openstack
/
releases
Code Issues Proposed changes

maintain team ownership info in the deliverable files 

Teams add and remove repositories, change owners of repositories, and
are themselves removed over time. We can't rely on the current version
of the governance repository to include the information about who owned
a deliverable, so we need to put that information here in the releases
repository where we can track changes over time.

This patch adds support for a required "team" field to the deliverable
file, using the values set in previous patches in the series, and
updates the rendering code to use it instead of the project list from
the governance repository.

Signed-off-by: Doug Hellmann <doug@doughellmann.com>
Change-Id: I5941072c641c05bf0983984cae29a34927246d86
This commit is contained in:
Doug Hellmann 2016-06-09 14:33:31 -04:00
parent 66a5aaeb43
commit 9d12f20416
5 changed files with 308 additions and 207 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
README.rst
View File

@ -176,6 +176,10 @@ Deliverables File Schema
The top level of a deliverable file is a mapping with keys:
``team``
The name of the team that owns the deliverable, as listed in the
governance repository data files.
``launchpad``
The slug name of the launchpad project, suitable for use in URLs.

8
openstack_releases/cmds/validate.py
View File

@ -121,6 +121,14 @@ def main():
else:
print('found')
# Look for the team name
if 'team' not in deliverable_info:
errors.append('No team name given in %s' % filename)
print('no team name given')
elif deliverable_info['team'] not in team_data:
warnings.append('Team %r in %s not in governance data' %
(deliverable_info['team'], filename))
# Look for an email address to receive release announcements
try:
announce_to = deliverable_info['send-announcements-to']

184
openstack_releases/deliverable.py Normal file
View File

@ -0,0 +1,184 @@
# All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
"""Class for manipulating all of the deliverable data.
"""
from __future__ import print_function
import collections
import glob
import os
import pbr.version
import yaml
def _safe_semver(v):
"""Get a SemanticVersion that closely represents the version string.
We can't always get a SemanticVersion instance because some of the
legacy tags don't comply with the parser. This method corrects
some of the more common mistakes in formatting to make it more
likely we can construct a SemanticVersion, even if the results
don't quite match the input.
"""
orig = v = str(v)
# Remove "v" prefixes.
v = v.lstrip('v')
# Remove any stray "." at the start or end, after the other
# cleanups.
v = v.strip('.')
# If we have a version with 4 positions that are all integers,
# drop the fourth.
parts = v.split('.')
if len(parts) > 3:
try:
int(parts[3])
except ValueError:
pass
else:
parts = parts[:3]
v = '.'.join(parts)
if v != orig:
print(' changed version %r to %r' % (orig, v))
return pbr.version.SemanticVersion.from_pip_string(v)
def _version_sort_key(release):
"""Return a value we can compare for sorting.
"""
return _safe_semver(release['version'])
def _collapse_deliverable_history(name, info):
"""Collapse pre-releases into their final release.
Edit the info dictionary in place.
"""
sorted_releases = sorted(
info.get('releases', []),
key=_version_sort_key,
)
# Collapse pre-releases into their final release.
releases = []
known_versions = set()
for r in reversed(sorted_releases):
try:
parsed_vers = pbr.version.SemanticVersion.from_pip_string(
str(r['version']))
vers_tuple = parsed_vers.version_tuple()
except:
# If we can't parse the version, it must be some sort
# of made up legacy tag. Ignore the parse error
# and include the value in our output.
releases.append(r)
else:
if len(vers_tuple) != 3:
# This is not a normal release, so assume it
# is a pre-release.
final = parsed_vers.brief_string()
if final in known_versions:
print('[deliverables] ignoring %s %s' %
(name, r['version']))
continue
releases.append(r)
known_versions.add(r['version'])
info['releases'] = list(reversed(releases))
class Deliverables(object):
def __init__(self, root_dir):
self._root_dir = root_dir
# Map team names to a list of all of their deliverables.
self._team_deliverables = collections.defaultdict(set)
# Map team names to a set of all the series in which they
# produced anything.
self._team_series = collections.defaultdict(set)
# Map both team and series name to a list of the deliverable
# files.
self._by_team_and_series = collections.defaultdict(list)
self._by_series = collections.defaultdict(list)
# Map filenames to parsed content.
self._by_filename = {}
self._load_deliverable_files(root_dir)
def _load_deliverable_files(self, root_dir):
deliverable_files = glob.glob(os.path.join(root_dir, '*/*.yaml'))
for filename in sorted(deliverable_files):
print('[deliverables] reading %s' % filename)
series = self._series_from_filename(filename)
deliverable = self._deliverable_from_filename(filename)
with open(filename, 'r') as f:
d_info = yaml.load(f.read())
_collapse_deliverable_history(deliverable, d_info)
team = d_info['team']
self._add_deliverable_file(
filename, series, team, deliverable, d_info,
)
@staticmethod
def _series_from_filename(filename):
return os.path.basename(os.path.dirname(filename))
@staticmethod
def _deliverable_from_filename(filename):
return os.path.splitext(os.path.basename(filename))[0]
def _add_deliverable_file(self, filename, series, team, deliverable,
d_info):
self._by_filename[filename] = d_info
self._by_team_and_series[(team, series)].append(filename)
self._by_series[series].append(filename)
self._team_deliverables[team].add(deliverable)
self._team_series[team].add(series)
def get_team_deliverables(self, team):
"Returns a list of deliverable names produced by the team."
return list(sorted(self._team_deliverables[team]))
def get_team_series(self, team):
"Return the names of the series in which the team produced anything."
return self._team_series[team]
def get_teams(self):
"Return all of the names of all of the teams seen."
return list(self._team_series.keys())
def get_deliverables(self, team, series):
"""Return a sequence of deliverable data for the team and series.
Return tuples containing team, series, deliverable, and parsed
deliverable file content.
If the team or series is None, treat that value as a wildcard.
"""
if team is None:
if series is None:
series = '_independent'
filenames = self._by_series[series]
else:
filenames = self._by_team_and_series[(team, series)]
for filename in filenames:
yield (
team,
self._series_from_filename(filename),
self._deliverable_from_filename(filename),
self._by_filename.get(filename, {}),
)

