4763.2.4
by John Arbash Meinel
merge bzr.2.1 in preparation for NEWS entry. |
1 |
# Copyright (C) 2007-2010 Canonical Ltd
|
2592.1.4
by Robert Collins
Create a GraphIndexBuilder. |
2 |
#
|
3 |
# This program is free software; you can redistribute it and/or modify
|
|
4 |
# it under the terms of the GNU General Public License as published by
|
|
5 |
# the Free Software Foundation; either version 2 of the License, or
|
|
6 |
# (at your option) any later version.
|
|
7 |
#
|
|
8 |
# This program is distributed in the hope that it will be useful,
|
|
9 |
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
10 |
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
11 |
# GNU General Public License for more details.
|
|
12 |
#
|
|
13 |
# You should have received a copy of the GNU General Public License
|
|
14 |
# along with this program; if not, write to the Free Software
|
|
4183.7.1
by Sabin Iacob
update FSF mailing address |
15 |
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
2592.1.4
by Robert Collins
Create a GraphIndexBuilder. |
16 |
|
17 |
"""Tests for indices."""
|
|
18 |
||
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
19 |
from bzrlib import ( |
20 |
errors, |
|
21 |
index, |
|
22 |
tests, |
|
23 |
transport, |
|
24 |
)
|
|
25 |
||
26 |
||
27 |
class TestGraphIndexBuilder(tests.TestCaseWithMemoryTransport): |
|
2592.1.4
by Robert Collins
Create a GraphIndexBuilder. |
28 |
|
29 |
def test_build_index_empty(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
30 |
builder = index.GraphIndexBuilder() |
2592.1.4
by Robert Collins
Create a GraphIndexBuilder. |
31 |
stream = builder.finish() |
32 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
33 |
self.assertEqual( |
34 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=1\nlen=0\n\n", |
|
35 |
contents) |
|
2592.1.6
by Robert Collins
Record the number of node reference lists a particular index has. |
36 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
37 |
def test_build_index_empty_two_element_keys(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
38 |
builder = index.GraphIndexBuilder(key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
39 |
stream = builder.finish() |
40 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
41 |
self.assertEqual( |
42 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=2\nlen=0\n\n", |
|
43 |
contents) |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
44 |
|
2592.1.6
by Robert Collins
Record the number of node reference lists a particular index has. |
45 |
def test_build_index_one_reference_list_empty(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
46 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2592.1.6
by Robert Collins
Record the number of node reference lists a particular index has. |
47 |
stream = builder.finish() |
48 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
49 |
self.assertEqual( |
50 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=1\nlen=0\n\n", |
|
51 |
contents) |
|
2592.1.4
by Robert Collins
Create a GraphIndexBuilder. |
52 |
|
2592.1.10
by Robert Collins
Make validate detect node reference parsing errors. |
53 |
def test_build_index_two_reference_list_empty(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
54 |
builder = index.GraphIndexBuilder(reference_lists=2) |
2592.1.10
by Robert Collins
Make validate detect node reference parsing errors. |
55 |
stream = builder.finish() |
56 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
57 |
self.assertEqual( |
58 |
"Bazaar Graph Index 1\nnode_ref_lists=2\nkey_elements=1\nlen=0\n\n", |
|
59 |
contents) |
|
2592.1.10
by Robert Collins
Make validate detect node reference parsing errors. |
60 |
|
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
61 |
def test_build_index_one_node_no_refs(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
62 |
builder = index.GraphIndexBuilder() |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
63 |
builder.add_node(('akey', ), 'data') |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
64 |
stream = builder.finish() |
65 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
66 |
self.assertEqual( |
67 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=1\nlen=1\n" |
|
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
68 |
"akey\x00\x00\x00data\n\n", contents) |
69 |
||
70 |
def test_build_index_one_node_no_refs_accepts_empty_reflist(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
71 |
builder = index.GraphIndexBuilder() |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
72 |
builder.add_node(('akey', ), 'data', ()) |
2592.1.12
by Robert Collins
Handle basic node adds. |
73 |
stream = builder.finish() |
74 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
75 |
self.assertEqual( |
76 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=1\nlen=1\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
77 |
"akey\x00\x00\x00data\n\n", contents) |
2592.1.12
by Robert Collins
Handle basic node adds. |
78 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
79 |
def test_build_index_one_node_2_element_keys(self): |
2624.2.11
by Robert Collins
Review comments. |
80 |
# multipart keys are separated by \x00 - because they are fixed length,
|
81 |
# not variable this does not cause any issues, and seems clearer to the
|
|
82 |
# author.
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
83 |
builder = index.GraphIndexBuilder(key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
84 |
builder.add_node(('akey', 'secondpart'), 'data') |
85 |
stream = builder.finish() |
|
86 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
87 |
self.assertEqual( |
88 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=2\nlen=1\n" |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
89 |
"akey\x00secondpart\x00\x00\x00data\n\n", contents) |
90 |
||
2592.1.21
by Robert Collins
Empty values are ok. |
91 |
def test_add_node_empty_value(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
92 |
builder = index.GraphIndexBuilder() |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
93 |
builder.add_node(('akey', ), '') |
2592.1.21
by Robert Collins
Empty values are ok. |
94 |
stream = builder.finish() |
95 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
96 |
self.assertEqual( |
97 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=1\nlen=1\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
98 |
"akey\x00\x00\x00\n\n", contents) |
2592.1.21
by Robert Collins
Empty values are ok. |
99 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
100 |
def test_build_index_nodes_sorted(self): |
2592.1.17
by Robert Collins
Multi node sort order is defined. |
101 |
# the highest sorted node comes first.
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
102 |
builder = index.GraphIndexBuilder() |
2592.1.17
by Robert Collins
Multi node sort order is defined. |
103 |
# use three to have a good chance of glitching dictionary hash
|
104 |
# lookups etc. Insert in randomish order that is not correct
|
|
105 |
# and not the reverse of the correct order.
|
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
106 |
builder.add_node(('2002', ), 'data') |
107 |
builder.add_node(('2000', ), 'data') |
|
108 |
builder.add_node(('2001', ), 'data') |
|
2592.1.17
by Robert Collins
Multi node sort order is defined. |
109 |
stream = builder.finish() |
110 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
111 |
self.assertEqual( |
112 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=1\nlen=3\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
113 |
"2000\x00\x00\x00data\n" |
114 |
"2001\x00\x00\x00data\n" |
|
115 |
"2002\x00\x00\x00data\n" |
|
2592.1.17
by Robert Collins
Multi node sort order is defined. |
116 |
"\n", contents) |
117 |
||
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
118 |
def test_build_index_2_element_key_nodes_sorted(self): |
119 |
# multiple element keys are sorted first-key, second-key.
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
120 |
builder = index.GraphIndexBuilder(key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
121 |
# use three values of each key element, to have a good chance of
|
122 |
# glitching dictionary hash lookups etc. Insert in randomish order that
|
|
123 |
# is not correct and not the reverse of the correct order.
|
|
124 |
builder.add_node(('2002', '2002'), 'data') |
|
125 |
builder.add_node(('2002', '2000'), 'data') |
|
126 |
builder.add_node(('2002', '2001'), 'data') |
|
127 |
builder.add_node(('2000', '2002'), 'data') |
|
128 |
builder.add_node(('2000', '2000'), 'data') |
|
129 |
builder.add_node(('2000', '2001'), 'data') |
|
130 |
builder.add_node(('2001', '2002'), 'data') |
|
131 |
builder.add_node(('2001', '2000'), 'data') |
|
132 |
builder.add_node(('2001', '2001'), 'data') |
|
133 |
stream = builder.finish() |
|
134 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
135 |
self.assertEqual( |
136 |
"Bazaar Graph Index 1\nnode_ref_lists=0\nkey_elements=2\nlen=9\n" |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
137 |
"2000\x002000\x00\x00\x00data\n" |
138 |
"2000\x002001\x00\x00\x00data\n" |
|
139 |
"2000\x002002\x00\x00\x00data\n" |
|
140 |
"2001\x002000\x00\x00\x00data\n" |
|
141 |
"2001\x002001\x00\x00\x00data\n" |
|
142 |
"2001\x002002\x00\x00\x00data\n" |
|
143 |
"2002\x002000\x00\x00\x00data\n" |
|
144 |
"2002\x002001\x00\x00\x00data\n" |
|
145 |
"2002\x002002\x00\x00\x00data\n" |
|
146 |
"\n", contents) |
|
147 |
||
2592.1.19
by Robert Collins
Node references are tab separated. |
148 |
def test_build_index_reference_lists_are_included_one(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
149 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
150 |
builder.add_node(('key', ), 'data', ([], )) |
2592.1.19
by Robert Collins
Node references are tab separated. |
151 |
stream = builder.finish() |
152 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
153 |
self.assertEqual( |
154 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=1\nlen=1\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
155 |
"key\x00\x00\x00data\n" |
2592.1.19
by Robert Collins
Node references are tab separated. |
156 |
"\n", contents) |
157 |
||
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
158 |
def test_build_index_reference_lists_with_2_element_keys(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
159 |
builder = index.GraphIndexBuilder(reference_lists=1, key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
160 |
builder.add_node(('key', 'key2'), 'data', ([], )) |
161 |
stream = builder.finish() |
|
162 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
163 |
self.assertEqual( |
164 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=2\nlen=1\n" |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
165 |
"key\x00key2\x00\x00\x00data\n" |
166 |
"\n", contents) |
|
167 |
||
2592.1.19
by Robert Collins
Node references are tab separated. |
168 |
def test_build_index_reference_lists_are_included_two(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
169 |
builder = index.GraphIndexBuilder(reference_lists=2) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
170 |
builder.add_node(('key', ), 'data', ([], [])) |
2592.1.19
by Robert Collins
Node references are tab separated. |
171 |
stream = builder.finish() |
172 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
173 |
self.assertEqual( |
174 |
"Bazaar Graph Index 1\nnode_ref_lists=2\nkey_elements=1\nlen=1\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
175 |
"key\x00\x00\t\x00data\n" |
2592.1.19
by Robert Collins
Node references are tab separated. |
176 |
"\n", contents) |
177 |
||
4744.2.7
by John Arbash Meinel
Add .clear_cache() members to GraphIndexBuilder and BTreeBuilder. |
178 |
def test_clear_cache(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
179 |
builder = index.GraphIndexBuilder(reference_lists=2) |
4744.2.7
by John Arbash Meinel
Add .clear_cache() members to GraphIndexBuilder and BTreeBuilder. |
180 |
# This is a no-op, but the api should exist
|
181 |
builder.clear_cache() |
|
182 |
||
2592.1.22
by Robert Collins
Node references are byte offsets. |
183 |
def test_node_references_are_byte_offsets(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
184 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
185 |
builder.add_node(('reference', ), 'data', ([], )) |
186 |
builder.add_node(('key', ), 'data', ([('reference', )], )) |
|
2592.1.22
by Robert Collins
Node references are byte offsets. |
187 |
stream = builder.finish() |
188 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
189 |
self.assertEqual( |
190 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=1\nlen=2\n" |
|
191 |
"key\x00\x0072\x00data\n" |
|
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
192 |
"reference\x00\x00\x00data\n" |
2592.1.22
by Robert Collins
Node references are byte offsets. |
193 |
"\n", contents) |
194 |
||
2592.1.23
by Robert Collins
node reference delimiting tested. |
195 |
def test_node_references_are_cr_delimited(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
196 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
197 |
builder.add_node(('reference', ), 'data', ([], )) |
198 |
builder.add_node(('reference2', ), 'data', ([], )) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
199 |
builder.add_node(('key', ), 'data', |
200 |
([('reference', ), ('reference2', )], )) |
|
2592.1.23
by Robert Collins
node reference delimiting tested. |
201 |
stream = builder.finish() |
202 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
203 |
self.assertEqual( |
204 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=1\nlen=3\n" |
|
205 |
"key\x00\x00077\r094\x00data\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
206 |
"reference\x00\x00\x00data\n" |
207 |
"reference2\x00\x00\x00data\n" |
|
2592.1.23
by Robert Collins
node reference delimiting tested. |
208 |
"\n", contents) |
209 |
||
2592.1.24
by Robert Collins
Delimiting of multiple reference lists is by \t |
210 |
def test_multiple_reference_lists_are_tab_delimited(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
211 |
builder = index.GraphIndexBuilder(reference_lists=2) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
212 |
builder.add_node(('keference', ), 'data', ([], [])) |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
213 |
builder.add_node(('rey', ), 'data', |
214 |
([('keference', )], [('keference', )])) |
|
2592.1.24
by Robert Collins
Delimiting of multiple reference lists is by \t |
215 |
stream = builder.finish() |
216 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
217 |
self.assertEqual( |
218 |
"Bazaar Graph Index 1\nnode_ref_lists=2\nkey_elements=1\nlen=2\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
219 |
"keference\x00\x00\t\x00data\n" |
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
220 |
"rey\x00\x0059\t59\x00data\n" |
2592.1.24
by Robert Collins
Delimiting of multiple reference lists is by \t |
221 |
"\n", contents) |
222 |
||
2592.1.25
by Robert Collins
Fix and tune node offset calculation. |
223 |
def test_add_node_referencing_missing_key_makes_absent(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
224 |
builder = index.GraphIndexBuilder(reference_lists=1) |
225 |
builder.add_node(('rey', ), 'data', |
|
226 |
([('beference', ), ('aeference2', )], )) |
|
2592.1.25
by Robert Collins
Fix and tune node offset calculation. |
227 |
stream = builder.finish() |
228 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
229 |
self.assertEqual( |
230 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=1\nlen=1\n" |
|
2592.1.43
by Robert Collins
Various index tweaks and test clarity from John's review. |
231 |
"aeference2\x00a\x00\x00\n" |
232 |
"beference\x00a\x00\x00\n" |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
