1 | #!/usr/bin/env python3 |
---|
2 | |
---|
3 | import sys |
---|
4 | import gc |
---|
5 | import time |
---|
6 | import pyregfi |
---|
7 | |
---|
8 | def usage(): |
---|
9 | sys.stderr.write("USAGE: pyregfi-smoketest.py hive1 [hive2 ...]\n") |
---|
10 | |
---|
11 | |
---|
12 | # helper function |
---|
13 | def getCurrentPath(key): |
---|
14 | if key == None: |
---|
15 | return '' |
---|
16 | path = [] |
---|
17 | p = key |
---|
18 | while p != None: |
---|
19 | path.append(p.name) |
---|
20 | if p.is_root(): |
---|
21 | break |
---|
22 | else: |
---|
23 | p = p.get_parent() |
---|
24 | path.reverse() |
---|
25 | del path[0] |
---|
26 | |
---|
27 | return path |
---|
28 | |
---|
29 | |
---|
30 | # Uses the HiveIterator to walk all keys |
---|
31 | # Gathers various (meaningless) statistics to exercise simple attribute access |
---|
32 | # and to hopefully smoke out any bugs that can be identified by changing stats |
---|
33 | def iterTallyNames(hive): |
---|
34 | key_count = 0 |
---|
35 | key_lens = 0 |
---|
36 | key_rawlens = 0 |
---|
37 | value_count = 0 |
---|
38 | value_lens = 0 |
---|
39 | value_rawlens = 0 |
---|
40 | |
---|
41 | for k in hive: |
---|
42 | key_count += 1 |
---|
43 | if k.name != None: |
---|
44 | key_lens += len(k.name) |
---|
45 | if k.name_raw != None: |
---|
46 | key_rawlens += len(k.name_raw) |
---|
47 | |
---|
48 | for v in k.values: |
---|
49 | value_count += 1 |
---|
50 | if v.name != None: |
---|
51 | value_lens += len(v.name) |
---|
52 | if v.name_raw != None: |
---|
53 | value_rawlens += len(v.name_raw) |
---|
54 | |
---|
55 | print(" Counts: keys=%d, values=%d" % (key_count, value_count)) |
---|
56 | print(" Total name length: keys=%d, values=%d" % (key_lens, value_lens)) |
---|
57 | print(" Total raw name lengths: keys=%d, values=%d" % (key_rawlens, value_rawlens)) |
---|
58 | |
---|
59 | |
---|
60 | # For each key in the hive, this traverses the parent links up to the root, |
---|
61 | # recording the path as it goes, and then uses the subtree/descend method |
---|
62 | # to find the same key again, verifying it is the same. This test is currently |
---|
63 | # very slow because no key caching is used. |
---|
64 | def iterParentWalk(hive): |
---|
65 | i = 1 |
---|
66 | for k in hive: |
---|
67 | path = getCurrentPath(k) |
---|
68 | try: |
---|
69 | hive_iter = hive.subtree(path) |
---|
70 | if hive_iter.current_key() != k: |
---|
71 | print("WARNING: k != current_key for path '%s'." % path) |
---|
72 | else: |
---|
73 | i += 1 |
---|
74 | except Exception as e: |
---|
75 | print("WARNING: Could not decend to path '%s'.\nError:\n %s\n%s" % (path,e.args,e)) |
---|
76 | print(" Successfully tested paths on %d keys." % i) |
---|
77 | |
---|
78 | |
---|
79 | # Uses the HiveIterator to walk all keys |
---|
80 | # Gathers various (meaningless) statistics about data/data_raw attributes |
---|
81 | def iterTallyData(hive): |
---|
82 | data_stat = 0.0 |
---|
83 | dataraw_stat = 0.0 |
---|
84 | |
---|
85 | for k in hive: |
---|
86 | for v in k.values: |
---|
87 | d = v.data |
---|
88 | if d == None: |
---|
89 | data_stat += 0.1 |
