYS file - only toplevel /dist/lucene/KEYS is used')
elif text == 'maven/':
mavenURL = subURL
elif text.startswith('changes'):
if text not in ('changes/', 'changes-%s/' % version):
raise RuntimeError('lucene: found %s vs expected changes-%s/' % (text, version))
changesURL = subURL
elif artifact is None:
artifact = text
artifactURL = subURL
expected = 'lucene-%s' % version
if not artifact.startswith(expected):
raise RuntimeError('lucene: unknown artifact %s: expected prefix %s' % (text, expected))
sigs = []
elif text.startswith(artifact + '.'):
sigs.append(text[len(artifact)+1:])
else:
if sigs != expectedSigs:
raise RuntimeError('lucene: artifact %s has wrong sigs: expected %s but got %s' % (artifact, expectedSigs, sigs))
artifacts.append((artifact, artifactURL))
artifact = text
artifactURL = subURL
sigs = []
if sigs != []:
artifacts.append((artifact, artifactURL))
if sigs != expectedSigs:
raise RuntimeError('lucene: artifact %s has wrong sigs: expected %s but got %s' % (artifact, expectedSigs, sigs))
expected = ['lucene-%s-src.tgz' % version,
'lucene-%s.tgz' % version]
actual = [x[0] for x in artifacts]
if expected != actual:
raise RuntimeError('lucene: wrong artifacts: expected %s but got %s' % (expected, actual))
# Set up clean gpg world; import keys file:
gpgHomeDir = '%s/lucene.gpg' % tmpDir
if os.path.exists(gpgHomeDir):
shutil.rmtree(gpgHomeDir)
os.makedirs(gpgHomeDir, 0o700)
run('gpg --homedir %s --import %s' % (gpgHomeDir, keysFile),
'%s/lucene.gpg.import.log' % tmpDir)
if mavenURL is None:
raise RuntimeError('lucene is missing maven')
if changesURL is None:
raise RuntimeError('lucene is missing changes-%s' % version)
testChanges(version, changesURL)
for artifact, urlString in artifacts:
print(' download %s...' % artifact)
scriptutil.download(artifact, urlString, tmpDir, force_clean=FORCE_CLEAN)
verifyDigests(artifact, urlString, tmpDir)
if isSigned:
print(' verify sig')
# Test sig (this is done with a clean brand-new GPG world)
scriptutil.download(artifact + '.asc', urlString + '.asc', tmpDir, force_clean=FORCE_CLEAN)
sigFile = '%s/%s.asc' % (tmpDir, artifact)
artifactFile = '%s/%s' % (tmpDir, artifact)
logFile = '%s/lucene.%s.gpg.verify.log' % (tmpDir, artifact)
run('gpg --homedir %s --display-charset utf-8 --verify %s %s' % (gpgHomeDir, sigFile, artifactFile),
logFile)
# Forward any GPG warnings, except the expected one (since it's a clean world)
with open(logFile) as f:
print("File: %s" % logFile)
for line in f.readlines():
if line.lower().find('warning') != -1 \
and line.find('WARNING: This key is not certified with a trusted signature') == -1:
print(' GPG: %s' % line.strip())
# Test trust (this is done with the real users config)
run('gpg --import %s' % (keysFile),
'%s/lucene.gpg.trust.import.log' % tmpDir)
print(' verify trust')
logFile = '%s/lucene.%s.gpg.trust.log' % (tmpDir, artifact)
run('gpg --display-charset utf-8 --verify %s %s' % (sigFile, artifactFile), logFile)
# Forward any GPG warnings:
with open(logFile) as f:
for line in f.readlines():
if line.lower().find('warning') != -1:
print(' GPG: %s' % line.strip())
def testChanges(version, changesURLString):
print(' check changes HTML...')
changesURL = None
for text, subURL in getDirEntries(changesURLString):
if text == 'Changes.html':
changesURL = subURL
if changesURL is None:
raise RuntimeError('did not see Changes.html link from %s' % changesURLString)
s = load(changesURL)
checkChangesContent(s, version, changesURL, True)
def testChangesText(dir, version):
"Checks all CHANGES.txt under this dir."
