summaryrefslogtreecommitdiff
path: root/deps/npm/node_modules/npm-registry-fetch/node_modules/cacache/lib/verify.js
blob: 6a01004c97c4a846c8c60f0c97296ea61b6bbfc3 (plain)
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
'use strict'

const BB = require('bluebird')

const contentPath = require('./content/path')
const finished = BB.promisify(require('mississippi').finished)
const fixOwner = require('./util/fix-owner')
const fs = require('graceful-fs')
const glob = BB.promisify(require('glob'))
const index = require('./entry-index')
const path = require('path')
const rimraf = BB.promisify(require('rimraf'))
const ssri = require('ssri')

BB.promisifyAll(fs)

module.exports = verify
function verify (cache, opts) {
  opts = opts || {}
  opts.log && opts.log.silly('verify', 'verifying cache at', cache)
  return BB.reduce([
    markStartTime,
    fixPerms,
    garbageCollect,
    rebuildIndex,
    cleanTmp,
    writeVerifile,
    markEndTime
  ], (stats, step, i) => {
    const label = step.name || `step #${i}`
    const start = new Date()
    return BB.resolve(step(cache, opts)).then(s => {
      s && Object.keys(s).forEach(k => {
        stats[k] = s[k]
      })
      const end = new Date()
      if (!stats.runTime) { stats.runTime = {} }
      stats.runTime[label] = end - start
      return stats
    })
  }, {}).tap(stats => {
    stats.runTime.total = stats.endTime - stats.startTime
    opts.log && opts.log.silly('verify', 'verification finished for', cache, 'in', `${stats.runTime.total}ms`)
  })
}

function markStartTime (cache, opts) {
  return { startTime: new Date() }
}

function markEndTime (cache, opts) {
  return { endTime: new Date() }
}

function fixPerms (cache, opts) {
  opts.log && opts.log.silly('verify', 'fixing cache permissions')
  return fixOwner.mkdirfix(cache, opts.uid, opts.gid).then(() => {
    // TODO - fix file permissions too
    return fixOwner.chownr(cache, opts.uid, opts.gid)
  }).then(() => null)
}

// Implements a naive mark-and-sweep tracing garbage collector.
//
// The algorithm is basically as follows:
// 1. Read (and filter) all index entries ("pointers")
// 2. Mark each integrity value as "live"
// 3. Read entire filesystem tree in `content-vX/` dir
// 4. If content is live, verify its checksum and delete it if it fails
// 5. If content is not marked as live, rimraf it.
//
function garbageCollect (cache, opts) {
  opts.log && opts.log.silly('verify', 'garbage collecting content')
  const indexStream = index.lsStream(cache)
  const liveContent = new Set()
  indexStream.on('data', entry => {
    if (opts && opts.filter && !opts.filter(entry)) { return }
    liveContent.add(entry.integrity.toString())
  })
  return finished(indexStream).then(() => {
    const contentDir = contentPath._contentDir(cache)
    return glob(path.join(contentDir, '**'), {
      follow: false,
      nodir: true,
      nosort: true
    }).then(files => {
      return BB.resolve({
        verifiedContent: 0,
        reclaimedCount: 0,
        reclaimedSize: 0,
        badContentCount: 0,
        keptSize: 0
      }).tap((stats) => BB.map(files, (f) => {
        const split = f.split(/[/\\]/)
        const digest = split.slice(split.length - 3).join('')
        const algo = split[split.length - 4]
        const integrity = ssri.fromHex(digest, algo)
        if (liveContent.has(integrity.toString())) {
          return verifyContent(f, integrity).then(info => {
            if (!info.valid) {
              stats.reclaimedCount++
              stats.badContentCount++
              stats.reclaimedSize += info.size
            } else {
              stats.verifiedContent++
              stats.keptSize += info.size
            }
            return stats
          })
        } else {
          // No entries refer to this content. We can delete.
          stats.reclaimedCount++
          return fs.statAsync(f).then(s => {
            return rimraf(f).then(() => {
              stats.reclaimedSize += s.size
              return stats
            })
          })
        }
      }, {concurrency: opts.concurrency || 20}))
    })
  })
}

function verifyContent (filepath, sri) {
  return fs.statAsync(filepath).then(stat => {
    const contentInfo = {
      size: stat.size,
      valid: true
    }
    return ssri.checkStream(
      fs.createReadStream(filepath),
      sri
    ).catch(err => {
      if (err.code !== 'EINTEGRITY') { throw err }
      return rimraf(filepath).then(() => {
        contentInfo.valid = false
      })
    }).then(() => contentInfo)
  }).catch({code: 'ENOENT'}, () => ({size: 0, valid: false}))
}

function rebuildIndex (cache, opts) {
  opts.log && opts.log.silly('verify', 'rebuilding index')
  return index.ls(cache).then(entries => {
    const stats = {
      missingContent: 0,
      rejectedEntries: 0,
      totalEntries: 0
    }
    const buckets = {}
    for (let k in entries) {
      if (entries.hasOwnProperty(k)) {
        const hashed = index._hashKey(k)
        const entry = entries[k]
        const excluded = opts && opts.filter && !opts.filter(entry)
        excluded && stats.rejectedEntries++
        if (buckets[hashed] && !excluded) {
          buckets[hashed].push(entry)
        } else if (buckets[hashed] && excluded) {
          // skip
        } else if (excluded) {
          buckets[hashed] = []
          buckets[hashed]._path = index._bucketPath(cache, k)
        } else {
          buckets[hashed] = [entry]
          buckets[hashed]._path = index._bucketPath(cache, k)
        }
      }
    }
    return BB.map(Object.keys(buckets), key => {
      return rebuildBucket(cache, buckets[key], stats, opts)
    }, {concurrency: opts.concurrency || 20}).then(() => stats)
  })
}

function rebuildBucket (cache, bucket, stats, opts) {
  return fs.truncateAsync(bucket._path).then(() => {
    // This needs to be serialized because cacache explicitly
    // lets very racy bucket conflicts clobber each other.
    return BB.mapSeries(bucket, entry => {
      const content = contentPath(cache, entry.integrity)
      return fs.statAsync(content).then(() => {
        return index.insert(cache, entry.key, entry.integrity, {
          uid: opts.uid,
          gid: opts.gid,
          metadata: entry.metadata
        }).then(() => { stats.totalEntries++ })
      }).catch({code: 'ENOENT'}, () => {
        stats.rejectedEntries++
        stats.missingContent++
      })
    })
  })
}

function cleanTmp (cache, opts) {
  opts.log && opts.log.silly('verify', 'cleaning tmp directory')
  return rimraf(path.join(cache, 'tmp'))
}

function writeVerifile (cache, opts) {
  const verifile = path.join(cache, '_lastverified')
  opts.log && opts.log.silly('verify', 'writing verifile to ' + verifile)
  return fs.writeFileAsync(verifile, '' + (+(new Date())))
}

module.exports.lastRun = lastRun
function lastRun (cache) {
  return fs.readFileAsync(
    path.join(cache, '_lastverified'), 'utf8'
  ).then(data => new Date(+data))
}