233 |
"rey\x00\x00074\r059\x00data\n" |
2592.1.25
by Robert Collins
Fix and tune node offset calculation. |
234 |
"\n", contents) |
235 |
||
2592.1.26
by Robert Collins
Test digit buffering is accurate. |
236 |
def test_node_references_three_digits(self): |
237 |
# test the node digit expands as needed.
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
238 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
239 |
references = [(str(val), ) for val in reversed(range(9))] |
240 |
builder.add_node(('2-key', ), '', (references, )) |
|
2592.1.26
by Robert Collins
Test digit buffering is accurate. |
241 |
stream = builder.finish() |
242 |
contents = stream.read() |
|
4789.28.2
by John Arbash Meinel
Get rid of the GraphIndexBuilder/BTreeBuilder._keys attribute. |
243 |
self.assertEqualDiff( |
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
244 |
"Bazaar Graph Index 1\nnode_ref_lists=1\nkey_elements=1\nlen=1\n" |
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
245 |
"0\x00a\x00\x00\n" |
246 |
"1\x00a\x00\x00\n" |
|
247 |
"2\x00a\x00\x00\n" |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
248 |
"2-key\x00\x00151\r145\r139\r133\r127\r121\r071\r065\r059\x00\n" |
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
249 |
"3\x00a\x00\x00\n" |
250 |
"4\x00a\x00\x00\n" |
|
251 |
"5\x00a\x00\x00\n" |
|
252 |
"6\x00a\x00\x00\n" |
|
253 |
"7\x00a\x00\x00\n" |
|
2592.1.26
by Robert Collins
Test digit buffering is accurate. |
254 |
"8\x00a\x00\x00\n" |
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
255 |
"\n", contents) |
256 |
||
257 |
def test_absent_has_no_reference_overhead(self): |
|
258 |
# the offsets after an absent record should be correct when there are
|
|
259 |
# >1 reference lists.
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
260 |
builder = index.GraphIndexBuilder(reference_lists=2) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
261 |
builder.add_node(('parent', ), '', ([('aail', ), ('zther', )], [])) |
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
262 |
stream = builder.finish() |
263 |
contents = stream.read() |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
264 |
self.assertEqual( |
265 |
"Bazaar Graph Index 1\nnode_ref_lists=2\nkey_elements=1\nlen=1\n" |
|
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
266 |
"aail\x00a\x00\x00\n" |
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
267 |
"parent\x00\x0059\r84\t\x00\n" |
2592.1.40
by Robert Collins
Reverse index ordering - we do not have date prefixed revids. |
268 |
"zther\x00a\x00\x00\n" |
2592.1.26
by Robert Collins
Test digit buffering is accurate. |
269 |
"\n", contents) |
270 |
||
2592.1.13
by Robert Collins
Handle mismatched numbers of reference lists. |
271 |
def test_add_node_bad_key(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
272 |
builder = index.GraphIndexBuilder() |
2592.1.14
by Robert Collins
Detect bad reference key values. |
273 |
for bad_char in '\t\n\x0b\x0c\r\x00 ': |
274 |
self.assertRaises(errors.BadIndexKey, builder.add_node, |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
275 |
('a%skey' % bad_char, ), 'data') |
276 |
self.assertRaises(errors.BadIndexKey, builder.add_node, |
|
277 |
('', ), 'data') |
|
278 |
self.assertRaises(errors.BadIndexKey, builder.add_node, |
|
279 |
'not-a-tuple', 'data') |
|
280 |
# not enough length
|
|
281 |
self.assertRaises(errors.BadIndexKey, builder.add_node, |
|
282 |
(), 'data') |
|
283 |
# too long
|
|
284 |
self.assertRaises(errors.BadIndexKey, builder.add_node, |
|
285 |
('primary', 'secondary'), 'data') |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
286 |
# secondary key elements get checked too:
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
287 |
builder = index.GraphIndexBuilder(key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
288 |
for bad_char in '\t\n\x0b\x0c\r\x00 ': |
289 |
self.assertRaises(errors.BadIndexKey, builder.add_node, |
|
290 |
('prefix', 'a%skey' % bad_char), 'data') |
|
2592.1.12
by Robert Collins
Handle basic node adds. |
291 |
|
2592.1.13
by Robert Collins
Handle mismatched numbers of reference lists. |
292 |
def test_add_node_bad_data(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
293 |
builder = index.GraphIndexBuilder() |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
294 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
295 |
'data\naa') |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
296 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
297 |
'data\x00aa') |
2592.1.12
by Robert Collins
Handle basic node adds. |
298 |
|
2592.1.13
by Robert Collins
Handle mismatched numbers of reference lists. |
299 |
def test_add_node_bad_mismatched_ref_lists_length(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
300 |
builder = index.GraphIndexBuilder() |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
301 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
302 |
'data aa', ([], )) |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
303 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
304 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
305 |
'data aa') |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
306 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
307 |
'data aa', (), ) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
308 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
309 |
'data aa', ([], [])) |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
310 |
builder = index.GraphIndexBuilder(reference_lists=2) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
311 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
312 |
'data aa') |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
313 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
314 |
'data aa', ([], )) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
315 |
self.assertRaises(errors.BadIndexValue, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
316 |
'data aa', ([], [], [])) |
2592.1.13
by Robert Collins
Handle mismatched numbers of reference lists. |
317 |
|
2592.1.14
by Robert Collins
Detect bad reference key values. |
318 |
def test_add_node_bad_key_in_reference_lists(self): |
319 |
# first list, first key - trivial
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
320 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
321 |
self.assertRaises(errors.BadIndexKey, builder.add_node, ('akey', ), |
322 |
'data aa', ([('a key', )], )) |
|
323 |
# references keys must be tuples too
|
|
324 |
self.assertRaises(errors.BadIndexKey, builder.add_node, ('akey', ), |
|
325 |
'data aa', (['not-a-tuple'], )) |
|
326 |
# not enough length
|
|
327 |
self.assertRaises(errors.BadIndexKey, builder.add_node, ('akey', ), |
|
328 |
'data aa', ([()], )) |
|
329 |
# too long
|
|
330 |
self.assertRaises(errors.BadIndexKey, builder.add_node, ('akey', ), |
|
331 |
'data aa', ([('primary', 'secondary')], )) |
|
2592.1.14
by Robert Collins
Detect bad reference key values. |
332 |
# need to check more than the first key in the list
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
333 |
self.assertRaises(errors.BadIndexKey, builder.add_node, ('akey', ), |
334 |
'data aa', ([('agoodkey', ), ('that is a bad key', )], )) |
|
2592.1.14
by Robert Collins
Detect bad reference key values. |
335 |
# and if there is more than one list it should be getting checked
|
336 |
# too
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
337 |
builder = index.GraphIndexBuilder(reference_lists=2) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
338 |
self.assertRaises(errors.BadIndexKey, builder.add_node, ('akey', ), |
2592.1.46
by Robert Collins
Make GraphIndex accept nodes as key, value, references, so that the method |
339 |
'data aa', ([], ['a bad key'])) |
2592.1.14
by Robert Collins
Detect bad reference key values. |
340 |
|
2592.1.15
by Robert Collins
Detect duplicate key insertion. |
341 |
def test_add_duplicate_key(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
342 |
builder = index.GraphIndexBuilder() |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
343 |
builder.add_node(('key', ), 'data') |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
344 |
self.assertRaises(errors.BadIndexDuplicateKey, |
345 |
builder.add_node, ('key', ), 'data') |
|
2592.1.15
by Robert Collins
Detect duplicate key insertion. |
346 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
347 |
def test_add_duplicate_key_2_elements(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
348 |
builder = index.GraphIndexBuilder(key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
349 |
builder.add_node(('key', 'key'), 'data') |
350 |
self.assertRaises(errors.BadIndexDuplicateKey, builder.add_node, |
|
351 |
('key', 'key'), 'data') |
|
352 |
||
2592.1.16
by Robert Collins
Can add keys after referencing them. |
353 |
def test_add_key_after_referencing_key(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
354 |
builder = index.GraphIndexBuilder(reference_lists=1) |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
355 |
builder.add_node(('key', ), 'data', ([('reference', )], )) |
356 |
builder.add_node(('reference', ), 'data', ([],)) |
|
2592.1.16
by Robert Collins
Can add keys after referencing them. |
357 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
358 |
def test_add_key_after_referencing_key_2_elements(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
359 |
builder = index.GraphIndexBuilder(reference_lists=1, key_elements=2) |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
360 |
builder.add_node(('k', 'ey'), 'data', ([('reference', 'tokey')], )) |
361 |
builder.add_node(('reference', 'tokey'), 'data', ([],)) |
|
362 |
||
3777.5.3
by John Arbash Meinel
Add Builder.set_optimize(for_size=True) for GraphIndexBuilder and BTreeBuilder. |
363 |
def test_set_optimize(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
364 |
builder = index.GraphIndexBuilder(reference_lists=1, key_elements=2) |
3777.5.3
by John Arbash Meinel
Add Builder.set_optimize(for_size=True) for GraphIndexBuilder and BTreeBuilder. |
365 |
builder.set_optimize(for_size=True) |
366 |
self.assertTrue(builder._optimize_for_size) |
|
367 |
builder.set_optimize(for_size=False) |
|
368 |
self.assertFalse(builder._optimize_for_size) |
|
369 |
||
2592.1.5
by Robert Collins
Trivial index reading. |
370 |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
371 |
class TestGraphIndex(tests.TestCaseWithMemoryTransport): |
2592.1.5
by Robert Collins
Trivial index reading. |
372 |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
373 |
def make_key(self, number): |
374 |
return (str(number) + 'X'*100,) |
|
375 |
||
376 |
def make_value(self, number): |
|
377 |
return str(number) + 'Y'*100 |
|
378 |
||
379 |
def make_nodes(self, count=64): |
|
380 |
# generate a big enough index that we only read some of it on a typical
|
|
381 |
# bisection lookup.
|
|
382 |
nodes = [] |
|
383 |
for counter in range(count): |
|
384 |
nodes.append((self.make_key(counter), self.make_value(counter), ())) |
|
385 |
return nodes |
|
386 |
||
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
387 |
def make_index(self, ref_lists=0, key_elements=1, nodes=[]): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
388 |
builder = index.GraphIndexBuilder(ref_lists, key_elements=key_elements) |
2624.2.17
by Robert Collins
Review feedback. |
389 |
for key, value, references in nodes: |
390 |
builder.add_node(key, value, references) |
|
2592.1.5
by Robert Collins
Trivial index reading. |
391 |
stream = builder.finish() |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
392 |
trans = transport.get_transport('trace+' + self.get_url()) |
2890.2.1
by Robert Collins
* ``bzrlib.index.GraphIndex`` now requires a size parameter to the |
393 |
size = trans.put_file('index', stream) |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
394 |
return index.GraphIndex(trans, 'index', size) |
2592.1.5
by Robert Collins
Trivial index reading. |
395 |
|
5074.4.3
by John Arbash Meinel
Actually implement offset support for GraphIndex. |
396 |
def make_index_with_offset(self, ref_lists=0, key_elements=1, nodes=[], |
397 |
offset=0): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
398 |
builder = index.GraphIndexBuilder(ref_lists, key_elements=key_elements) |
5074.4.3
by John Arbash Meinel
Actually implement offset support for GraphIndex. |
399 |
for key, value, references in nodes: |
400 |
builder.add_node(key, value, references) |
|
401 |
content = builder.finish().read() |
|
402 |
size = len(content) |
|
403 |
trans = self.get_transport() |
|
404 |
trans.put_bytes('index', (' '*offset) + content) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
405 |
return index.GraphIndex(trans, 'index', size, offset=offset) |
5074.4.3
by John Arbash Meinel
Actually implement offset support for GraphIndex. |
406 |
|
4744.2.6
by John Arbash Meinel
Start exposing an GraphIndex.clear_cache() member. |
407 |
def test_clear_cache(self): |
408 |
index = self.make_index() |
|
409 |
# For now, we just want to make sure the api is available. As this is
|
|
410 |
# old code, we don't really worry if it *does* anything.
|
|
411 |
index.clear_cache() |
|
412 |
||
2592.1.7
by Robert Collins
A validate that goes boom. |
413 |
def test_open_bad_index_no_error(self): |
414 |
trans = self.get_transport() |
|
415 |
trans.put_bytes('name', "not an index\n") |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
416 |
idx = index.GraphIndex(trans, 'name', 13) |
2592.1.7
by Robert Collins
A validate that goes boom. |
417 |
|
5074.4.3
by John Arbash Meinel
Actually implement offset support for GraphIndex. |
418 |
def test_with_offset(self): |
419 |
nodes = self.make_nodes(200) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
420 |
idx = self.make_index_with_offset(offset=1234567, nodes=nodes) |
421 |
self.assertEqual(200, idx.key_count()) |
|
5074.4.3
by John Arbash Meinel
Actually implement offset support for GraphIndex. |
422 |
|
423 |
def test_buffer_all_with_offset(self): |
|
424 |
nodes = self.make_nodes(200) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
425 |
idx = self.make_index_with_offset(offset=1234567, nodes=nodes) |
426 |
idx._buffer_all() |
|
427 |
self.assertEqual(200, idx.key_count()) |
|
5074.4.3
by John Arbash Meinel
Actually implement offset support for GraphIndex. |
428 |
|
429 |
def test_side_effect_buffering_with_offset(self): |
|
430 |
nodes = self.make_nodes(20) |
|
431 |
index = self.make_index_with_offset(offset=1234567, nodes=nodes) |
|
432 |
index._transport.recommended_page_size = lambda:64*1024 |
|
433 |
subset_nodes = [nodes[0][0], nodes[10][0], nodes[19][0]] |
|
434 |
entries = [n[1] for n in index.iter_entries(subset_nodes)] |
|
435 |
self.assertEqual(sorted(subset_nodes), sorted(entries)) |
|
436 |
self.assertEqual(20, index.key_count()) |
|
5074.4.2
by John Arbash Meinel
Add 'offset=' to the GraphIndex api, but refuse to let it be nonzero for now. |
437 |
|
2890.2.2
by Robert Collins
Opening an index creates a map for the parsed bytes. |
438 |
def test_open_sets_parsed_map_empty(self): |
439 |
index = self.make_index() |
|
440 |
self.assertEqual([], index._parsed_byte_map) |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