293
openstack_releases/sphinxext.py
View File

@ -12,19 +12,17 @@
# License for the specific language governing permissions and limitations
# under the License.
import collections
import glob
import itertools
import operator
import os.path
from docutils import nodes
from docutils.parsers import rst
from docutils.parsers.rst import directives
from docutils.statemachine import ViewList
import pbr.version
from sphinx.util.nodes import nested_parse_with_titles
import yaml
from openstack_releases import deliverable
from openstack_releases import governance
@ -58,120 +56,33 @@ def _list_table(add, headers, data, title='', columns=None):
add('')
def _get_deliverable_type(deliverable_types, name):
if (name.startswith('python-') and not name.endswith('client')):
name = name[7:]
def _get_deliverable_type(name, data):
if (name.startswith('python-') and name.endswith('client')):
return 'type:library'
if name in deliverable_types:
return deliverable_types[name]
no_dashes = name.replace('-', '_')
if no_dashes in deliverable_types:
return deliverable_types[no_dashes]
return 'type:other'
def _safe_semver(v):
"""Get a SemanticVersion that closely represents the version string.
We can't always get a SemanticVersion instance because some of the
legacy tags don't comply with the parser. This method corrects
some of the more common mistakes in formatting to make it more
likely we can construct a SemanticVersion, even if the results
don't quite match the input.
"""
orig = v = str(v)
# Remove "v" prefixes.
v = v.lstrip('v')
# Remove any stray "." at the start or end, after the other
# cleanups.
v = v.strip('.')
# If we have a version with 4 positions that are all integers,
# drop the fourth.
parts = v.split('.')
if len(parts) > 3:
try:
int(parts[3])
except ValueError:
pass
else:
parts = parts[:3]
v = '.'.join(parts)
if v != orig:
print(' changed version %r to %r' % (orig, v))
return pbr.version.SemanticVersion.from_pip_string(v)
def _version_sort_key(release):
"""Return a value we can compare for sorting.
"""
return _safe_semver(release['version'])
def _collapse_deliverable_history(app, name, info):
"""Collapse pre-releases into their final release.
Edit the info dictionary in place.
"""
sorted_releases = sorted(
info.get('releases', []),
key=_version_sort_key,
)
# Collapse pre-releases into their final release.
releases = []
known_versions = set()
for r in reversed(sorted_releases):
try:
parsed_vers = pbr.version.SemanticVersion.from_pip_string(
str(r['version']))
vers_tuple = parsed_vers.version_tuple()
except:
# If we can't parse the version, it must be some sort
# of made up legacy tag. Ignore the parse error
# and include the value in our output.
releases.append(r)
else:
if len(vers_tuple) != 3:
# This is not a normal release, so assume it
# is a pre-release.
final = parsed_vers.brief_string()
if final in known_versions:
app.info('[deliverables] ignoring %s %s' %
(name, r['version']))
continue
releases.append(r)
known_versions.add(r['version'])
info['releases'] = list(reversed(releases))
_cached_deliverable_files = {}
def _get_deliverable_file_content(app, deliverable_name, filename):
if filename in _cached_deliverable_files:
return _cached_deliverable_files[filename]
app.info('[deliverables] reading %s' % filename)
with open(filename, 'r') as f:
d_info = yaml.load(f.read())
_collapse_deliverable_history(app, deliverable_name, d_info)
_cached_deliverable_files[filename] = d_info
return d_info
for tag in data.get('tags', []):
if tag == 'release:cycle-trailing':
return tag
if tag.startswith('type:'):
return tag
return _DEFAULT_TYPE
_DEFAULT_TYPE = 'type:other'
_deliverables = None
_all_teams = {}
_deliverable_files_by_team = {}
_all_deliverable_types = {}
def _initialize_team_data(app):
global _deliverables
global _all_teams
_deliverables = deliverable.Deliverables('deliverables')
team_data = governance.get_team_data()
for team in (governance.Team(n, i) for n, i in team_data.items()):
_all_teams[team.name] = team
_deliverable_files_by_team[team.name] = list(itertools.chain(
*(glob.glob('deliverables/*/%s.yaml' % dn)
