view tests/test-check-module-imports.t @ 36360:c25290b98190

perfbranchmap: allow to select the filter to benchmark

Running the branchmap computation on all filter levels can be expensive.
Narrowing the run to some specific filters can speed up benchmarking time when
working only on a subset of filter levels.

#require test-repo

  $ . "$TESTDIR/helpers-testrepo.sh"
  $ import_checker="$TESTDIR"/../contrib/import-checker.py

  $ cd "$TESTDIR"/..

There are a handful of cases here that require renaming a module so it
doesn't overlap with a stdlib module name. There are also some cycles
here that we should still endeavor to fix, and some cycles will be
hidden by deduplication algorithm in the cycle detector, so fixing
these may expose other cycles.

Known-bad files are excluded by -X as some of them would produce unstable
outputs, which should be fixed later.

  $ testrepohg locate 'set:**.py or grep(r"^#!.*?python")' \
  > 'tests/**.t' \
  > -X hgweb.cgi \
  > -X setup.py \
  > -X contrib/debugshell.py \
  > -X contrib/hgweb.fcgi \
  > -X contrib/python-zstandard/ \
  > -X contrib/win32/hgwebdir_wsgi.py \
  > -X doc/gendoc.py \
  > -X doc/hgmanpage.py \
  > -X i18n/posplit \
  > -X mercurial/thirdparty \
  > -X tests/hypothesishelpers.py \
  > -X tests/test-commit-interactive.t \
  > -X tests/test-contrib-check-code.t \
  > -X tests/test-demandimport.py \
  > -X tests/test-extension.t \
  > -X tests/test-hghave.t \
  > -X tests/test-hgweb-auth.py \
  > -X tests/test-hgweb-no-path-info.t \
  > -X tests/test-hgweb-no-request-uri.t \
  > -X tests/test-hgweb-non-interactive.t \
  > -X tests/test-hook.t \
  > -X tests/test-import.t \
  > -X tests/test-imports-checker.t \
  > -X tests/test-lock.py \
  > -X tests/test-verify-repo-operations.py \
  > | sed 's-\\-/-g' | $PYTHON "$import_checker" -