441 |
self.assertEqual([], index._parsed_key_map) |
442 |
||
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
443 |
def test_key_count_buffers(self): |
444 |
index = self.make_index(nodes=self.make_nodes(2)) |
|
445 |
# reset the transport log
|
|
446 |
del index._transport._activity[:] |
|
447 |
self.assertEqual(2, index.key_count()) |
|
448 |
# We should have requested reading the header bytes
|
|
449 |
self.assertEqual([ |
|
450 |
('readv', 'index', [(0, 200)], True, index._size), |
|
451 |
],
|
|
452 |
index._transport._activity) |
|
453 |
# And that should have been enough to trigger reading the whole index
|
|
454 |
# with buffering
|
|
455 |
self.assertIsNot(None, index._nodes) |
|
456 |
||
457 |
def test_lookup_key_via_location_buffers(self): |
|
458 |
index = self.make_index() |
|
459 |
# reset the transport log
|
|
460 |
del index._transport._activity[:] |
|
461 |
# do a _lookup_keys_via_location call for the middle of the file, which
|
|
462 |
# is what bisection uses.
|
|
463 |
result = index._lookup_keys_via_location( |
|
464 |
[(index._size // 2, ('missing', ))]) |
|
465 |
# this should have asked for a readv request, with adjust_for_latency,
|
|
466 |
# and two regions: the header, and half-way into the file.
|
|
467 |
self.assertEqual([ |
|
468 |
('readv', 'index', [(30, 30), (0, 200)], True, 60), |
|
469 |
],
|
|
470 |
index._transport._activity) |
|
471 |
# and the result should be that the key cannot be present, because this
|
|
472 |
# is a trivial index.
|
|
473 |
self.assertEqual([((index._size // 2, ('missing', )), False)], |
|
474 |
result) |
|
475 |
# And this should have caused the file to be fully buffered
|
|
476 |
self.assertIsNot(None, index._nodes) |
|
477 |
self.assertEqual([], index._parsed_byte_map) |
|
478 |
||
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
479 |
def test_first_lookup_key_via_location(self): |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
480 |
# We need enough data so that the _HEADER_READV doesn't consume the
|
481 |
# whole file. We always read 800 bytes for every key, and the local
|
|
482 |
# transport natural expansion is 4096 bytes. So we have to have >8192
|
|
483 |
# bytes or we will trigger "buffer_all".
|
|
484 |
# We also want the 'missing' key to fall within the range that *did*
|
|
485 |
# read
|
|
486 |
nodes = [] |
|
487 |
index = self.make_index(nodes=self.make_nodes(64)) |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
488 |
# reset the transport log
|
489 |
del index._transport._activity[:] |
|
2890.2.18
by Robert Collins
Review feedback. |
490 |
# do a _lookup_keys_via_location call for the middle of the file, which
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
491 |
# is what bisection uses.
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
492 |
start_lookup = index._size // 2 |
2890.2.18
by Robert Collins
Review feedback. |
493 |
result = index._lookup_keys_via_location( |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
494 |
[(start_lookup, ('40missing', ))]) |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
495 |
# this should have asked for a readv request, with adjust_for_latency,
|
496 |
# and two regions: the header, and half-way into the file.
|
|
497 |
self.assertEqual([ |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
498 |
('readv', 'index', |
499 |
[(start_lookup, 800), (0, 200)], True, index._size), |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
500 |
],
|
501 |
index._transport._activity) |
|
502 |
# and the result should be that the key cannot be present, because this
|
|
503 |
# is a trivial index.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
504 |
self.assertEqual([((start_lookup, ('40missing', )), False)], |
505 |
result) |
|
506 |
# And this should not have caused the file to be fully buffered
|
|
507 |
self.assertIs(None, index._nodes) |
|
508 |
# And the regions of the file that have been parsed should be in the
|
|
509 |
# parsed_byte_map and the parsed_key_map
|
|
510 |
self.assertEqual([(0, 4008), (5046, 8996)], index._parsed_byte_map) |
|
511 |
self.assertEqual([(None, self.make_key(26)), |
|
512 |
(self.make_key(31), self.make_key(48))], |
|
513 |
index._parsed_key_map) |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
514 |
|
515 |
def test_parsing_non_adjacent_data_trims(self): |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
516 |
index = self.make_index(nodes=self.make_nodes(64)) |
2890.2.18
by Robert Collins
Review feedback. |
517 |
result = index._lookup_keys_via_location( |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
518 |
[(index._size // 2, ('40', ))]) |
519 |
# and the result should be that the key cannot be present, because key is
|
|
520 |
# in the middle of the observed data from a 4K read - the smallest transport
|
|
521 |
# will do today with this api.
|
|
522 |
self.assertEqual([((index._size // 2, ('40', )), False)], |
|
523 |
result) |
|
524 |
# and we should have a parse map that includes the header and the
|
|
525 |
# region that was parsed after trimming.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
526 |
self.assertEqual([(0, 4008), (5046, 8996)], index._parsed_byte_map) |
527 |
self.assertEqual([(None, self.make_key(26)), |
|
528 |
(self.make_key(31), self.make_key(48))], |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
529 |
index._parsed_key_map) |
530 |
||
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
531 |
def test_parsing_data_handles_parsed_contained_regions(self): |
532 |
# the following patten creates a parsed region that is wholly within a
|
|
533 |
# single result from the readv layer:
|
|
534 |
# .... single-read (readv-minimum-size) ...
|
|
535 |
# which then trims the start and end so the parsed size is < readv
|
|
536 |
# miniumum.
|
|
537 |
# then a dual lookup (or a reference lookup for that matter) which
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
538 |
# abuts or overlaps the parsed region on both sides will need to
|
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
539 |
# discard the data in the middle, but parse the end as well.
|
540 |
#
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
541 |
# we test this by doing a single lookup to seed the data, then
|
542 |
# a lookup for two keys that are present, and adjacent -
|
|
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
543 |
# we except both to be found, and the parsed byte map to include the
|
544 |
# locations of both keys.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
545 |
index = self.make_index(nodes=self.make_nodes(128)) |
2890.2.18
by Robert Collins
Review feedback. |
546 |
result = index._lookup_keys_via_location( |
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
547 |
[(index._size // 2, ('40', ))]) |
548 |
# and we should have a parse map that includes the header and the
|
|
549 |
# region that was parsed after trimming.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
550 |
self.assertEqual([(0, 4045), (11759, 15707)], index._parsed_byte_map) |
551 |
self.assertEqual([(None, self.make_key(116)), |
|
552 |
(self.make_key(35), self.make_key(51))], |
|
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
553 |
index._parsed_key_map) |
554 |
# now ask for two keys, right before and after the parsed region
|
|
2890.2.18
by Robert Collins
Review feedback. |
555 |
result = index._lookup_keys_via_location( |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
556 |
[(11450, self.make_key(34)), (15707, self.make_key(52))]) |
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
557 |
self.assertEqual([ |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
558 |
((11450, self.make_key(34)), |
559 |
(index, self.make_key(34), self.make_value(34))), |
|
560 |
((15707, self.make_key(52)), |
|
561 |
(index, self.make_key(52), self.make_value(52))), |
|
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
562 |
],
|
563 |
result) |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
564 |
self.assertEqual([(0, 4045), (9889, 17993)], index._parsed_byte_map) |
2890.2.14
by Robert Collins
Parse more than one segment of data from a single readv response if needed. |
565 |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
566 |
def test_lookup_missing_key_answers_without_io_when_map_permits(self): |
567 |
# generate a big enough index that we only read some of it on a typical
|
|
568 |
# bisection lookup.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
569 |
index = self.make_index(nodes=self.make_nodes(64)) |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
570 |
# lookup the keys in the middle of the file
|
2890.2.18
by Robert Collins
Review feedback. |
571 |
result =index._lookup_keys_via_location( |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
572 |
[(index._size // 2, ('40', ))]) |
573 |
# check the parse map, this determines the test validity
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
574 |
self.assertEqual([(0, 4008), (5046, 8996)], index._parsed_byte_map) |
575 |
self.assertEqual([(None, self.make_key(26)), |
|
576 |
(self.make_key(31), self.make_key(48))], |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
577 |
index._parsed_key_map) |
578 |
# reset the transport log
|
|
579 |
del index._transport._activity[:] |
|
580 |
# now looking up a key in the portion of the file already parsed should
|
|
581 |
# not create a new transport request, and should return False (cannot
|
|
582 |
# be in the index) - even when the byte location we ask for is outside
|
|
583 |
# the parsed region
|
|
2890.2.18
by Robert Collins
Review feedback. |
584 |
result = index._lookup_keys_via_location( |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
585 |
[(4000, ('40', ))]) |
586 |
self.assertEqual([((4000, ('40', )), False)], |
|
587 |
result) |
|
588 |
self.assertEqual([], index._transport._activity) |
|
589 |
||
590 |
def test_lookup_present_key_answers_without_io_when_map_permits(self): |
|
591 |
# generate a big enough index that we only read some of it on a typical
|
|
592 |
# bisection lookup.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
593 |
index = self.make_index(nodes=self.make_nodes(64)) |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
594 |
# lookup the keys in the middle of the file
|
2890.2.18
by Robert Collins
Review feedback. |
595 |
result =index._lookup_keys_via_location( |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
596 |
[(index._size // 2, ('40', ))]) |
597 |
# check the parse map, this determines the test validity
|
|
598 |
self.assertEqual([(0, 4008), (5046, 8996)], index._parsed_byte_map) |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
599 |
self.assertEqual([(None, self.make_key(26)), |
600 |
(self.make_key(31), self.make_key(48))], |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
601 |
index._parsed_key_map) |
602 |
# reset the transport log
|
|
603 |
del index._transport._activity[:] |
|
604 |
# now looking up a key in the portion of the file already parsed should
|
|
605 |
# not create a new transport request, and should return False (cannot
|
|
606 |
# be in the index) - even when the byte location we ask for is outside
|
|
607 |
# the parsed region
|
|
3943.8.1
by Marius Kruger
remove all trailing whitespace from bzr source |
608 |
#
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
609 |
result = index._lookup_keys_via_location([(4000, self.make_key(40))]) |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
610 |
self.assertEqual( |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
611 |
[((4000, self.make_key(40)), |
612 |
(index, self.make_key(40), self.make_value(40)))], |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
613 |
result) |
614 |
self.assertEqual([], index._transport._activity) |
|
615 |
||
616 |
def test_lookup_key_below_probed_area(self): |
|
617 |
# generate a big enough index that we only read some of it on a typical
|
|
618 |
# bisection lookup.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
619 |
index = self.make_index(nodes=self.make_nodes(64)) |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
620 |
# ask for the key in the middle, but a key that is located in the
|
621 |
# unparsed region before the middle.
|
|
2890.2.18
by Robert Collins
Review feedback. |
622 |
result =index._lookup_keys_via_location( |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
623 |
[(index._size // 2, ('30', ))]) |
624 |
# check the parse map, this determines the test validity
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
625 |
self.assertEqual([(0, 4008), (5046, 8996)], index._parsed_byte_map) |
626 |
self.assertEqual([(None, self.make_key(26)), |
|
627 |
(self.make_key(31), self.make_key(48))], |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
628 |
index._parsed_key_map) |
629 |
self.assertEqual([((index._size // 2, ('30', )), -1)], |
|
630 |
result) |
|
631 |
||
632 |
def test_lookup_key_above_probed_area(self): |
|
633 |
# generate a big enough index that we only read some of it on a typical
|
|
634 |
# bisection lookup.
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
635 |
index = self.make_index(nodes=self.make_nodes(64)) |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
