Rooba
/
perkeep
mirror of https://github.com/perkeep/perkeep.git
1
1
Fork 0
Code Issues Packages Projects Releases Wiki Activity
5,909 Commits 44 Branches 11 Tags 59 MiB
Commit Graph

15 Commits

Author SHA1 Message Date
Tamás Gulácsi 9f7e1df32b Remove pkg/context 
Use golang.org/x/net/context instead.

Fixes #648.

Change-Id: I676a2f25458be97610a49d6f954f2102cbd373fa
 2015-12-12 23:09:02 +01:00
Daniel Erat 82429eeedb index,search: Return media tags for describe requests. 
Change-Id: Ie983465554e1dc9f5475753a074e48b363f25013
 2014-04-15 21:11:33 -07:00
Brad Fitzpatrick 328706b8c6 index: bunch of TODOs and planning for edge tracking 
Also, change to storing a pointer to claims so those pointers can be re-used elsewhere.

Change-Id: Ib0979d622ba48f1e2389575d07d7a0930264ebea
 2013-12-09 17:15:34 +04:00
Brad Fitzpatrick bf4d2721ba search: efficient permanode queries, sorted, and with a limit. 
Change-Id: I7f8edfe3636ff3f1344ebb41e27d23073c630ff5
 2013-12-02 20:01:37 -08:00
Brad Fitzpatrick a918fac281 search: let GetRecentPermanodes take a before param. 
For infinite scroll.  Will be done better later, but this works for now.

Change-Id: I88dc7793b643226a4a5e3a583c5de5d4c7fa4819
 2013-11-25 20:35:59 -08:00
Brad Fitzpatrick e8603b1293 Put claims in memory too for in-memory search. Required index schema version bump. 
Change-Id: I194d65476bddea111277cd0b1472c56b5527226b
 2013-11-17 16:52:51 -08:00
Brad Fitzpatrick 0b07171c0c index: change GetOwnerClaims interface for efficiency, rename to AppendClaims 
Change-Id: Icec235961ed8a09256df87c5fe1455cbad81ae3f
 2013-11-17 14:54:30 -08:00
Brad Fitzpatrick 8abaddfa80 index: rearrange interface definition. 
more stable stuff at the top. stuff that needs a re-think below.

Change-Id: Id1682b393718a2e9928088d38a4ddea477698e98
 2013-11-17 13:44:49 -08:00
Brad Fitzpatrick 04a116fed0 TODOs 
Change-Id: I6a7866d889f7446bc89c25a19c57d0c8220bf99f
 2013-11-17 11:01:09 -08:00
Brad Fitzpatrick d82f02db50 Search optimization: keep blobs arranged by camliType in memory; use when possible. 
Change-Id: I8b3ebcf7dfece21d17074da32dc20e4d600e4c50
 2013-11-17 10:52:37 -08:00
Brad Fitzpatrick 3eb493599e in-memory search: better structure for keeping memory corpus and kv 
index in sync, both at start-up and while running and receiving blobs.
They both use the same mechanism now.

Also adds KeyId to the index and Corpus, as the next step. Plenty more
row types remain...

Change-Id: Id79955ba25dc79d5fbd94b0e5248d33dcf71d97e
 2013-11-17 09:41:45 -08:00
Brad Fitzpatrick 272db6545a index: change GetBlobMIMEType to GetBlobMeta. mime types on a blob was wrong anyway. 
And some more of in-memory search, which this makes easier / less garbage-y.

Change-Id: If1fa7dd44842244ae5e5440f84e92d66dc3321fb
 2013-11-16 19:40:14 -08:00
Brad Fitzpatrick e26eacfa2f index: return FileInfo and ImageInfo as values, not pointers. 
Change-Id: I6aba3a74ce233cc834550b561f597e549fb2bfab
 2013-11-16 17:49:28 -08:00
Brad Fitzpatrick 2984897ac7 search: more in-memory search work. 
keep blob metadata in memory, and start of testing all search queries in three modes:
classic index.Storage scanning, all in-memory with corpus scanned from the index.Storage,
and the in-memory corpus built up over time as blobs arrive.

Change-Id: I40536e498a63bece5bd4897cdbbd0cef78085f44
 2013-11-16 17:24:02 -08:00
Brad Fitzpatrick 705107ad80 search/index: invert depedency. search now depends on index. 
creates new package types/camtypes for misc types needed by both. might eventually go away as
search matures.

Change-Id: Ib771ead7bea39936ba478b7e5d58de997060861b
 2013-11-16 15:00:30 -08:00