@@ -63,7 +63,7 @@ if __name__ == '__main__':
for resultline in results.splitlines():
if testid in resultline:
- result, _ = resultline.split(' ', 1)
+ result, _ = resultline.split(':', 1)
if expected_result.upper() == "FAIL" and result.upper() == "FAIL":
xfailcount = xfailcount + 1
@@ -5,22 +5,22 @@
# SPDX-License-Identifier: GPL-2.0
import base
-import re
+import pyparsing
class Author(base.Base):
- auh_email = '<auh@auh.yoctoproject.org>'
+ auh_email = 'auh@auh.yoctoproject.org'
- invalids = [re.compile("^Upgrade Helper.+"),
- re.compile(re.escape(auh_email)),
- re.compile("uh@not\.set"),
- re.compile("\S+@example\.com")]
+ invalids = [pyparsing.Regex("^Upgrade Helper.+"),
+ pyparsing.Regex(auh_email),
+ pyparsing.Regex("uh@not\.set"),
+ pyparsing.Regex("\S+@example\.com")]
def test_author_valid(self):
for commit in self.commits:
for invalid in self.invalids:
- if invalid.search(commit.author):
+ if invalid.search_string(commit.author):
self.fail('Invalid author %s. Resend the series with a valid patch author' % commit.author, commit=commit)
def test_non_auh_upgrade(self):
@@ -4,17 +4,17 @@
#
# SPDX-License-Identifier: GPL-2.0
-import re
+import pyparsing
import base
class Bugzilla(base.Base):
- rexp_detect = re.compile("\[\s?YOCTO.*\]", re.IGNORECASE)
- rexp_validation = re.compile("\[(\s?YOCTO\s?#\s?(\d+)\s?,?)+\]", re.IGNORECASE)
+ rexp_detect = pyparsing.Regex('\[\s?YOCTO.*\]')
+ rexp_validation = pyparsing.Regex('\[(\s?YOCTO\s?#\s?(\d+)\s?,?)+\]')
def test_bugzilla_entry_format(self):
for commit in Bugzilla.commits:
- for line in commit.commit_message.splitlines():
- if self.rexp_detect.match(line):
- if not self.rexp_validation.match(line):
- self.fail('Bugzilla issue ID is not correctly formatted - specify it with format: "[YOCTO #<bugzilla ID>]"', commit=commit)
+ if not self.rexp_detect.search_string(commit.commit_message):
+ self.skip("No bug ID found")
+ elif not self.rexp_validation.search_string(commit.commit_message):
+ self.fail('Bugzilla issue ID is not correctly formatted - specify it with format: "[YOCTO #<bugzilla ID>]"', commit=commit)
@@ -20,12 +20,13 @@
import base
import os
import parse_cve_tags
-import re
+import pyparsing
class CVE(base.Base):
- revert_shortlog_regex = re.compile('Revert\s+".*"')
+ revert_shortlog_regex = pyparsing.Regex('Revert\s+".*"')
prog = parse_cve_tags.cve_tag
+ patch_prog = parse_cve_tags.patch_cve_tag
def setUp(self):
if self.unidiff_parse_error:
@@ -34,15 +35,17 @@ class CVE(base.Base):