636 |
# ask for the key in the middle, but a key that is located in the
|
637 |
# unparsed region after the middle.
|
|
2890.2.18
by Robert Collins
Review feedback. |
638 |
result =index._lookup_keys_via_location( |
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
639 |
[(index._size // 2, ('50', ))]) |
640 |
# check the parse map, this determines the test validity
|
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
641 |
self.assertEqual([(0, 4008), (5046, 8996)], index._parsed_byte_map) |
642 |
self.assertEqual([(None, self.make_key(26)), |
|
643 |
(self.make_key(31), self.make_key(48))], |
|
2890.2.5
by Robert Collins
Create a content lookup function for bisection in GraphIndex. |
644 |
index._parsed_key_map) |
645 |
self.assertEqual([((index._size // 2, ('50', )), +1)], |
|
646 |
result) |
|
2890.2.2
by Robert Collins
Opening an index creates a map for the parsed bytes. |
647 |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
648 |
def test_lookup_key_resolves_references(self): |
649 |
# generate a big enough index that we only read some of it on a typical
|
|
650 |
# bisection lookup.
|
|
651 |
nodes = [] |
|
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
652 |
for counter in range(99): |
653 |
nodes.append((self.make_key(counter), self.make_value(counter), |
|
654 |
((self.make_key(counter + 20),),) )) |
|
655 |
index = self.make_index(ref_lists=1, nodes=nodes) |
|
656 |
# lookup a key in the middle that does not exist, so that when we can
|
|
657 |
# check that the referred-to-keys are not accessed automatically.
|
|
658 |
index_size = index._size |
|
659 |
index_center = index_size // 2 |
|
660 |
result = index._lookup_keys_via_location( |
|
661 |
[(index_center, ('40', ))]) |
|
662 |
# check the parse map - only the start and middle should have been
|
|
663 |
# parsed.
|
|
664 |
self.assertEqual([(0, 4027), (10198, 14028)], index._parsed_byte_map) |
|
665 |
self.assertEqual([(None, self.make_key(17)), |
|
666 |
(self.make_key(44), self.make_key(5))], |
|
667 |
index._parsed_key_map) |
|
668 |
# and check the transport activity likewise.
|
|
669 |
self.assertEqual( |
|
670 |
[('readv', 'index', [(index_center, 800), (0, 200)], True, |
|
671 |
index_size)], |
|
672 |
index._transport._activity) |
|
673 |
# reset the transport log for testing the reference lookup
|
|
674 |
del index._transport._activity[:] |
|
675 |
# now looking up a key in the portion of the file already parsed should
|
|
676 |
# only perform IO to resolve its key references.
|
|
677 |
result = index._lookup_keys_via_location([(11000, self.make_key(45))]) |
|
678 |
self.assertEqual( |
|
679 |
[((11000, self.make_key(45)), |
|
680 |
(index, self.make_key(45), self.make_value(45), |
|
681 |
((self.make_key(65),),)))], |
|
682 |
result) |
|
683 |
self.assertEqual([('readv', 'index', [(15093, 800)], True, index_size)], |
|
684 |
index._transport._activity) |
|
685 |
||
686 |
def test_lookup_key_can_buffer_all(self): |
|
687 |
nodes = [] |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
688 |
for counter in range(64): |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
689 |
nodes.append((self.make_key(counter), self.make_value(counter), |
690 |
((self.make_key(counter + 20),),) )) |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
691 |
index = self.make_index(ref_lists=1, nodes=nodes) |
692 |
# lookup a key in the middle that does not exist, so that when we can
|
|
693 |
# check that the referred-to-keys are not accessed automatically.
|
|
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
694 |
index_size = index._size |
695 |
index_center = index_size // 2 |
|
696 |
result = index._lookup_keys_via_location([(index_center, ('40', ))]) |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
697 |
# check the parse map - only the start and middle should have been
|
698 |
# parsed.
|
|
699 |
self.assertEqual([(0, 3890), (6444, 10274)], index._parsed_byte_map) |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
700 |
self.assertEqual([(None, self.make_key(25)), |
701 |
(self.make_key(37), self.make_key(52))], |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
702 |
index._parsed_key_map) |
703 |
# and check the transport activity likewise.
|
|
704 |
self.assertEqual( |
|
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
705 |
[('readv', 'index', [(index_center, 800), (0, 200)], True, |
706 |
index_size)], |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
707 |
index._transport._activity) |
708 |
# reset the transport log for testing the reference lookup
|
|
709 |
del index._transport._activity[:] |
|
710 |
# now looking up a key in the portion of the file already parsed should
|
|
711 |
# only perform IO to resolve its key references.
|
|
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
712 |
result = index._lookup_keys_via_location([(7000, self.make_key(40))]) |
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
713 |
self.assertEqual( |
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
714 |
[((7000, self.make_key(40)), |
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
715 |
(index, self.make_key(40), self.make_value(40), |
716 |
((self.make_key(60),),)))], |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
717 |
result) |
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
718 |
# Resolving the references would have required more data read, and we
|
719 |
# are already above the 50% threshold, so it triggered a _buffer_all
|
|
720 |
self.assertEqual([('get', 'index')], index._transport._activity) |
|
2890.2.6
by Robert Collins
Add support for key references to the index lookup_keys_via_location bisection interface. |
721 |
|
2592.1.5
by Robert Collins
Trivial index reading. |
722 |
def test_iter_all_entries_empty(self): |
723 |
index = self.make_index() |
|
724 |
self.assertEqual([], list(index.iter_all_entries())) |
|
725 |
||
2592.1.27
by Robert Collins
Test missing end lines with non-empty indices. |
726 |
def test_iter_all_entries_simple(self): |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
727 |
index = self.make_index(nodes=[(('name', ), 'data', ())]) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
728 |
self.assertEqual([(index, ('name', ), 'data')], |
2592.1.27
by Robert Collins
Test missing end lines with non-empty indices. |
729 |
list(index.iter_all_entries())) |
730 |
||
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
731 |
def test_iter_all_entries_simple_2_elements(self): |
732 |
index = self.make_index(key_elements=2, |
|
733 |
nodes=[(('name', 'surname'), 'data', ())]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
734 |
self.assertEqual([(index, ('name', 'surname'), 'data')], |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
735 |
list(index.iter_all_entries())) |
736 |
||
2592.1.28
by Robert Collins
Basic two pass iter_all_entries. |
737 |
def test_iter_all_entries_references_resolved(self): |
738 |
index = self.make_index(1, nodes=[ |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
739 |
(('name', ), 'data', ([('ref', )], )), |
740 |
(('ref', ), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
741 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),)), |
742 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
2592.1.28
by Robert Collins
Basic two pass iter_all_entries. |
743 |
set(index.iter_all_entries())) |
744 |
||
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
745 |
def test_iter_entries_buffers_once(self): |
746 |
index = self.make_index(nodes=self.make_nodes(2)) |
|
747 |
# reset the transport log
|
|
748 |
del index._transport._activity[:] |
|
749 |
self.assertEqual(set([(index, self.make_key(1), self.make_value(1))]), |
|
750 |
set(index.iter_entries([self.make_key(1)]))) |
|
751 |
# We should have requested reading the header bytes
|
|
752 |
# But not needed any more than that because it would have triggered a
|
|
753 |
# buffer all
|
|
754 |
self.assertEqual([ |
|
755 |
('readv', 'index', [(0, 200)], True, index._size), |
|
756 |
],
|
|
757 |
index._transport._activity) |
|
758 |
# And that should have been enough to trigger reading the whole index
|
|
759 |
# with buffering
|
|
760 |
self.assertIsNot(None, index._nodes) |
|
761 |
||
3665.3.3
by John Arbash Meinel
If we read more than 50% of the whole index, |
762 |
def test_iter_entries_buffers_by_bytes_read(self): |
763 |
index = self.make_index(nodes=self.make_nodes(64)) |
|
764 |
list(index.iter_entries([self.make_key(10)])) |
|
765 |
# The first time through isn't enough to trigger a buffer all
|
|
766 |
self.assertIs(None, index._nodes) |
|
767 |
self.assertEqual(4096, index._bytes_read) |
|
768 |
# Grabbing a key in that same page won't trigger a buffer all, as we
|
|
769 |
# still haven't read 50% of the file
|
|
770 |
list(index.iter_entries([self.make_key(11)])) |
|
771 |
self.assertIs(None, index._nodes) |
|
772 |
self.assertEqual(4096, index._bytes_read) |
|
773 |
# We haven't read more data, so reading outside the range won't trigger
|
|
774 |
# a buffer all right away
|
|
775 |
list(index.iter_entries([self.make_key(40)])) |
|
776 |
self.assertIs(None, index._nodes) |
|
777 |
self.assertEqual(8192, index._bytes_read) |
|
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
778 |
# On the next pass, we will not trigger buffer all if the key is
|
779 |
# available without reading more
|
|
3665.3.3
by John Arbash Meinel
If we read more than 50% of the whole index, |
780 |
list(index.iter_entries([self.make_key(32)])) |
3665.3.5
by John Arbash Meinel
Move the point at which we 'buffer_all' if we've read >50% of the index. |
781 |
self.assertIs(None, index._nodes) |
782 |
# But if we *would* need to read more to resolve it, then we will
|
|
783 |
# buffer all.
|
|
784 |
list(index.iter_entries([self.make_key(60)])) |
|
3665.3.3
by John Arbash Meinel
If we read more than 50% of the whole index, |
785 |
self.assertIsNot(None, index._nodes) |
786 |
||
2890.2.10
by Robert Collins
Add test coverage to ensure \r's are not mangled by bisection parsing. |
787 |
def test_iter_entries_references_resolved(self): |
788 |
index = self.make_index(1, nodes=[ |
|
789 |
(('name', ), 'data', ([('ref', ), ('ref', )], )), |
|
790 |
(('ref', ), 'refdata', ([], ))]) |
|
791 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),('ref',)),)), |
|
792 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
793 |
set(index.iter_entries([('name',), ('ref',)]))) |
|
794 |
||
795 |
def test_iter_entries_references_2_refs_resolved(self): |
|
796 |
index = self.make_index(2, nodes=[ |
|
797 |
(('name', ), 'data', ([('ref', )], [('ref', )])), |
|
798 |
(('ref', ), 'refdata', ([], []))]) |
|
799 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),), (('ref',),))), |
|
800 |
(index, ('ref', ), 'refdata', ((), ()))]), |
|
801 |
set(index.iter_entries([('name',), ('ref',)]))) |
|
802 |
||
2592.1.30
by Robert Collins
Absent entries are not yeilded. |
803 |
def test_iteration_absent_skipped(self): |
804 |
index = self.make_index(1, nodes=[ |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
805 |
(('name', ), 'data', ([('ref', )], ))]) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
806 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),))]), |
2592.1.30
by Robert Collins
Absent entries are not yeilded. |
807 |
set(index.iter_all_entries())) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
808 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),))]), |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
809 |
set(index.iter_entries([('name', )]))) |
810 |
self.assertEqual([], list(index.iter_entries([('ref', )]))) |
|
2592.1.30
by Robert Collins
Absent entries are not yeilded. |
811 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
812 |
def test_iteration_absent_skipped_2_element_keys(self): |
813 |
index = self.make_index(1, key_elements=2, nodes=[ |
|
814 |
(('name', 'fin'), 'data', ([('ref', 'erence')], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
815 |
self.assertEqual(set([(index, ('name', 'fin'), 'data', ((('ref', 'erence'),),))]), |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
816 |
set(index.iter_all_entries())) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
817 |
self.assertEqual(set([(index, ('name', 'fin'), 'data', ((('ref', 'erence'),),))]), |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
818 |
set(index.iter_entries([('name', 'fin')]))) |
819 |
self.assertEqual([], list(index.iter_entries([('ref', 'erence')]))) |
|
820 |
||
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
821 |
def test_iter_all_keys(self): |
2592.1.29
by Robert Collins
Basic iter_entries working. |
822 |
index = self.make_index(1, nodes=[ |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
823 |
(('name', ), 'data', ([('ref', )], )), |
824 |
(('ref', ), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
825 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),)), |
826 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
827 |
set(index.iter_entries([('name', ), ('ref', )]))) |
2592.1.29
by Robert Collins
Basic iter_entries working. |
828 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
829 |
def test_iter_nothing_empty(self): |
2592.1.9
by Robert Collins
Iterating no keys should work too. |
830 |
index = self.make_index() |
831 |
self.assertEqual([], list(index.iter_entries([]))) |
|
832 |
||
2592.1.5
by Robert Collins
Trivial index reading. |
833 |
def test_iter_missing_entry_empty(self): |
834 |
index = self.make_index() |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
835 |
self.assertEqual([], list(index.iter_entries([('a', )]))) |
2592.1.7
by Robert Collins
A validate that goes boom. |
836 |
|
2890.2.8
by Robert Collins
Make the size of the index optionally None for the pack-names index. |
837 |
def test_iter_missing_entry_empty_no_size(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
838 |
idx = self.make_index() |
839 |
idx = index.GraphIndex(idx._transport, 'index', None) |
|
840 |
self.assertEqual([], list(idx.iter_entries([('a', )]))) |
|
2890.2.8
by Robert Collins
Make the size of the index optionally None for the pack-names index. |
841 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
842 |
def test_iter_key_prefix_1_element_key_None(self): |
843 |
index = self.make_index() |
|
844 |
self.assertRaises(errors.BadIndexKey, list, |
|
845 |
index.iter_entries_prefix([(None, )])) |
|
846 |
||
847 |
def test_iter_key_prefix_wrong_length(self): |
|
848 |
index = self.make_index() |
|
849 |
self.assertRaises(errors.BadIndexKey, list, |
|
850 |
index.iter_entries_prefix([('foo', None)])) |
|
851 |
index = self.make_index(key_elements=2) |
|
852 |
self.assertRaises(errors.BadIndexKey, list, |
|
853 |
index.iter_entries_prefix([('foo', )])) |
|
854 |
self.assertRaises(errors.BadIndexKey, list, |
|
855 |
index.iter_entries_prefix([('foo', None, None)])) |
|
856 |
||
857 |
def test_iter_key_prefix_1_key_element_no_refs(self): |
|
858 |
index = self.make_index( nodes=[ |
|
859 |
(('name', ), 'data', ()), |
|
860 |
(('ref', ), 'refdata', ())]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
861 |
self.assertEqual(set([(index, ('name', ), 'data'), |
862 |
(index, ('ref', ), 'refdata')]), |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
863 |
set(index.iter_entries_prefix([('name', ), ('ref', )]))) |
864 |
||
865 |
def test_iter_key_prefix_1_key_element_refs(self): |
|
866 |
index = self.make_index(1, nodes=[ |
|
867 |
(('name', ), 'data', ([('ref', )], )), |
|
868 |
(('ref', ), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
869 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),)), |
870 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
871 |
set(index.iter_entries_prefix([('name', ), ('ref', )]))) |
872 |
||
873 |
def test_iter_key_prefix_2_key_element_no_refs(self): |
|
874 |
index = self.make_index(key_elements=2, nodes=[ |
|
875 |
(('name', 'fin1'), 'data', ()), |
|
876 |
(('name', 'fin2'), 'beta', ()), |
|
877 |
(('ref', 'erence'), 'refdata', ())]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
878 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data'), |
879 |
(index, ('ref', 'erence'), 'refdata')]), |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
880 |
set(index.iter_entries_prefix([('name', 'fin1'), ('ref', 'erence')]))) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
881 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data'), |
882 |
(index, ('name', 'fin2'), 'beta')]), |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
883 |
set(index.iter_entries_prefix([('name', None)]))) |
884 |
||
885 |
def test_iter_key_prefix_2_key_element_refs(self): |
|
886 |
index = self.make_index(1, key_elements=2, nodes=[ |
|
887 |
(('name', 'fin1'), 'data', ([('ref', 'erence')], )), |
|
888 |
(('name', 'fin2'), 'beta', ([], )), |
|
889 |
(('ref', 'erence'), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
890 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data', ((('ref', 'erence'),),)), |
891 |
(index, ('ref', 'erence'), 'refdata', ((), ))]), |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
892 |
set(index.iter_entries_prefix([('name', 'fin1'), ('ref', 'erence')]))) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
893 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data', ((('ref', 'erence'),),)), |
894 |
(index, ('name', 'fin2'), 'beta', ((), ))]), |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
