José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 1 | #!/usr/bin/env python |
| 2 | ########################################################################## |
| 3 | # |
| 4 | # Copyright 2011 Jose Fonseca |
| 5 | # All Rights Reserved. |
| 6 | # |
| 7 | # Permission is hereby granted, free of charge, to any person obtaining a copy |
| 8 | # of this software and associated documentation files (the "Software"), to deal |
| 9 | # in the Software without restriction, including without limitation the rights |
| 10 | # to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| 11 | # copies of the Software, and to permit persons to whom the Software is |
| 12 | # furnished to do so, subject to the following conditions: |
| 13 | # |
| 14 | # The above copyright notice and this permission notice shall be included in |
| 15 | # all copies or substantial portions of the Software. |
| 16 | # |
| 17 | # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 18 | # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 19 | # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 20 | # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 21 | # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 22 | # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN |
| 23 | # THE SOFTWARE. |
| 24 | # |
| 25 | ##########################################################################/ |
| 26 | |
| 27 | |
| 28 | import json |
| 29 | import sys |
| 30 | |
| 31 | |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 32 | def strip_object_hook(obj): |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 33 | if '__class__' in obj: |
| 34 | return None |
| 35 | for name in obj.keys(): |
| 36 | if name.startswith('__') and name.endswith('__'): |
| 37 | del obj[name] |
| 38 | return obj |
| 39 | |
| 40 | |
| 41 | class Visitor: |
| 42 | |
| 43 | def visit(self, node, *args, **kwargs): |
| 44 | if isinstance(node, dict): |
| 45 | return self.visit_object(node, *args, **kwargs) |
| 46 | elif isinstance(node, list): |
| 47 | return self.visit_array(node, *args, **kwargs) |
| 48 | else: |
| 49 | return self.visit_value(node, *args, **kwargs) |
| 50 | |
| 51 | def visit_object(self, node, *args, **kwargs): |
| 52 | pass |
| 53 | |
| 54 | def visit_array(self, node, *args, **kwargs): |
| 55 | pass |
| 56 | |
| 57 | def visit_value(self, node, *args, **kwargs): |
| 58 | pass |
| 59 | |
| 60 | |
| 61 | class Dumper(Visitor): |
| 62 | |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 63 | def __init__(self, stream = sys.stdout): |
| 64 | self.stream = stream |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 65 | self.level = 0 |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 66 | |
| 67 | def _write(self, s): |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 68 | self.stream.write(s) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 69 | |
| 70 | def _indent(self): |
| 71 | self._write(' '*self.level) |
| 72 | |
| 73 | def _newline(self): |
| 74 | self._write('\n') |
| 75 | |
| 76 | def visit_object(self, node): |
| 77 | self.enter_object() |
| 78 | |
| 79 | members = node.keys() |
| 80 | members.sort() |
| 81 | for i in range(len(members)): |
| 82 | name = members[i] |
| 83 | value = node[name] |
| 84 | self.enter_member(name) |
| 85 | self.visit(value) |
| 86 | if i: |
| 87 | self._write(',') |
| 88 | self.leave_member() |
| 89 | self.leave_object() |
| 90 | |
| 91 | def enter_object(self): |
| 92 | self._write('{') |
| 93 | self._newline() |
| 94 | self.level += 1 |
| 95 | |
| 96 | def enter_member(self, name): |
| 97 | self._indent() |
| 98 | self._write('%s: ' % name) |
| 99 | |
| 100 | def leave_member(self): |
| 101 | self._newline() |
| 102 | |
| 103 | def leave_object(self): |
| 104 | self.level -= 1 |
| 105 | self._indent() |
| 106 | self._write('}') |
| 107 | |
| 108 | def visit_array(self, node): |
| 109 | self.enter_array() |
| 110 | for i in range(len(node)): |
| 111 | value = node[i] |
| 112 | self._indent() |
| 113 | self.visit(value) |
| 114 | if i: |
| 115 | self._write(',') |
| 116 | self._newline() |
| 117 | self.leave_array() |
| 118 | |
| 119 | def enter_array(self): |
| 120 | self._write('[') |
| 121 | self._newline() |
| 122 | self.level += 1 |
| 123 | |
| 124 | def leave_array(self): |
| 125 | self.level -= 1 |
| 126 | self._indent() |
| 127 | self._write(']') |
| 128 | |
| 129 | def visit_value(self, node): |
| 130 | self._write(repr(node)) |
| 131 | |
| 132 | |
| 133 | |
| 134 | class Comparer(Visitor): |
| 135 | |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 136 | def __init__(self, ignore_added = False): |
| 137 | self.ignore_added = ignore_added |
| 138 | |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 139 | def visit_object(self, a, b): |
| 140 | if not isinstance(b, dict): |
