feat: store pins in datastore instead of a DAG (#2771)

Adds a `.pins` datastore to `ipfs-repo` and uses that to store pins as cbor binary keyed by multihash.

### Format

As stored in the datastore, each pin has several fields:

```javascript
{
  codec: // optional Number, the codec from the CID that this multihash was pinned with, if omitted, treated as 'dag-pb'
  version: // optional Number, the version number from the CID that this multihash was pinned with, if omitted, treated as v0
  depth: // Number Infinity = recursive pin, 0 = direct, 1+ = pinned to a depth
  comments: // optional String user-friendly description of the pin
  metadata: // optional Object, user-defined data for the pin
}
```

Notes:

`.codec` and `.version` are stored so we can recreate the original CID when listing pins.

### Metadata

The intention is for us to be able to add extra fields that have technical meaning to the root of the object, and the user can store application-specific data in the `metadata` field.

### CLI

```console
$ ipfs pin add bafyfoo --metadata key1=value1,key2=value2
$ ipfs pin add bafyfoo --metadata-format=json --metadata '{"key1":"value1","key2":"value2"}'

$ ipfs pin list
bafyfoo

$ ipfs pin list -l
CID      Name    Type       Metadata
bafyfoo  My pin  Recursive  {"key1":"value1","key2":"value2"}

$ ipfs pin metadata Qmfoo --format=json
{"key1":"value1","key2":"value2"}
```

### HTTP API

* '/api/v0/pin/add' route adds new `metadata` argument, accepts a json string
* '/api/v0/pin/metadata' returns metadata as json

### Core API

* `ipfs.pin.addAll` accepts and returns an async iterator
* `ipfs.pin.rmAll` accepts and returns an async iterator

```javascript
// pass a cid or IPFS Path with options
const { cid } = await ipfs.pin.add(new CID('/ipfs/Qmfoo'), {
  recursive: false,
  metadata: {
    key: 'value
  },
  timeout: 2000
}))

// pass an iterable of CIDs
const [{ cid: cid1 }, { cid: cid2 }] = await all(ipfs.pin.addAll([
  new CID('/ipfs/Qmfoo'),
  new CID('/ipfs/Qmbar')
], { timeout: '2s' }))

// pass an iterable of objects with options
const [{ cid: cid1 }, { cid: cid2 }] = await all(ipfs.pin.addAll([
  { cid: new CID('/ipfs/Qmfoo'), recursive: true, comments: 'A recursive pin' },
  { cid: new CID('/ipfs/Qmbar'), recursive: false, comments: 'A direct pin' }
], { timeout: '2s' }))
```

* ipfs.pin.rmAll accepts and returns an async generator (other input types are available)

```javascript
// pass an IPFS Path or CID
const { cid } = await ipfs.rm(new CID('/ipfs/Qmfoo/file.txt'))

// pass options
const { cid } = await all(ipfs.rm(new CID('/ipfs/Qmfoo'), { recursive: true }))

// pass an iterable of CIDs or objects with options
const [{ cid }] = await all(ipfs.rmAll([{ cid: new CID('/ipfs/Qmfoo'), recursive: true }]))
```

Bonus: Lets us pipe the output of one command into another:

```javascript
await pipe(
	ipfs.pin.ls({ type: 'recursive' }),
    (source) => ipfs.pin.rmAll(source)
)

// or
await all(ipfs.pin.rmAll(ipfs.pin.ls({ type: 'recursive'})))
```

BREAKING CHANGES:

* pins are now stored in a datastore, a repo migration will occur on startup
* All deps of this module now use Uint8Arrays in place of node Buffers

Author: achingbrain

docs/BROWSERS.md

@@ -66,7 +66,7 @@ document.addEventListener('DOMContentLoaded', async () => {
   const cid = results[0].hash
   console.log('CID created via ipfs.add:', cid)
   const data = await node.cat(cid)
-  console.log('Data read back via ipfs.cat:', data.toString())
+  console.log('Data read back via ipfs.cat:', new TextDecoder().decode(data))
 })
 </script>
 ```

docs/MIGRATION-TO-ASYNC-AWAIT.md

@@ -171,9 +171,10 @@ e.g.
 
 ```js
 const readable = ipfs.catReadableStream('QmHash')
+const decoder = new TextDecoder()
 
 readable.on('data', chunk => {
-  console.log(chunk.toString())
+  console.log(decoder.decode(chunk))
 })
 
 readable.on('end', () => {
@@ -185,9 +186,10 @@ Becomes:
 
 ```js
 const source = ipfs.cat('QmHash')
+const decoder = new TextDecoder()
 
 for await (const chunk of source) {
-  console.log(chunk.toString())
+  console.log(decoder.decode(chunk))
 }
 
 console.log('done')
@@ -201,9 +203,10 @@ e.g.
 