895 |
set(index.iter_entries_prefix([('name', None)]))) |
896 |
||
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
897 |
def test_key_count_empty(self): |
898 |
index = self.make_index() |
|
899 |
self.assertEqual(0, index.key_count()) |
|
900 |
||
901 |
def test_key_count_one(self): |
|
902 |
index = self.make_index(nodes=[(('name', ), '', ())]) |
|
903 |
self.assertEqual(1, index.key_count()) |
|
904 |
||
905 |
def test_key_count_two(self): |
|
906 |
index = self.make_index(nodes=[ |
|
907 |
(('name', ), '', ()), (('foo', ), '', ())]) |
|
908 |
self.assertEqual(2, index.key_count()) |
|
909 |
||
3665.3.3
by John Arbash Meinel
If we read more than 50% of the whole index, |
910 |
def test_read_and_parse_tracks_real_read_value(self): |
911 |
index = self.make_index(nodes=self.make_nodes(10)) |
|
912 |
del index._transport._activity[:] |
|
913 |
index._read_and_parse([(0, 200)]) |
|
914 |
self.assertEqual([ |
|
915 |
('readv', 'index', [(0, 200)], True, index._size), |
|
916 |
],
|
|
917 |
index._transport._activity) |
|
918 |
# The readv expansion code will expand the initial request to 4096
|
|
919 |
# bytes, which is more than enough to read the entire index, and we
|
|
920 |
# will track the fact that we read that many bytes.
|
|
921 |
self.assertEqual(index._size, index._bytes_read) |
|
922 |
||
923 |
def test_read_and_parse_triggers_buffer_all(self): |
|
3665.3.1
by John Arbash Meinel
Updates to GraphIndex processing. |
924 |
index = self.make_index(key_elements=2, nodes=[ |
925 |
(('name', 'fin1'), 'data', ()), |
|
926 |
(('name', 'fin2'), 'beta', ()), |
|
927 |
(('ref', 'erence'), 'refdata', ())]) |
|
928 |
self.assertTrue(index._size > 0) |
|
929 |
self.assertIs(None, index._nodes) |
|
930 |
index._read_and_parse([(0, index._size)]) |
|
931 |
self.assertIsNot(None, index._nodes) |
|
932 |
||
2592.1.7
by Robert Collins
A validate that goes boom. |
933 |
def test_validate_bad_index_errors(self): |
934 |
trans = self.get_transport() |
|
935 |
trans.put_bytes('name', "not an index\n") |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
936 |
idx = index.GraphIndex(trans, 'name', 13) |
937 |
self.assertRaises(errors.BadIndexFormatSignature, idx.validate) |
|
2592.1.8
by Robert Collins
Empty files should validate ok. |
938 |
|
2592.1.10
by Robert Collins
Make validate detect node reference parsing errors. |
939 |
def test_validate_bad_node_refs(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
940 |
idx = self.make_index(2) |
2592.1.10
by Robert Collins
Make validate detect node reference parsing errors. |
941 |
trans = self.get_transport() |
942 |
content = trans.get_bytes('index') |
|
943 |
# change the options line to end with a rather than a parseable number
|
|
944 |
new_content = content[:-2] + 'a\n\n' |
|
945 |
trans.put_bytes('index', new_content) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
946 |
self.assertRaises(errors.BadIndexOptions, idx.validate) |
2592.1.10
by Robert Collins
Make validate detect node reference parsing errors. |
947 |
|
2592.1.27
by Robert Collins
Test missing end lines with non-empty indices. |
948 |
def test_validate_missing_end_line_empty(self): |
2592.1.11
by Robert Collins
Detect truncated indices. |
949 |
index = self.make_index(2) |
950 |
trans = self.get_transport() |
|
951 |
content = trans.get_bytes('index') |
|
952 |
# truncate the last byte
|
|
953 |
trans.put_bytes('index', content[:-1]) |
|
954 |
self.assertRaises(errors.BadIndexData, index.validate) |
|
955 |
||
2592.1.27
by Robert Collins
Test missing end lines with non-empty indices. |
956 |
def test_validate_missing_end_line_nonempty(self): |
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
957 |
index = self.make_index(2, nodes=[(('key', ), '', ([], []))]) |
2592.1.27
by Robert Collins
Test missing end lines with non-empty indices. |
958 |
trans = self.get_transport() |
959 |
content = trans.get_bytes('index') |
|
960 |
# truncate the last byte
|
|
961 |
trans.put_bytes('index', content[:-1]) |
|
962 |
self.assertRaises(errors.BadIndexData, index.validate) |
|
963 |
||
2592.1.8
by Robert Collins
Empty files should validate ok. |
964 |
def test_validate_empty(self): |
965 |
index = self.make_index() |
|
966 |
index.validate() |
|
2592.1.12
by Robert Collins
Handle basic node adds. |
967 |
|
968 |
def test_validate_no_refs_content(self): |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
969 |
index = self.make_index(nodes=[(('key', ), 'value', ())]) |
2592.1.12
by Robert Collins
Handle basic node adds. |
970 |
index.validate() |
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
971 |
|
4011.5.3
by Andrew Bennetts
Implement and test external_references on GraphIndex and BTreeGraphIndex. |
972 |
# XXX: external_references tests are duplicated in test_btree_index. We
|
973 |
# probably should have per_graph_index tests...
|
|
974 |
def test_external_references_no_refs(self): |
|
975 |
index = self.make_index(ref_lists=0, nodes=[]) |
|
976 |
self.assertRaises(ValueError, index.external_references, 0) |
|
977 |
||
978 |
def test_external_references_no_results(self): |
|
979 |
index = self.make_index(ref_lists=1, nodes=[ |
|
980 |
(('key',), 'value', ([],))]) |
|
981 |
self.assertEqual(set(), index.external_references(0)) |
|
982 |
||
983 |
def test_external_references_missing_ref(self): |
|
984 |
missing_key = ('missing',) |
|
985 |
index = self.make_index(ref_lists=1, nodes=[ |
|
986 |
(('key',), 'value', ([missing_key],))]) |
|
987 |
self.assertEqual(set([missing_key]), index.external_references(0)) |
|
988 |
||
989 |
def test_external_references_multiple_ref_lists(self): |
|
990 |
missing_key = ('missing',) |
|
991 |
index = self.make_index(ref_lists=2, nodes=[ |
|
992 |
(('key',), 'value', ([], [missing_key]))]) |
|
993 |
self.assertEqual(set([]), index.external_references(0)) |
|
994 |
self.assertEqual(set([missing_key]), index.external_references(1)) |
|
995 |
||
996 |
def test_external_references_two_records(self): |
|
997 |
index = self.make_index(ref_lists=1, nodes=[ |
|
998 |
(('key-1',), 'value', ([('key-2',)],)), |
|
999 |
(('key-2',), 'value', ([],)), |
|
1000 |
])
|
|
1001 |
self.assertEqual(set([]), index.external_references(0)) |
|
1002 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1003 |
def test__find_ancestors(self): |
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1004 |
key1 = ('key-1',) |
1005 |
key2 = ('key-2',) |
|
1006 |
index = self.make_index(ref_lists=1, key_elements=1, nodes=[ |
|
1007 |
(key1, 'value', ([key2],)), |
|
1008 |
(key2, 'value', ([],)), |
|
1009 |
])
|
|
1010 |
parent_map = {} |
|
1011 |
missing_keys = set() |
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1012 |
search_keys = index._find_ancestors([key1], 0, parent_map, missing_keys) |
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1013 |
self.assertEqual({key1: (key2,)}, parent_map) |
1014 |
self.assertEqual(set(), missing_keys) |
|
1015 |
self.assertEqual(set([key2]), search_keys) |
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1016 |
search_keys = index._find_ancestors(search_keys, 0, parent_map, |
1017 |
missing_keys) |
|
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1018 |
self.assertEqual({key1: (key2,), key2: ()}, parent_map) |
1019 |
self.assertEqual(set(), missing_keys) |
|
1020 |
self.assertEqual(set(), search_keys) |
|
1021 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1022 |
def test__find_ancestors_w_missing(self): |
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1023 |
key1 = ('key-1',) |
1024 |
key2 = ('key-2',) |
|
1025 |
key3 = ('key-3',) |
|
1026 |
index = self.make_index(ref_lists=1, key_elements=1, nodes=[ |
|
1027 |
(key1, 'value', ([key2],)), |
|
1028 |
(key2, 'value', ([],)), |
|
1029 |
])
|
|
1030 |
parent_map = {} |
|
1031 |
missing_keys = set() |
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1032 |
search_keys = index._find_ancestors([key2, key3], 0, parent_map, |
1033 |
missing_keys) |
|
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1034 |
self.assertEqual({key2: ()}, parent_map) |
1035 |
self.assertEqual(set([key3]), missing_keys) |
|
1036 |
self.assertEqual(set(), search_keys) |
|
1037 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1038 |
def test__find_ancestors_dont_search_known(self): |
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1039 |
key1 = ('key-1',) |
1040 |
key2 = ('key-2',) |
|
1041 |
key3 = ('key-3',) |
|
1042 |
index = self.make_index(ref_lists=1, key_elements=1, nodes=[ |
|
1043 |
(key1, 'value', ([key2],)), |
|
1044 |
(key2, 'value', ([key3],)), |
|
1045 |
(key3, 'value', ([],)), |
|
1046 |
])
|
|
1047 |
# We already know about key2, so we won't try to search for key3
|
|
1048 |
parent_map = {key2: (key3,)} |
|
1049 |
missing_keys = set() |
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1050 |
search_keys = index._find_ancestors([key1], 0, parent_map, |
1051 |
missing_keys) |
|
4593.4.7
by John Arbash Meinel
Basic implementation of a conforming interface for GraphIndex. |
1052 |
self.assertEqual({key1: (key2,), key2: (key3,)}, parent_map) |
1053 |
self.assertEqual(set(), missing_keys) |
|
1054 |
self.assertEqual(set(), search_keys) |
|
1055 |
||
4634.71.1
by John Arbash Meinel
Work around bug #402623 by allowing BTreeGraphIndex(...,unlimited_cache=True). |
1056 |
def test_supports_unlimited_cache(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1057 |
builder = index.GraphIndexBuilder(0, key_elements=1) |
4634.71.1
by John Arbash Meinel
Work around bug #402623 by allowing BTreeGraphIndex(...,unlimited_cache=True). |
1058 |
stream = builder.finish() |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1059 |
trans = transport.get_transport(self.get_url()) |
4634.71.1
by John Arbash Meinel
Work around bug #402623 by allowing BTreeGraphIndex(...,unlimited_cache=True). |
1060 |
size = trans.put_file('index', stream) |
1061 |
# It doesn't matter what unlimited_cache does here, just that it can be
|
|
1062 |
# passed
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1063 |
idx = index.GraphIndex(trans, 'index', size, unlimited_cache=True) |
1064 |
||
1065 |
||
1066 |
class TestCombinedGraphIndex(tests.TestCaseWithMemoryTransport): |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1067 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
1068 |
def make_index(self, name, ref_lists=0, key_elements=1, nodes=[]): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1069 |
builder = index.GraphIndexBuilder(ref_lists, key_elements=key_elements) |
2624.2.17
by Robert Collins
Review feedback. |
1070 |
for key, value, references in nodes: |
1071 |
builder.add_node(key, value, references) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1072 |
stream = builder.finish() |
1073 |
trans = self.get_transport() |
|
2890.2.1
by Robert Collins
* ``bzrlib.index.GraphIndex`` now requires a size parameter to the |
1074 |
size = trans.put_file(name, stream) |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1075 |
return index.GraphIndex(trans, name, size) |
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1076 |
|
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1077 |
def make_combined_index_with_missing(self, missing=['1', '2']): |
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1078 |
"""Create a CombinedGraphIndex which will have missing indexes.
|
1079 |
||
1080 |
This creates a CGI which thinks it has 2 indexes, however they have
|
|
1081 |
been deleted. If CGI._reload_func() is called, then it will repopulate
|
|
1082 |
with a new index.
