Commit graph

173 commits

Author SHA1 Message Date
Jim Posen
3425d33506 blockchain/indexers: Add methods on CfIndex for batch retrieval.
Fetching all items in one db transaction will save time when
responding to getcfilters or getcfheaders requests.
2018-05-23 16:46:15 -07:00
Jim Posen
e617483b44 blockchain/indexers: Store filter hashes with cfindex. 2018-05-23 16:46:15 -07:00
Jim Posen
d07fd2f333 blockchain/indexers: Reduce duplication in cfindex.
The index will hold three types of entries for each filter type, block
pair: filter, header, and hash. Since they all have similar methods
and implementations, refactor to reduce duplication.
2018-05-23 16:46:15 -07:00
Jim Posen
185577f4c2 blockchain: Implement IntervalBlockHashes method.
This will be used to respond to getcfcheckpt queries.
2018-05-23 16:46:15 -07:00
Jim Posen
07393c0dab blockchain: HeightToHashRange function for pulling range of hashes.
This functionality is required to implement BIP 157.
2018-05-23 16:46:15 -07:00
Olaoluwa Osuntokun
10432160d5 multi: rebase to upstream, update API's 2018-05-23 16:46:15 -07:00
Alex
f4060b107c blockchain/indexers: check for ErrNoData in extended filters
Since the tx hash has moved to the basic filter, generating an extended filter
can result in `ErrNoData`. This is handled by writing a nil filter and giving
it a zero hash.
2018-05-23 16:46:15 -07:00
Alex
c7e7acc7fd multi: use hidden varint for cftypes count; make filter type enum, not uint8 2018-05-23 16:46:15 -07:00
Alex
621c73dad1 multi: change cfilter Extended bool to FilterType uint8
The cfilter BIP specifies that the filter type is a uint8. The
current code encodes it correctly on the wire, but everywhere else,
it's treated as a boolean (false for basic filter, true for
extended). This commit corrects that to account for possible
additional filter types in the future. All package changes are
done in one commit as they're all interdependent. The following
packages are updated:

