|
| 1 | +"""Helpers for repository package catalog, metrics, and rebuild collapse.""" |
| 2 | + |
| 3 | +from collections import defaultdict |
| 4 | + |
| 5 | +from django.db.models import CharField, Func, Max, Min, Q, Value |
| 6 | +from packaging.version import InvalidVersion, Version |
| 7 | + |
| 8 | +from pulp_python.app.models import PythonPackageContent |
| 9 | + |
| 10 | +# POSIX regex for REGEXP_REPLACE. PostgreSQL does not treat ``\d`` as digits. |
| 11 | +BUILD_SUFFIX_PG_REGEX = r"\.[a-zA-Z]+-[0-9]+$" |
| 12 | + |
| 13 | + |
| 14 | +def base_version_annotation(field_name="version"): |
| 15 | + """SQL expression that strips a trailing rebuild suffix from ``version``. |
| 16 | +
|
| 17 | + PostgreSQL POSIX regex does not treat ``\\d`` as digits, so the SQL pattern |
| 18 | + uses ``[0-9]`` while the Python pattern in ``versions.strip_build_suffix`` uses ``\\d``. |
| 19 | + Implemented with ``REGEXP_REPLACE`` so it does not depend on Django's |
| 20 | + ``RegexpReplace`` (not present in every Django 4.2/5.2 packaging Pulp uses). |
| 21 | + """ |
| 22 | + return Func( |
| 23 | + field_name, |
| 24 | + Value(BUILD_SUFFIX_PG_REGEX), |
| 25 | + Value(""), |
| 26 | + function="REGEXP_REPLACE", |
| 27 | + output_field=CharField(), |
| 28 | + ) |
| 29 | + |
| 30 | + |
| 31 | +def collapse_python_builds(queryset): |
| 32 | + """Keep one content unit per ``(name_normalized, base_version)``. |
| 33 | +
|
| 34 | + ``base_version`` is ``version`` with a trailing rebuild suffix stripped. |
| 35 | + The unit with the latest ``pulp_created`` is kept. Callers that want one |
| 36 | + row per logical version (not per wheel/sdist) should also filter |
| 37 | + ``packagetype``. |
| 38 | + """ |
| 39 | + return ( |
| 40 | + queryset.prefetch_related(None) |
| 41 | + .annotate(_collapse_base_version=base_version_annotation()) |
| 42 | + .order_by("name_normalized", "_collapse_base_version", "-pulp_created") |
| 43 | + .distinct("name_normalized", "_collapse_base_version") |
| 44 | + ) |
| 45 | + |
| 46 | + |
| 47 | +def python_packages_in_version(repository_version): |
| 48 | + """Python package content contained in ``repository_version``.""" |
| 49 | + if repository_version is None: |
| 50 | + return PythonPackageContent.objects.none() |
| 51 | + return PythonPackageContent.objects.filter(pk__in=repository_version.content) |
| 52 | + |
| 53 | + |
| 54 | +def apply_package_prefix_filters(queryset, name_normalized_prefix=None, name_prefix=None): |
| 55 | + """Apply case-insensitive prefix filters used by the package index.""" |
| 56 | + if name_normalized_prefix: |
| 57 | + queryset = queryset.filter(name_normalized__istartswith=name_normalized_prefix) |
| 58 | + if name_prefix: |
| 59 | + queryset = queryset.filter(name__istartswith=name_prefix) |
| 60 | + return queryset |
| 61 | + |
| 62 | + |
| 63 | +def distinct_package_names_qs(content_qs): |
| 64 | + """One row per distinct ``name_normalized``, ordered for stable pagination.""" |
| 65 | + return ( |
| 66 | + content_qs.order_by() |
| 67 | + .values("name_normalized") |
| 68 | + .annotate(name=Max("name")) |
| 69 | + .order_by("name_normalized") |
| 70 | + ) |
| 71 | + |
| 72 | + |
| 73 | +def _version_sort_key(version): |
| 74 | + try: |
| 75 | + return (0, Version(version)) |
| 76 | + except InvalidVersion: |
| 77 | + return (1, version) |
| 78 | + |
| 79 | + |
| 80 | +def assemble_package_index(content_qs, name_rows, repository, repository_version): |
