[cleaner] Add support for Insights client archives

Adds a new type of `SoSObfuscationArchive` to add support for
obfuscating archives generated by the Insights project.

Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>

3 files changed, 46 insertions, 1 deletions

diff --git a/man/en/sos-clean.1 b/man/en/sos-clean.1
index 54026713..358ec0cb 100644
--- a/man/en/sos-clean.1
+++ b/man/en/sos-clean.1
@@ -105,6 +105,7 @@ The following are accepted values for this option:
     \fBauto\fR          Automatically detect the archive type
     \fBreport\fR        An archive generated by \fBsos report\fR
     \fBcollect\fR       An archive generated by \fBsos collect\fR
+    \fBinsights\fR      An archive generated by the \fBinsights-client\fR package
 
 The following may also be used, however note that these do not attempt to pre-load
 any information from the archives into the parsers. This means that, among other limitations,
diff --git a/sos/cleaner/__init__.py b/sos/cleaner/__init__.py
index 6d2eb483..3e08aa28 100644
--- a/sos/cleaner/__init__.py
+++ b/sos/cleaner/__init__.py
@@ -29,6 +29,7 @@ from sos.cleaner.archives.sos import (SoSReportArchive, SoSReportDirectory,
                                       SoSCollectorArchive,
                                       SoSCollectorDirectory)
 from sos.cleaner.archives.generic import DataDirArchive, TarballArchive
+from sos.cleaner.archives.insights import InsightsArchive
 from sos.utilities import get_human_readable
 from textwrap import fill
 
@@ -100,6 +101,7 @@ class SoSCleaner(SoSComponent):
             SoSReportArchive,
             SoSCollectorDirectory,
             SoSCollectorArchive,
+            InsightsArchive,
             # make sure these two are always last as they are fallbacks
             DataDirArchive,
             TarballArchive
@@ -194,7 +196,7 @@ third party.
                                help='The directory or archive to obfuscate')
         clean_grp.add_argument('--archive-type', default='auto',
                                choices=['auto', 'report', 'collect',
-                                        'data-dir', 'tarball'],
+                                        'insights', 'data-dir', 'tarball'],
                                help=('Specify what kind of archive the target '
                                      'was generated as'))
         clean_grp.add_argument('--domains', action='extend', default=[],
diff --git a/sos/cleaner/archives/insights.py b/sos/cleaner/archives/insights.py
new file mode 100644
index 00000000..dab48b16
--- /dev/null
+++ b/sos/cleaner/archives/insights.py
@@ -0,0 +1,42 @@
+# Copyright 2021 Red Hat, Inc. Jake Hunsaker <jhunsake@redhat.com>
+
+# This file is part of the sos project: https://github.com/sosreport/sos
+#
+# This copyrighted material is made available to anyone wishing to use,
+# modify, copy, or redistribute it subject to the terms and conditions of
+# version 2 of the GNU General Public License.
+#
+# See the LICENSE file in the source distribution for further information.
+
+
+from sos.cleaner.archives import SoSObfuscationArchive
+
+import tarfile
+
+
+class InsightsArchive(SoSObfuscationArchive):
+    """This class represents archives generated by the insights-client utility
+    for RHEL systems.
+    """
+
+    type_name = 'insights'
+    description = 'insights-client archive'
+
+    prep_files = {
+        'hostname': 'data/insights_commands/hostname_-f',
+        'ip': 'data/insights_commands/ip_addr',
+        'mac': 'data/insights_commands/ip_addr'
+    }
+
+    @classmethod
+    def check_is_type(cls, arc_path):
+        try:
+            return tarfile.is_tarfile(arc_path) and 'insights-' in arc_path
+        except Exception:
+            return False
+
+    def get_archive_root(self):
+        top = self.archive_path.split('/')[-1].split('.tar')[0]
+        if self.tarobj.firstmember.name == '.':
+            top = './' + top
+        return top