update-python-libraries: commit updates and specify update kind

This commit introduces two new features:

1. specify with --target whether major, minor or patch updates should be made
2. use --commit to create commits for each of the updates

1 files changed, 138 insertions, 22 deletions

diff --git a/maintainers/scripts/update-python-libraries b/maintainers/scripts/update-python-libraries
index 3ddc8c23a79..ec2691ff617 100755
--- a/maintainers/scripts/update-python-libraries
+++ b/maintainers/scripts/update-python-libraries
@@ -1,5 +1,5 @@
 #! /usr/bin/env nix-shell
-#! nix-shell -i python3 -p 'python3.withPackages(ps: with ps; [ requests toolz ])'
+#! nix-shell -i python3 -p 'python3.withPackages(ps: with ps; [ packaging requests toolz ])' -p git
 
 """
 Update a Python package expression by passing in the `.nix` file, or the directory containing it.
@@ -18,7 +18,12 @@ import os
 import re
 import requests
 import toolz
-from concurrent.futures import ThreadPoolExecutor as pool
+from concurrent.futures import ThreadPoolExecutor as Pool
+from packaging.version import Version as _Version
+from packaging.version import InvalidVersion
+from packaging.specifiers import SpecifierSet
+import collections
+import subprocess
 
 INDEX = "https://pypi.io/pypi"
 """url of PyPI"""
@@ -26,10 +31,30 @@ INDEX = "https://pypi.io/pypi"
 EXTENSIONS = ['tar.gz', 'tar.bz2', 'tar', 'zip', '.whl']
 """Permitted file extensions. These are evaluated from left to right and the first occurance is returned."""
 
+PRERELEASES = False
+
 import logging
 logging.basicConfig(level=logging.INFO)
 
 
+class Version(_Version, collections.abc.Sequence):
+
+    def __init__(self, version):
+        super().__init__(version)
+        # We cannot use `str(Version(0.04.21))` because that becomes `0.4.21`
+        # https://github.com/avian2/unidecode/issues/13#issuecomment-354538882
+        self.raw_version = version
+
+    def __getitem__(self, i):
+        return self._version.release[i]
+
+    def __len__(self):
+        return len(self._version.release)
+
+    def __iter__(self):
+        yield from self._version.release
+
+
 def _get_values(attribute, text):
     """Match attribute in text and return all matches.
 
@@ -82,13 +107,59 @@ def _fetch_page(url):
     else:
         raise ValueError("request for {} failed".format(url))
 
-def _get_latest_version_pypi(package, extension):
+
+SEMVER = {
+    'major' : 0,
+    'minor' : 1,
+    'patch' : 2,
+}
+
+
+def _determine_latest_version(current_version, target, versions):
+    """Determine latest version, given `target`.
+    """
+    current_version = Version(current_version)
+
+    def _parse_versions(versions):
+        for v in versions:
+            try:
+                yield Version(v)
+            except InvalidVersion:
+                pass
+
+    versions = _parse_versions(versions)
+
+    index = SEMVER[target]
+
+    ceiling = list(current_version[0:index])
+    if len(ceiling) == 0:
+        ceiling = None
+    else:
+        ceiling[-1]+=1
+        ceiling = Version(".".join(map(str, ceiling)))
+
+    # We do not want prereleases
+    versions = SpecifierSet(prereleases=PRERELEASES).filter(versions)
+
+    if ceiling is not None:
+        versions = SpecifierSet(f"<{ceiling}").filter(versions)
+
+    return (max(sorted(versions))).raw_version
+
+
+def _get_latest_version_pypi(package, extension, current_version, target):
     """Get latest version and hash from PyPI."""
     url = "{}/{}/json".format(INDEX, package)
     json = _fetch_page(url)
 
-    version = json['info']['version']
-    for release in json['releases'][version]:
+    versions = json['releases'].keys()
+    version = _determine_latest_version(current_version, target, versions)
+
+    try:
+        releases = json['releases'][version]
+    except KeyError as e:
+        raise KeyError('Could not find version {} for {}'.format(version, package)) from e
+    for release in releases:
         if release['filename'].endswith(extension):
             # TODO: In case of wheel we need to do further checks!
             sha256 = release['digests']['sha256']
@@ -98,7 +169,7 @@ def _get_latest_version_pypi(package, extension):
     return version, sha256
 
 
-def _get_latest_version_github(package, extension):
+def _get_latest_version_github(package, extension, current_version, target):
     raise ValueError("updating from GitHub is not yet supported.")
 
