ccmlib: Remove usages of distutils

This change removes all uses of distutils modules and replaces them with
either direct replacements or extractions from the library.
LooseVersion is extracted into ccmlib.utils.version and dir_util.copy_tree
is replaced by shutil.copytree.

Fixes scylladb#537

README.md

@@ -90,7 +90,7 @@ Windows only:
 Installation
 ------------
 
-ccm uses python distutils so from the source directory run:
+ccm uses python setuptools (with distutils fallback) so from the source directory run:
 
  sudo ./setup.py install
 

ccmlib/repository.py

@@ -11,10 +11,11 @@
 import tarfile
 import tempfile
 import time
+from typing import List
 import urllib
-from distutils.version import LooseVersion
 
 
+from ccmlib.utils.version import LooseVersion as Version
 from ccmlib.common import (ArgumentError, CCMError, get_default_path,
  platform_binary, rmdirs, validate_install_dir,
  assert_jdk_valid_for_cassandra_version, get_version_from_build)
@@ -373,7 +374,7 @@ def get_tagged_version_numbers(series='stable'):
  """Retrieve git tags and find version numbers for a release series
 
  series - 'stable', 'oldstable', or 'testing'"""
- releases = []
+ releases: List[Version] = []
  if series == 'testing':
  # Testing releases always have a hyphen after the version number:
  tag_regex = re.compile(r'^refs/tags/cassandra-([0-9]+\.[0-9]+\.[0-9]+-.*$)')
@@ -385,15 +386,15 @@ def get_tagged_version_numbers(series='stable'):
  for ref in (i.get('ref', '') for i in json.loads(tag_url.read())):
  m = tag_regex.match(ref)
  if m:
- releases.append(LooseVersion(m.groups()[0]))
+ releases.append(Version(m.groups()[0]))
 
  # Sort by semver:
  releases.sort(reverse=True)
 
- stable_major_version = LooseVersion(str(releases[0].version[0]) + "." + str(releases[0].version[1]))
+ stable_major_version = Version(str(releases[0].version[0]) + "." + str(releases[0].version[1]))
  stable_releases = [r for r in releases if r >= stable_major_version]
  oldstable_releases = [r for r in releases if r not in stable_releases]
- oldstable_major_version = LooseVersion(str(oldstable_releases[0].version[0]) + "." + str(oldstable_releases[0].version[1]))
+ oldstable_major_version = Version(str(oldstable_releases[0].version[0]) + "." + str(oldstable_releases[0].version[1]))
  oldstable_releases = [r for r in oldstable_releases if r >= oldstable_major_version]
 
  if series == 'testing':

ccmlib/scylla_cluster.py

@@ -8,14 +8,14 @@
 import uuid
 import datetime
 
-from distutils.version import LooseVersion
 
 from ccmlib import common
 from ccmlib.cluster import Cluster
 from ccmlib.scylla_node import ScyllaNode
 from ccmlib.node import NodeError
 from ccmlib import scylla_repository
 from ccmlib.utils.sni_proxy import stop_sni_proxy
+from ccmlib.utils.version import LooseVersion as Version
 
 SNITCH = 'org.apache.cassandra.locator.GossipingPropertyFileSnitch'
 
@@ -294,7 +294,7 @@ def __init__(self, scylla_cluster, install_dir=None):
  def version(self):
  stdout, _ = self.sctool(["version"], ignore_exit_status=True)
  version_string = stdout[stdout.find(": ") + 2:].strip() # Removing unnecessary information
- version_code = LooseVersion(version_string)
+ version_code = Version(version_string)
  return version_code
 
