Source code for insights.parsers.yum_list_installed

YumListInstalled - Command ``yum list installed``

.. warning::
    This module is deprecated, please import the parsers in
    :py:mod:`insights.parsers.yum_list` instead.
from collections import defaultdict

from insights import CommandParser, parser, SkipComponent
from insights.specs import Specs
from insights.parsers.installed_rpms import InstalledRpm, RpmList
from insights.util import deprecated

[docs]class YumInstalledRpm(InstalledRpm): """ The same as :py:class:`insights.parsers.installed_rpms.InstalledRpm` but with an additional ``.repo`` attribute. """ def __init__(self, data): deprecated(YumInstalledRpm, "Import YumListRpm from insights.parsers.yum_list instead") self.repo = None """str: yum / dnf repository name, if available.""" super(YumInstalledRpm, self).__init__(data)
[docs]@parser(Specs.yum_list_installed) class YumListInstalled(CommandParser, RpmList): """ ``YumListInstalled`` shares the :py:class:`insights.parsers.installed_rpms.RpmList` interface with :py:class:`insights.parsers.installed_rpms.InstalledRpms`. The only difference is ``YumListInstalled`` takes the output of ``yum list installed`` as its source data, and the :py:class:`YumInstalledRpm` instances it produces contain a ``.repo`` attribute. """ def __init__(self, context): deprecated(YumListInstalled, "Import YumListBase from insights.parsers.yum_list instead") self.expired_cache = False """bool: Indicates if the yum repo cache is expired.""" super(YumListInstalled, self).__init__(context) def _find_start(self, content): for i, c in enumerate(content): if 'Repodata is over 2 weeks old' in c: self.expired_cache = True elif c == "Installed Packages": break return i + 1 def _get_rows(self, content): """ Yields: a list per row of the following form: .. code:: python [ <name.arch>, <[epoch:]version-release>, <repo or @installed-from-repo> ] """ start = self._find_start(content) if start == len(content): raise SkipComponent() # join hanging wrapped lines together into a single line. # see cur = [] for line in content[start:]: if not cur: cur.append(line.strip()) elif line.startswith(" "): cur.append(line.strip()) else: yield " ".join(cur).split() cur = [line.strip()] if cur: yield " ".join(cur).split() def _make_record(self, package, ver_rel, repo): """ Given the fields of a ``yum list installed`` row, return a dictionary of name, version, release, epoch, arch, and repo. """ name, _, arch = package.rpartition(".") repo = repo.lstrip("@") # Kept as string in InstalledRpm. Duplicating here for consistency. epoch = "0" if ":" in ver_rel: epoch, ver_rel = ver_rel.split(":", 1) version, release = ver_rel.split("-") # This is special cased for InstalledRpm. Duplicating here for # consistency. if name.startswith('oracleasm') and name.endswith('.el5'): name, version2 = name.split('-', 1) version = version2 + '-' + version return {"name": name, "version": version, "release": release, "epoch": epoch, "arch": arch, "repo": repo} def _unknown_row(self, row): """ Heuristic to tell us we've hit the bottom of the Installed Packages stanza. """ return len(row) != 3 or row[:2] == ["Loaded", "plugins:"]
[docs] def parse_content(self, content): """ ``yum list installed`` output is basically tabular with an ignorable set of rows at the top and a line "Installed Packages" that designates the following rows as data. Each column has a maximum width, and if any column overflows, the following columns wrap to the next line and indent to their usual starting positions. It's also possible for the data rows to be followed by more lines that should be ignored. Since ``yum list installed`` is for human consumption, the footer lines can be syntactically ambiguous with data lines. We use heuristics to check for an invalid row to signal the end of data. """ packages = defaultdict(list) for row in self._get_rows(content): if self._unknown_row(row): break rec = self._make_record(*row) packages[rec["name"]].append(YumInstalledRpm(rec)) self.packages = dict(packages)