forked from elliotchance/phpserialize
-
Notifications
You must be signed in to change notification settings - Fork 0
/
consume.go
375 lines (305 loc) · 9.46 KB
/
consume.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
package phpserialize
import (
"errors"
"reflect"
"strconv"
)
// The internal consume functions work as the parser/lexer when reading
// individual items off the serialized stream.
// consumeStringUntilByte will return a string that includes all characters
// after the given offset, but only up until (and not including) a found byte.
//
// This function will only work with a plain, non-encoded series of bytes. It
// should not be used to capture anything other that ASCII data that is
// terminated by a single byte.
func consumeStringUntilByte(data []byte, lookingFor byte, offset int) (s string, newOffset int) {
newOffset = findByte(data, lookingFor, offset)
if newOffset < 0 {
return "", -1
}
s = string(data[offset:newOffset])
return
}
func consumeInt(data []byte, offset int) (int64, int, error) {
if !checkType(data, 'i', offset) {
return 0, -1, errors.New("not an integer")
}
alphaNumber, newOffset := consumeStringUntilByte(data, ';', offset+2)
i, err := strconv.Atoi(alphaNumber)
if err != nil {
return 0, -1, err
}
// The +1 is to skip over the final ';'
return int64(i), newOffset + 1, nil
}
func consumeFloat(data []byte, offset int) (float64, int, error) {
if !checkType(data, 'd', offset) {
return 0, -1, errors.New("not a float")
}
alphaNumber, newOffset := consumeStringUntilByte(data, ';', offset+2)
v, err := strconv.ParseFloat(alphaNumber, 64)
if err != nil {
return 0, -1, err
}
return v, newOffset + 1, nil
}
func consumeString(data []byte, offset int) (string, int, error) {
if !checkType(data, 's', offset) {
return "", -1, errors.New("not a string")
}
return consumeStringRealPart(data, offset+2)
}
// consumeIntPart will consume an integer followed by and including a colon.
// This is used in many places to describe the number of elements or an upcoming
// length.
func consumeIntPart(data []byte, offset int) (int, int, error) {
rawValue, newOffset := consumeStringUntilByte(data, ':', offset)
value, err := strconv.Atoi(rawValue)
if err != nil {
return 0, -1, err
}
// The +1 is to skip over the ':'
return value, newOffset + 1, nil
}
func consumeStringRealPart(data []byte, offset int) (string, int, error) {
length, newOffset, err := consumeIntPart(data, offset)
if err != nil {
return "", -1, err
}
// Skip over the '"' at the start of the string. I'm not sure why they
// decided to wrap the string in double quotes since it's totally
// redundant.
offset = newOffset + 1
s := DecodePHPString(data[offset : length+offset])
// The +2 is to skip over the final '";'
return s, offset + length + 2, nil
}
func consumeNil(data []byte, offset int) (interface{}, int, error) {
if !checkType(data, 'N', offset) {
return nil, -1, errors.New("not null")
}
return nil, offset + 2, nil
}
func consumeBool(data []byte, offset int) (bool, int, error) {
if !checkType(data, 'b', offset) {
return false, -1, errors.New("not a boolean")
}
return data[offset+2] == '1', offset + 4, nil
}
func consumeObjectAsMap(data []byte, offset int) (
map[interface{}]interface{}, int, error) {
result := map[interface{}]interface{}{}
// Read the class name. The class name follows the same format as a
// string. We could just ignore the length and hope that no class name
// ever had a non-ascii characters in it, but this is safer - and
// probably easier.
_, offset, err := consumeStringRealPart(data, offset+2)
if err != nil {
return nil, -1, err
}
// Read the number of elements in the object.
length, offset, err := consumeIntPart(data, offset)
if err != nil {
return nil, -1, err
}
// Skip over the '{'
offset++
// Read the elements
for i := 0; i < length; i++ {
var key string
var value interface{}
// The key should always be a string. I am not completely sure
// about this.
key, offset, err = consumeString(data, offset)
if err != nil {
return nil, -1, err
}
// If the next item is an object we can't simply consume it,
// rather we send the reflect.Value back through consumeObject
// so the recursion can be handled correctly.
