michaelh/zephyr
1
0
Fork 0
Code Releases Activity

twister: More restrictive regex for log markers

Browse source 
Some simulators - like simics - may end up adding extraneous suffixes to
logged lines. This may cause some regex that match too much fail. This
patch fixes two such cases:

  - regex to find RunID changed to only match valid hexadecimal
    characters;
  - regex to match start of testsuite changed to only match valid word
    characters (0-9A-Za-z_).

Signed-off-by: Ederson de Souza <ederson.desouza@intel.com>
This commit is contained in:
Ederson de Souza 2025-02-07 17:21:53 -08:00 • committed by Benjamin Cabé
commit 9741c25715
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
scripts/pylib/twister/twisterlib/harness.py
View file

@ -38,7 +38,7 @@ class Harness:
FAULT = "ZEPHYR FATAL ERROR"
RUN_PASSED = "PROJECT EXECUTION SUCCESSFUL"
RUN_FAILED = "PROJECT EXECUTION FAILED"
run_id_pattern = r"RunID: (?P<run_id>.*)"
run_id_pattern = r"RunID: (?P<run_id>[0-9A-Fa-f]+)"
def __init__(self):
self._status = TwisterStatus.NONE
@ -789,7 +789,7 @@ class Test(Harness):
# Ztest log patterns don't require to match the line start exactly: there are platforms
# where there is some logging prefix at each console line whereas on other platforms
# without prefixes the leading space is stripped.
test_suite_start_pattern = re.compile(r"Running TESTSUITE (?P<suite_name>\S*)")
test_suite_start_pattern = re.compile(r"Running TESTSUITE (?P<suite_name>\w*)")
test_suite_end_pattern = re.compile(
r"TESTSUITE (?P<suite_name>\S*)\s+(?P<suite_status>succeeded|failed)"
)