  def _install(self, install_dir):
@@ -314,8 +314,8 @@ def _update_config(self, install_dir=None):
  data['database'] = {}
  data['database']['hosts'] = [self.scylla_cluster.get_node_ip(1)]
  data['database']['replication_factor'] = 3
- if install_dir and (self.version < LooseVersion("2.5") or
- LooseVersion('666') < self.version < LooseVersion('666.dev-0.20210430.2217cc84')):
+ if install_dir and (self.version < Version("2.5") or
+ Version('666') < self.version < Version('666.dev-0.20210430.2217cc84')):
  data['database']['migrate_dir'] = os.path.join(install_dir, 'schema', 'cql')
  if 'https' in data:
  del data['https']
@@ -328,7 +328,7 @@ def _update_config(self, install_dir=None):
  data['logger']['mode'] = 'stderr'
  if not 'repair' in data:
  data['repair'] = {}
- if self.version < LooseVersion("2.2"):
+ if self.version < Version("2.2"):
  data['repair']['segments_per_repair'] = 16
  data['prometheus'] = f"{self.scylla_cluster.get_node_ip(1)}:56091"
  # Changing port to 56091 since the manager and the first node share the same ip and 56090 is already in use

ccmlib/utils/sni_proxy.py

@@ -8,7 +8,7 @@
 from contextlib import contextmanager
 import tempfile
 from textwrap import dedent
-import distutils.dir_util
+from shutil import copytree
 from dataclasses import dataclass
 
 import yaml
@@ -198,7 +198,7 @@ def refresh_certs(cluster, nodes_info):
  dns_names = ['cql.cluster-id.scylla.com'] + \
  [f'{node.host_id}.cql.cluster-id.scylla.com' for node in nodes_info]
  generate_ssl_stores(tmp_dir, dns_names=dns_names)
- distutils.dir_util.copy_tree(tmp_dir, cluster.get_path())
+ copytree(tmp_dir, cluster.get_path(), dirs_exist_ok=True)
 
 
 if __name__ == "__main__":

ccmlib/utils/version.py

@@ -1,5 +1,74 @@
+import re
 from packaging.version import parse, Version
 
+class LooseVersion:
+ component_re = re.compile(r'(\d+ | [a-z]+ | \.)', re.VERBOSE)
+
+ def __init__ (self, vstring=None):
+ if vstring:
+ self.parse(vstring)
+
+ def parse (self, vstring):
+ self.vstring = vstring
+ components = [x for x in self.component_re.split(vstring)
+ if x and x != '.']
+ for i, obj in enumerate(components):
+ try:
+ components[i] = int(obj)
+ except ValueError:
+ pass
+
+ self.version = components
+
+ def _cmp (self, other):
+ if isinstance(other, str):
+ other = LooseVersion(other)
+ elif not isinstance(other, LooseVersion):
+ return NotImplemented
+
+ if self.version == other.version:
+ return 0
+ if self.version < other.version:
+ return -1
+ if self.version > other.version:
+ return 1
+
+ def __str__ (self):
+ return self.vstring
+
+ def __repr__ (self):
+ return "LooseVersion ('%s')" % str(self)
+
+ def __eq__(self, other):
+ c = self._cmp(other)
+ if c is NotImplemented:
+ return c
+ return c == 0
+
+ def __lt__(self, other):
+ c = self._cmp(other)
+ if c is NotImplemented:
+ return c
+ return c < 0
+
+ def __le__(self, other):
+ c = self._cmp(other)
+ if c is NotImplemented:
+ return c
+ return c <= 0
+
+ def __gt__(self, other):
+ c = self._cmp(other)
+ if c is NotImplemented:
+ return c
+ return c > 0
+
+ def __ge__(self, other):
+ c = self._cmp(other)
+ if c is NotImplemented:
+ return c
+ return c >= 0
+
 
 def parse_version(v: str) -> Version:
  v = v.replace('~', '-')

tests/test_config.py

@@ -1,7 +1,17 @@
-import distutils.util
 import os
 
-DEV_MODE = bool(distutils.util.strtobool(os.environ.get("DEV_MODE", "False")))
+
+def strtobool(val: str):
+ normalized_val = val.lower()
+ if normalized_val in ('y', 'yes', 't', 'true', 'on', '1'):
+ return True
+ if normalized_val in ('n', 'no', 'f', 'false', 'off', '0'):
+ return False
+ else:
+ raise ValueError(f"invalid truth value {val}")
+
+
+DEV_MODE = strtobool(os.environ.get("DEV_MODE", "False"))
 RESULTS_DIR = "test_results"
 TEST_ID = os.environ.get("CCM_TEST_ID", None)
 SCYLLA_DOCKER_IMAGE = os.environ.get(