/
githubmirror
/
cmssw
Обзор
Документация
Войти
/
githubmirror
/
cmssw
Код
Запросы
0
Пакеты
0
Релизы
0
Аналитика
Безопасность
master
DQMServices/Components/scripts/dqm-ls
275 строк
9 KB
Marco Rovere
Fix dqm-* scripts for python3.
07 июн 2023, 17:54
Не верифицирован
07 июн 2023, 17:54
4cb474b
Код
Авторство
О чём код?
#!/usr/bin/env python3 """Usage: dqm-ls [-s SERVER] [-n CONNECTIONS] Parse ROOT file contents listings on a DQM GUI server. In order to authenticate to the target server, standard grid certificate environment must be available. Typically this would be X509_CERT_DIR and either X509_USER_PROXY or X509_USER_CERT and X509_USER_KEY environment variables. If these variables are not set, the following defaults are checked for existence. Note that if the script falls back on using a key rather than a proxy, it will prompt for the key password. - $X509_CERT_DIR: /etc/grid-security/certificates - $X509_USER_KEY: $HOME/.globus/userkey.pem - $X509_USER_CERT: $HOME/.globus/usercert.pem """ from DQMServices.Components.HTTP import RequestManager from DQMServices.Components.X509 import SSLOptions import os, sys, re, pycurl, urllib from optparse import OptionParser from time import time, strptime, strftime, gmtime from calendar import timegm from datetime import datetime from urllib.parse import urlparse, quote from tempfile import mkstemp from traceback import print_exc from functools import cmp_to_key # Object types. DIR = 0 FILE = 1 # HTTP protocol `User-agent` identification string. ident = "DQMLS/1.0 python/%s.%s.%s" % sys.version_info[:3] # SSL/X509 options. ssl_opts = None # HTTP request manager for content requests. reqman = None # Number of HTTP requests made for content. nfetched = 0 # Found objects. found = [] def logme(msg, *args): """Generate agent log message.""" procid = "[%s/%d]" % (__file__.rsplit("/", 1)[-1], os.getpid()) print(datetime.now(), procid, msg % args) def myumask(): """Get the current process umask.""" val = os.umask(0) os.umask(val) return val def handle_init(c): """Prepare custom properties on download handles.""" c.temp_file = None c.temp_path = None c.local_path = None def request_init(c, options, path, kind): """`RequestManager` callback to initialise directory contents request.""" c.setopt(pycurl.URL, options.server + quote(path) + ((path != "/" and "/") or "")) assert c.temp_file == None assert c.temp_path == None assert c.local_path == None # If this is file, prepare temporary destination file in the target # directory for possible download. parse_dir_and_files() will finish this off. if kind == FILE and options.download: try: (fd, tmp) = mkstemp() fp = os.fdopen(fd, 'wb') c.setopt(pycurl.WRITEFUNCTION, fp.write) c.temp_file = fp c.temp_path = tmp c.local_path = path.strip('/') c.buffer = None except Exception as e: logme("ERROR: %s: %s", path, str(e)) print_exc() def cleanup(c): """Clean up file copy operation, usually after any failures.""" if c.temp_file: try: c.temp_file.close() except: pass if c.temp_path: try: os.remove(c.temp_path) except: pass if c.local_path: try: os.remove(c.local_path) except: pass c.temp_file = None c.temp_path = None c.local_path = None c.buffer = None def report_error(c, task, errmsg, errno): """`RequestManager` callback to report directory contents request errors.""" print("FAILED to retrieve %s: %s (%d)" % (task, errmsg, errno), file=sys.stderr) global nfetched; nfetched += 1 def parse_dir_and_files(c): """`RequestManager` callback to handle directory content response. This gets called once per every directory which has been successfully retrieved from the server. It parses the HTML response and turns it into object listing with all the file meta information. If verbosity has been requested, also shows simple progress bar on the search progress, one dot for every ten directories retrieved.""" options, path, kind = c.task root_url = urlparse(options.server).path.rstrip('/') if kind == FILE and options.download: assert c.local_path, "Expected local path property to be set" assert c.temp_file, "Exepected temporary file property to be set" try: c.setopt(pycurl.WRITEFUNCTION, lambda *args: None) c.temp_file.close() c.temp_file = None os.chmod(c.temp_path, 0o0666 & ~UMASK) os.system('mv %s %s' % (c.temp_path, c.local_path)) c.local_path = None logme("INFO: downloaded %s", path) except Exception as e: logme("ERROR: downloading %s into %s failed: %s", path, c.local_path, str(e)) print_exc() finally: cleanup(c) elif kind == DIR: items = re.findall(r"<tr><td><a href='(.