Date
1 - 3 of 3
[master][PATCH v2] python3-inotify: add ptest
|
Qiu, Zheng
Add ptest provided from the source code, and run it using pytest.
Since the tarball provided with pypi does not include ptest, switch the source to be its git repo. ptest result: PASS FAIL SKIP qemux86-64 5 3 1 qemuarm64 5 3 1 This issue exists since inotify adds unit test support, and it is reported here: https://github.com/dsoprea/PyInotify/issues/95 Signed-off-by: Zheng Qiu <zheng.qiu@...> --- .../python/python3-inotify/run-ptest | 3 ++ .../python/python3-inotify_0.2.10.bb | 14 -------- .../python/python3-inotify_git.bb | 35 +++++++++++++++++++ 3 files changed, 38 insertions(+), 14 deletions(-) create mode 100644 meta-python/recipes-devtools/python/python3-inotify/run-ptest delete mode 100644 meta-python/recipes-devtools/python/python3-inotify_0.2.10.bb create mode 100644 meta-python/recipes-devtools/python/python3-inotify_git.bb diff --git a/meta-python/recipes-devtools/python/python3-inotify/run-ptest b/meta-python/recipes-devtools/python/python3-inotify/run-ptest new file mode 100644 index 000000000..3385d6893 --- /dev/null +++ b/meta-python/recipes-devtools/python/python3-inotify/run-ptest @@ -0,0 +1,3 @@ +#!/bin/sh + +pytest -o log_cli=true -o log_cli_level=INFO | sed -e 's/\[...%\]//g'| sed -e 's/PASSED/PASS/g'| sed -e 's/FAILED/FAIL/g'|sed -e 's/SKIPED/SKIP/g'| awk '{if ($NF=="PASS" || $NF=="FAIL" || $NF=="SKIP" || $NF=="XFAIL" || $NF=="XPASS"){printf "%s: %s\n", $NF, $0}else{print}}'| awk '{if ($NF=="PASS" || $NF=="FAIL" || $NF=="SKIP" || $NF=="XFAIL" || $NF=="XPASS") {$NF="";print $0}else{print}}' diff --git a/meta-python/recipes-devtools/python/python3-inotify_0.2.10.bb b/meta-python/recipes-devtools/python/python3-inotify_0.2.10.bb deleted file mode 100644 index 524c3a0d2..000000000 --- a/meta-python/recipes-devtools/python/python3-inotify_0.2.10.bb +++ /dev/null @@ -1,14 +0,0 @@ -SUMMARY = "An adapter to Linux kernel support for inotify directory-watching." -AUTHOR = "Dustin Oprea" -HOMEPAGE = "https://pypi.org/project/inotify/" -LICENSE = "GPL-2.0-only" -LIC_FILES_CHKSUM = "file://setup.py;md5=79b7ec72aa0d446a552d3cceb5799e41;beginline=28;endline=28" - -SRC_URI[sha256sum] = "974a623a338482b62e16d4eb705fb863ed33ec178680fc3e96ccdf0df6c02a07" - -inherit pypi setuptools3 - -RDEPENDS:${PN} += " \ - python3-ctypes \ - python3-logging \ -" diff --git a/meta-python/recipes-devtools/python/python3-inotify_git.bb b/meta-python/recipes-devtools/python/python3-inotify_git.bb new file mode 100644 index 000000000..22615a97e --- /dev/null +++ b/meta-python/recipes-devtools/python/python3-inotify_git.bb @@ -0,0 +1,35 @@ +SUMMARY = "An adapter to Linux kernel support for inotify directory-watching." +AUTHOR = "Dustin Oprea" +HOMEPAGE = "https://pypi.org/project/inotify/" +LICENSE = "GPL-2.0-only" +LIC_FILES_CHKSUM = "file://setup.py;md5=79b7ec72aa0d446a552d3cceb5799e41;beginline=28;endline=28" + +SRC_URI[sha256sum] = "974a623a338482b62e16d4eb705fb863ed33ec178680fc3e96ccdf0df6c02a07" + +SRC_URI = " \ + git://github.com/dsoprea/pyinotify.git;branch=master;protocol=https \ + file://run-ptest \ +" + +SRCREV = "288a2b01fd4a726dc68d960f8351630aee788a13" + +S = "${WORKDIR}/git" + +inherit setuptools3 ptest +PIP_INSTALL_PACKAGE = "inotify" +PIP_INSTALL_DIST_PATH = "${S}/dist" + + +RDEPENDS:${PN} += " \ + python3-ctypes \ + python3-logging \ +" + +RDEPENDS:${PN}-ptest += " \ + ${PYTHON_PN}-pytest \ +" + +do_install_ptest() { + install -d ${D}${PTEST_PATH}/tests + cp -rf ${S}/tests/* ${D}${PTEST_PATH}/tests/ +} -- 2.33.0 |
|
|
|
On 2022-11-25 14:56, Zheng Qiu wrote:
Add ptest provided from the source code, and run it using pytest.As you mentioned elsewhere: The two of the test are trying to make notifification using the package when a new directory has been created or a directory has been renamed. The issue occurs because the result in a list (i.e. ['IN_ISDIR', 'IN_CREATE']) is in different order but contains same components. The other test that is always failing is a test that makes a directory and trying to access those directories. The test is failing because a bunch of “width” variable is not as expected. so while the tests should be fixed, they don't seem to problems with the underlying functionality. We'll send fixes to upstream should they ever come out of open source hibernation. I still think we should use the git/HEAD as I'm suggested in other email. ../Randy Signed-off-by: Zheng Qiu <zheng.qiu@...>-- # Randy MacLeod # Wind River Linux |
|
|
|
Qiu, Zheng
While further investigating it, I start doubting if the tests are failing because of some issue in the base code.
Furthermore, if I remember correctly, there are actually more failing tests with later commits.
I will double-check to find the best strategy.
|
|
|