e7b6b37c18
Tests in test_runner.py can intermittently fail because test_parse_url was using case.BASE_TEST without first copying it. Therefore changes that were made in test_parse_url could be reflected elsewhere. This points to a generic problem in gabbi itself: There needs to be better management of global config-like items. test-limit.sh was failing because the test it uses to signal limiting was moved from driver to suitemaker. The string used in a grep had to be corrected.
16 lines
580 B
Bash
Executable File
16 lines
580 B
Bash
Executable File
#!/bin/bash
|
|
# Run a test which is limited to just one request from a file that
|
|
# contains many requests and confirm that only one was run and that
|
|
# it did actually run.
|
|
#
|
|
# This covers a situation where the change of intercepts to fixtures
|
|
# broke limiting tests and we never knew.
|
|
|
|
GREP_TEST_MATCH='gabbi.suitemaker.test_intercept_self_checklimit.test_request ... ok'
|
|
GREP_COUNT_MATCH='Ran 1 '
|
|
|
|
python setup.py testr --testr-args="checklimit" && \
|
|
testr last --subunit | subunit2pyunit 2>&1 | \
|
|
grep "${GREP_TEST_MATCH}" && \
|
|
testr last | grep "${GREP_COUNT_MATCH}"
|