|
|
1083 |
||
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1084 |
:param missing: The underlying indexes to delete
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1085 |
:return: (CombinedGraphIndex, reload_counter)
|
1086 |
"""
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1087 |
idx1 = self.make_index('1', nodes=[(('1',), '', ())]) |
1088 |
idx2 = self.make_index('2', nodes=[(('2',), '', ())]) |
|
1089 |
idx3 = self.make_index('3', nodes=[ |
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1090 |
(('1',), '', ()), |
1091 |
(('2',), '', ())]) |
|
1092 |
||
1093 |
# total_reloads, num_changed, num_unchanged
|
|
1094 |
reload_counter = [0, 0, 0] |
|
1095 |
def reload(): |
|
1096 |
reload_counter[0] += 1 |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1097 |
new_indices = [idx3] |
1098 |
if idx._indices == new_indices: |
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1099 |
reload_counter[2] += 1 |
1100 |
return False |
|
1101 |
reload_counter[1] += 1 |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1102 |
idx._indices[:] = new_indices |
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1103 |
return True |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1104 |
idx = index.CombinedGraphIndex([idx1, idx2], reload_func=reload) |
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1105 |
trans = self.get_transport() |
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1106 |
for fname in missing: |
1107 |
trans.delete(fname) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1108 |
return idx, reload_counter |
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1109 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1110 |
def test_open_missing_index_no_error(self): |
1111 |
trans = self.get_transport() |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1112 |
idx1 = index.GraphIndex(trans, 'missing', 100) |
1113 |
idx = index.CombinedGraphIndex([idx1]) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1114 |
|
2592.1.37
by Robert Collins
Add CombinedGraphIndex.insert_index. |
1115 |
def test_add_index(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1116 |
idx = index.CombinedGraphIndex([]) |
1117 |
idx1 = self.make_index('name', 0, nodes=[(('key', ), '', ())]) |
|
1118 |
idx.insert_index(0, idx1) |
|
1119 |
self.assertEqual([(idx1, ('key', ), '')], |
|
1120 |
list(idx.iter_all_entries())) |
|
2592.1.37
by Robert Collins
Add CombinedGraphIndex.insert_index. |
1121 |
|
4744.2.6
by John Arbash Meinel
Start exposing an GraphIndex.clear_cache() member. |
1122 |
def test_clear_cache(self): |
1123 |
log = [] |
|
1124 |
||
1125 |
class ClearCacheProxy(object): |
|
1126 |
||
1127 |
def __init__(self, index): |
|
1128 |
self._index = index |
|
1129 |
||
1130 |
def __getattr__(self, name): |
|
1131 |
return getattr(self._index) |
|
1132 |
||
1133 |
def clear_cache(self): |
|
1134 |
log.append(self._index) |
|
1135 |
return self._index.clear_cache() |
|
1136 |
||
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1137 |
idx = index.CombinedGraphIndex([]) |
1138 |
idx1 = self.make_index('name', 0, nodes=[(('key', ), '', ())]) |
|
1139 |
idx.insert_index(0, ClearCacheProxy(idx1)) |
|
1140 |
idx2 = self.make_index('name', 0, nodes=[(('key', ), '', ())]) |
|
1141 |
idx.insert_index(1, ClearCacheProxy(idx2)) |
|
4744.2.6
by John Arbash Meinel
Start exposing an GraphIndex.clear_cache() member. |
1142 |
# CombinedGraphIndex should call 'clear_cache()' on all children
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1143 |
idx.clear_cache() |
1144 |
self.assertEqual(sorted([idx1, idx2]), sorted(log)) |
|
4744.2.6
by John Arbash Meinel
Start exposing an GraphIndex.clear_cache() member. |
1145 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1146 |
def test_iter_all_entries_empty(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1147 |
idx = index.CombinedGraphIndex([]) |
1148 |
self.assertEqual([], list(idx.iter_all_entries())) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1149 |
|
1150 |
def test_iter_all_entries_children_empty(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1151 |
idx1 = self.make_index('name') |
1152 |
idx = index.CombinedGraphIndex([idx1]) |
|
1153 |
self.assertEqual([], list(idx.iter_all_entries())) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1154 |
|
1155 |
def test_iter_all_entries_simple(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1156 |
idx1 = self.make_index('name', nodes=[(('name', ), 'data', ())]) |
1157 |
idx = index.CombinedGraphIndex([idx1]) |
|
1158 |
self.assertEqual([(idx1, ('name', ), 'data')], |
|
1159 |
list(idx.iter_all_entries())) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1160 |
|
1161 |
def test_iter_all_entries_two_indices(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1162 |
idx1 = self.make_index('name1', nodes=[(('name', ), 'data', ())]) |
1163 |
idx2 = self.make_index('name2', nodes=[(('2', ), '', ())]) |
|
1164 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1165 |
self.assertEqual([(idx1, ('name', ), 'data'), |
|
1166 |
(idx2, ('2', ), '')], |
|
1167 |
list(idx.iter_all_entries())) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1168 |
|
2592.1.39
by Robert Collins
CombinedGraphIndex.iter_entries does not need to see all entries. |
1169 |
def test_iter_entries_two_indices_dup_key(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1170 |
idx1 = self.make_index('name1', nodes=[(('name', ), 'data', ())]) |
1171 |
idx2 = self.make_index('name2', nodes=[(('name', ), 'data', ())]) |
|
1172 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1173 |
self.assertEqual([(idx1, ('name', ), 'data')], |
|
1174 |
list(idx.iter_entries([('name', )]))) |
|
2592.1.39
by Robert Collins
CombinedGraphIndex.iter_entries does not need to see all entries. |
1175 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1176 |
def test_iter_all_entries_two_indices_dup_key(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1177 |
idx1 = self.make_index('name1', nodes=[(('name', ), 'data', ())]) |
1178 |
idx2 = self.make_index('name2', nodes=[(('name', ), 'data', ())]) |
|
1179 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1180 |
self.assertEqual([(idx1, ('name', ), 'data')], |
|
1181 |
list(idx.iter_all_entries())) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1182 |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
1183 |
def test_iter_key_prefix_2_key_element_refs(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1184 |
idx1 = self.make_index('1', 1, key_elements=2, nodes=[ |
1185 |
(('name', 'fin1'), 'data', ([('ref', 'erence')], ))]) |
|
1186 |
idx2 = self.make_index('2', 1, key_elements=2, nodes=[ |
|
1187 |
(('name', 'fin2'), 'beta', ([], )), |
|
1188 |
(('ref', 'erence'), 'refdata', ([], ))]) |
|
1189 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1190 |
self.assertEqual(set([(idx1, ('name', 'fin1'), 'data', |
|
1191 |
((('ref', 'erence'),),)), |
|
1192 |
(idx2, ('ref', 'erence'), 'refdata', ((), ))]), |
|
1193 |
set(idx.iter_entries_prefix([('name', 'fin1'), |
|
1194 |
('ref', 'erence')]))) |
|
1195 |
self.assertEqual(set([(idx1, ('name', 'fin1'), 'data', |
|
1196 |
((('ref', 'erence'),),)), |
|
1197 |
(idx2, ('name', 'fin2'), 'beta', ((), ))]), |
|
1198 |
set(idx.iter_entries_prefix([('name', None)]))) |
|
2624.2.9
by Robert Collins
Introduce multiple component keys, which is what is needed to combine multiple knit indices into one. |
1199 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1200 |
def test_iter_nothing_empty(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1201 |
idx = index.CombinedGraphIndex([]) |
1202 |
self.assertEqual([], list(idx.iter_entries([]))) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1203 |
|
1204 |
def test_iter_nothing_children_empty(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1205 |
idx1 = self.make_index('name') |
1206 |
idx = index.CombinedGraphIndex([idx1]) |
|
1207 |
self.assertEqual([], list(idx.iter_entries([]))) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1208 |
|
1209 |
def test_iter_all_keys(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1210 |
idx1 = self.make_index('1', 1, nodes=[(('name', ), 'data', |
1211 |
([('ref', )], ))]) |
|
1212 |
idx2 = self.make_index('2', 1, nodes=[(('ref', ), 'refdata', ((), ))]) |
|
1213 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1214 |
self.assertEqual(set([(idx1, ('name', ), 'data', ((('ref', ), ), )), |
|
1215 |
(idx2, ('ref', ), 'refdata', ((), ))]), |
|
1216 |
set(idx.iter_entries([('name', ), ('ref', )]))) |
|
3943.8.1
by Marius Kruger
remove all trailing whitespace from bzr source |
1217 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1218 |
def test_iter_all_keys_dup_entry(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1219 |
idx1 = self.make_index('1', 1, nodes=[(('name', ), 'data', |
1220 |
([('ref', )], )), |
|
1221 |
(('ref', ), 'refdata', ([], ))]) |
|
1222 |
idx2 = self.make_index('2', 1, nodes=[(('ref', ), 'refdata', ([], ))]) |
|
1223 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1224 |
self.assertEqual(set([(idx1, ('name', ), 'data', ((('ref',),),)), |
|
1225 |
(idx1, ('ref', ), 'refdata', ((), ))]), |
|
1226 |
set(idx.iter_entries([('name', ), ('ref', )]))) |
|
3943.8.1
by Marius Kruger
remove all trailing whitespace from bzr source |
1227 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1228 |
def test_iter_missing_entry_empty(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1229 |
idx = index.CombinedGraphIndex([]) |
1230 |
self.assertEqual([], list(idx.iter_entries([('a', )]))) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1231 |
|
1232 |
def test_iter_missing_entry_one_index(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1233 |
idx1 = self.make_index('1') |
1234 |
idx = index.CombinedGraphIndex([idx1]) |
|
1235 |
self.assertEqual([], list(idx.iter_entries([('a', )]))) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1236 |
|
1237 |
def test_iter_missing_entry_two_index(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1238 |
idx1 = self.make_index('1') |
1239 |
idx2 = self.make_index('2') |
|
1240 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1241 |
self.assertEqual([], list(idx.iter_entries([('a', )]))) |
|
3943.8.1
by Marius Kruger
remove all trailing whitespace from bzr source |
1242 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1243 |
def test_iter_entry_present_one_index_only(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1244 |
idx1 = self.make_index('1', nodes=[(('key', ), '', ())]) |
1245 |
idx2 = self.make_index('2', nodes=[]) |
|
1246 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1247 |
self.assertEqual([(idx1, ('key', ), '')], |
|
1248 |
list(idx.iter_entries([('key', )]))) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1249 |
# and in the other direction
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1250 |
idx = index.CombinedGraphIndex([idx2, idx1]) |
1251 |
self.assertEqual([(idx1, ('key', ), '')], |
|
1252 |
list(idx.iter_entries([('key', )]))) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1253 |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
1254 |
def test_key_count_empty(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1255 |
idx1 = self.make_index('1', nodes=[]) |
1256 |
idx2 = self.make_index('2', nodes=[]) |
|
1257 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1258 |
self.assertEqual(0, idx.key_count()) |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
1259 |
|
1260 |
def test_key_count_sums_index_keys(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1261 |
idx1 = self.make_index('1', nodes=[ |
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
1262 |
(('1',), '', ()), |
1263 |
(('2',), '', ())]) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1264 |
idx2 = self.make_index('2', nodes=[(('1',), '', ())]) |
1265 |
idx = index.CombinedGraphIndex([idx1, idx2]) |
|
1266 |
self.assertEqual(3, idx.key_count()) |
|
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
1267 |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1268 |
def test_validate_bad_child_index_errors(self): |
1269 |
trans = self.get_transport() |
|
1270 |
trans.put_bytes('name', "not an index\n") |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1271 |
idx1 = index.GraphIndex(trans, 'name', 13) |
1272 |
idx = index.CombinedGraphIndex([idx1]) |
|
1273 |
self.assertRaises(errors.BadIndexFormatSignature, idx.validate) |
|
2592.1.31
by Robert Collins
Build a combined graph index to use multiple indices at once. |
1274 |
|
1275 |
def test_validate_empty(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1276 |
idx = index.CombinedGraphIndex([]) |
1277 |
idx.validate() |
|
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1278 |
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1279 |
def test_key_count_reloads(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1280 |
idx, reload_counter = self.make_combined_index_with_missing() |
1281 |
self.assertEqual(2, idx.key_count()) |
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1282 |
self.assertEqual([1, 1, 0], reload_counter) |
1283 |
||
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1284 |
def test_key_count_no_reload(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1285 |