for root, dirs, files in os.walk(dir):
# NOTE: O(N) but N should be smallish:
if 'CHANGES.txt' in files:
fullPath = '%s/CHANGES.txt' % root
#print 'CHECK %s' % fullPath
checkChangesContent(open(fullPath, encoding='UTF-8').read(), version, fullPath, False)
reChangesSectionHREF = re.compile('<a id="(.*?)".*?>(.*?)</a>', re.IGNORECASE)
reUnderbarNotDashHTML = re.compile(r'<li>(\s*(LUCENE)_\d\d\d\d+)')
reUnderbarNotDashTXT = re.compile(r'\s+((LUCENE)_\d\d\d\d+)', re.MULTILINE)
def checkChangesContent(s, version, name, isHTML):
currentVersionTuple = versionToTuple(version, name)
if isHTML and s.find('Release %s' % version) == -1:
raise RuntimeError('did not see "Release %s" in %s' % (version, name))
if isHTML:
r = reUnderbarNotDashHTML
else:
r = reUnderbarNotDashTXT
m = r.search(s)
if m is not None:
raise RuntimeError('incorrect issue (_ instead of -) in %s: %s' % (name, m.group(1)))
if s.lower().find('not yet released') != -1:
raise RuntimeError('saw "not yet released" in %s' % name)
if not isHTML:
sub = 'Lucene %s' % version
if s.find(sub) == -1:
# benchmark never seems to include release info:
if name.find('/benchmark/') == -1:
raise RuntimeError('did not see "%s" in %s' % (sub, name))
if isHTML:
# Make sure that a section only appears once under each release,
# and that each release is not greater than the current version
seenIDs = set()
seenText = set()
release = None
for id, text in reChangesSectionHREF.findall(s):
if text.lower().startswith('release '):
release = text[8:].strip()
seenText.clear()
releaseTuple = versionToTuple(release, name)
if releaseTuple > currentVersionTuple:
raise RuntimeError('Future release %s is greater than %s in %s' % (release, version, name))
if id in seenIDs:
raise RuntimeError('%s has duplicate section "%s" under release "%s"' % (name, text, release))
seenIDs.add(id)
if text in seenText:
raise RuntimeError('%s has duplicate section "%s" under release "%s"' % (name, text, release))
seenText.add(text)
reVersion = re.compile(r'(\d+)\.(\d+)(?:\.(\d+))?\s*(-alpha|-beta|final|RC\d+)?\s*(?:\[.*\])?', re.IGNORECASE)
def versionToTuple(version, name):
versionMatch = reVersion.match(version)
if versionMatch is None:
raise RuntimeError('Version %s in %s cannot be parsed' % (version, name))
versionTuple = versionMatch.groups()
while versionTuple[-1] is None or versionTuple[-1] == '':
versionTuple = versionTuple[:-1]
if versionTuple[-1].lower() == '-alpha':
versionTuple = versionTuple[:-1] + ('0',)
elif versionTuple[-1].lower() == '-beta':
versionTuple = versionTuple[:-1] + ('1',)
elif versionTuple[-1].lower() == 'final':
versionTuple = versionTuple[:-2] + ('100',)
elif versionTuple[-1].lower()[:2] == 'rc':
versionTuple = versionTuple[:-2] + (versionTuple[-1][2:],)
return tuple(int(x) if x is not None and x.isnumeric() else x for x in versionTuple)
reUnixPath = re.compile(r'\b[a-zA-Z_]+=(?:"(?:\\"|[^"])*"' + '|(?:\\\\.|[^"\'\\s])*' + r"|'(?:\\'|[^'])*')" \
+ r'|(/(?:\\.|[^"\'\s])*)' \
+ r'|("/(?:\\.|[^"])*")' \
+ r"|('/(?:\\.|[^'])*')")
def unix2win(matchobj):
if matchobj.group(1) is not None: return cygwinWindowsRoot + matchobj.group()
if matchobj.group(2) is not None: return '"%s%s' % (cygwinWindowsRoot, matchobj.group().lstrip('"'))
if matchobj.group(3) is not None: return "'%s%s" % (cygwinWindowsRoot, matchobj.group().lstrip("'"))
return matchobj.group()
def cygwinifyPaths(command):
# The problem: Native Windows applications running under Cygwin can't
# handle Cygwin's Unix-style paths. However, environment variable
# values are automatically converted, so only paths outside of
# environment variable values should be converted to Windows paths.
# Assumption: all paths will be absolute.
if '; gradlew ' in command: command = reUnixPath.sub(unix2win, command)
return command
def printFileContents(fileName):
# Assume log file was written in system's default encoding, but
# even if we are wrong, we replace errors ... the ASCII chars
# (which is what we mostly care about eg for the test seed) should
# still survive:
txt = codecs.open(fileName, 'r', encoding=sys.getdefaultencoding(), errors='replace').read()
# Encode to our output encoding (likely also system's default
# encoding):
bytes = txt.encode(sys.stdout.encoding, errors='replace')
# Decode back to string and print... we should hit no exception here
# since all errors have been replaced:
print(codecs.getdecoder(sys.stdout.encoding)(bytes)[0])
print()
def run(command, logFile):
if cygwin: command = cygwinifyPaths(command)
if os.system('%s > %s 2>&1' % (command, logFile)):
logPath = os.path.abspath(logFile)
print('\ncommand "%s" failed:' % command)
printFileContents(logFile)
raise RuntimeError('command "%s" failed; see log file %s' % (command, logPath))
def verifyDigests(artifact, urlString, tmpDir):
print(' verify sha512 digest')