if data[offset] == 'O' {
var subMap interface{}
subMap, offset, err = consumeObjectAsMap(data, offset)
if err != nil {
return nil, -1, err
}
result[key] = subMap
} else {
value, offset, err = consumeNext(data, offset)
if err != nil {
return nil, -1, err
}
result[key] = value
}
}
// The +1 is for the final '}'
return result, offset + 1, nil
}
func setField(structFieldValue reflect.Value, value interface{}) error {
if !structFieldValue.IsValid() {
return nil
}
val := reflect.ValueOf(value)
switch structFieldValue.Type().Kind() {
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
structFieldValue.SetInt(val.Int())
case reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
structFieldValue.SetUint(val.Uint())
case reflect.Float32, reflect.Float64:
structFieldValue.SetFloat(val.Float())
case reflect.Struct:
m := val.Interface().(map[interface{}]interface{})
fillStruct(structFieldValue, m)
case reflect.Slice:
l := val.Len()
arrayOfObjects := reflect.MakeSlice(structFieldValue.Type(), l, l)
for i := 0; i < l; i++ {
if m, ok := val.Index(i).Interface().(map[interface{}]interface{}); ok {
obj := arrayOfObjects.Index(i)
fillStruct(obj, m)
} else {
switch arrayOfObjects.Index(i).Kind() {
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
arrayOfObjects.Index(i).SetInt(val.Index(i).Elem().Int())
case reflect.Float32, reflect.Float64:
arrayOfObjects.Index(i).SetFloat(val.Index(i).Elem().Float())
default:
arrayOfObjects.Index(i).Set(val.Index(i).Elem())
}
}
}
structFieldValue.Set(arrayOfObjects)
default:
structFieldValue.Set(val)
}
return nil
}
// https://stackoverflow.com/questions/26744873/converting-map-to-struct
func fillStruct(obj reflect.Value, m map[interface{}]interface{}) error {
tt := obj.Type()
for i := 0; i < obj.NumField(); i++ {
field := obj.Field(i)
if !field.CanSet() {
continue
}
var key string
if tag := tt.Field(i).Tag.Get("php"); tag == "-" {
continue
} else if tag != "" {
key = tag
} else {
key = lowerCaseFirstLetter(tt.Field(i).Name)
}
if v, ok := m[key]; ok {
setField(field, v)
}
}
return nil
}
func consumeObject(data []byte, offset int, v reflect.Value) (int, error) {
if !checkType(data, 'O', offset) {
return -1, errors.New("not an object")
}
m, offset, err := consumeObjectAsMap(data, offset)
if err != nil {
return -1, err
}
return offset, fillStruct(v, m)
}
func consumeNext(data []byte, offset int) (interface{}, int, error) {
if offset >= len(data) {
return nil, -1, errors.New("corrupt")
}
switch data[offset] {
case 'a':
return consumeIndexedOrAssociativeArray(data, offset)
case 'b':
return consumeBool(data, offset)
case 'd':
return consumeFloat(data, offset)
case 'i':
return consumeInt(data, offset)
case 's':
return consumeString(data, offset)
case 'N':
return consumeNil(data, offset)
case 'O':
return consumeObjectAsMap(data, offset)
}
return nil, -1, errors.New("can not consume type: " +
string(data[offset:]))
}
func consumeIndexedOrAssociativeArray(data []byte, offset int) (interface{}, int, error) {
// Sometimes we don't know if the array is going to be indexed or
// associative until we have already started to consume it.
originalOffset := offset
// Try to consume it as an indexed array first.
arr, offset, err := consumeIndexedArray(data, originalOffset)
if err == nil {
return arr, offset, err
}
// Fallback to consuming an associative array
return consumeAssociativeArray(data, originalOffset)
}
func consumeAssociativeArray(data []byte, offset int) (map[interface{}]interface{}, int, error) {
if !checkType(data, 'a', offset) {
return map[interface{}]interface{}{}, -1, errors.New("not an array")
}
// Skip over the "a:"
offset += 2
rawLength, offset := consumeStringUntilByte(data, ':', offset)
length, err := strconv.Atoi(rawLength)
if err != nil {
return map[interface{}]interface{}{}, -1, err
}
// Skip over the ":{"
offset += 2
result := map[interface{}]interface{}{}
for i := 0; i < length; i++ {
var key interface{}
key, offset, err = consumeNext(data, offset)
if err != nil {
return map[interface{}]interface{}{}, -1, err
}
result[key], offset, err = consumeNext(data, offset)
if err != nil {
return map[interface{}]interface{}{}, -1, err
}
}
return result, offset + 1, nil
}
func consumeIndexedArray(data []byte, offset int) ([]interface{}, int, error) {
if !checkType(data, 'a', offset) {
return []interface{}{}, -1, errors.New("not an array")
}
rawLength, offset := consumeStringUntilByte(data, ':', offset+2)
length, err := strconv.Atoi(rawLength)
if err != nil {
return []interface{}{}, -1, err
}
// Skip over the ":{"
offset += 2
result := make([]interface{}, length)
for i := 0; i < length; i++ {
// Even non-associative arrays (arrays that are zero-indexed)
// still have their keys serialized. We need to read these
// indexes to make sure we are actually decoding a slice and not
// a map.
var index int64
index, offset, err = consumeInt(data, offset)
if err != nil {
return []interface{}{}, -1, err
}
if index != int64(i) {
return []interface{}{}, -1,
errors.New("cannot decode map as slice")
}
// Now we consume the value
result[i], offset, err = consumeNext(data, offset)
if err != nil {
return []interface{}{}, -1, err
}
}
// The +1 is for the final '}'
return result, offset + 1, nil
}