idx, reload_counter = self.make_combined_index_with_missing() |
1286 |
idx._reload_func = None |
|
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1287 |
# Without a _reload_func we just raise the exception
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1288 |
self.assertRaises(errors.NoSuchFile, idx.key_count) |
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1289 |
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1290 |
def test_key_count_reloads_and_fails(self): |
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1291 |
# We have deleted all underlying indexes, so we will try to reload, but
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1292 |
# still fail. This is mostly to test we don't get stuck in an infinite
|
1293 |
# loop trying to reload
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1294 |
idx, reload_counter = self.make_combined_index_with_missing( |
1295 |
['1', '2', '3']) |
|
1296 |
self.assertRaises(errors.NoSuchFile, idx.key_count) |
|
3789.1.3
by John Arbash Meinel
CombinedGraphIndex can now reload when calling key_count(). |
1297 |
self.assertEqual([2, 1, 1], reload_counter) |
1298 |
||
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1299 |
def test_iter_entries_reloads(self): |
1300 |
index, reload_counter = self.make_combined_index_with_missing() |
|
1301 |
result = list(index.iter_entries([('1',), ('2',), ('3',)])) |
|
1302 |
index3 = index._indices[0] |
|
1303 |
self.assertEqual([(index3, ('1',), ''), (index3, ('2',), '')], |
|
1304 |
result) |
|
1305 |
self.assertEqual([1, 1, 0], reload_counter) |
|
1306 |
||
1307 |
def test_iter_entries_reloads_midway(self): |
|
1308 |
# The first index still looks present, so we get interrupted mid-way
|
|
1309 |
# through
|
|
1310 |
index, reload_counter = self.make_combined_index_with_missing(['2']) |
|
1311 |
index1, index2 = index._indices |
|
1312 |
result = list(index.iter_entries([('1',), ('2',), ('3',)])) |
|
1313 |
index3 = index._indices[0] |
|
3789.1.5
by John Arbash Meinel
CombinedGraphIndex.iter_all_entries() can now reload when needed. |
1314 |
# We had already yielded '1', so we just go on to the next, we should
|
3789.1.6
by John Arbash Meinel
CombinedGraphIndex.iter_entries_prefix can now reload when needed. |
1315 |
# not yield '1' twice.
|
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1316 |
self.assertEqual([(index1, ('1',), ''), (index3, ('2',), '')], |
1317 |
result) |
|
1318 |
self.assertEqual([1, 1, 0], reload_counter) |
|
1319 |
||
1320 |
def test_iter_entries_no_reload(self): |
|
1321 |
index, reload_counter = self.make_combined_index_with_missing() |
|
1322 |
index._reload_func = None |
|
1323 |
# Without a _reload_func we just raise the exception
|
|
1324 |
self.assertListRaises(errors.NoSuchFile, index.iter_entries, [('3',)]) |
|
1325 |
||
1326 |
def test_iter_entries_reloads_and_fails(self): |
|
1327 |
index, reload_counter = self.make_combined_index_with_missing( |
|
1328 |
['1', '2', '3']) |
|
1329 |
self.assertListRaises(errors.NoSuchFile, index.iter_entries, [('3',)]) |
|
1330 |
self.assertEqual([2, 1, 1], reload_counter) |
|
1331 |
||
3789.1.5
by John Arbash Meinel
CombinedGraphIndex.iter_all_entries() can now reload when needed. |
1332 |
def test_iter_all_entries_reloads(self): |
1333 |
index, reload_counter = self.make_combined_index_with_missing() |
|
1334 |
result = list(index.iter_all_entries()) |
|
1335 |
index3 = index._indices[0] |
|
1336 |
self.assertEqual([(index3, ('1',), ''), (index3, ('2',), '')], |
|
1337 |
result) |
|
1338 |
self.assertEqual([1, 1, 0], reload_counter) |
|
1339 |
||
1340 |
def test_iter_all_entries_reloads_midway(self): |
|
1341 |
index, reload_counter = self.make_combined_index_with_missing(['2']) |
|
1342 |
index1, index2 = index._indices |
|
1343 |
result = list(index.iter_all_entries()) |
|
1344 |
index3 = index._indices[0] |
|
1345 |
# We had already yielded '1', so we just go on to the next, we should
|
|
3789.1.6
by John Arbash Meinel
CombinedGraphIndex.iter_entries_prefix can now reload when needed. |
1346 |
# not yield '1' twice.
|
3789.1.5
by John Arbash Meinel
CombinedGraphIndex.iter_all_entries() can now reload when needed. |
1347 |
self.assertEqual([(index1, ('1',), ''), (index3, ('2',), '')], |
1348 |
result) |
|
1349 |
self.assertEqual([1, 1, 0], reload_counter) |
|
1350 |
||
1351 |
def test_iter_all_entries_no_reload(self): |
|
1352 |
index, reload_counter = self.make_combined_index_with_missing() |
|
1353 |
index._reload_func = None |
|
1354 |
self.assertListRaises(errors.NoSuchFile, index.iter_all_entries) |
|
1355 |
||
1356 |
def test_iter_all_entries_reloads_and_fails(self): |
|
1357 |
index, reload_counter = self.make_combined_index_with_missing( |
|
1358 |
['1', '2', '3']) |
|
1359 |
self.assertListRaises(errors.NoSuchFile, index.iter_all_entries) |
|
3789.1.4
by John Arbash Meinel
CombinedGraphIndex.iter_entries() is now able to reload on request. |
1360 |
|
3789.1.6
by John Arbash Meinel
CombinedGraphIndex.iter_entries_prefix can now reload when needed. |
1361 |
def test_iter_entries_prefix_reloads(self): |
1362 |
index, reload_counter = self.make_combined_index_with_missing() |
|
1363 |
result = list(index.iter_entries_prefix([('1',)])) |
|
1364 |
index3 = index._indices[0] |
|
1365 |
self.assertEqual([(index3, ('1',), '')], result) |
|
1366 |
self.assertEqual([1, 1, 0], reload_counter) |
|
1367 |
||
1368 |
def test_iter_entries_prefix_reloads_midway(self): |
|
1369 |
index, reload_counter = self.make_combined_index_with_missing(['2']) |
|
1370 |
index1, index2 = index._indices |
|
1371 |
result = list(index.iter_entries_prefix([('1',)])) |
|
1372 |
index3 = index._indices[0] |
|
1373 |
# We had already yielded '1', so we just go on to the next, we should
|
|
1374 |
# not yield '1' twice.
|
|
1375 |
self.assertEqual([(index1, ('1',), '')], result) |
|
1376 |
self.assertEqual([1, 1, 0], reload_counter) |
|
1377 |
||
1378 |
def test_iter_entries_prefix_no_reload(self): |
|
1379 |
index, reload_counter = self.make_combined_index_with_missing() |
|
1380 |
index._reload_func = None |
|
1381 |
self.assertListRaises(errors.NoSuchFile, index.iter_entries_prefix, |
|
1382 |
[('1',)]) |
|
1383 |
||
1384 |
def test_iter_entries_prefix_reloads_and_fails(self): |
|
1385 |
index, reload_counter = self.make_combined_index_with_missing( |
|
1386 |
['1', '2', '3']) |
|
1387 |
self.assertListRaises(errors.NoSuchFile, index.iter_entries_prefix, |
|
1388 |
[('1',)]) |
|
1389 |
||
5086.7.7
by Andrew Bennetts
Add another test for propagating reorderings. |
1390 |
|
1391 |
def make_index_with_simple_nodes(self, name, num_nodes=1): |
|
1392 |
"""Make an index named after 'name', with keys named after 'name' too.
|
|
1393 |
||
1394 |
Nodes will have a value of '' and no references.
|
|
1395 |
"""
|
|
1396 |
nodes = [ |
|
1397 |
(('index-%s-key-%s' % (name, n),), '', ()) |
|
1398 |
for n in range(1, num_nodes+1)] |
|
1399 |
return self.make_index('index-%s' % name, 0, nodes=nodes) |
|
1400 |
||
5086.7.5
by Andrew Bennetts
Add test for CombinedGraphIndex's auto-reordering behaviour. |
1401 |
def test_reorder_after_iter_entries(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1402 |
# Four indices: [key1] in idx1, [key2,key3] in idx2, [] in idx3,
|
1403 |
# [key4] in idx4.
|
|
1404 |
idx = index.CombinedGraphIndex([]) |
|
1405 |
idx.insert_index(0, self.make_index_with_simple_nodes('1'), '1') |
|
1406 |
idx.insert_index(1, self.make_index_with_simple_nodes('2'), '2') |
|
1407 |
idx.insert_index(2, self.make_index_with_simple_nodes('3'), '3') |
|
1408 |
idx.insert_index(3, self.make_index_with_simple_nodes('4'), '4') |
|
1409 |
idx1, idx2, idx3, idx4 = idx._indices |
|
1410 |
# Query a key from idx4 and idx2.
|
|
1411 |
self.assertLength(2, list(idx.iter_entries( |
|
5086.7.7
by Andrew Bennetts
Add another test for propagating reorderings. |
1412 |
[('index-4-key-1',), ('index-2-key-1',)]))) |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1413 |
# Now idx2 and idx4 should be moved to the front (and idx1 should
|
1414 |
# still be before idx3).
|
|
1415 |
self.assertEqual([idx2, idx4, idx1, idx3], idx._indices) |
|
1416 |
self.assertEqual(['2', '4', '1', '3'], idx._index_names) |
|
5086.7.5
by Andrew Bennetts
Add test for CombinedGraphIndex's auto-reordering behaviour. |
1417 |
|
5086.7.7
by Andrew Bennetts
Add another test for propagating reorderings. |
1418 |
def test_reorder_propagates_to_siblings(self): |
1419 |
# Two CombinedGraphIndex objects, with the same number of indicies with
|
|
1420 |
# matching names.
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1421 |
cgi1 = index.CombinedGraphIndex([]) |
1422 |
cgi2 = index.CombinedGraphIndex([]) |
|
5086.7.7
by Andrew Bennetts
Add another test for propagating reorderings. |
1423 |
cgi1.insert_index(0, self.make_index_with_simple_nodes('1-1'), 'one') |
1424 |
cgi1.insert_index(1, self.make_index_with_simple_nodes('1-2'), 'two') |
|
1425 |
cgi2.insert_index(0, self.make_index_with_simple_nodes('2-1'), 'one') |
|
1426 |
cgi2.insert_index(1, self.make_index_with_simple_nodes('2-2'), 'two') |
|
1427 |
index2_1, index2_2 = cgi2._indices |
|
1428 |
cgi1.set_sibling_indices([cgi2]) |
|
1429 |
# Trigger a reordering in cgi1. cgi2 will be reordered as well.
|
|
1430 |
list(cgi1.iter_entries([('index-1-2-key-1',)])) |
|
1431 |
self.assertEqual([index2_2, index2_1], cgi2._indices) |
|
1432 |
self.assertEqual(['two', 'one'], cgi2._index_names) |
|
1433 |
||
3789.1.7
by John Arbash Meinel
CombinedGraphIndex.validate() will now reload. |
1434 |
def test_validate_reloads(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1435 |
idx, reload_counter = self.make_combined_index_with_missing() |
1436 |
idx.validate() |
|
3789.1.7
by John Arbash Meinel
CombinedGraphIndex.validate() will now reload. |
1437 |
self.assertEqual([1, 1, 0], reload_counter) |
1438 |
||
1439 |
def test_validate_reloads_midway(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1440 |
idx, reload_counter = self.make_combined_index_with_missing(['2']) |
1441 |
idx.validate() |
|
3789.1.7
by John Arbash Meinel
CombinedGraphIndex.validate() will now reload. |
1442 |
|
1443 |
def test_validate_no_reload(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1444 |
idx, reload_counter = self.make_combined_index_with_missing() |
1445 |
idx._reload_func = None |
|
1446 |
self.assertRaises(errors.NoSuchFile, idx.validate) |
|
3789.1.7
by John Arbash Meinel
CombinedGraphIndex.validate() will now reload. |
1447 |
|
1448 |
def test_validate_reloads_and_fails(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1449 |
idx, reload_counter = self.make_combined_index_with_missing( |
1450 |
['1', '2', '3']) |
|
1451 |
self.assertRaises(errors.NoSuchFile, idx.validate) |
|
3789.1.7
by John Arbash Meinel
CombinedGraphIndex.validate() will now reload. |
1452 |
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1453 |
def test_find_ancestors_across_indexes(self): |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1454 |
key1 = ('key-1',) |
1455 |
key2 = ('key-2',) |
|
1456 |
key3 = ('key-3',) |
|
1457 |
key4 = ('key-4',) |
|
1458 |
index1 = self.make_index('12', ref_lists=1, nodes=[ |
|
1459 |
(key1, 'value', ([],)), |
|
1460 |
(key2, 'value', ([key1],)), |
|
1461 |
])
|
|
1462 |
index2 = self.make_index('34', ref_lists=1, nodes=[ |
|
1463 |
(key3, 'value', ([key2],)), |
|
1464 |
(key4, 'value', ([key3],)), |
|
1465 |
])
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1466 |
c_index = index.CombinedGraphIndex([index1, index2]) |
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1467 |
parent_map, missing_keys = c_index.find_ancestry([key1], 0) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1468 |
self.assertEqual({key1: ()}, parent_map) |
1469 |
self.assertEqual(set(), missing_keys) |
|
1470 |
# Now look for a key from index2 which requires us to find the key in
|
|
1471 |
# the second index, and then continue searching for parents in the
|
|
1472 |
# first index
|
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1473 |
parent_map, missing_keys = c_index.find_ancestry([key3], 0) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1474 |
self.assertEqual({key1: (), key2: (key1,), key3: (key2,)}, parent_map) |
1475 |
self.assertEqual(set(), missing_keys) |
|
1476 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1477 |
def test_find_ancestors_missing_keys(self): |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1478 |
key1 = ('key-1',) |
1479 |
key2 = ('key-2',) |
|
1480 |
key3 = ('key-3',) |
|
1481 |
key4 = ('key-4',) |
|
1482 |
index1 = self.make_index('12', ref_lists=1, nodes=[ |
|
1483 |
(key1, 'value', ([],)), |
|
1484 |
(key2, 'value', ([key1],)), |
|
1485 |
])
|
|
1486 |
index2 = self.make_index('34', ref_lists=1, nodes=[ |
|
1487 |
(key3, 'value', ([key2],)), |
|
1488 |
])
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1489 |
c_index = index.CombinedGraphIndex([index1, index2]) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1490 |
# Searching for a key which is actually not present at all should
|
1491 |
# eventually converge
|
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1492 |
parent_map, missing_keys = c_index.find_ancestry([key4], 0) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1493 |
self.assertEqual({}, parent_map) |
1494 |
self.assertEqual(set([key4]), missing_keys) |
|
1495 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1496 |
def test_find_ancestors_no_indexes(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1497 |
c_index = index.CombinedGraphIndex([]) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1498 |
key1 = ('key-1',) |
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1499 |
parent_map, missing_keys = c_index.find_ancestry([key1], 0) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1500 |
self.assertEqual({}, parent_map) |
1501 |
self.assertEqual(set([key1]), missing_keys) |
|
1502 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1503 |
def test_find_ancestors_ghost_parent(self): |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1504 |
key1 = ('key-1',) |
1505 |
key2 = ('key-2',) |
|
1506 |
key3 = ('key-3',) |
|
1507 |
key4 = ('key-4',) |
|
1508 |
index1 = self.make_index('12', ref_lists=1, nodes=[ |
|
1509 |
(key1, 'value', ([],)), |
|
1510 |
(key2, 'value', ([key1],)), |
|
1511 |
])
|
|
1512 |
index2 = self.make_index('34', ref_lists=1, nodes=[ |
|
1513 |
(key4, 'value', ([key2, key3],)), |
|
1514 |
])
|
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1515 |
c_index = index.CombinedGraphIndex([index1, index2]) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1516 |
# Searching for a key which is actually not present at all should
|
1517 |
# eventually converge
|
|
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1518 |
parent_map, missing_keys = c_index.find_ancestry([key4], 0) |
4593.4.8
by John Arbash Meinel
Implement CombinedGraphIndex.get_ancestry() |
1519 |
self.assertEqual({key4: (key2, key3), key2: (key1,), key1: ()}, |
1520 |
parent_map) |
|
1521 |
self.assertEqual(set([key3]), missing_keys) |
|
1522 |
||
4593.4.12
by John Arbash Meinel
Name the specific index api _find_ancestors, and the public CombinedGraphIndex api find_ancestry() |
1523 |
def test__find_ancestors_empty_index(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1524 |
idx = self.make_index('test', ref_lists=1, key_elements=1, nodes=[]) |
4593.4.11
by John Arbash Meinel
Snapshot the work in progress. |
1525 |
parent_map = {} |
1526 |
missing_keys = set() |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1527 |
search_keys = idx._find_ancestors([('one',), ('two',)], 0, parent_map, |
1528 |
missing_keys) |
|
4593.4.11