sha512Expected, t = load(urlString + '.sha512').strip().split()
if t != '*'+artifact:
raise RuntimeError('SHA512 %s.sha512 lists artifact %s but expected *%s' % (urlString, t, artifact))
s512 = hashlib.sha512()
f = open('%s/%s' % (tmpDir, artifact), 'rb')
while True:
x = f.read(65536)
if len(x) == 0:
break
s512.update(x)
f.close()
sha512Actual = s512.hexdigest()
if sha512Actual != sha512Expected:
raise RuntimeError('SHA512 digest mismatch for %s: expected %s but got %s' % (artifact, sha512Expected, sha512Actual))
def getDirEntries(urlString):
if urlString.startswith('file:/') and not urlString.startswith('file://'):
# stupid bogus ant URI
urlString = "file:///" + urlString[6:]
if urlString.startswith('file://'):
path = urlString[7:]
if path.endswith('/'):
path = path[:-1]
if cygwin: # Convert Windows path to Cygwin path
path = re.sub(r'^/([A-Za-z]):/', r'/cygdrive/\1/', path)
l = []
for ent in os.listdir(path):
entPath = '%s/%s' % (path, ent)
if os.path.isdir(entPath):
entPath += '/'
ent += '/'
l.append((ent, 'file://%s' % entPath))
l.sort()
return l
else:
links = getHREFs(urlString)
for i, (text, subURL) in enumerate(links):
if text == 'Parent Directory' or text == '..':
return links[(i+1):]
def unpackAndVerify(java, tmpDir, artifact, gitRevision, version, testArgs):
destDir = '%s/unpack' % tmpDir
if os.path.exists(destDir):
shutil.rmtree(destDir)
os.makedirs(destDir)
os.chdir(destDir)
print(' unpack %s...' % artifact)
unpackLogFile = '%s/lucene-unpack-%s.log' % (tmpDir, artifact)
if artifact.endswith('.tar.gz') or artifact.endswith('.tgz'):
run('tar xzf %s/%s' % (tmpDir, artifact), unpackLogFile)
elif artifact.endswith('.zip'):
run('unzip %s/%s' % (tmpDir, artifact), unpackLogFile)
# make sure it unpacks to proper subdir
l = os.listdir(destDir)
expected = 'lucene-%s' % version
if l != [expected]:
raise RuntimeError('unpack produced entries %s; expected only %s' % (l, expected))
unpackPath = '%s/%s' % (destDir, expected)
verifyUnpacked(java, artifact, unpackPath, gitRevision, version, testArgs)
return unpackPath
LUCENE_NOTICE = None
LUCENE_LICENSE = None
def is_in_list(in_folder, files, indent=4):
for fileName in files:
print("%sChecking %s" % (" "*indent, fileName))
found = False
for f in [fileName, fileName + '.txt', fileName + '.md']:
if f in in_folder:
in_folder.remove(f)
found = True
if not found:
raise RuntimeError('file "%s" is missing' % fileName)
def verifyUnpacked(java, artifact, unpackPath, gitRevision, version, testArgs):
global LUCENE_NOTICE
global LUCENE_LICENSE
os.chdir(unpackPath)
isSrc = artifact.find('-src') != -1
# Check text files in release
print(" %s" % artifact)
in_root_folder = list(filter(lambda x: x[0] != '.', os.listdir(unpackPath)))
in_lucene_folder = []
if isSrc:
in_lucene_folder.extend(os.listdir(os.path.join(unpackPath, 'lucene')))
is_in_list(in_root_folder, ['LICENSE', 'NOTICE', 'README'])
is_in_list(in_lucene_folder, ['JRE_VERSION_MIGRATION', 'CHANGES', 'MIGRATE', 'SYSTEM_REQUIREMENTS'])
else:
is_in_list(in_root_folder, ['LICENSE', 'NOTICE', 'README', 'JRE_VERSION_MIGRATION', 'CHANGES',
'MIGRATE', 'SYSTEM_REQUIREMENTS'])
if LUCENE_NOTICE is None:
LUCENE_NOTICE = open('%s/NOTICE.txt' % unpackPath, encoding='UTF-8').read()
if LUCENE_LICENSE is None:
LUCENE_LICENSE = open('%s/LICENSE.txt' % unpackPath, encoding='UTF-8').read()
# if not isSrc:
# # TODO: we should add verifyModule/verifySubmodule (e.g. analysis) here and recurse through
# expectedJARs = ()
#
# for fileName in expectedJARs:
# fileName += '.jar'
# if fileName not in l:
# raise RuntimeError('lucene: file "%s" is missing from artifact %s' % (fileName, artifact))
# in_root_folder.remove(fileName)
expected_folders = ['analysis', 'backward-codecs', 'benchmark', 'classification', 'codecs', 'core',
'demo', 'expressions', 'facet', 'grouping', 'highlighter', 'join',
'luke', 'memory', 'misc', 'monitor', 'queries', 'queryparser', 'replicator',
'sandbox', 'spatial-extras', 'spatial3d', 'suggest', 'test-framework', 'licenses']
if isSrc:
expected_src_root_files = ['build.gradle', 'buildSrc', 'dev-docs', 'dev-tools', 'gradle', 'gradlew',
'gradlew.bat', 'help', 'lucene', 'settings.gradle', 'versions.lock', 'versions.props']
expected_src_lucene_files = ['build.gradle', 'documentation', 'distribution', 'dev-docs']
is_in_list(in_root_folder, expected_src_root_files)
is_in_list(in_lucene_folder, expected_folders)
is_in_list(in_lucene_folder, expected_src_lucene_files)
if len(in_lucene_folder) > 0:
raise RuntimeError('lucene: unexpected files/dirs in artifact %s lucene/ folder: %s' % (artifact, in_lucene_folder))
else:
is_in_list(in_root_folder, ['bin', 'docs', 'licenses', 'modules', 'modules-thirdparty'])
if len(in_root_folder) > 0:
raise RuntimeError('lucene: unexpected files/dirs in artifact %s: %s' % (artifact, in_root_folder))
if isSrc:
print(' make sure no JARs/WARs in src dist...')