* blockchain/indexers
* btcjson
* peer
* wire
* main (server.go and rpcserver.go)
2018-05-23 16:46:15 -07:00
Olaoluwa Osuntokun
3d1caa2f83 multi: update to point to roasbeef forks 2018-05-23 16:46:15 -07:00
Olaoluwa Osuntokun
9b9ef42f8a blockchain/indexers: proper handling of empty filters 2018-05-23 16:46:15 -07:00
Olaoluwa Osuntokun
e0943a84bd blockchain/indexers: add a bit more line spacing to cfindex.go 2018-05-23 16:46:15 -07:00
Alex
1c5f25bbf2 blockchain/cfindex: Factor out filter/header calc. 2018-05-23 16:46:15 -07:00
Alex
860100019f blockchain: fix --dropcfindex 2018-05-23 16:46:15 -07:00
Alex
ba7b5f3308 Use NBytes() instead of Bytes() in CFIndex 2018-05-23 16:46:15 -07:00
Alex
9780ef5997 Skip TxIn on coinbase transaction when indexing 2018-05-23 16:46:15 -07:00
Alex
6102e129c5 Fixed a couple of bugs and added --dropcfindex option 2018-05-23 16:46:15 -07:00
pedro martelletto
a5bf8941d5 Define FilterHeaderByBlockHash() 2018-05-23 16:46:15 -07:00
pedro martelletto
e620538343 Generate and store filter headers 2018-05-23 16:46:15 -07:00
pedro martelletto
b53c42f5dc Define dbFetch{Basic,Extended}Header() 2018-05-23 16:46:15 -07:00
pedro martelletto
7673859108 Create db buckets for committed filter hashes 2018-05-23 16:46:15 -07:00
pedro martelletto
ddfaed7f6f Delete extended filters in DisconnectBlock() 2018-05-23 16:46:15 -07:00
pedro martelletto
0a841fefcf Generate and store extended filters in ConnectBlock() 2018-05-23 16:46:15 -07:00
pedro martelletto
71c421db66 Differentiate between basic/extended filters in p2p/RPC 2018-05-23 16:46:15 -07:00
pedro martelletto
f703e18652 Populate filter #1 (extended) 2018-05-23 16:46:15 -07:00
pedro martelletto
f16da156c9 Include data pushes from pkscripts 2018-05-23 16:46:15 -07:00
pedro martelletto
396d28955c better separation between filters; comments 2018-05-23 16:46:15 -07:00
pedro martelletto
57995fd111 Start preparing the ground for layer {0,1} filters 2018-05-23 16:46:15 -07:00
pedro martelletto
472141f88d Switch to new GCS builder interface 2018-05-23 16:46:15 -07:00
pedro martelletto
6d6677b797 zap errCFEntry 2018-05-23 16:46:15 -07:00
pedro martelletto
92c8ec4094 cbfindex.go -> cfindex.go 2018-05-23 16:46:15 -07:00
pedro martelletto
6e5f650be9 CBFilter -> CFilter, discussed with davec@ 2018-05-23 16:46:15 -07:00
pedro martelletto
a77b1e00d5 FilterByBlockHash(), pointed out by davec@ 2018-05-23 16:46:15 -07:00
pedro martelletto
9f02951b0e CloneBytes() -> [:], pointed out by davec@ 2018-05-23 16:46:15 -07:00
pedro martelletto
8b8c7bcf05 Implement DisconnectBlock() for the CBF Indexer 2018-05-23 16:46:15 -07:00
pedro martelletto
43bf8db793 Look up filters from p2p GetCBF message 2018-05-23 16:46:15 -07:00
pedro martelletto
76926f8904 Roll in a dbFetchCBFIndexEntry() 2018-05-23 16:46:15 -07:00
pedro martelletto
9809f4ffdd Instrument basic logging 2018-05-23 16:46:15 -07:00
pedro martelletto
cf74c8c3ca Abstract filter generation to a separate function 2018-05-23 16:46:15 -07:00
pedro martelletto
05d2fdeb11 Rename a variable in CBF's ConnectBlock() 2018-05-23 16:46:15 -07:00
pedro martelletto
620ad5b6fb Store filter in a db block index 2018-05-23 16:46:15 -07:00
pedro martelletto
b2990e7999 Move filter generating code to CBF indexer 2018-05-23 16:46:15 -07:00
pedro martelletto
95cbe2a911 Add the skeleton of a CBF indexer 2018-05-23 16:46:15 -07:00
Jim Posen
0ef0d8c59b blockchain: Initialize database with genesis block.
This fixes a bug introduced by #1066.
2018-02-09 11:34:30 -05:00
Alex Manuskin
50de9da05b blockchain: Fix typo in example_test.go. 2018-01-31 02:40:46 -06:00
Jim Posen
9aa9e79ebf blockchain: Enhance migration to v2 block index bucket.
This updates the block index migration to also store entries with
valid status bytes.
2018-01-28 23:34:56 -06:00
Jim Posen
52cddc19cd blockchain: Persist block status changes to disk.
The block index now tracks the set of dirty block nodes with status
changes that haven't been persisted and flushes the changes to the DB
at the appropriate times.
2018-01-28 23:34:56 -06:00
Jim Posen
31444f5890 blockchain: Add parent to blockNode constructor. 2018-01-28 23:34:56 -06:00
Jim Posen
a4d22d8384 blockchain: Delete block headers from old block index bucket.
As part of the migration to the new block index bucket, also delete
the redundant data from the old bucket.
2018-01-28 23:34:56 -06:00
Jim Posen
74fb6e56da blockchain: Database migration to populate block index bucket.
This creates a migration function that populates the block index
bucket using data from the ffldb block index bucket if it does not
exist.
2018-01-28 23:34:56 -06:00