# we are just interested in series that introduce CVE patches, thus discard other
# possibilities: modification to current CVEs, patch directly introduced into the
# recipe, upgrades already including the CVE, etc.
- new_cves = [p for p in self.patchset if p.path.endswith('.patch') and p.is_added_file]
- if not new_cves:
- self.skip('No new CVE patches introduced')
+ new_patches = [p for p in self.patchset if p.path.endswith('.patch') and p.is_added_file]
+ if not new_patches:
+ self.skip('No new patches introduced')
def test_cve_presence_in_commit_message(self):
for commit in CVE.commits:
# skip those patches that revert older commits, these do not required the tag presence
- if self.revert_shortlog_regex.match(commit.shortlog):
+ if self.revert_shortlog_regex.search_string(commit.shortlog):
continue
- if not self.prog.search_string(commit.payload):
+ if not self.patch_prog.search_string(commit.payload):
+ self.skip("No CVE tag in added patch, so not needed in mbox")
+ elif not self.prog.search_string(commit.payload):
self.fail('Missing or incorrectly formatted CVE tag in mbox. Correct or include the CVE tag in the mbox with format: "CVE: CVE-YYYY-XXXX"',
commit=commit)
@@ -5,11 +5,10 @@
# SPDX-License-Identifier: GPL-2.0
import base
-import re
class MboxFormat(base.Base):
def test_mbox_format(self):
if self.unidiff_parse_error:
self.fail('Series cannot be parsed correctly due to malformed diff lines. Create the series again using git-format-patch and ensure it can be applied using git am',
- data=[('Diff line', re.sub('^.+:\s(?<!$)','',self.unidiff_parse_error))])
+ data=[('Diff line',self.unidiff_parse_error)])
@@ -7,7 +7,7 @@
import subprocess
import collections
import base
-import re
+import pyparsing
from data import PatchTestInput
class MailingList(base.Base):
@@ -39,9 +39,9 @@ class MailingList(base.Base):
# a meta project may be indicted in the message subject, if this is the case, just fail
# TODO: there may be other project with no-meta prefix, we also need to detect these
- project_regex = re.compile("\[(?P<project>meta-.+)\]")
+ project_regex = pyparsing.Regex("\[(?P<project>meta-.+)\]")
for commit in MailingList.commits:
- match = project_regex.match(commit.subject)
+ match = project_regex.search_string(commit.subject)
if match:
self.fail('Series sent to the wrong mailing list. Check the project\'s README (%s) and send the patch to the indicated list' % match.group('project'),
commit=commit)
@@ -6,11 +6,11 @@
import base
import parse_signed_off_by
-import re
+import pyparsing
class SignedOffBy(base.Base):
- revert_shortlog_regex = re.compile('Revert\s+".*"')
+ revert_shortlog_regex = pyparsing.Regex('Revert\s+".*"')
@classmethod
def setUpClassLocal(cls):
@@ -20,7 +20,7 @@ class SignedOffBy(base.Base):
def test_signed_off_by_presence(self):
for commit in SignedOffBy.commits:
# skip those patches that revert older commits, these do not required the tag presence
- if self.revert_shortlog_regex.match(commit.shortlog):
+ if self.revert_shortlog_regex.search_string(commit.shortlog):
continue
if not SignedOffBy.prog.search_string(commit.payload):
self.fail('Mbox is missing Signed-off-by. Add it manually or with "git commit --amend -s"',
@@ -5,7 +5,7 @@
# SPDX-License-Identifier: GPL-2.0
import base
-import re
+import pyparsing
from data import PatchTestInput, PatchTestDataStore
class LicFilesChkSum(base.Metadata):
@@ -13,7 +13,7 @@ class LicFilesChkSum(base.Metadata):
license = 'LICENSE'
closed = 'CLOSED'
lictag = 'License-Update'
- lictag_re = re.compile("^%s:" % lictag, re.MULTILINE)
+ lictag_re = pyparsing.Regex("^%s:" % lictag)
def setUp(self):
# these tests just make sense on patches that can be merged
@@ -73,7 +73,7 @@ class LicFilesChkSum(base.Metadata):