lines = os.popen('find . -name \\*.jar').readlines()
if len(lines) != 0:
print(' FAILED:')
for line in lines:
print(' %s' % line.strip())
raise RuntimeError('source release has JARs...')
lines = os.popen('find . -name \\*.war').readlines()
if len(lines) != 0:
print(' FAILED:')
for line in lines:
print(' %s' % line.strip())
raise RuntimeError('source release has WARs...')
validateCmd = './gradlew --no-daemon check -p lucene/documentation'
print(' run "%s"' % validateCmd)
java.run_java11(validateCmd, '%s/validate.log' % unpackPath)
print(" run tests w/ Java 11 and testArgs='%s'..." % testArgs)
java.run_java11('./gradlew --no-daemon test %s' % testArgs, '%s/test.log' % unpackPath)
print(" compile jars w/ Java 11")
java.run_java11('./gradlew --no-daemon jar -Dversion.release=%s' % version, '%s/compile.log' % unpackPath)
testDemo(java.run_java11, isSrc, version, '11')
if java.run_java17:
print(" run tests w/ Java 17 and testArgs='%s'..." % testArgs)
java.run_java17('./gradlew --no-daemon test %s' % testArgs, '%s/test.log' % unpackPath)
print(" compile jars w/ Java 17")
java.run_java17('./gradlew --no-daemon jar -Dversion.release=%s' % version, '%s/compile.log' % unpackPath)
testDemo(java.run_java17, isSrc, version, '17')
print(' confirm all releases have coverage in TestBackwardsCompatibility')
confirmAllReleasesAreTestedForBackCompat(version, unpackPath)
else:
checkAllJARs(os.getcwd(), gitRevision, version)
testDemo(java.run_java11, isSrc, version, '11')
if java.run_java17:
testDemo(java.run_java17, isSrc, version, '17')
testChangesText('.', version)
def testDemo(run_java, isSrc, version, jdk):
if os.path.exists('index'):
shutil.rmtree('index') # nuke any index from any previous iteration
print(' test demo with %s...' % jdk)
sep = ';' if cygwin else ':'
if isSrc:
# For source release, use the classpath for each module.
classPath = ['lucene/core/build/libs/lucene-core-%s.jar' % version,
'lucene/demo/build/libs/lucene-demo-%s.jar' % version,
'lucene/analysis/common/build/libs/lucene-analyzers-common-%s.jar' % version,
'lucene/queryparser/build/libs/lucene-queryparser-%s.jar' % version]
cp = sep.join(classPath)
docsDir = 'lucene/core/src'
checkIndexCmd = 'java -ea -cp "%s" org.apache.lucene.index.CheckIndex index' % cp
indexFilesCmd = 'java -cp "%s" -Dsmoketester=true org.apache.lucene.demo.IndexFiles -index index -docs %s' % (cp, docsDir)
searchFilesCmd = 'java -cp "%s" org.apache.lucene.demo.SearchFiles -index index -query lucene' % cp
else:
# For binary release, set up classpath as modules.
cp = "--module-path modules"
docsDir = 'docs'
checkIndexCmd = 'java -ea %s --module org.apache.lucene.core/org.apache.lucene.index.CheckIndex index' % cp
indexFilesCmd = 'java -Dsmoketester=true %s --module org.apache.lucene.demo/org.apache.lucene.demo.IndexFiles -index index -docs %s' % (cp, docsDir)
searchFilesCmd = 'java %s --module org.apache.lucene.demo/org.apache.lucene.demo.SearchFiles -index index -query lucene' % cp
run_java(indexFilesCmd, 'index.log')
run_java(searchFilesCmd, 'search.log')
reMatchingDocs = re.compile('(\d+) total matching documents')
m = reMatchingDocs.search(open('search.log', encoding='UTF-8').read())
if m is None:
raise RuntimeError('lucene demo\'s SearchFiles found no results')
else:
numHits = int(m.group(1))
if numHits < 100:
raise RuntimeError('lucene demo\'s SearchFiles found too few results: %s' % numHits)
print(' got %d hits for query "lucene"' % numHits)
print(' checkindex with %s...' % jdk)
run_java(checkIndexCmd, 'checkindex.log')
s = open('checkindex.log').read()
m = re.search(r'^\s+version=(.*?)$', s, re.MULTILINE)
if m is None:
raise RuntimeError('unable to locate version=NNN output from CheckIndex; see checkindex.log')
actualVersion = m.group(1)
if removeTrailingZeros(actualVersion) != removeTrailingZeros(version):
raise RuntimeError('wrong version from CheckIndex: got "%s" but expected "%s"' % (actualVersion, version))
def removeTrailingZeros(version):
return re.sub(r'(\.0)*$', '', version)
def checkMaven(baseURL, tmpDir, gitRevision, version, isSigned, keysFile):
print(' download artifacts')
artifacts = []
artifactsURL = '%s/lucene/maven/org/apache/lucene/' % baseURL
targetDir = '%s/maven/org/apache/lucene' % tmpDir
if not os.path.exists(targetDir):
os.makedirs(targetDir)
crawl(artifacts, artifactsURL, targetDir)
print()
verifyPOMperBinaryArtifact(artifacts, version)
verifyMavenDigests(artifacts)