by John Arbash Meinel
Snapshot the work in progress. |
1529 |
self.assertEqual(set(), search_keys) |
1530 |
self.assertEqual({}, parent_map) |
|
1531 |
self.assertEqual(set([('one',), ('two',)]), missing_keys) |
|
1532 |
||
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1533 |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1534 |
class TestInMemoryGraphIndex(tests.TestCaseWithMemoryTransport): |
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1535 |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1536 |
def make_index(self, ref_lists=0, key_elements=1, nodes=[]): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1537 |
result = index.InMemoryGraphIndex(ref_lists, key_elements=key_elements) |
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1538 |
result.add_nodes(nodes) |
1539 |
return result |
|
1540 |
||
2624.2.1
by Robert Collins
InMemoryGraphIndex.add_nodes was inconsistent with other metods for non-node-reference indices. |
1541 |
def test_add_nodes_no_refs(self): |
1542 |
index = self.make_index(0) |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1543 |
index.add_nodes([(('name', ), 'data')]) |
1544 |
index.add_nodes([(('name2', ), ''), (('name3', ), '')]) |
|
2624.2.1
by Robert Collins
InMemoryGraphIndex.add_nodes was inconsistent with other metods for non-node-reference indices. |
1545 |
self.assertEqual(set([ |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1546 |
(index, ('name', ), 'data'), |
1547 |
(index, ('name2', ), ''), |
|
1548 |
(index, ('name3', ), ''), |
|
2624.2.1
by Robert Collins
InMemoryGraphIndex.add_nodes was inconsistent with other metods for non-node-reference indices. |
1549 |
]), set(index.iter_all_entries())) |
1550 |
||
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1551 |
def test_add_nodes(self): |
1552 |
index = self.make_index(1) |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1553 |
index.add_nodes([(('name', ), 'data', ([],))]) |
1554 |
index.add_nodes([(('name2', ), '', ([],)), (('name3', ), '', ([('r', )],))]) |
|
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1555 |
self.assertEqual(set([ |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1556 |
(index, ('name', ), 'data', ((),)), |
1557 |
(index, ('name2', ), '', ((),)), |
|
1558 |
(index, ('name3', ), '', ((('r', ), ), )), |
|
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1559 |
]), set(index.iter_all_entries())) |
1560 |
||
1561 |
def test_iter_all_entries_empty(self): |
|
1562 |
index = self.make_index() |
|
1563 |
self.assertEqual([], list(index.iter_all_entries())) |
|
1564 |
||
1565 |
def test_iter_all_entries_simple(self): |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1566 |
index = self.make_index(nodes=[(('name', ), 'data')]) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1567 |
self.assertEqual([(index, ('name', ), 'data')], |
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1568 |
list(index.iter_all_entries())) |
1569 |
||
1570 |
def test_iter_all_entries_references(self): |
|
1571 |
index = self.make_index(1, nodes=[ |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1572 |
(('name', ), 'data', ([('ref', )], )), |
1573 |
(('ref', ), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1574 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref', ),),)), |
1575 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1576 |
set(index.iter_all_entries())) |
1577 |
||
1578 |
def test_iteration_absent_skipped(self): |
|
1579 |
index = self.make_index(1, nodes=[ |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1580 |
(('name', ), 'data', ([('ref', )], ))]) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1581 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),))]), |
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1582 |
set(index.iter_all_entries())) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1583 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),))]), |
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1584 |
set(index.iter_entries([('name', )]))) |
1585 |
self.assertEqual([], list(index.iter_entries([('ref', )]))) |
|
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1586 |
|
1587 |
def test_iter_all_keys(self): |
|
1588 |
index = self.make_index(1, nodes=[ |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1589 |
(('name', ), 'data', ([('ref', )], )), |
1590 |
(('ref', ), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1591 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),)), |
1592 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1593 |
set(index.iter_entries([('name', ), ('ref', )]))) |
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1594 |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1595 |
def test_iter_key_prefix_1_key_element_no_refs(self): |
1596 |
index = self.make_index( nodes=[ |
|
1597 |
(('name', ), 'data'), |
|
1598 |
(('ref', ), 'refdata')]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1599 |
self.assertEqual(set([(index, ('name', ), 'data'), |
1600 |
(index, ('ref', ), 'refdata')]), |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1601 |
set(index.iter_entries_prefix([('name', ), ('ref', )]))) |
1602 |
||
1603 |
def test_iter_key_prefix_1_key_element_refs(self): |
|
1604 |
index = self.make_index(1, nodes=[ |
|
1605 |
(('name', ), 'data', ([('ref', )], )), |
|
1606 |
(('ref', ), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1607 |
self.assertEqual(set([(index, ('name', ), 'data', ((('ref',),),)), |
1608 |
(index, ('ref', ), 'refdata', ((), ))]), |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1609 |
set(index.iter_entries_prefix([('name', ), ('ref', )]))) |
1610 |
||
1611 |
def test_iter_key_prefix_2_key_element_no_refs(self): |
|
1612 |
index = self.make_index(key_elements=2, nodes=[ |
|
1613 |
(('name', 'fin1'), 'data'), |
|
1614 |
(('name', 'fin2'), 'beta'), |
|
1615 |
(('ref', 'erence'), 'refdata')]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1616 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data'), |
1617 |
(index, ('ref', 'erence'), 'refdata')]), |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1618 |
set(index.iter_entries_prefix([('name', 'fin1'), ('ref', 'erence')]))) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1619 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data'), |
1620 |
(index, ('name', 'fin2'), 'beta')]), |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1621 |
set(index.iter_entries_prefix([('name', None)]))) |
1622 |
||
1623 |
def test_iter_key_prefix_2_key_element_refs(self): |
|
1624 |
index = self.make_index(1, key_elements=2, nodes=[ |
|
1625 |
(('name', 'fin1'), 'data', ([('ref', 'erence')], )), |
|
1626 |
(('name', 'fin2'), 'beta', ([], )), |
|
1627 |
(('ref', 'erence'), 'refdata', ([], ))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1628 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data', ((('ref', 'erence'),),)), |
1629 |
(index, ('ref', 'erence'), 'refdata', ((), ))]), |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1630 |
set(index.iter_entries_prefix([('name', 'fin1'), ('ref', 'erence')]))) |
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1631 |
self.assertEqual(set([(index, ('name', 'fin1'), 'data', ((('ref', 'erence'),),)), |
1632 |
(index, ('name', 'fin2'), 'beta', ((), ))]), |
|
2624.2.10
by Robert Collins
Also add iter_key_prefix support to InMemoryGraphIndex. |
1633 |
set(index.iter_entries_prefix([('name', None)]))) |
1634 |
||
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1635 |
def test_iter_nothing_empty(self): |
1636 |
index = self.make_index() |
|
1637 |
self.assertEqual([], list(index.iter_entries([]))) |
|
1638 |
||
1639 |
def test_iter_missing_entry_empty(self): |
|
1640 |
index = self.make_index() |
|
1641 |
self.assertEqual([], list(index.iter_entries(['a']))) |
|
1642 |
||
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
1643 |
def test_key_count_empty(self): |
1644 |
index = self.make_index() |
|
1645 |
self.assertEqual(0, index.key_count()) |
|
1646 |
||
1647 |
def test_key_count_one(self): |
|
1648 |
index = self.make_index(nodes=[(('name', ), '')]) |
|
1649 |
self.assertEqual(1, index.key_count()) |
|
1650 |
||
1651 |
def test_key_count_two(self): |
|
1652 |
index = self.make_index(nodes=[(('name', ), ''), (('foo', ), '')]) |
|
1653 |
self.assertEqual(2, index.key_count()) |
|
1654 |
||
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1655 |
def test_validate_empty(self): |
1656 |
index = self.make_index() |
|
1657 |
index.validate() |
|
1658 |
||
1659 |
def test_validate_no_refs_content(self): |
|
2624.2.5
by Robert Collins
Change bzrlib.index.Index keys to be 1-tuples, not strings. |
1660 |
index = self.make_index(nodes=[(('key', ), 'value')]) |
2592.1.38
by Robert Collins
Create an InMemoryGraphIndex for temporary indexing. |
1661 |
index.validate() |
1662 |
||
1663 |
||
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1664 |
class TestGraphIndexPrefixAdapter(tests.TestCaseWithMemoryTransport): |
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1665 |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1666 |
def make_index(self, ref_lists=1, key_elements=2, nodes=[], |
1667 |
add_callback=False): |
|
1668 |
result = index.InMemoryGraphIndex(ref_lists, key_elements=key_elements) |
|
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1669 |
result.add_nodes(nodes) |
2624.2.13
by Robert Collins
Implement add_node/add_nodes to the GraphIndexPrefixAdapter. |
1670 |
if add_callback: |
2624.2.17
by Robert Collins
Review feedback. |
1671 |
add_nodes_callback = result.add_nodes |
2624.2.13
by Robert Collins
Implement add_node/add_nodes to the GraphIndexPrefixAdapter. |
1672 |
else: |
2624.2.17
by Robert Collins
Review feedback. |
1673 |
add_nodes_callback = None |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1674 |
adapter = index.GraphIndexPrefixAdapter( |
1675 |
result, ('prefix', ), key_elements - 1, |
|
2624.2.13
by Robert Collins
Implement add_node/add_nodes to the GraphIndexPrefixAdapter. |
1676 |
add_nodes_callback=add_nodes_callback) |
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1677 |
return result, adapter |
1678 |
||
2624.2.13
by Robert Collins
Implement add_node/add_nodes to the GraphIndexPrefixAdapter. |
1679 |
def test_add_node(self): |
1680 |
index, adapter = self.make_index(add_callback=True) |
|
1681 |
adapter.add_node(('key',), 'value', ((('ref',),),)) |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1682 |
self.assertEqual(set([(index, ('prefix', 'key'), 'value', |
1683 |
((('prefix', 'ref'),),))]), |
|
2624.2.13
by Robert Collins
Implement add_node/add_nodes to the GraphIndexPrefixAdapter. |
1684 |
set(index.iter_all_entries())) |
1685 |
||
1686 |
def test_add_nodes(self): |
|
1687 |
index, adapter = self.make_index(add_callback=True) |
|
1688 |
adapter.add_nodes(( |
|
1689 |
(('key',), 'value', ((('ref',),),)), |
|
1690 |
(('key2',), 'value2', ((),)), |
|
1691 |
))
|
|
1692 |
self.assertEqual(set([ |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1693 |
(index, ('prefix', 'key2'), 'value2', ((),)), |
1694 |
(index, ('prefix', 'key'), 'value', ((('prefix', 'ref'),),)) |
|
2624.2.13
by Robert Collins
Implement add_node/add_nodes to the GraphIndexPrefixAdapter. |
1695 |
]),
|
1696 |
set(index.iter_all_entries())) |
|
1697 |
||
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1698 |
def test_construct(self): |
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1699 |
idx = index.InMemoryGraphIndex() |
1700 |
adapter = index.GraphIndexPrefixAdapter(idx, ('prefix', ), 1) |
|
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1701 |
|
1702 |
def test_construct_with_callback(self): |
|
5273.1.7
by Vincent Ladeuil
No more use of the get_transport imported *symbol*, all uses are through |
1703 |
idx = index.InMemoryGraphIndex() |
1704 |
adapter = index.GraphIndexPrefixAdapter(idx, ('prefix', ), 1, |
|
1705 |
idx.add_nodes) |
|
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1706 |
|
1707 |
def test_iter_all_entries_cross_prefix_map_errors(self): |
|
1708 |
index, adapter = self.make_index(nodes=[ |
|
1709 |
(('prefix', 'key1'), 'data1', ((('prefixaltered', 'key2'),),))]) |
|
1710 |
self.assertRaises(errors.BadIndexData, list, adapter.iter_all_entries()) |
|
1711 |
||
1712 |
def test_iter_all_entries(self): |
|
1713 |
index, adapter = self.make_index(nodes=[ |
|
1714 |
(('notprefix', 'key1'), 'data', ((), )), |
|
1715 |
(('prefix', 'key1'), 'data1', ((), )), |
|
1716 |
(('prefix', 'key2'), 'data2', ((('prefix', 'key1'),),))]) |
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1717 |
self.assertEqual(set([(index, ('key1', ), 'data1', ((),)), |
1718 |
(index, ('key2', ), 'data2', ((('key1',),),))]), |
|
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1719 |
set(adapter.iter_all_entries())) |
1720 |
||
1721 |
def test_iter_entries(self): |
|
1722 |
index, adapter = self.make_index(nodes=[ |
|
1723 |
(('notprefix', 'key1'), 'data', ((), )), |
|
1724 |
(('prefix', 'key1'), 'data1', ((), )), |
|
1725 |
(('prefix', 'key2'), 'data2', ((('prefix', 'key1'),),))]) |
|
1726 |
# ask for many - get all
|
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1727 |
self.assertEqual(set([(index, ('key1', ), 'data1', ((),)), |
1728 |
(index, ('key2', ), 'data2', ((('key1', ),),))]), |
|
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1729 |
set(adapter.iter_entries([('key1', ), ('key2', )]))) |
1730 |
# ask for one, get one
|
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1731 |
self.assertEqual(set([(index, ('key1', ), 'data1', ((),))]), |
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1732 |
set(adapter.iter_entries([('key1', )]))) |
1733 |
# ask for missing, get none
|
|
1734 |
self.assertEqual(set(), |
|
1735 |
set(adapter.iter_entries([('key3', )]))) |
|
1736 |
||
1737 |
def test_iter_entries_prefix(self): |
|
1738 |
index, adapter = self.make_index(key_elements=3, nodes=[ |
|
1739 |
(('notprefix', 'foo', 'key1'), 'data', ((), )), |
|
1740 |
(('prefix', 'prefix2', 'key1'), 'data1', ((), )), |
|
1741 |
(('prefix', 'prefix2', 'key2'), 'data2', ((('prefix', 'prefix2', 'key1'),),))]) |
|
1742 |
# ask for a prefix, get the results for just that prefix, adjusted.
|
|
2624.2.14
by Robert Collins
Add source index to the index iteration API to allow mapping back to the origin of retrieved data. |
1743 |
self.assertEqual(set([(index, ('prefix2', 'key1', ), 'data1', ((),)), |
1744 |
(index, ('prefix2', 'key2', ), 'data2', ((('prefix2', 'key1', ),),))]), |
|
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1745 |
set(adapter.iter_entries_prefix([('prefix2', None)]))) |
1746 |
||
2624.2.16
by Robert Collins
Add a key_count method to GraphIndex and friends, allowing optimisation of length calculations by the index. |
1747 |
def test_key_count_no_matching_keys(self): |
1748 |
index, adapter = self.make_index(nodes=[ |
|
1749 |
(('notprefix', 'key1'), 'data', ((), ))]) |
|
1750 |
self.assertEqual(0, adapter.key_count()) |
|
1751 |
||
1752 |
def test_key_count_some_keys(self): |
|
1753 |
index, adapter = self.make_index(nodes=[ |
|
1754 |
(('notprefix', 'key1'), 'data', ((), )), |
|
1755 |
(('prefix', 'key1'), 'data1', ((), )), |
|
1756 |
(('prefix', 'key2'), 'data2', ((('prefix', 'key1'),),))]) |
|
1757 |
self.assertEqual(2, adapter.key_count()) |
|
1758 |
||
2624.2.12
by Robert Collins
Create an adapter between indices with differing key lengths. |
1759 |
def test_validate(self): |
1760 |
index, adapter = self.make_index() |
|
1761 |
calls = [] |
|
1762 |
def validate(): |
|
1763 |
calls.append('called') |
|
1764 |
index.validate = validate |
|
1765 |
adapter.validate() |
|
1766 |
self.assertEqual(['called'], calls) |