More merging adjustments.

Looks like this should be in pretty good shape and has passed some
of the basic backwards compat. merging tests that I added.
This commit is contained in:
Joshua Harlow 2013-05-03 14:41:28 -07:00
parent 7ea735dd35
commit 2ec7b1fe37
10 changed files with 201 additions and 34 deletions

View File

@ -25,7 +25,7 @@ from cloudinit import type_utils
NAME_MTCH = re.compile(r"(^[a-zA-Z_][A-Za-z0-9_]*)\((.*?)\)$")
LOG = logging.getLogger(__name__)
DEF_MERGE_TYPE = "list()+dict()"
DEF_MERGE_TYPE = "list()+dict()+str()"
MERGER_PREFIX = 'm_'
MERGER_ATTR = 'Merger'

View File

@ -16,21 +16,32 @@
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
DEF_MERGE_TYPE = 'no_replace'
MERGE_TYPES = ('replace', DEF_MERGE_TYPE,)
def _has_any(what, *keys):
for k in keys:
if k in what:
return True
return False
class Merger(object):
def __init__(self, merger, opts):
self._merger = merger
# Affects merging behavior...
self._method = 'replace'
for m in ['replace', 'no_replace']:
self._method = DEF_MERGE_TYPE
for m in MERGE_TYPES:
if m in opts:
self._method = m
break
# Affect how recursive merging is done on other primitives
# Affect how recursive merging is done on other primitives.
self._recurse_str = 'recurse_str' in opts
self._recurse_dict = True
self._recurse_array = 'recurse_array' in opts
self._recurse_array = _has_any(opts, 'recurse_array', 'recurse_list')
self._allow_delete = 'allow_delete' in opts
# Backwards compat require this to be on.
self._recurse_dict = True
def __str__(self):
s = ('DictMerger: (method=%s,recurse_str=%s,'
@ -42,14 +53,14 @@ class Merger(object):
self._allow_delete)
return s
def _do_dict_replace(self, value, merge_with, do_replace=True):
def _do_dict_replace(self, value, merge_with, do_replace):
def merge_same_key(old_v, new_v):
if do_replace:
return new_v
if isinstance(new_v, (list, tuple)) and self._recurse_array:
return self._merger.merge(old_v, new_v)
if isinstance(new_v, (str, basestring)) and self._recurse_str:
if isinstance(new_v, (basestring)) and self._recurse_str:
return self._merger.merge(old_v, new_v)
if isinstance(new_v, (dict)) and self._recurse_dict:
return self._merger.merge(old_v, new_v)
@ -70,7 +81,7 @@ class Merger(object):
if not isinstance(merge_with, (dict)):
return value
if self._method == 'replace':
merged = self._do_dict_replace(dict(value), merge_with)
merged = self._do_dict_replace(dict(value), merge_with, True)
elif self._method == 'no_replace':
merged = self._do_dict_replace(dict(value), merge_with, False)
else:

View File

@ -32,7 +32,8 @@ class Merger(object):
self._recurse_array = 'recurse_array' in opts
def __str__(self):
return 'ListMerger: (m=%s,rs=%s,rd=%s,ra=%s)' % (self._method,
return ('ListMerger: (method=%s,recurse_str=%s,'
'recurse_dict=%s,recurse_array=%s)') % (self._method,
self._recurse_str,
self._recurse_dict,
self._recurse_array)

View File

@ -0,0 +1,44 @@
# -*- coding: utf-8 -*-
# vi: ts=4 expandtab
#
# Copyright (C) 2012 Yahoo! Inc.
#
# Author: Joshua Harlow <harlowja@yahoo-inc.com>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License version 3, as
# published by the Free Software Foundation.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
class Merger(object):
def __init__(self, _merger, opts):
self._append = 'append' in opts
def __str__(self):
return 'StringMerger: (append=%s)' % (self._append)
# On encountering a unicode object to merge value with
# we will for now just proxy into the string method to let it handle it.
def _on_unicode(self, value, merge_with):
return self._on_str(value, merge_with)
# On encountering a string object to merge with we will
# perform the following action, if appending we will
# merge them together, otherwise we will just return value.
def _on_str(self, value, merge_with):
if not isinstance(value, (basestring)):
return merge_with
if not self._append:
return merge_with
if isinstance(value, unicode):
return value + unicode(merge_with)
else:
return value + str(merge_with)

View File

@ -1,3 +1,3 @@
Blah: 3
Blah: 1
Blah2: 2
Blah3: [1]
Blah3: 3

View File

@ -0,0 +1,7 @@
#cloud-config
Blah: 3
Blah2: 2
Blah3: [1]

View File

@ -0,0 +1,6 @@
#cloud-config
Blah: 1
Blah2: 2
Blah3: 3

View File

@ -0,0 +1,8 @@
#cloud-config
Blah: 3
Blah2: 2
Blah3: [1]
merge_how: 'dict(replace)+list(append)'