checkJavadocAndSourceArtifacts(artifacts, version)
verifyDeployedPOMsCoordinates(artifacts, version)
if isSigned:
verifyMavenSigs(tmpDir, artifacts, keysFile)
distFiles = getBinaryDistFiles(tmpDir, version, baseURL)
checkIdenticalMavenArtifacts(distFiles, artifacts, version)
checkAllJARs('%s/maven/org/apache/lucene' % tmpDir, gitRevision, version)
def getBinaryDistFiles(tmpDir, version, baseURL):
distribution = 'lucene-%s.tgz' % version
if not os.path.exists('%s/%s' % (tmpDir, distribution)):
distURL = '%s/lucene/%s' % (baseURL, distribution)
print(' download %s...' % distribution, end=' ')
scriptutil.download(distribution, distURL, tmpDir, force_clean=FORCE_CLEAN)
destDir = '%s/unpack-lucene-getBinaryDistFiles' % tmpDir
if os.path.exists(destDir):
shutil.rmtree(destDir)
os.makedirs(destDir)
os.chdir(destDir)
print(' unpack %s...' % distribution)
unpackLogFile = '%s/unpack-%s-getBinaryDistFiles.log' % (tmpDir, distribution)
run('tar xzf %s/%s' % (tmpDir, distribution), unpackLogFile)
distributionFiles = []
for root, dirs, files in os.walk(destDir):
distributionFiles.extend([os.path.join(root, file) for file in files])
return distributionFiles
def checkJavadocAndSourceArtifacts(artifacts, version):
print(' check for javadoc and sources artifacts...')
for artifact in artifacts:
if artifact.endswith(version + '.jar'):
javadocJar = artifact[:-4] + '-javadoc.jar'
if javadocJar not in artifacts:
raise RuntimeError('missing: %s' % javadocJar)
sourcesJar = artifact[:-4] + '-sources.jar'
if sourcesJar not in artifacts:
raise RuntimeError('missing: %s' % sourcesJar)
def getZipFileEntries(fileName):
entries = []
with zipfile.ZipFile(fileName) as zf:
for zi in zf.infolist():
entries.append(zi.filename)
# Sort by name:
entries.sort()
return entries
def checkIdenticalMavenArtifacts(distFiles, artifacts, version):
print(' verify that Maven artifacts are same as in the binary distribution...')
reJarWar = re.compile(r'%s\.[wj]ar$' % version) # exclude *-javadoc.jar and *-sources.jar
distFilenames = dict()
for file in distFiles:
baseName = os.path.basename(file)
distFilenames[baseName] = file
for artifact in artifacts:
if reJarWar.search(artifact):
artifactFilename = os.path.basename(artifact)
if artifactFilename not in distFilenames:
raise RuntimeError('Maven artifact %s is not present in lucene binary distribution' % artifact)
else:
identical = filecmp.cmp(artifact, distFilenames[artifactFilename], shallow=False)
if not identical:
raise RuntimeError('Maven artifact %s is not identical to %s in lucene binary distribution'
% (artifact, distFilenames[artifactFilename]))
def verifyMavenDigests(artifacts):
print(" verify Maven artifacts' md5/sha1 digests...")
reJarWarPom = re.compile(r'\.(?:[wj]ar|pom)$')
for artifactFile in [a for a in artifacts if reJarWarPom.search(a)]:
if artifactFile + '.md5' not in artifacts:
raise RuntimeError('missing: MD5 digest for %s' % artifactFile)
if artifactFile + '.sha1' not in artifacts:
raise RuntimeError('missing: SHA1 digest for %s' % artifactFile)
with open(artifactFile + '.md5', encoding='UTF-8') as md5File:
md5Expected = md5File.read().strip()
with open(artifactFile + '.sha1', encoding='UTF-8') as sha1File:
sha1Expected = sha1File.read().strip()
md5 = hashlib.md5()
sha1 = hashlib.sha1()
inputFile = open(artifactFile, 'rb')
while True:
bytes = inputFile.read(65536)
if len(bytes) == 0:
break
md5.update(bytes)
sha1.update(bytes)
inputFile.close()
md5Actual = md5.hexdigest()
sha1Actual = sha1.hexdigest()
if md5Actual != md5Expected:
raise RuntimeError('MD5 digest mismatch for %s: expected %s but got %s'
% (artifactFile, md5Expected, md5Actual))
if sha1Actual != sha1Expected:
raise RuntimeError('SHA1 digest mismatch for %s: expected %s but got %s'
% (artifactFile, sha1Expected, sha1Actual))
def getPOMcoordinate(treeRoot):
namespace = '{http://maven.apache.org/POM/4.0.0}'
groupId = treeRoot.find('%sgroupId' % namespace)
if groupId is None:
groupId = treeRoot.find('{0}parent/{0}groupId'.format(namespace))
groupId = groupId.text.strip()
artifactId = treeRoot.find('%sartifactId' % namespace).text.strip()
version = treeRoot.find('%sversion' % namespace)
if version is None:
version = treeRoot.find('{0}parent/{0}version'.format(namespace))
version = version.text.strip()
packaging = treeRoot.find('%spackaging' % namespace)
packaging = 'jar' if packaging is None else packaging.text.strip()
