aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* make fmtBryan Newbold2021-11-161-1/+1
|
* SPNv2: make 'resources' optionalBryan Newbold2021-11-161-1/+1
| | | | | | | | This was always present previously. A change was made to SPNv2 API recently that borked it a bit, though in theory should be present on new captures. I'm not seeing it for some captures, so pushing this work around. It seems like we don't actually use this field anyways, at least for ingest pipeline.
* grobid: handle XML parsing errors, and have them recorded in sandcrawler-dbBryan Newbold2021-11-121-1/+5
|
* ingest_file: more efficient GROBID metadata copyBryan Newbold2021-11-121-3/+3
|
* wrap up crossref refs backfill notesBryan Newbold2021-11-101-0/+47
|
* grobid_tool: helper to process a single fileBryan Newbold2021-11-101-0/+15
|
* ingest: start re-processing GROBID with newer versionBryan Newbold2021-11-101-2/+6
|
* simple persist worker/tool to backfill grobid_refsBryan Newbold2021-11-102-0/+62
|
* grobid: extract more metadata in document TEI-XMLBryan Newbold2021-11-101-0/+5
|
* grobid: update 'TODO' comment based on reviewBryan Newbold2021-11-041-3/+0
|
* update crossref/grobid refs generation notesBryan Newbold2021-11-041-4/+96
|
* crossref grobid refs: another error case (ReadTimeout)Bryan Newbold2021-11-042-5/+11
| | | | | With this last exception handled, was about to get through millions of rows of references, with only a few dozen errors (mostly invalid XML).
* db (postgrest): actually use an HTTP sessionBryan Newbold2021-11-041-12/+24
| | | | Not as important with GET as POST, I think, but still best practice.
* grobid: use requests sessionBryan Newbold2021-11-041-3/+4
| | | | | | This should fix an embarassing bug with exhausting local ports: requests.exceptions.ConnectionError: HTTPConnectionPool(host='wbgrp-svc096.us.archive.org', port=8070): Max retries exceeded with url: /api/processCitationList (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f8dfc24e250>: Failed to establish a new connection: [Errno 99] Cannot assign requested address'))
* grobid crossref refs: try to handle HTTP 5xx and XML parse errorsBryan Newbold2021-11-042-5/+33
|
* grobid: handle weird whitespace unstructured from crossrefBryan Newbold2021-11-041-1/+10
| | | | See also: https://github.com/kermitt2/grobid/issues/849
* crossref persist: batch size depends on whether parsing refsBryan Newbold2021-11-042-2/+8
|
* sql: grobid_refs table JSON as 'JSON' not 'JSONB'Bryan Newbold2021-11-042-3/+3
| | | | | I keep flip-flopping on this, but our disk usage is really large, and if 'JSON' is smaller than 'JSONB' in postgresql at all it is worth it.
* grobid refs backfill progressBryan Newbold2021-11-041-1/+43
|
* record SQL table sizes at start of crossref re-ingestBryan Newbold2021-11-041-0/+19
|
* start notes on crossref refs backfillBryan Newbold2021-11-041-0/+54
|
* crossref persist: make GROBID ref parsing an option (not default)Bryan Newbold2021-11-043-9/+33
|
* add grobid_refs and crossref_with_refs to sandcrawler-db SQL schemaBryan Newbold2021-11-041-0/+21
|
* glue, utils, and worker code for crossref and grobid_refsBryan Newbold2021-11-044-5/+212
|
* update grobid refs proposalBryan Newbold2021-11-041-10/+72
|
* iterated GROBID citation cleaning and processingBryan Newbold2021-11-041-27/+45
| | | | Switched to using just 'key'/'id' for downstream matching.
* grobid citations: first pass at cleaning unstructuredBryan Newbold2021-11-041-2/+34
|
* initial proposal for GROBID refs table and pipelineBryan Newbold2021-11-041-0/+63
|
* initial crossref-refs via GROBID helper routineBryan Newbold2021-11-047-6/+839
|
* pipenv: bump grobid_tei_xml version to 0.1.2Bryan Newbold2021-11-042-11/+11
|
* pdftrio client: use HTTP session for POSTsBryan Newbold2021-11-031-1/+1
|
* workers: use HTTP session for archive.org fetchesBryan Newbold2021-11-031-3/+3
|
* IA (wayback): actually use an HTTP session for replay fetchesBryan Newbold2021-11-031-2/+3
| | | | | | | | I am embarassed this wasn't actually the case already! Looks like I had even instantiated a session but wasn't using it. Hopefully this change, which adds extra retries and better backoff behavior, will improve sandcrawler ingest throughput.
* SPN reingest: 6 hour minimum, 6 month maxBryan Newbold2021-11-031-2/+2
|
* sql: fix typo in quarterly (not weekly) scriptBryan Newbold2021-11-031-1/+1
|
* sql: fixes to ingest_fileset_platform schema (from table creation)Bryan Newbold2021-11-012-12/+12
|
* updates/corrections to old small.json GROBID metadata example fileBryan Newbold2021-10-271-6/+1
|
* remove grobid2json helper file, replace with grobid_tei_xmlBryan Newbold2021-10-277-224/+22
|
* small type annotation things from additional packagesBryan Newbold2021-10-272-5/+14
|
* toolchain config updatesBryan Newbold2021-10-273-10/+6
|
* make fmt (black 21.9b0)Bryan Newbold2021-10-2757-3126/+3991
|
* pipenv: flipflop from yapf back to black; more type packages; bump ↵Bryan Newbold2021-10-272-27/+112
| | | | grobid_tei_xml
* fileset: refactor out tables of helpersBryan Newbold2021-10-273-21/+19
| | | | | | | Having these objects invoked in tables resulted in a whole bunch of objects (including children) getting initialized, which seems like the wrong thing to do. Defer this until the actual ingest fileset worker is initialized.
* gitlab-ci: copy env var in to place for testsBryan Newbold2021-10-271-0/+1
|
* fix type annotations for petabox body fetch helperBryan Newbold2021-10-265-8/+11
|
* small type annotation hackBryan Newbold2021-10-261-1/+1
|
* fileset: fix field renaming bug (caught by mypy)Bryan Newbold2021-10-261-2/+2
|
* fileset ingest: fix table name typo (via mypy)Bryan Newbold2021-10-261-1/+1
|
* update 'XXX' notes from fileset ingest developmentBryan Newbold2021-10-262-9/+6
|
* bugfix: setting html_biblio on ingest resultsBryan Newbold2021-10-262-2/+2
| | | | This was caught during lint cleanup