parse.js
13.9 KB
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
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
'use strict'
// this[BUFFER] is the remainder of a chunk if we're waiting for
// the full 512 bytes of a header to come in. We will Buffer.concat()
// it to the next write(), which is a mem copy, but a small one.
//
// this[QUEUE] is a Yallist of entries that haven't been emitted
// yet this can only get filled up if the user keeps write()ing after
// a write() returns false, or does a write() with more than one entry
//
// We don't buffer chunks, we always parse them and either create an
// entry, or push it into the active entry. The ReadEntry class knows
// to throw data away if .ignore=true
//
// Shift entry off the buffer when it emits 'end', and emit 'entry' for
// the next one in the list.
//
// At any time, we're pushing body chunks into the entry at WRITEENTRY,
// and waiting for 'end' on the entry at READENTRY
//
// ignored entries get .resume() called on them straight away
const warner = require('./warn-mixin.js')
const Header = require('./header.js')
const EE = require('events')
const Yallist = require('yallist')
const maxMetaEntrySize = 1024 * 1024
const Entry = require('./read-entry.js')
const Pax = require('./pax.js')
const zlib = require('minizlib')
const gzipHeader = Buffer.from([0x1f, 0x8b])
const STATE = Symbol('state')
const WRITEENTRY = Symbol('writeEntry')
const READENTRY = Symbol('readEntry')
const NEXTENTRY = Symbol('nextEntry')
const PROCESSENTRY = Symbol('processEntry')
const EX = Symbol('extendedHeader')
const GEX = Symbol('globalExtendedHeader')
const META = Symbol('meta')
const EMITMETA = Symbol('emitMeta')
const BUFFER = Symbol('buffer')
const QUEUE = Symbol('queue')
const ENDED = Symbol('ended')
const EMITTEDEND = Symbol('emittedEnd')
const EMIT = Symbol('emit')
const UNZIP = Symbol('unzip')
const CONSUMECHUNK = Symbol('consumeChunk')
const CONSUMECHUNKSUB = Symbol('consumeChunkSub')
const CONSUMEBODY = Symbol('consumeBody')
const CONSUMEMETA = Symbol('consumeMeta')
const CONSUMEHEADER = Symbol('consumeHeader')
const CONSUMING = Symbol('consuming')
const BUFFERCONCAT = Symbol('bufferConcat')
const MAYBEEND = Symbol('maybeEnd')
const WRITING = Symbol('writing')
const ABORTED = Symbol('aborted')
const DONE = Symbol('onDone')
const SAW_VALID_ENTRY = Symbol('sawValidEntry')
const SAW_NULL_BLOCK = Symbol('sawNullBlock')
const SAW_EOF = Symbol('sawEOF')
const noop = _ => true
module.exports = warner(class Parser extends EE {
constructor (opt) {
opt = opt || {}
super(opt)
this.file = opt.file || ''
// set to boolean false when an entry starts. 1024 bytes of \0
// is technically a valid tarball, albeit a boring one.
this[SAW_VALID_ENTRY] = null
// these BADARCHIVE errors can't be detected early. listen on DONE.
this.on(DONE, _ => {
if (this[STATE] === 'begin' || this[SAW_VALID_ENTRY] === false) {
// either less than 1 block of data, or all entries were invalid.
// Either way, probably not even a tarball.
this.warn('TAR_BAD_ARCHIVE', 'Unrecognized archive format')
}
})
if (opt.ondone)
this.on(DONE, opt.ondone)
else {
this.on(DONE, _ => {
this.emit('prefinish')
this.emit('finish')
this.emit('end')
this.emit('close')
})
}
this.strict = !!opt.strict
this.maxMetaEntrySize = opt.maxMetaEntrySize || maxMetaEntrySize
this.filter = typeof opt.filter === 'function' ? opt.filter : noop
// have to set this so that streams are ok piping into it
this.writable = true
this.readable = false
this[QUEUE] = new Yallist()
this[BUFFER] = null
this[READENTRY] = null
this[WRITEENTRY] = null
this[STATE] = 'begin'
this[META] = ''
this[EX] = null
this[GEX] = null
this[ENDED] = false
this[UNZIP] = null
this[ABORTED] = false
this[SAW_NULL_BLOCK] = false
this[SAW_EOF] = false
if (typeof opt.onwarn === 'function')
this.on('warn', opt.onwarn)
if (typeof opt.onentry === 'function')
this.on('entry', opt.onentry)
}
[CONSUMEHEADER] (chunk, position) {
if (this[SAW_VALID_ENTRY] === null)
this[SAW_VALID_ENTRY] = false
let header
try {
header = new Header(chunk, position, this[EX], this[GEX])
} catch (er) {
return this.warn('TAR_ENTRY_INVALID', er)
}
if (header.nullBlock) {
if (this[SAW_NULL_BLOCK]) {
this[SAW_EOF] = true
// ending an archive with no entries. pointless, but legal.