return groupId, artifactId, packaging, version
def verifyMavenSigs(tmpDir, artifacts, keysFile):
print(' verify maven artifact sigs', end=' ')
# Set up clean gpg world; import keys file:
gpgHomeDir = '%s/lucene.gpg' % tmpDir
if os.path.exists(gpgHomeDir):
shutil.rmtree(gpgHomeDir)
os.makedirs(gpgHomeDir, 0o700)
run('gpg --homedir %s --import %s' % (gpgHomeDir, keysFile),
'%s/lucene.gpg.import.log' % tmpDir)
reArtifacts = re.compile(r'\.(?:pom|[jw]ar)$')
for artifactFile in [a for a in artifacts if reArtifacts.search(a)]:
artifact = os.path.basename(artifactFile)
sigFile = '%s.asc' % artifactFile
# Test sig (this is done with a clean brand-new GPG world)
logFile = '%s/lucene.%s.gpg.verify.log' % (tmpDir, artifact)
run('gpg --display-charset utf-8 --homedir %s --verify %s %s' % (gpgHomeDir, sigFile, artifactFile),
logFile)
# Forward any GPG warnings, except the expected one (since it's a clean world)
print_warnings_in_file(logFile)
# Test trust (this is done with the real users config)
run('gpg --import %s' % keysFile,
'%s/lucene.gpg.trust.import.log' % tmpDir)
logFile = '%s/lucene.%s.gpg.trust.log' % (tmpDir, artifact)
run('gpg --display-charset utf-8 --verify %s %s' % (sigFile, artifactFile), logFile)
# Forward any GPG warnings:
print_warnings_in_file(logFile)
sys.stdout.write('.')
print()
def print_warnings_in_file(file):
with open(file) as f:
for line in f.readlines():
if line.lower().find('warning') != -1 \
and line.find('WARNING: This key is not certified with a trusted signature') == -1 \
and line.find('WARNING: using insecure memory') == -1:
print(' GPG: %s' % line.strip())
def verifyPOMperBinaryArtifact(artifacts, version):
print(' verify that each binary artifact has a deployed POM...')
reBinaryJarWar = re.compile(r'%s\.[jw]ar$' % re.escape(version))
for artifact in [a for a in artifacts if reBinaryJarWar.search(a)]:
POM = artifact[:-4] + '.pom'
if POM not in artifacts:
raise RuntimeError('missing: POM for %s' % artifact)
def verifyDeployedPOMsCoordinates(artifacts, version):
"""
verify that each POM's coordinate (drawn from its content) matches
its filepath, and verify that the corresponding artifact exists.
"""
print(" verify deployed POMs' coordinates...")
for POM in [a for a in artifacts if a.endswith('.pom')]:
treeRoot = ET.parse(POM).getroot()
groupId, artifactId, packaging, POMversion = getPOMcoordinate(treeRoot)
POMpath = '%s/%s/%s/%s-%s.pom' \
% (groupId.replace('.', '/'), artifactId, version, artifactId, version)
if not POM.endswith(POMpath):
raise RuntimeError("Mismatch between POM coordinate %s:%s:%s and filepath: %s"
% (groupId, artifactId, POMversion, POM))
# Verify that the corresponding artifact exists
artifact = POM[:-3] + packaging
if artifact not in artifacts:
raise RuntimeError('Missing corresponding .%s artifact for POM %s' % (packaging, POM))
def crawl(downloadedFiles, urlString, targetDir, exclusions=set()):
for text, subURL in getDirEntries(urlString):
if text not in exclusions:
path = os.path.join(targetDir, text)
if text.endswith('/'):
if not os.path.exists(path):
os.makedirs(path)
crawl(downloadedFiles, subURL, path, exclusions)
else:
if not os.path.exists(path) or FORCE_CLEAN:
scriptutil.download(text, subURL, targetDir, quiet=True, force_clean=FORCE_CLEAN)
downloadedFiles.append(path)
sys.stdout.write('.')
def make_java_config(parser, java17_home):
def _make_runner(java_home, version):
print('Java %s JAVA_HOME=%s' % (version, java_home))
if cygwin:
java_home = subprocess.check_output('cygpath -u "%s"' % java_home, shell=True).decode('utf-8').strip()
cmd_prefix = 'export JAVA_HOME="%s" PATH="%s/bin:$PATH" JAVACMD="%s/bin/java"' % \
(java_home, java_home, java_home)
s = subprocess.check_output('%s; java -version' % cmd_prefix,
shell=True, stderr=subprocess.STDOUT).decode('utf-8')
if s.find(' version "%s' % version) == -1:
parser.error('got wrong version for java %s:\n%s' % (version, s))
def run_java(cmd, logfile):
run('%s; %s' % (cmd_prefix, cmd), logfile)
return run_java
java11_home = os.environ.get('JAVA_HOME')
if java11_home is None:
parser.error('JAVA_HOME must be set')
run_java11 = _make_runner(java11_home, '11')
run_java17 = None
if java17_home is not None:
run_java17 = _make_runner(java17_home, '17')
jc = namedtuple('JavaConfig', 'run_java11 java11_home run_java17 java17_home')
return jc(run_java11, java11_home, run_java17, java17_home)
version_re = re.compile(r'(\d+\.\d+\.\d+(-ALPHA|-BETA)?)')
