summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* finish fatcat-api => fatcat-api-spec refactorBryan Newbold2018-09-118-17/+17
|
* fix fatcat-api-spec depsBryan Newbold2018-09-112-6/+6
|
* new rustfmtBryan Newbold2018-09-112-26/+35
|
* refactor fatcat-api => fatcat-api-specBryan Newbold2018-09-1123-67/+112
|
* update TODO lists (september plan)Bryan Newbold2018-09-113-90/+86
|
* rust clippy (lint) tweaksBryan Newbold2018-09-114-14/+15
|
* use JSONB over JSONBryan Newbold2018-09-102-26/+26
| | | | | | Sort of an impulsive "why not" decision. From earlier tests, expect this to consume a few % more disk (but not index)... adds a lot more query flexibility if we want that in the future. Seems worth the price.
* sql: add an index from cockroach branchBryan Newbold2018-09-101-0/+2
|
* add import timing notes from weekendBryan Newbold2018-09-101-0/+249
|
* update TODOBryan Newbold2018-09-101-8/+2
|
* many small API cleanupsBryan Newbold2018-09-105-193/+148
| | | | | | - use FatCatId much more often (though not everywhere yet) - more consistent types - remove redundant error handling code in wrappers
* re-gen rust codeBryan Newbold2018-09-107-61/+133
|
* make API types more consistentBryan Newbold2018-09-101-6/+14
|
* continue CRUD refactor, removing CUD handlersBryan Newbold2018-09-104-276/+60
| | | | | | | | Can't remove the "get" handlers because of the expand parameter, which is only implemented for releases and in the handler (not in the CRUD trait, yet). Also didn't do the batch handler stuff yet.
* squelch unused macro warningBryan Newbold2018-09-101-0/+2
|
* database_entity_crud -> api_entity_crudBryan Newbold2018-09-105-5/+5
|
* fix python import of ORCIDs ending 'X'Bryan Newbold2018-09-104-1/+9
|
* TODO updateBryan Newbold2018-09-101-0/+7
|
* remove 'IS NOT NULL' identifier index constraintsBryan Newbold2018-09-101-14/+14
| | | | | | | | | | | These seemed to be resulting in table scans on 404s in QA with postgres 10, despite the adding "IS NOT NULL" WHERE clauses earlier. Query time was very significant, even for the creator table (1.2 seconds or so on SSD). I looked at using hash indices (which have improved in postgres 10), which could save index size (disk and RAM) and potentially be faster for these trivial exact lookups, but didn't go for it at this time.
* generic edit accept, and per-row variantBryan Newbold2018-09-092-44/+158
| | | | The per-row variant is for use with cockroach.
* api_server: give up and use star schema importBryan Newbold2018-09-091-6/+1
|
* update TODOBryan Newbold2018-09-091-9/+11
|
* parallelize abstracts insertionBryan Newbold2018-09-091-15/+20
|
* TODO and commentBryan Newbold2018-09-075-1/+1102
|
* fixes after autoaccept mergeBryan Newbold2018-09-074-49/+41
|
* re-codegenBryan Newbold2018-09-071-1/+11
|
* fmtBryan Newbold2018-09-075-207/+364
|
* Merge branch 'autoaccept' into http-verbsBryan Newbold2018-09-0720-75/+628
|\ | | | | | | | | | | | | | | | | Started resolving conflicts in: TODO notes/cloud_instances.txt rust/fatcat-api/README.md rust/src/api_server.rs
| * autoaccept notesBryan Newbold2018-08-301-0/+31
| |
| * "true" is the only truthy value (oh no)Bryan Newbold2018-08-303-3/+3
| |
| * importers: create editgroup and passBryan Newbold2018-08-305-9/+31
| |
| * editgroup IDs are fcid, not uuidBryan Newbold2018-08-301-1/+1
| |
| * minimal autoaccept testsBryan Newbold2018-08-301-0/+41
| |
| * editgroup parameter for batch insertsBryan Newbold2018-08-302-6/+21
| |
| * WIP: autoacceptBryan Newbold2018-08-2411-64/+269
| |
| * cost notes on hosting a fatcat mirrorBryan Newbold2018-08-201-0/+8
| |
| * notes on recent bulk importBryan Newbold2018-08-201-0/+292
| |
| * codegen pythonBryan Newbold2018-08-201-5/+25
| |
| * fix bug in new batch create python importerBryan Newbold2018-08-201-1/+1
| |
| * python autoaccept implementationBryan Newbold2018-08-204-6/+3
| | | | | | | | | | | | More to review: - duplicate containers getting created in batch mode? eg, if multiple releases with same container. should push+autoaccept those anyways?
| * implement autoacceptBryan Newbold2018-08-202-20/+46
| | | | | | | | | | | | Need to review: - changelog creation - clobbering of existing editgroup flag
| * codegenBryan Newbold2018-08-208-51/+223
| |
| * autoaccept flag for batch insertsBryan Newbold2018-08-201-0/+29
| |
* | fix/replace a lot of expect() callsBryan Newbold2018-09-073-47/+48
| |
* | mostly done with CRUD refactorBryan Newbold2018-09-073-625/+745
| | | | | | | | One failing test in this commit.
* | batch inserts by defaultBryan Newbold2018-09-073-27/+95
| |
* | generic create (multiple inserts)Bryan Newbold2018-09-071-26/+24
| | | | | | | | | | | | | | This does away entirely with the fancy CTE-based custom SQL inserts. Because performance is currently only important for batch inserts, I think this is acceptable, and will refactor to do batch inserts as actual batch SQL operations next.
* | crud: refactor to split out rev insertBryan Newbold2018-09-071-43/+51
| |
* | major CRUD refactorBryan Newbold2018-09-075-85/+336
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is the start of a large refactor to move all entity CRUD (create, read, update, delete) model/database code into it's own file. HACKING has been updated with an overview of what happens in each file. Next steps: - split rev (and sub-table) insertion in to db_rev_insert and make create/update generic - inserts should be batch (vector) by default - move all other entities into this new trait framework - bypass api_server wrappers and call into CRUD from api_wrappers for entity ops (should be a big cleanup)
* | update TODOBryan Newbold2018-09-061-0/+7
| |
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-24 02:42:45 +0000