for dn in team.deliverables)
))
for tn, td in team_data.items():
_all_teams[tn] = td
for dn, dd in td['deliverables'].items():
_all_deliverable_types[dn] = _get_deliverable_type(dn, dd)
class DeliverableDirectiveBase(rst.Directive):
@ -192,81 +103,70 @@ class DeliverableDirectiveBase(rst.Directive):
env = self.state.document.settings.env
app = env.app
# The series value is optional for some directives.
series = self.options.get('series')
# The series value is optional for some directives. If it is
# present but an empty string, convert to None so the
# Deliverables class will treat it like a wildcard.
series = self.options.get('series') or None
# If the user specifies a team, track only the deliverables
# for that team.
self.team_name = self.options.get('team')
self.team_deliverables = []
if self.team_name:
deliverables = _all_teams[self.team_name].deliverables
self.team_deliverables = list(deliverables.keys())
else:
deliverables = {}
for team in _all_teams.values():
deliverables.update(team.deliverables)
# Pre-populate the mapping between deliverable names and their
# types.
deliverable_types = {}
for dn, di in deliverables.items():
for tag in di.tags:
# Treat the cycle-trailing model as a separate "type"
# so those items are all grouped together in the
# output.
if tag == 'release:cycle-trailing':
deliverable_types[dn] = tag
break
if tag.startswith('type:'):
deliverable_types[dn] = tag
break
self.team_name = self.options.get('team') or None
result = ViewList()
# Assemble all of the deliverable data to be displayed and
# build the RST representation.
# get_deliverables() -> (team, series, deliverable, info)
if self.team_name:
deliverables = []
for filename in sorted(self._get_deliverables_files(series)):
deliverable_name = os.path.basename(filename)[:-5] # strip .yaml
d_info = _get_deliverable_file_content(
app, deliverable_name, filename,
)
deliverables.append(
(deliverable_name,
filename,
d_info))
self._add_deliverables(
None,
deliverables,
series,
app,
result,
# All deliverables are shown, in alphabetical order. They
# are organized by series but not type.
d_source = itertools.groupby(
sorted(_deliverables.get_deliverables(self.team_name, series)),
key=operator.itemgetter(1) # the series
)
for s, d in d_source:
self._add_deliverables(
None,
((i[2], i[3]) for i in d), # only name and info
s,
app,
result,
)
else:
deliverables = collections.defaultdict(list)
for filename in sorted(self._get_deliverables_files(series)):
deliverable_name = os.path.basename(filename)[:-5] # strip .yaml
deliverable_type = _get_deliverable_type(
deliverable_types,
deliverable_name,
# Only the deliverables for the given series are
# shown. They are organized by type. The type is only
# available from the governance data, so we have to add it
# to the raw data before sorting and grouping.
raw_deliverables = (
(_all_deliverable_types.get(d[2], _DEFAULT_TYPE), d[2], d[3])
for d in _deliverables.get_deliverables(
self.team_name,
series,
)
d_info = _get_deliverable_file_content(
app, deliverable_name, filename,
)
deliverables[deliverable_type].append(
(deliverable_name,
filename,
d_info))
)
raw_deliverables = list(raw_deliverables)
grouped = itertools.groupby(
sorted(raw_deliverables),
key=operator.itemgetter(0), # the deliverable type
)
# Convert the grouping iterators to a dictionary mapping
# type to the list of tuples with deliverable name and
# parsed deliverable info that _add_deliverables() needs.
by_type = {}
for deliverable_type, deliverables in grouped:
by_type[deliverable_type] = [
(d[1], d[2])
for d in deliverables
]
for type_tag in self._TYPE_ORDER:
if type_tag not in by_type:
app.info('No %r for %s' % (type_tag, (self.team_name, series)))
continue
self._add_deliverables(
type_tag,
deliverables[type_tag],
by_type[type_tag],
series,
app,
result,
@ -310,6 +210,7 @@ class DeliverableDirectiveBase(rst.Directive):
def _add_deliverables(self, type_tag, deliverables, series, app, result):
source_name = '<' + __name__ + '>'
deliverables = list(deliverables) # expand any generators passed in