 ```js
 const readable = ipfs.catReadableStream('QmHash')
+const decoder = new TextDecoder()
 
 readable.on('data', chunk => {
-  console.log(chunk.toString())
+  console.log(decoder.decode(chunk))
 })
 
 readable.on('end', () => {
@@ -216,9 +219,10 @@ Becomes:
 ```js
 const toStream = require('it-to-stream')
 const readable = toStream.readable(ipfs.cat('QmHash'))
+const decoder = new TextDecoder()
 
 readable.on('data', chunk => {
-  console.log(chunk.toString())
+  console.log(decoder.decode(chunk))
 })
 
 readable.on('end', () => {
@@ -238,11 +242,12 @@ e.g.
 
 ```js
 const { pipeline, Writable } = require('stream')
+const decoder = new TextDecoder()
 
-let data = Buffer.alloc(0)
+let data = new Uint8Array(0)
 const concat = new Writable({
   write (chunk, enc, cb) {
-    data = Buffer.concat([data, chunk])
+    data = uint8ArrayConcat([data, chunk])
     cb()
   }
 })
@@ -251,7 +256,7 @@ pipeline(
   ipfs.catReadableStream('QmHash'),
   concat,
   err => {
-    console.log(data.toString())
+    console.log(decoder.decode(chunk))
   }
 )
 ```
@@ -260,11 +265,12 @@ Becomes:
 
 ```js
 const pipe = require('it-pipe')
+const decoder = new TextDecoder()
 
-let data = Buffer.alloc(0)
+let data = new Uint8Array(0)
 const concat = async source => {
   for await (const chunk of source) {
-    data = Buffer.concat([data, chunk])
+    data = uint8ArrayConcat([data, chunk])
   }
 }
 
@@ -273,15 +279,16 @@ const data = await pipe(
   concat
 )
 
-console.log(data.toString())
+console.log(decoder.decode(data))
 ```
 
 ...which, by the way, could more succinctly be written as:
 
 ```js
 const toBuffer = require('it-to-buffer')
+const decoder = new TextDecoder()
 const data = await toBuffer(ipfs.cat('QmHash'))
-console.log(data.toString())
+console.log(decoder.decode(data))
 ```
 
 **Impact 🍏**
@@ -292,11 +299,12 @@ e.g.
 
 ```js
 const { pipeline, Writable } = require('stream')
+const decoder = new TextDecoder()
 
-let data = Buffer.alloc(0)
+let data = new Uint8Array(0)
 const concat = new Writable({
   write (chunk, enc, cb) {
-    data = Buffer.concat([data, chunk])
+    data = uint8ArrayConcat([data, chunk])
     cb()
   }
 })
@@ -305,7 +313,7 @@ pipeline(
   ipfs.catReadableStream('QmHash'),
   concat,
   err => {
-    console.log(data.toString())
+    console.log(decoder.decode(data))
   }
 )
 ```
@@ -315,11 +323,12 @@ Becomes:
 ```js
 const toStream = require('it-to-stream')
 const { pipeline, Writable } = require('stream')
+const decoder = new TextDecoder()
 
-let data = Buffer.alloc(0)
+let data = new Uint8Array(0)
 const concat = new Writable({
   write (chunk, enc, cb) {
-    data = Buffer.concat([data, chunk])
+    data = uint8ArrayConcat([data, chunk])
     cb()
   }
 })
@@ -328,7 +337,7 @@ pipeline(
   toStream.readable(ipfs.cat('QmHash')),
   concat,
   err => {
-    console.log(data.toString())
+    console.log(decoder.decode(data))
   }
 )
 ```
@@ -472,10 +481,12 @@ Use a [for/await](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Refere
 e.g.
 
 ```js
+const decoder = new TextDecoder()
+
 pull(
   ipfs.catPullStream('QmHash'),
   pull.through(chunk => {
-    console.log(chunk.toString())
+    console.log(decoder.decode(data))
   }),
   pull.onEnd(err => {
     console.log('done')
@@ -486,8 +497,10 @@ pull(
 Becomes:
 
 ```js
+const decoder = new TextDecoder()
+
 for await (const chunk of ipfs.cat('QmHash')) {
-  console.log(chunk.toString())
+  console.log(decoder.decode(data))
 }
 
 console.log('done')
@@ -500,10 +513,12 @@ Convert the async iterable to a pull stream.
 e.g.
 