View File

@ -11,14 +11,39 @@ import glob
import os
import re
SOURCE_PAT = "source*.*yaml"
EXPECTED_PAT = "expected%s.yaml"
def _old_mergedict(src, cand):
"""
Merge values from C{cand} into C{src}.
If C{src} has a key C{cand} will not override.
Nested dictionaries are merged recursively.
"""
if isinstance(src, dict) and isinstance(cand, dict):
for (k, v) in cand.iteritems():
if k not in src:
src[k] = v
else:
src[k] = _old_mergedict(src[k], v)
return src
def _old_mergemanydict(*args):
out = {}
for a in args:
out = _old_mergedict(out, a)
return out
class TestSimpleRun(helpers.ResourceUsingTestCase):
def _load_merge_files(self, data_dir):
merge_root = self.resourceLocation(data_dir)
def _load_merge_files(self):
merge_root = self.resourceLocation('merge_sources')
tests = []
source_ids = collections.defaultdict(list)
expected_files = {}
for fn in glob.glob(os.path.join(merge_root, "source*.*yaml")):
for fn in glob.glob(os.path.join(merge_root, SOURCE_PAT)):
base_fn = os.path.basename(fn)
file_id = re.match(r"source(\d+)\-(\d+)[.]yaml", base_fn)
if not file_id:
@ -26,31 +51,97 @@ class TestSimpleRun(helpers.ResourceUsingTestCase):
% (fn))
file_id = int(file_id.group(1))
source_ids[file_id].append(fn)
expected_fn = os.path.join(merge_root,
"expected%s.yaml" % (file_id))
expected_fn = os.path.join(merge_root, EXPECTED_PAT % (file_id))
if not os.path.isfile(expected_fn):
raise IOError("No expected file found at %s" % (expected_fn))
expected_files[file_id] = expected_fn
for id in sorted(source_ids.keys()):
for i in sorted(source_ids.keys()):
source_file_contents = []
for fn in sorted(source_ids[id]):
source_file_contents.append(util.load_file(fn))
expected = util.load_yaml(util.load_file(expected_files[id]))
tests.append((source_file_contents, expected))
for fn in sorted(source_ids[i]):
source_file_contents.append([fn, util.load_file(fn)])
expected = util.load_yaml(util.load_file(expected_files[i]))
entry = [source_file_contents, [expected, expected_files[i]]]
tests.append(entry)
return tests
def test_merge_samples(self):
tests = self._load_merge_files('merge_sources')
tests = self._load_merge_files()
paths = c_helpers.Paths({})
cc_handler = cloud_config.CloudConfigPartHandler(paths)
cc_handler.cloud_fn = None
for (payloads, expected_merge) in tests:
for (payloads, (expected_merge, expected_fn)) in tests:
cc_handler.handle_part(None, CONTENT_START, None,
None, None, None)
for (i, p) in enumerate(payloads):
cc_handler.handle_part(None, None, "t-%s.yaml" % (i + 1),
p, None, {})
merging_fns = []
for (fn, contents) in payloads:
cc_handler.handle_part(None, None, "%s.yaml" % (fn),
contents, None, {})
merging_fns.append(fn)
merged_buf = cc_handler.cloud_buf
cc_handler.handle_part(None, CONTENT_END, None,
None, None, None)
self.assertEquals(expected_merge, merged_buf)
fail_msg = "Equality failure on checking %s with %s: %s != %s"
fail_msg = fail_msg % (expected_fn,
",".join(merging_fns), merged_buf,
expected_merge)
self.assertEquals(expected_merge, merged_buf, msg=fail_msg)
def test_compat_merges_dict(self):
a = {
'1': '2',
'b': 'c',
}
b = {
'b': 'e',
}
c = _old_mergedict(a, b)
d = util.mergemanydict([a, b])
self.assertEquals(c, d)
def test_compat_merges_list(self):
a = {'b': [1, 2, 3]}
b = {'b': [4, 5]}
c = {'b': [6, 7]}
e = _old_mergemanydict(a, b, c)
f = util.mergemanydict([a, b, c])
self.assertEquals(e, f)
def test_compat_merges_str(self):
a = {'b': "hi"}
b = {'b': "howdy"}
c = {'b': "hallo"}
e = _old_mergemanydict(a, b, c)
f = util.mergemanydict([a, b, c])
self.assertEquals(e, f)
def test_compat_merge_sub_dict(self):
a = {
'1': '2',
'b': {
'f': 'g',
}
}
b = {
'b': {
'e': 'c',
}
}
c = _old_mergedict(a, b)
d = util.mergemanydict([a, b])
self.assertEquals(c, d)
def test_compat_merge_sub_list(self):
a = {
'1': '2',
'b': {
'f': ['1'],
}
}
b = {
'b': {
'f': [],
}
}
c = _old_mergedict(a, b)
d = util.mergemanydict([a, b])
self.assertEquals(c, d)

View File

@ -60,7 +60,6 @@ run:
- c
'''
message1 = MIMEBase("text", "cloud-config")
message1['Merge-Type'] = 'dict()+list(extend)+str(append)'
message1.set_payload(blob)
blob2 = '''
@ -72,7 +71,8 @@ run:
- morestuff
'''
message2 = MIMEBase("text", "cloud-config")
message2['X-Merge-Type'] = 'dict()+list(extend)+str()'
message2['X-Merge-Type'] = ('dict(recurse_array,'
'recurse_str)+list(append)+str(append)')
message2.set_payload(blob2)
blob3 = '''
@ -84,7 +84,6 @@ e:
p: 1
'''
message3 = MIMEBase("text", "cloud-config")
message3['Merge-Type'] = 'dict()+list()+str()'
message3.set_payload(blob3)
messages = [message1, message2, message3]