if not deliverables:
# There are no deliverables of this type, and that's OK.
return
@ -326,7 +227,7 @@ class DeliverableDirectiveBase(rst.Directive):
# deliverable.
if not self.team_name:
most_recent = []
for deliverable_name, filename, deliverable_info in deliverables:
for deliverable_name, deliverable_info in deliverables:
earliest_version = deliverable_info.get('releases', {})[0].get(
'version', 'unreleased')
recent_version = deliverable_info.get('releases', {})[-1].get(
@ -342,22 +243,25 @@ class DeliverableDirectiveBase(rst.Directive):
)
else:
notes_link = '`release notes <%s>`__' % release_notes
most_recent.append((ref, earliest_version, recent_version, notes_link))
most_recent.append(
(ref, earliest_version, recent_version, notes_link)
)
_list_table(
lambda t: result.append(t, source_name),
['Deliverable', 'Earliest Version', 'Most Recent Version', 'Notes'],
['Deliverable', 'Earliest Version',
'Most Recent Version', 'Notes'],
most_recent,
title='Release Summary',
)
# Show the detailed history of the deliverables within the series.
for deliverable_name, filename, deliverable_info in deliverables:
for deliverable_name, deliverable_info in deliverables:
# These closures need to be redefined in each iteration of
# the loop because they use the filename.
# the loop because they use the deliverable name.
def _add(text):
result.append(text, filename)
result.append(text, '%s/%s' % (series, deliverable_name))
def _title(text, underline):
text = str(text) # version numbers might be seen as floats
@ -403,19 +307,6 @@ class DeliverableDirectiveBase(rst.Directive):
class DeliverableDirective(DeliverableDirectiveBase):
def _get_deliverables_files(self, series):
if self.team_name:
# Only show the deliverables associated with the team
# specified.
return itertools.chain(
*(glob.glob('deliverables/%s/%s.yaml' % (series, dn))
for dn in self.team_deliverables)
)
else:
# Show all of the deliverables for all teams producing
# anything in the series.
return glob.glob('deliverables/%s/*.yaml' % series)
def run(self):
# Require a series value.
series = self.options.get('series')
@ -430,9 +321,7 @@ class DeliverableDirective(DeliverableDirectiveBase):
class IndependentDeliverablesDirective(DeliverableDirectiveBase):
def _get_deliverables_files(self, series):
return glob.glob('deliverables/_independent/*.yaml')
pass
class TeamDirective(rst.Directive):
@ -453,20 +342,13 @@ class TeamDirective(rst.Directive):
line=self.lineno)
return [error]
if self.team_name not in _all_teams:
error = self.state_machine.reporter.error(
'Team %r not found in governance data' % self.team_name,
nodes.literal_block(self.block_text, self.block_text),
line=self.lineno)
return [error]
team = _all_teams[self.team_name]
self.team_deliverables = list(team.deliverables.keys())
self.team_deliverables = _deliverables.get_team_deliverables(
self.team_name
)
deliverable_files = _deliverable_files_by_team[self.team_name]
all_series = reversed(sorted(set(
os.path.basename(os.path.dirname(df))
for df in deliverable_files
)))
all_series = reversed(sorted(
_deliverables.get_team_series(self.team_name)
))
result = ViewList()
@ -493,10 +375,7 @@ class TeamDirective(rst.Directive):
def _generate_team_pages(app):
teams_with_deliverables = []
for team_name in sorted(_all_teams.keys()):
if _deliverable_files_by_team.get(team_name):
teams_with_deliverables.append(team_name)
teams_with_deliverables = list(sorted(_deliverables.get_teams()))
for team_name in teams_with_deliverables:
app.info('[team page] %s' % team_name)
slug = team_name.lower().replace('-', '_').replace(' ', '_')

26
tools/add_teams_to_deliverable_files.sh Executable file
View File

@ -0,0 +1,26 @@
#!/bin/sh
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
# Try to add team fields to deliverable files that don't have them.
list-deliverables | while read team deliverable; do
for filename in deliverables/*/${deliverable}.yaml; do
if [ -f $filename ]; then
if grep -q '^team:' $filename; then
continue
fi
echo $filename
sed -i "/^launchpad:/a team: $team" $filename
fi
done
done