if pretest != test:
# if any patch on the series contain reference on the metadata, fail
for commit in self.commits:
- if self.lictag_re.search(commit.commit_message):
+ if self.lictag_re.search_string(commit.commit_message):
break
else:
self.fail('LIC_FILES_CHKSUM changed on target %s but there is no "%s" tag in commit message. Include it with a brief description' % (pn, self.lictag),
@@ -5,10 +5,10 @@
# SPDX-License-Identifier: GPL-2.0
import base
-import re
+import pyparsing
class MaxLength(base.Base):
- add_mark = re.compile('\+ ')
+ add_mark = pyparsing.Regex('\+ ')
max_length = 200
def test_max_line_length(self):
@@ -18,7 +18,7 @@ class MaxLength(base.Base):
continue
payload = str(patch)
for line in payload.splitlines():
- if self.add_mark.match(line):
+ if self.add_mark.search_string(line):
current_line_length = len(line[1:])
if current_line_length > self.max_length:
self.fail('Patch line too long (current length %s, maximum is %s)' % (current_line_length, self.max_length),
@@ -6,8 +6,8 @@
import subprocess
import base
-import re
import os
+import pyparsing
from data import PatchTestInput, PatchTestDataStore
class SrcUri(base.Metadata):
@@ -15,7 +15,7 @@ class SrcUri(base.Metadata):
metadata = 'SRC_URI'
md5sum = 'md5sum'
sha256sum = 'sha256sum'
- git_regex = re.compile('^git\:\/\/.*')
+ git_regex = pyparsing.Regex('^git\:\/\/.*')
def setUp(self):
# these tests just make sense on patches that can be merged
@@ -19,12 +19,12 @@
import base
import os
-import re
+import pyparsing
class CVE(base.Base):
- re_cve_pattern = re.compile("CVE\-\d{4}\-\d+", re.IGNORECASE)
- re_cve_payload_tag = re.compile("\+CVE:(\s+CVE\-\d{4}\-\d+)+")
+ re_cve_pattern = pyparsing.Regex("CVE\-\d{4}\-\d+")
+ re_cve_payload_tag = pyparsing.Regex("\+CVE:(\s+CVE\-\d{4}\-\d+)+")
def setUp(self):
if self.unidiff_parse_error:
@@ -39,10 +39,10 @@ class CVE(base.Base):
def test_cve_tag_format(self):
for commit in CVE.commits:
- if self.re_cve_pattern.search(commit.shortlog) or self.re_cve_pattern.search(commit.commit_message):
+ if self.re_cve_pattern.search_string(commit.shortlog) or self.re_cve_pattern.search_string(commit.commit_message):
tag_found = False
for line in commit.payload.splitlines():
- if self.re_cve_payload_tag.match(line):
+ if self.re_cve_payload_tag.search_string(line):
tag_found = True
break
if not tag_found:
@@ -6,7 +6,6 @@
import base
import parse_signed_off_by
-import re
class PatchSignedOffBy(base.Base):
Various tweaks to make the test suite cleaner and more efficient: - Replace use of "re" module with "pyparsing" in tests (but not base.py) - Make test_mbox_cve only check for CVE tags in the commit if the added patch has them - Make test_mbox_cve SKIP instead of PASS if there's no CVE tag - Simplify the bugzilla tag checking test now that pyparsing is used - Modify the selftest script to correctly parse the new result output Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com> --- meta/lib/patchtest/selftest/selftest | 2 +- meta/lib/patchtest/tests/test_mbox_author.py | 14 +++++++------- meta/lib/patchtest/tests/test_mbox_bugzilla.py | 14 +++++++------- meta/lib/patchtest/tests/test_mbox_cve.py | 17 ++++++++++------- meta/lib/patchtest/tests/test_mbox_format.py | 3 +-- .../patchtest/tests/test_mbox_mailinglist.py | 6 +++--- .../patchtest/tests/test_mbox_signed_off_by.py | 6 +++--- .../tests/test_metadata_lic_files_chksum.py | 6 +++--- .../patchtest/tests/test_metadata_max_length.py | 6 +++--- .../patchtest/tests/test_metadata_src_uri.py | 4 ++-- meta/lib/patchtest/tests/test_patch_cve.py | 10 +++++----- .../patchtest/tests/test_patch_signed_off_by.py | 1 - 12 files changed, 45 insertions(+), 44 deletions(-)