*?)'>(.*?)</a></td><td>(\d+| |-)</td>" r"<td>( |\d\d\d\d-\d\d-\d\d \d\d:\d\d:\d\d UTC)</td>", c.buffer.getvalue().decode('utf-8')) for path, name, size, date in items: assert path.startswith(root_url) path = path[len(root_url):] if date == " ": date = -1 else: date = timegm(strptime(date, "%Y-%m-%d %H:%M:%S %Z")) if size == " " or size == "-": size = -1 else: size = int(size) if path.endswith("/"): assert size == -1 path = path[:-1] obj = {'task': c.task, 'kind': DIR, 'name': name, 'size': size, 'date': date, 'path': path} found.append(obj) if options.recursive: reqman.put((options, path, DIR)) else: assert size >= 0 obj = {'task': c.task, 'kind': FILE, 'name': name, 'size': size, 'date': date, 'path': path} if options.match_file: if name == options.match_file: found.append(obj) if options.download: reqman.put((options, path, FILE)) else: found.append(obj) if options.download: reqman.put((options, path, FILE)) global nfetched nfetched += 1 if options.verbose and nfetched % 10 == 0: sys.stdout.write(".") sys.stdout.flush() if nfetched % 750 == 0: print() def objcmp(a, b): diff = 0 diff = a['date'] - b['date'] if diff: return diff diff = a['size'] - b['size'] if diff: return diff return diff # Parse command line options. op = OptionParser(usage = __doc__) op.add_option("-s", "--server", dest = "server", type = "string", action = "store", metavar = "SERVER", default = "https://cmsweb.cern.ch/dqm/offline/data/browse", help = "Pull content from SERVER") op.add_option("-n", "--connections", dest = "connections", type = "int", action = "store", metavar = "NUM", default = 10, help = "Use NUM concurrent connections") op.add_option("-v", "--verbose", dest = "verbose", action = "store_true", default = False, help = "Show verbose scan information") op.add_option("-r", "--recursive", dest = "recursive", action = "store_true", default = False, help = "Perform a recursive scan starting from the root directory.") op.add_option("-d", "--download", dest = "download", action = "store_true", default = False, help = "Download all touched ROOT files locally. To be used with extreme care.") op.add_option("-m", "--match_file", dest = "match_file", type= "string", action = "store", default = "", help = "Filter results based on exact file name matching.") options, args = op.parse_args() if args: print("Too many arguments", file=sys.stderr) sys.exit(1) if not options.server: print("Server contact string required", sys.stderr) sys.exit(1) UMASK = myumask() # In case a user specifies a root file as address, remove it from the # address and return results that match only that name. gr = re.match("(.*)/(DQM_V.*\.root)$", options.server) if gr: options.server = gr.group(1) options.match_file = gr.group(2) # Get SSL X509 parametres. ssl_opts = SSLOptions() if options.verbose: print("Using SSL cert dir", ssl_opts.ca_path) print("Using SSL private key", ssl_opts.key_file) print("Using SSL public key", ssl_opts.cert_file) # Start a request manager for contents. reqman = RequestManager(num_connections = options.connections, ssl_opts = ssl_opts, user_agent = ident, request_init = request_init, request_respond = parse_dir_and_files, request_error = report_error, handle_init = handle_init) # Process from root directory. start = time() reqman.put((options, "/", DIR)) reqman.process() end = time() for x in sorted(found, key=cmp_to_key(objcmp)): print("%20s\t%s\t%s" % (x['size'], strftime("%Y-%m-%d %H:%M:%S %Z", gmtime(x['date'])), ((options.recursive and x['path']) or x['name']))) if options.verbose: print("\nFound %d directories, %d objects in %.3f seconds" % (nfetched, len(found), end - start))