| 81 | + """Build package-index dicts for ``name_rows``. |
| 82 | +
|
| 83 | + ``created_at`` is the earliest repository-membership time |
| 84 | + (``RepositoryContent.pulp_created``) of any file of that logical version |
| 85 | + in ``repository_version``, falling back to the content unit's ``pulp_created``. |
| 86 | + """ |
| 87 | + if not name_rows or repository_version is None: |
| 88 | + return [] |
| 89 | + |
| 90 | + names = [row["name_normalized"] for row in name_rows] |
| 91 | + name_by_normalized = {row["name_normalized"]: row["name"] for row in name_rows} |
| 92 | + |
| 93 | + in_this_version = Q( |
| 94 | + version_memberships__repository=repository, |
| 95 | + version_memberships__version_added__number__lte=repository_version.number, |
| 96 | + ) & ( |
| 97 | + Q(version_memberships__version_removed__isnull=True) |
| 98 | + | Q(version_memberships__version_removed__number__gt=repository_version.number) |
| 99 | + ) |
| 100 | + |
| 101 | + release_rows = ( |
| 102 | + content_qs.filter(name_normalized__in=names) |
| 103 | + .annotate(_base_version=base_version_annotation()) |
| 104 | + .values("name_normalized", "_base_version") |
| 105 | + .annotate( |
| 106 | + membership_created=Min( |
| 107 | + "version_memberships__pulp_created", |
| 108 | + filter=in_this_version, |
| 109 | + ), |
| 110 | + unit_created=Min("pulp_created"), |
| 111 | + ) |
| 112 | + ) |
| 113 | + |
| 114 | + releases_by_name = defaultdict(list) |
| 115 | + for rel in release_rows: |
| 116 | + releases_by_name[rel["name_normalized"]].append(rel) |
| 117 | + |
| 118 | + result = [] |
| 119 | + for row in name_rows: |
| 120 | + normalized = row["name_normalized"] |
| 121 | + rels = sorted( |
| 122 | + releases_by_name.get(normalized, []), |
| 123 | + key=lambda item: _version_sort_key(item["_base_version"]), |
| 124 | + ) |
| 125 | + versions = [item["_base_version"] for item in rels] |
| 126 | + latest_releases = [ |
| 127 | + { |
| 128 | + "version": item["_base_version"], |
| 129 | + "release": "", |
| 130 | + "created_at": item["membership_created"] or item["unit_created"], |
| 131 | + } |
| 132 | + for item in rels |
| 133 | + ] |
| 134 | + result.append( |
| 135 | + { |
| 136 | + "name": name_by_normalized[normalized], |
| 137 | + "name_normalized": normalized, |
| 138 | + "versions": versions, |
| 139 | + "latest_releases": latest_releases, |
| 140 | + } |
| 141 | + ) |
| 142 | + return result |
| 143 | + |
| 144 | + |
| 145 | +def repository_metrics(content_qs): |
| 146 | + """Distinct package / logical-version / build counts for package content. |
| 147 | +
|
| 148 | + Identity is always ``PythonPackageContent`` (not filtered by packagetype): |
| 149 | +
|
| 150 | + * ``package_count``: distinct ``name_normalized`` |
| 151 | + * ``version_count``: distinct ``(name_normalized, base_version)`` |
| 152 | + * ``build_count``: distinct ``(name_normalized, version)`` |
| 153 | +
|
| 154 | + Until rebuild suffixes exist, ``version_count`` equals ``build_count``. |
| 155 | + """ |
| 156 | + content_qs = content_qs.order_by() |
| 157 | + return { |
| 158 | + "package_count": content_qs.values("name_normalized").distinct().count(), |
| 159 | + "version_count": ( |
| 160 | + content_qs.annotate(_base_version=base_version_annotation()) |
| 161 | + .values("name_normalized", "_base_version") |
| 162 | + .distinct() |
| 163 | + .count() |
| 164 | + ), |
| 165 | + "build_count": content_qs.values("name_normalized", "version").distinct().count(), |
| 166 | + } |
0 commit comments