revision_re = re.compile(r'rev-([a-f\d]+)')
def parse_config():
epilogue = textwrap.dedent('''
Example usage:
python3 -u dev-tools/scripts/smokeTestRelease.py https://dist.apache.org/repos/dist/dev/lucene/lucene-9.0.0-RC1-rev-c7510a0...
''')
description = 'Utility to test a release.'
parser = argparse.ArgumentParser(description=description, epilog=epilogue,
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument('--tmp-dir', metavar='PATH',
help='Temporary directory to test inside, defaults to /tmp/smoke_lucene_$version_$revision')
parser.add_argument('--not-signed', dest='is_signed', action='store_false', default=True,
help='Indicates the release is not signed')
parser.add_argument('--local-keys', metavar='PATH',
help='Uses local KEYS file instead of fetching from https://archive.apache.org/dist/lucene/KEYS')
parser.add_argument('--revision',
help='GIT revision number that release was built with, defaults to that in URL')
parser.add_argument('--version', metavar='X.Y.Z(-ALPHA|-BETA)?',
help='Version of the release, defaults to that in URL')
parser.add_argument('--test-java17', metavar='java17_home',
help='Path to Java17 home directory, to run tests with if specified')
parser.add_argument('--download-only', action='store_true', default=False,
help='Only perform download and sha hash check steps')
parser.add_argument('url', help='Url pointing to release to test')
parser.add_argument('test_args', nargs=argparse.REMAINDER,
help='Arguments to pass to gradle for testing, e.g. -Dwhat=ever.')
c = parser.parse_args()
if c.version is not None:
if not version_re.match(c.version):
parser.error('version "%s" does not match format X.Y.Z[-ALPHA|-BETA]' % c.version)
else:
version_match = version_re.search(c.url)
if version_match is None:
parser.error('Could not find version in URL')
c.version = version_match.group(1)
if c.revision is None:
revision_match = revision_re.search(c.url)
if revision_match is None:
parser.error('Could not find revision in URL')
c.revision = revision_match.group(1)
print('Revision: %s' % c.revision)
if c.local_keys is not None and not os.path.exists(c.local_keys):
parser.error('Local KEYS file "%s" not found' % c.local_keys)
c.java = make_java_config(parser, c.test_java17)
if c.tmp_dir:
c.tmp_dir = os.path.abspath(c.tmp_dir)
else:
tmp = '/tmp/smoke_lucene_%s_%s' % (c.version, c.revision)
c.tmp_dir = tmp
i = 1
while os.path.exists(c.tmp_dir):
c.tmp_dir = tmp + '_%d' % i
i += 1
return c
reVersion1 = re.compile(r'\>(\d+)\.(\d+)\.(\d+)(-alpha|-beta)?/\<', re.IGNORECASE)
reVersion2 = re.compile(r'-(\d+)\.(\d+)\.(\d+)(-alpha|-beta)?\.', re.IGNORECASE)
def getAllLuceneReleases():
s = load('https://archive.apache.org/dist/lucene/java')
releases = set()
for r in reVersion1, reVersion2:
for tup in r.findall(s):
if tup[-1].lower() == '-alpha':
tup = tup[:3] + ('0',)
elif tup[-1].lower() == '-beta':
tup = tup[:3] + ('1',)
elif tup[-1] == '':
tup = tup[:3]
else:
raise RuntimeError('failed to parse version: %s' % tup[-1])
releases.add(tuple(int(x) for x in tup))
l = list(releases)
l.sort()
return l
def confirmAllReleasesAreTestedForBackCompat(smokeVersion, unpackPath):
print(' find all past Lucene releases...')
allReleases = getAllLuceneReleases()
#for tup in allReleases:
# print(' %s' % '.'.join(str(x) for x in tup))
testedIndices = set()
os.chdir(unpackPath)
print(' run TestBackwardsCompatibility..')
command = './gradlew --no-daemon test -p lucene/backward-codecs --tests TestBackwardsCompatibility --max-workers=1 ' \
'-Dtests.verbose=true '
p = subprocess.Popen(command, shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
stdout, stderr = p.communicate()
if p.returncode != 0:
# Not good: the test failed!
raise RuntimeError('%s failed:\n%s' % (command, stdout))
stdout = stdout.decode('utf-8',errors='replace').replace('\r\n','\n')
if stderr is not None:
# Should not happen since we redirected stderr to stdout:
raise RuntimeError('stderr non-empty')
reIndexName = re.compile(r'TEST: index[\s*=\s*](.*?)(-cfs|-nocfs)$', re.MULTILINE)
for name, cfsPart in reIndexName.findall(stdout):
# Fragile: decode the inconsistent naming schemes we've used in TestBWC's indices:
#print('parse name %s' % name)
tup = tuple(name.split('.'))