if (this[STATE] === 'begin')
this[STATE] = 'header'
this[EMIT]('eof')
} else {
this[SAW_NULL_BLOCK] = true
this[EMIT]('nullBlock')
}
} else {
this[SAW_NULL_BLOCK] = false
if (!header.cksumValid)
this.warn('TAR_ENTRY_INVALID', 'checksum failure', {header})
else if (!header.path)
this.warn('TAR_ENTRY_INVALID', 'path is required', {header})
else {
const type = header.type
if (/^(Symbolic)?Link$/.test(type) && !header.linkpath)
this.warn('TAR_ENTRY_INVALID', 'linkpath required', {header})
else if (!/^(Symbolic)?Link$/.test(type) && header.linkpath)
this.warn('TAR_ENTRY_INVALID', 'linkpath forbidden', {header})
else {
const entry = this[WRITEENTRY] = new Entry(header, this[EX], this[GEX])
// we do this for meta & ignored entries as well, because they
// are still valid tar, or else we wouldn't know to ignore them
if (!this[SAW_VALID_ENTRY]) {
if (entry.remain) {
// this might be the one!
const onend = () => {
if (!entry.invalid)
this[SAW_VALID_ENTRY] = true
}
entry.on('end', onend)
} else
this[SAW_VALID_ENTRY] = true
}
if (entry.meta) {
if (entry.size > this.maxMetaEntrySize) {
entry.ignore = true
this[EMIT]('ignoredEntry', entry)
this[STATE] = 'ignore'
entry.resume()
} else if (entry.size > 0) {
this[META] = ''
entry.on('data', c => this[META] += c)
this[STATE] = 'meta'
}
} else {
this[EX] = null
entry.ignore = entry.ignore || !this.filter(entry.path, entry)
if (entry.ignore) {
// probably valid, just not something we care about
this[EMIT]('ignoredEntry', entry)
this[STATE] = entry.remain ? 'ignore' : 'header'
entry.resume()
} else {
if (entry.remain)
this[STATE] = 'body'
else {
this[STATE] = 'header'
entry.end()
}
if (!this[READENTRY]) {
this[QUEUE].push(entry)
this[NEXTENTRY]()
} else
this[QUEUE].push(entry)
}
}
}
}
}
}
[PROCESSENTRY] (entry) {
let go = true
if (!entry) {
this[READENTRY] = null
go = false
} else if (Array.isArray(entry))
this.emit.apply(this, entry)
else {
this[READENTRY] = entry
this.emit('entry', entry)
if (!entry.emittedEnd) {
entry.on('end', _ => this[NEXTENTRY]())
go = false
}
}
return go
}
[NEXTENTRY] () {
do {} while (this[PROCESSENTRY](this[QUEUE].shift()))
if (!this[QUEUE].length) {
// At this point, there's nothing in the queue, but we may have an
// entry which is being consumed (readEntry).
// If we don't, then we definitely can handle more data.
// If we do, and either it's flowing, or it has never had any data
// written to it, then it needs more.
// The only other possibility is that it has returned false from a
// write() call, so we wait for the next drain to continue.
const re = this[READENTRY]
const drainNow = !re || re.flowing || re.size === re.remain
if (drainNow) {
if (!this[WRITING])
this.emit('drain')
} else
re.once('drain', _ => this.emit('drain'))
}
}
[CONSUMEBODY] (chunk, position) {
// write up to but no more than writeEntry.blockRemain
const entry = this[WRITEENTRY]
const br = entry.blockRemain
const c = (br >= chunk.length && position === 0) ? chunk
: chunk.slice(position, position + br)
entry.write(c)
if (!entry.blockRemain) {
this[STATE] = 'header'
this[WRITEENTRY] = null
entry.end()
}
return c.length
}
[CONSUMEMETA] (chunk, position) {
const entry = this[WRITEENTRY]
const ret = this[CONSUMEBODY](chunk, position)
// if we finished, then the entry is reset
if (!this[WRITEENTRY])
this[EMITMETA](entry)
return ret
}
[EMIT] (ev, data, extra) {
if (!this[QUEUE].length && !this[READENTRY])
this.emit(ev, data, extra)
else
this[QUEUE].push([ev, data, extra])
}
[EMITMETA] (entry) {
this[EMIT]('meta', this[META])
switch (entry.type) {
case 'ExtendedHeader':
case 'OldExtendedHeader':
this[EX] = Pax.parse(this[META], this[EX], false)
break
case 'GlobalExtendedHeader':
this[GEX] = Pax.parse(this[META], this[GEX], true)
break
case 'NextFileHasLongPath':
case 'OldGnuLongPath':