 
@@ -141,9 +212,9 @@ def _determine_extension(text, fetcher):
     """
     if fetcher == 'fetchPypi':
         try:
-            format = _get_unique_value('format', text)
+            src_format = _get_unique_value('format', text)
         except ValueError as e:
-            format = None   # format was not given
+            src_format = None   # format was not given
 
         try:
             extension = _get_unique_value('extension', text)
@@ -151,9 +222,11 @@ def _determine_extension(text, fetcher):
             extension = None    # extension was not given
 
         if extension is None:
-            if format is None:
-                format = 'setuptools'
-            extension = FORMATS[format]
+            if src_format is None:
+                src_format = 'setuptools'
+            elif src_format == 'flit':
+                raise ValueError("Don't know how to update a Flit package.")
+            extension = FORMATS[src_format]
 
     elif fetcher == 'fetchurl':
         url = _get_unique_value('url', text)
@@ -167,9 +240,7 @@ def _determine_extension(text, fetcher):
     return extension
 
 
-def _update_package(path):
-
-
+def _update_package(path, target):
 
     # Read the expression
     with open(path, 'r') as f:
@@ -186,11 +257,13 @@ def _update_package(path):
 
     extension = _determine_extension(text, fetcher)
 
-    new_version, new_sha256 = _get_latest_version_pypi(pname, extension)
+    new_version, new_sha256 = FETCHERS[fetcher](pname, extension, version, target)
 
     if new_version == version:
         logging.info("Path {}: no update available for {}.".format(path, pname))
         return False
+    elif new_version <= version:
+        raise ValueError("downgrade for {}.".format(pname))
     if not new_sha256:
         raise ValueError("no file available for {}.".format(pname))
 
@@ -202,10 +275,19 @@ def _update_package(path):
 
         logging.info("Path {}: updated {} from {} to {}".format(path, pname, version, new_version))
 
-    return True
+    result = {
+        'path'  : path,
+        'target': target,
+        'pname': pname,
+        'old_version'   : version,
+        'new_version'   : new_version,
+        #'fetcher'       : fetcher,
+        }
 
+    return result
 
-def _update(path):
+
+def _update(path, target):
 
     # We need to read and modify a Nix expression.
     if os.path.isdir(path):
@@ -222,24 +304,58 @@ def _update(path):
         return False
 
     try:
-        return _update_package(path)
+        return _update_package(path, target)
     except ValueError as e:
         logging.warning("Path {}: {}".format(path, e))
         return False
 
+
+def _commit(path, pname, old_version, new_version, **kwargs):
+    """Commit result.
+    """
+
+    msg = f'python: {pname}: {old_version} -> {new_version}'
+
+    try:
+        subprocess.check_call(['git', 'add', path])
+        subprocess.check_call(['git', 'commit', '-m', msg])
+    except subprocess.CalledProcessError as e:
+        subprocess.check_call(['git', 'checkout', path])
+        raise subprocess.CalledProcessError(f'Could not commit {path}') from e
+
+    return True
+
+
 def main():
 
     parser = argparse.ArgumentParser()
     parser.add_argument('package', type=str, nargs='+')
+    parser.add_argument('--target', type=str, choices=SEMVER.keys(), default='major')
+    parser.add_argument('--commit', action='store_true', help='Create a commit for each package update')
 
     args = parser.parse_args()
+    target = args.target
+
+    packages = list(map(os.path.abspath, args.package))
+
+    logging.info("Updating packages...")
+
+    # Use threads to update packages concurrently
+    with Pool() as p:
+        results = list(p.map(lambda pkg: _update(pkg, target), packages))
+
+    logging.info("Finished updating packages.")
+
+    # Commits are created sequentially.
+    if args.commit:
+        logging.info("Committing updates...")
+        list(map(lambda x: _commit(**x), filter(bool, results)))
+        logging.info("Finished committing updates")
 
-    packages = map(os.path.abspath, args.package)
+    count = sum(map(bool, results))
+    logging.info("{} package(s) updated".format(count))
 
-    with pool() as p:
-        count = list(p.map(_update, packages))
 
-    logging.info("{} package(s) updated".format(sum(count)))
 
 if __name__ == '__main__':
     main()
\ No newline at end of file