---|
90 | elif hasattr(d, "__len__"): |
---|
91 | data_stat += len(d) |
---|
92 | else: |
---|
93 | data_stat += d/2.0**64 |
---|
94 | |
---|
95 | d = v.data_raw |
---|
96 | if d == None: |
---|
97 | dataraw_stat += 0.1 |
---|
98 | else: |
---|
99 | dataraw_stat += len(d) |
---|
100 | |
---|
101 | print(" Data stat: %f" % data_stat) |
---|
102 | print(" Raw data stat: %f" % dataraw_stat) |
---|
103 | |
---|
104 | |
---|
105 | recurseKey_stat = 0.0 |
---|
106 | recurseValue_stat = 0.0 |
---|
107 | def checkValues(key): |
---|
108 | global recurseKey_stat |
---|
109 | global recurseValue_stat |
---|
110 | recurseKey_stat += (key.mtime.low^key.mtime.high - key.max_bytes_subkeyname) * key.flags |
---|
111 | for v in key.values: |
---|
112 | recurseValue_stat += (v.data_off - v.data_size) / (1.0 + v.flags) + v.data_in_offset |
---|
113 | value = key.values[v.name] |
---|
114 | if v != value: |
---|
115 | print("WARNING: iterator value '%s' does not match dictionary value '%s'." |
---|
116 | % (v.name, value.name)) |
---|
117 | |
---|
118 | def recurseTree(cur, operation): |
---|
119 | for k in cur.subkeys: |
---|
120 | key = cur.subkeys[k.name] |
---|
121 | if k != key: |
---|
122 | print("WARNING: iterator subkey '%s' does not match dictionary subkey '%s'." |
---|
123 | % (k.name, key.name)) |
---|
124 | del key |
---|
125 | operation(k) |
---|
126 | recurseTree(k, operation) |
---|
127 | |
---|
128 | # Retrieves all keys by recursion, rather than the iterator, and validates |
---|
129 | # list dictionary access. Also builds nonsensical statistics as an excuse |
---|
130 | # to access various base structure attributes. |
---|
131 | def recurseKeyTally(hive): |
---|
132 | root = hive.get_root() |
---|
133 | checkValues(root) |
---|
134 | recurseTree(root, checkValues) |
---|
135 | print(" Key stat: %f" % recurseKey_stat) |
---|
136 | print(" Value stat: %f" % recurseValue_stat) |
---|
137 | |
---|
138 | |
---|
139 | if len(sys.argv) < 2: |
---|
140 | usage() |
---|
141 | sys.exit(1) |
---|
142 | |
---|
143 | |
---|
144 | #tests = [("iterTallyNames",iterTallyNames),("iterParentWalk",iterParentWalk),("iterTallyData",iterTallyData),] |
---|
145 | tests = [("recurseKeyTally",recurseKeyTally),] |
---|
146 | |
---|
147 | files = [] |
---|
148 | for f in sys.argv[1:]: |
---|
149 | files.append((f, open(f,"r+b"))) |
---|
150 | |
---|
151 | |
---|
152 | start_time = time.time() |
---|
153 | for hname,fh in files: |
---|
154 | hive = pyregfi.Hive(fh) |
---|
155 | for tname,t in tests: |
---|
156 | teststart = time.time() |
---|
157 | tstr = "'%s' on '%s'" % (tname,hname) |
---|
158 | print("##BEGIN %s:" % tstr) |
---|
159 | t(hive) |
---|
160 | print("##END %s; runtime=%f; messages:" % (tstr, time.time() - teststart)) |
---|
161 | print(pyregfi.GetLogMessages()) |
---|
162 | print |
---|
163 | sys.stdout.flush() |
---|
164 | |
---|
165 | hive = None |
---|
166 | files = None |
---|
167 | tests = None |
---|
168 | gc.collect() |
---|
169 | print("### Tests Completed, runtime: %f ###" % (time.time() - start_time)) |
---|
170 | #print(gc.garbage) |
---|