| 141 | return False |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 142 | if len(a) != len(b) and not self.ignore_added: |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 143 | return False |
| 144 | ak = a.keys() |
| 145 | bk = b.keys() |
| 146 | ak.sort() |
| 147 | bk.sort() |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 148 | if ak != bk and not self.ignore_added: |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 149 | return False |
| 150 | for k in ak: |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 151 | ae = a[k] |
| 152 | try: |
| 153 | be = b[k] |
| 154 | except KeyError: |
| 155 | return False |
| 156 | if not self.visit(ae, be): |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 157 | return False |
| 158 | return True |
| 159 | |
| 160 | def visit_array(self, a, b): |
| 161 | if not isinstance(b, list): |
| 162 | return False |
| 163 | if len(a) != len(b): |
| 164 | return False |
| 165 | for ae, be in zip(a, b): |
| 166 | if not self.visit(ae, be): |
| 167 | return False |
| 168 | return True |
| 169 | |
| 170 | def visit_value(self, a, b): |
| 171 | return a == b |
| 172 | |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 173 | |
| 174 | |
| 175 | class Differ(Visitor): |
| 176 | |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 177 | def __init__(self, stream = sys.stdout, ignore_added = False): |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 178 | self.dumper = Dumper(stream) |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 179 | self.comparer = Comparer(ignore_added = ignore_added) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 180 | |
| 181 | def visit(self, a, b): |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 182 | if self.comparer.visit(a, b): |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 183 | return |
| 184 | Visitor.visit(self, a, b) |
| 185 | |
| 186 | def visit_object(self, a, b): |
| 187 | if not isinstance(b, dict): |
| 188 | self.replace(a, b) |
| 189 | else: |
| 190 | self.dumper.enter_object() |
| 191 | names = set(a.keys()) |
| 192 | names.update(b.keys()) |
| 193 | names = list(names) |
| 194 | names.sort() |
| 195 | |
| 196 | for name in names: |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 197 | try: |
| 198 | ae = a[name] |
| 199 | except KeyError: |
| 200 | if self.comparer.ignore_added: |
| 201 | continue |
| 202 | else: |
| 203 | ae = None |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 204 | be = b.get(name, None) |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 205 | if not self.comparer.visit(ae, be): |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 206 | self.dumper.enter_member(name) |
| 207 | self.visit(ae, be) |
| 208 | self.dumper.leave_member() |
| 209 | |
| 210 | self.dumper.leave_object() |
| 211 | |
| 212 | def visit_array(self, a, b): |
| 213 | if not isinstance(b, list): |
| 214 | self.replace(a, b) |
| 215 | else: |
| 216 | self.dumper.enter_array() |
| 217 | for i in range(max(len(a), len(b))): |
| 218 | try: |
| 219 | ae = a[i] |
José Fonseca | 250b36b | 2011-04-13 13:59:00 +0100 | [diff] [blame] | 220 | except IndexError: |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 221 | ae = None |
| 222 | try: |
| 223 | be = b[i] |
José Fonseca | 250b36b | 2011-04-13 13:59:00 +0100 | [diff] [blame] | 224 | except IndexError: |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 225 | be = None |
| 226 | self.dumper._indent() |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 227 | if self.comparer.visit(ae, be): |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 228 | self.dumper.visit(ae) |
| 229 | else: |
| 230 | self.visit(ae, be) |
| 231 | self.dumper._newline() |
| 232 | |
| 233 | self.dumper.leave_array() |
| 234 | |
| 235 | def visit_value(self, a, b): |
| 236 | if a != b: |
| 237 | self.replace(a, b) |
| 238 | |
| 239 | def replace(self, a, b): |
| 240 | self.dumper.visit(a) |
| 241 | self.dumper._write(' -> ') |
| 242 | self.dumper.visit(b) |
| 243 | |
| 244 | |
José Fonseca | e147b61 | 2011-12-04 15:32:03 +0000 | [diff] [blame] | 245 | def load(stream, strip = True): |
| 246 | if strip: |
| 247 | object_hook = strip_object_hook |
| 248 | else: |
| 249 | object_hook = None |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 250 | return json.load(stream, strict=False, object_hook = object_hook) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 251 | |
| 252 | |
| 253 | def main(): |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 254 | a = load(open(sys.argv[1], 'rt')) |
| 255 | b = load(open(sys.argv[2], 'rt')) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 256 | |
| 257 | #dumper = Dumper() |
| 258 | #dumper.visit(a) |
| 259 | |
| 260 | differ = Differ() |
| 261 | differ.visit(a, b) |
| 262 | |
| 263 | |
| 264 | if __name__ == '__main__': |
| 265 | main() |