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 | |
| 32 | def object_hook(obj): |
| 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 | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 65 | self.level = 0; |
| 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 | |
| 136 | def visit_object(self, a, b): |
| 137 | if not isinstance(b, dict): |
| 138 | return False |
| 139 | if len(a) != len(b): |
| 140 | return False |
| 141 | ak = a.keys() |
| 142 | bk = b.keys() |
| 143 | ak.sort() |
| 144 | bk.sort() |
| 145 | if ak != bk: |
| 146 | return False |
| 147 | for k in ak: |
| 148 | if not self.visit(a[k], b[k]): |
| 149 | return False |
| 150 | return True |
| 151 | |
| 152 | def visit_array(self, a, b): |
| 153 | if not isinstance(b, list): |
| 154 | return False |
| 155 | if len(a) != len(b): |
| 156 | return False |
| 157 | for ae, be in zip(a, b): |
| 158 | if not self.visit(ae, be): |
| 159 | return False |
| 160 | return True |
| 161 | |
| 162 | def visit_value(self, a, b): |
| 163 | return a == b |
| 164 | |
| 165 | comparer = Comparer() |
| 166 | |
| 167 | |
| 168 | class Differ(Visitor): |
| 169 | |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 170 | def __init__(self, stream = sys.stdout): |
| 171 | self.dumper = Dumper(stream) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 172 | |
| 173 | def visit(self, a, b): |
| 174 | if comparer.visit(a, b): |
| 175 | return |
| 176 | Visitor.visit(self, a, b) |
| 177 | |
| 178 | def visit_object(self, a, b): |
| 179 | if not isinstance(b, dict): |
| 180 | self.replace(a, b) |
| 181 | else: |
| 182 | self.dumper.enter_object() |
| 183 | names = set(a.keys()) |
| 184 | names.update(b.keys()) |
| 185 | names = list(names) |
| 186 | names.sort() |
| 187 | |
| 188 | for name in names: |
| 189 | ae = a.get(name, None) |
| 190 | be = b.get(name, None) |
| 191 | if not comparer.visit(ae, be): |
| 192 | self.dumper.enter_member(name) |
| 193 | self.visit(ae, be) |
| 194 | self.dumper.leave_member() |
| 195 | |
| 196 | self.dumper.leave_object() |
| 197 | |
| 198 | def visit_array(self, a, b): |
| 199 | if not isinstance(b, list): |
| 200 | self.replace(a, b) |
| 201 | else: |
| 202 | self.dumper.enter_array() |
| 203 | for i in range(max(len(a), len(b))): |
| 204 | try: |
| 205 | ae = a[i] |
José Fonseca | 250b36b | 2011-04-13 13:59:00 +0100 | [diff] [blame] | 206 | except IndexError: |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 207 | ae = None |
| 208 | try: |
| 209 | be = b[i] |
José Fonseca | 250b36b | 2011-04-13 13:59:00 +0100 | [diff] [blame] | 210 | except IndexError: |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 211 | be = None |
| 212 | self.dumper._indent() |
| 213 | if comparer.visit(ae, be): |
| 214 | self.dumper.visit(ae) |
| 215 | else: |
| 216 | self.visit(ae, be) |
| 217 | self.dumper._newline() |
| 218 | |
| 219 | self.dumper.leave_array() |
| 220 | |
| 221 | def visit_value(self, a, b): |
| 222 | if a != b: |
| 223 | self.replace(a, b) |
| 224 | |
| 225 | def replace(self, a, b): |
| 226 | self.dumper.visit(a) |
| 227 | self.dumper._write(' -> ') |
| 228 | self.dumper.visit(b) |
| 229 | |
| 230 | |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 231 | def load(stream): |
| 232 | return json.load(stream, strict=False, object_hook = object_hook) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 233 | |
| 234 | |
| 235 | def main(): |
José Fonseca | 0b956fd | 2011-06-04 22:51:45 +0100 | [diff] [blame] | 236 | a = load(open(sys.argv[1], 'rt')) |
| 237 | b = load(open(sys.argv[2], 'rt')) |
José Fonseca | 0badbf0 | 2011-04-12 09:17:32 +0100 | [diff] [blame] | 238 | |
| 239 | #dumper = Dumper() |
| 240 | #dumper.visit(a) |
| 241 | |
| 242 | differ = Differ() |
| 243 | differ.visit(a, b) |
| 244 | |
| 245 | |
| 246 | if __name__ == '__main__': |
| 247 | main() |