if len(tup) == 3:
# ok
tup = tuple(int(x) for x in tup)
elif tup == ('4', '0', '0', '1'):
# CONFUSING: this is the 4.0.0-alpha index??
tup = 4, 0, 0, 0
elif tup == ('4', '0', '0', '2'):
# CONFUSING: this is the 4.0.0-beta index??
tup = 4, 0, 0, 1
elif name == '5x-with-4x-segments':
# Mixed version test case; ignore it for our purposes because we only
# tally up the "tests single Lucene version" indices
continue
elif name == '5.0.0.singlesegment':
tup = 5, 0, 0
else:
raise RuntimeError('could not parse version %s' % name)
testedIndices.add(tup)
l = list(testedIndices)
l.sort()
if False:
for release in l:
print(' %s' % '.'.join(str(x) for x in release))
allReleases = set(allReleases)
for x in testedIndices:
if x not in allReleases:
# Curious: we test 1.9.0 index but it's not in the releases (I think it was pulled because of nasty bug?)
if x != (1, 9, 0):
raise RuntimeError('tested version=%s but it was not released?' % '.'.join(str(y) for y in x))
notTested = []
for x in allReleases:
if x not in testedIndices:
releaseVersion = '.'.join(str(y) for y in x)
if releaseVersion in ('1.4.3', '1.9.1', '2.3.1', '2.3.2'):
# Exempt the dark ages indices
continue
if x >= tuple(int(y) for y in smokeVersion.split('.')):
# Exempt versions not less than the one being smoke tested
print(' Backcompat testing not required for release %s because it\'s not less than %s'
% (releaseVersion, smokeVersion))
continue
notTested.append(x)
if len(notTested) > 0:
notTested.sort()
print('Releases that don\'t seem to be tested:')
failed = True
for x in notTested:
print(' %s' % '.'.join(str(y) for y in x))
raise RuntimeError('some releases are not tested by TestBackwardsCompatibility?')
else:
print(' success!')
def main():
c = parse_config()
# Pick <major>.<minor> part of version and require script to be from same branch
scriptVersion = re.search(r'((\d+).(\d+)).(\d+)', scriptutil.find_current_version()).group(1).strip()
if not c.version.startswith(scriptVersion + '.'):
raise RuntimeError('smokeTestRelease.py for %s.X is incompatible with a %s release.' % (scriptVersion, c.version))
print('NOTE: output encoding is %s' % sys.stdout.encoding)
smokeTest(c.java, c.url, c.revision, c.version, c.tmp_dir, c.is_signed, c.local_keys, ' '.join(c.test_args),
downloadOnly=c.download_only)
def smokeTest(java, baseURL, gitRevision, version, tmpDir, isSigned, local_keys, testArgs, downloadOnly=False):
startTime = datetime.datetime.now()
# Tests annotated @Nightly are more resource-intensive but often cover
# important code paths. They're disabled by default to preserve a good
# developer experience, but we enable them for smoke tests where we want good
# coverage.
testArgs = '-Dtests.nigthly=true %s' % testArgs
if FORCE_CLEAN:
if os.path.exists(tmpDir):
raise RuntimeError('temp dir %s exists; please remove first' % tmpDir)
if not os.path.exists(tmpDir):
os.makedirs(tmpDir)
lucenePath = None
print()
print('Load release URL "%s"...' % baseURL)
newBaseURL = unshortenURL(baseURL)
if newBaseURL != baseURL:
print(' unshortened: %s' % newBaseURL)
baseURL = newBaseURL
for text, subURL in getDirEntries(baseURL):
if text.lower().find('lucene') != -1:
lucenePath = subURL
if lucenePath is None:
raise RuntimeError('could not find lucene subdir')
print()
print('Get KEYS...')
if local_keys is not None:
print(" Using local KEYS file %s" % local_keys)
keysFile = local_keys
else:
keysFileURL = "https://archive.apache.org/dist/lucene/KEYS"
print(" Downloading online KEYS file %s" % keysFileURL)
scriptutil.download('KEYS', keysFileURL, tmpDir, force_clean=FORCE_CLEAN)
keysFile = '%s/KEYS' % (tmpDir)
print()
print('Test Lucene...')
checkSigs(lucenePath, version, tmpDir, isSigned, keysFile)
if not downloadOnly:
unpackAndVerify(java, tmpDir, 'lucene-%s.tgz' % version, gitRevision, version, testArgs)
unpackAndVerify(java, tmpDir, 'lucene-%s-src.tgz' % version, gitRevision, version, testArgs)
print()
print('Test Maven artifacts...')
checkMaven(baseURL, tmpDir, gitRevision, version, isSigned, keysFile)
else:
print("\nLucene test done (--download-only specified)")
print('\nSUCCESS! [%s]\n' % (datetime.datetime.now() - startTime))
if __name__ == '__main__':
try:
main()
except KeyboardInterrupt:
print('Keyboard interrupt...exiting')
Reference in New Issue View Git Blame Copy Permalink