this[EX] = this[EX] || Object.create(null)
this[EX].path = this[META].replace(/\0.*/, '')
break
case 'NextFileHasLongLinkpath':
this[EX] = this[EX] || Object.create(null)
this[EX].linkpath = this[META].replace(/\0.*/, '')
break
/* istanbul ignore next */
default: throw new Error('unknown meta: ' + entry.type)
}
}
abort (error) {
this[ABORTED] = true
this.emit('abort', error)
// always throws, even in non-strict mode
this.warn('TAR_ABORT', error, { recoverable: false })
}
write (chunk) {
if (this[ABORTED])
return
// first write, might be gzipped
if (this[UNZIP] === null && chunk) {
if (this[BUFFER]) {
chunk = Buffer.concat([this[BUFFER], chunk])
this[BUFFER] = null
}
if (chunk.length < gzipHeader.length) {
this[BUFFER] = chunk
return true
}
for (let i = 0; this[UNZIP] === null && i < gzipHeader.length; i++) {
if (chunk[i] !== gzipHeader[i])
this[UNZIP] = false
}
if (this[UNZIP] === null) {
const ended = this[ENDED]
this[ENDED] = false
this[UNZIP] = new zlib.Unzip()
this[UNZIP].on('data', chunk => this[CONSUMECHUNK](chunk))
this[UNZIP].on('error', er => this.abort(er))
this[UNZIP].on('end', _ => {
this[ENDED] = true
this[CONSUMECHUNK]()
})
this[WRITING] = true
const ret = this[UNZIP][ended ? 'end' : 'write'](chunk)
this[WRITING] = false
return ret
}
}
this[WRITING] = true
if (this[UNZIP])
this[UNZIP].write(chunk)
else
this[CONSUMECHUNK](chunk)
this[WRITING] = false
// return false if there's a queue, or if the current entry isn't flowing
const ret =
this[QUEUE].length ? false :
this[READENTRY] ? this[READENTRY].flowing :
true
// if we have no queue, then that means a clogged READENTRY
if (!ret && !this[QUEUE].length)
this[READENTRY].once('drain', _ => this.emit('drain'))
return ret
}
[BUFFERCONCAT] (c) {
if (c && !this[ABORTED])
this[BUFFER] = this[BUFFER] ? Buffer.concat([this[BUFFER], c]) : c
}
[MAYBEEND] () {
if (this[ENDED] &&
!this[EMITTEDEND] &&
!this[ABORTED] &&
!this[CONSUMING]) {
this[EMITTEDEND] = true
const entry = this[WRITEENTRY]
if (entry && entry.blockRemain) {
// truncated, likely a damaged file
const have = this[BUFFER] ? this[BUFFER].length : 0
this.warn('TAR_BAD_ARCHIVE', `Truncated input (needed ${
entry.blockRemain} more bytes, only ${have} available)`, {entry})
if (this[BUFFER])
entry.write(this[BUFFER])
entry.end()
}
this[EMIT](DONE)
}
}
[CONSUMECHUNK] (chunk) {
if (this[CONSUMING])
this[BUFFERCONCAT](chunk)
else if (!chunk && !this[BUFFER])
this[MAYBEEND]()
else {
this[CONSUMING] = true
if (this[BUFFER]) {
this[BUFFERCONCAT](chunk)
const c = this[BUFFER]
this[BUFFER] = null
this[CONSUMECHUNKSUB](c)
} else
this[CONSUMECHUNKSUB](chunk)
while (this[BUFFER] &&
this[BUFFER].length >= 512 &&
!this[ABORTED] &&
!this[SAW_EOF]) {
const c = this[BUFFER]
this[BUFFER] = null
this[CONSUMECHUNKSUB](c)
}
this[CONSUMING] = false
}
if (!this[BUFFER] || this[ENDED])
this[MAYBEEND]()
}
[CONSUMECHUNKSUB] (chunk) {
// we know that we are in CONSUMING mode, so anything written goes into
// the buffer. Advance the position and put any remainder in the buffer.
let position = 0
const length = chunk.length
while (position + 512 <= length && !this[ABORTED] && !this[SAW_EOF]) {
switch (this[STATE]) {
case 'begin':
case 'header':
this[CONSUMEHEADER](chunk, position)
position += 512
break
case 'ignore':
case 'body':
position += this[CONSUMEBODY](chunk, position)
break
case 'meta':
position += this[CONSUMEMETA](chunk, position)
break
/* istanbul ignore next */
default:
throw new Error('invalid state: ' + this[STATE])
}
}
if (position < length) {
if (this[BUFFER])
this[BUFFER] = Buffer.concat([chunk.slice(position), this[BUFFER]])
else
this[BUFFER] = chunk.slice(position)
}
}
end (chunk) {
if (!this[ABORTED]) {
if (this[UNZIP])
this[UNZIP].end(chunk)
else {
this[ENDED] = true
this.write(chunk)
}
}
}
})