diff --git a/selftest/report_test/report_test.py b/selftest/report_test/report_test.py
new file mode 100755
index 0000000..57e3a89
--- /dev/null
+++ b/selftest/report_test/report_test.py
@@ -0,0 +1,133 @@
+#!/usr/bin/env python3
+import _prep
+
+from osmo_gsm_tester.core import report
+from osmo_gsm_tester.core import log
+from osmo_gsm_tester.core import util
+from osmo_gsm_tester.core import test
+from osmo_gsm_tester.core import suite
+from osmo_gsm_tester.core import config
+
+import os
+import sys
+import shutil
+import difflib
+import xml.etree.ElementTree as et
+
+class FakeTrial(log.Origin):
+    def __init__(self):
+        super().__init__(log.C_TST, 'trial')
+        self.dir = util.Dir(example_trial_dir)
+        self._run_dir = None
+        self.suites = []
+
+    def get_all_inst_hash_info(self):
+        return { 'foobar/potato': '1234', 'orange': 'abcd' }
+
+    def get_run_dir(self):
+        if self._run_dir is not None:
+            return self._run_dir
+        self._run_dir = util.Dir(self.dir.new_child('test_run'))
+        self._run_dir.mkdir()
+        return self._run_dir
+
+class FakeSuiteDefinition(log.Origin):
+        def __init__(self, name, num_tests):
+            super().__init__(log.C_TST, name)
+            self.test_basenames = [name + '-' + str(tid) for tid in range(num_tests) ]
+            self.conf = {}
+            self.suite_dir = util.Dir(example_trial_dir).new_child('suitedef' + name)
+
+
+def fake_run_test(test_obj, status, duration, sysout=None):
+    test_obj.status = status
+    test_obj.duration = duration
+    if sysout is not None:
+        test_obj.set_report_stdout(sysout)
+    if status == test.Test.FAIL:
+        test_obj.fail_type = 'fake_fail_type'
+        test_obj.fail_message = 'fake_fail_message'
+        test_obj.fail_tb = 'system stderr fake content'
+
+def fake_run_suite(suite_obj, duration):
+    suite_obj.duration = duration
+    suite_obj.determine_status()
+
+config.override_conf = os.path.join(os.path.dirname(sys.argv[0]), 'main.conf')
+
+example_trial_dir = os.path.join('test_trial_tmp')
+
+trial = FakeTrial()
+
+# Suite passes with 2 tests passing
+s_def = FakeSuiteDefinition('suiteA', 2)
+s = suite.SuiteRun(trial, s_def.name(), s_def)
+trial.suites.append(s)
+fake_run_test(s.tests[0], test.Test.PASS, 30)
+fake_run_test(s.tests[1], test.Test.PASS, 10, 'yay this is a test-applied stdout')
+#fake_run_test(suiteA.tests[0], test.Test.UNKNOWN, 20)
+fake_run_suite(s, 50)
+
+# Suite passes first test but next ones are not ececuted
+s_def = FakeSuiteDefinition('suiteB', 3)
+s = suite.SuiteRun(trial, s_def.name(), s_def)
+trial.suites.append(s)
+fake_run_test(s.tests[0], test.Test.PASS, 10)
+fake_run_suite(s, 20)
+
+# Suite passes one test selected, others are skipped
+s_def = FakeSuiteDefinition('suiteC', 3)
+s = suite.SuiteRun(trial, s_def.name(), s_def)
+trial.suites.append(s)
+s.tests[0].set_skip()
+fake_run_test(s.tests[1], test.Test.PASS, 10)
+s.tests[2].set_skip()
+fake_run_suite(s, 12)
+
+# Suite fails due to one of its tests failing
+s_def = FakeSuiteDefinition('suiteD', 2)
+s = suite.SuiteRun(trial, s_def.name(), s_def)
+trial.suites.append(s)
+fake_run_test(s.tests[0], test.Test.FAIL, 12)
+fake_run_test(s.tests[1], test.Test.PASS, 10)
+fake_run_suite(s, 20)
+
+element = report.trial_to_junit(trial)
+
+def indent(elem, level=0):
+    i = "\n" + level*"  "
+    if len(elem):
+        if not elem.text or not elem.text.strip():
+            elem.text = i + "  "
+        if not elem.tail or not elem.tail.strip():
+            elem.tail = i
+        for elem in elem:
+            indent(elem, level+1)
+        if not elem.tail or not elem.tail.strip():
+            elem.tail = i
+    else:
+        if level and (not elem.tail or not elem.tail.strip()):
+            elem.tail = i
+
+def udiff(expect, got, expect_path):
+    expect = expect.splitlines(1)
+    got =  got.splitlines(1)
+    for line in difflib.unified_diff(expect, got,
+                                     fromfile=expect_path, tofile='got'):
+        sys.stderr.write(line)
+        if not line.endswith('\n'):
+            sys.stderr.write('[no-newline]\n')
+
+indent(element)
+#canonicalize() is only available in python3.8+, and we need it to have reliable string output:
+if hasattr(et, 'canonicalize'):
+    got = et.canonicalize(et.tostring(element)).rstrip()
+    exp_path = os.path.join(os.path.dirname(sys.argv[0]), 'expected_junit_output.xml')
+    with open(exp_path, 'r') as f:
+        exp = f.read().rstrip()
+    udiff(exp, got, exp_path)
+
+#deleting generated tmp trial dir:
+shutil.rmtree(example_trial_dir, ignore_errors=True)
+
+# vim: expandtab tabstop=4 shiftwidth=4