 ```js
+const decoder = new TextDecoder()
+
 pull(
   ipfs.catPullStream('QmHash'),
   pull.through(chunk => {
-    console.log(chunk.toString())
+    console.log(decoder.decode(data))
   }),
   pull.onEnd(err => {
     console.log('done')
@@ -515,11 +530,12 @@ Becomes:
 
 ```js
 const toPull = require('async-iterator-to-pull-stream')
+const decoder = new TextDecoder()
 
 pull(
   toPull.source(ipfs.cat('QmHash')),
   pull.through(chunk => {
-    console.log(chunk.toString())
+    console.log(decoder.decode(data))
   }),
   pull.onEnd(err => {
     console.log('done')
@@ -538,10 +554,12 @@ Use `it-pipe` and `it-concat` concat data from an async iterable.
 e.g.
 
 ```js
+const decoder = new TextDecoder()
+
 pull(
   ipfs.catPullStream('QmHash'),
   pull.collect((err, chunks) => {
-    console.log(Buffer.concat(chunks).toString())
+    console.log(decoder.decode(uint8ArrayConcat(chunks)))
   })
 )
 ```
@@ -551,13 +569,14 @@ Becomes:
 ```js
 const pipe = require('it-pipe')
 const concat = require('it-concat')
+const decoder = new TextDecoder()
 
 const data = await pipe(
   ipfs.cat('QmHash'),
   concat
 )
 
-console.log(data.toString())
+console.log(decoder.decode(data))
 ```
 
 #### Transform Pull Streams
@@ -640,8 +659,8 @@ e.g.
 
 ```js
 const results = await ipfs.addAll([
-  { path: 'root/1.txt', content: Buffer.from('one') },
-  { path: 'root/2.txt', content: Buffer.from('two') }
+  { path: 'root/1.txt', content: 'one' },
+  { path: 'root/2.txt', content: 'two' }
 ])
 
 // Note that ALL files have already been added to IPFS
@@ -654,8 +673,8 @@ Becomes:
 
 ```js
 const addSource = ipfs.addAll([
-  { path: 'root/1.txt', content: Buffer.from('one') },
-  { path: 'root/2.txt', content: Buffer.from('two') }
+  { path: 'root/1.txt', content: 'one' },
+  { path: 'root/2.txt', content: 'two' }
 ])
 
 for await (const file of addSource) {
@@ -669,8 +688,8 @@ Alternatively you can buffer up the results using the `it-all` utility:
 const all = require('it-all')
 
 const results = await all(ipfs.addAll([
-  { path: 'root/1.txt', content: Buffer.from('one') },
-  { path: 'root/2.txt', content: Buffer.from('two') }
+  { path: 'root/1.txt', content: 'one' },
+  { path: 'root/2.txt', content: 'two' }
 ]))
 
 results.forEach(file => {
@@ -682,8 +701,8 @@ Often you just want the last item (the root directory entry) when adding multipl
 
 ```js
 const results = await ipfs.addAll([
-  { path: 'root/1.txt', content: Buffer.from('one') },
-  { path: 'root/2.txt', content: Buffer.from('two') }
+  { path: 'root/1.txt', content: 'one' },
+  { path: 'root/2.txt', content: 'two' }
 ])
 
 const lastResult = results[results.length - 1]
@@ -695,8 +714,8 @@ Becomes:
 
 ```js
 const addSource = ipfs.addAll([
-  { path: 'root/1.txt', content: Buffer.from('one') },
-  { path: 'root/2.txt', content: Buffer.from('two') }
+  { path: 'root/1.txt', content: 'one' },
+  { path: 'root/2.txt', content: 'two' }
 ])
 
 let lastResult
@@ -711,8 +730,8 @@ Alternatively you can use the `it-last` utility:
 
 ```js
 const lastResult = await last(ipfs.addAll([
-  { path: 'root/1.txt', content: Buffer.from('one') },
-  { path: 'root/2.txt', content: Buffer.from('two') }
+  { path: 'root/1.txt', content: 'one' },
+  { path: 'root/2.txt', content: 'two' }
 ]))
 
 console.log(lastResult)

docs/core-api/BLOCK.md

@@ -92,11 +92,12 @@ An optional object which may have the following keys:
 
 ```JavaScript
 // Defaults
-const buf = Buffer.from('a serialized object')
+const buf = new TextEncoder().encode('a serialized object')
+const decoder = new TextDecoder()
 
 const block = await ipfs.block.put(buf)
 
-console.log(block.data.toString())
+console.log(decoder.decode(block.data))
 // Logs:
 // a serialized object
 console.log(block.cid.toString())
@@ -105,12 +106,12 @@ console.log(block.cid.toString())
 
 // With custom format and hashtype through CID
 const CID = require('cids')
-const buf = Buffer.from('another serialized object')
+const buf = new TextEncoder().encode('another serialized object')
 const cid = new CID(1, 'dag-pb', multihash)
 
 const block = await ipfs.block.put(blob, cid)
 
-console.log(block.data.toString())
+console.log(decoder.decode(block.data))
 // Logs:
 // a serialized object
 console.log(block.cid.toString())