aboutsummaryrefslogtreecommitdiffstats
path: root/python/fatcat_import.py
Commit message (Collapse)AuthorAgeFilesLines
* refactor fatcat_import kafka group namesBryan Newbold2020-01-211-13/+54
| | | | | | | | | | | | | My current understanding is that consumer group names should be one-to-one with topic names. I previously though offsets were stored on a {topic, group} key, but they seem to be mixed and having too many workers in the same group is bad. In particular, we don't want cross-talk or load between QA and prod. All these topics are caught up in prod, so deploying this change and restarting workers should be safe. This commit does not update the elasticsearch or entity updates workers.
* fix trivial one-character typo in fatcat_import.pyBryan Newbold2020-01-171-1/+1
| | | | Should have run tests before pushing!
* actually control pubmed updates with a flagBryan Newbold2020-01-171-0/+4
|
* add missing sentry/raven tagsBryan Newbold2020-01-101-0/+6
| | | | | | Good to have exceptions tracked and stored even for commands run from the command line. But in particular the importer runs as a kafka worker and should be tracking excpetions.
* Merge branch 'martin-datacite-import'Martin Czygan2020-01-081-0/+43
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pipfile.lock is broken. * martin-datacite-import: (68 commits) datacite: pass in doi into factored out method datacite: reformat test cases and use jq . --sort-keys datacite: factor out contributor handling datacite: catch type mismatch in language detection datacite: adjust tests for release_month datacite: name extra.month, extra.release_month datacite: mark additional files as stub datacite: CCDC are entries, mostly datacite: use more specific release_type, if possible datacite: ignore certain names datacite: over 3% records have the same title: stub datacite: fill a few more release_type gaps datacite: adding datacite-specific extra metadata datacite: apply pylint suggestions datacite: fix typos datacite: set release_stage to published by default datacite: month field should be top-level datacite: include month in extra datacite: indicate mismatched file in test datacite: clean abstracts, use unknown value tokens ...
| * datacite: fix typosMartin Czygan2020-01-071-1/+1
| |
| * datacite: remove --lang-detect flagMartin Czygan2020-01-031-4/+0
| | | | | | | | Estimated time for a single call is in the order of 50ms.
| * datacite: use specific auth varMartin Czygan2019-12-281-1/+1
| |
| * datacite: add missing --extid-map-file flagMartin Czygan2019-12-281-0/+4
| |
| * improve datacite field mapping and importMartin Czygan2019-12-281-1/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Current version succeeded to import a random sample of 100000 records (0.5%) from datacite. The --debug (write JSON to stdout) and --insert-log-file (log batch before committing to db) flags are temporary added to help debugging. Add few unit tests. Some edge cases: a) Existing keys without value requires a slightly awkward: ``` titles = attributes.get('titles', []) or [] ``` b) There can be 0, 1, or more (first one wins) titles. c) Date handling is probably not ideal. Datacite has a potentiall fine grained list of dates. The test case (tests/files/datacite_sample.jsonl) refers to https://ssl.fao.org/glis/doi/10.18730/8DYM9, which has date (main descriptor) 1986. The datacite record contains: 2017 (publicationYear, probably the year of record creation with reference system), 1978-06-03 (collected, e.g. experimental sample), 1986 ("Accepted"). The online version of the resource knows even one more date (2019-06-05 10:14:43 by WIEWS update).
| * datacite: importer skeletonMartin Czygan2019-12-281-0/+30
| | | | | | | | | | | | * contributors, title, date, publisher, container, license Field and value analysis via https://github.com/miku/indigo.
* | importers: control update behavior with more-standard flagBryan Newbold2020-01-061-1/+5
|/
* savepapernow result importerBryan Newbold2019-12-121-0/+24
| | | | Based on ingest-file-results importer
* improve argparse usageBryan Newbold2019-12-111-18/+30
| | | | | | | | | | | | | | Use --fatcat-api-url instead of (ambiguous) --host-url for commands that aren't deployed/running via systemd. TODO: update the other --host-url usage, and either roll-out change consistently or support the old arg as an alias during cut-over Use argparse.ArgumentDefaultsHelpFormatter (thanks Martin!) Add help messages for all sub-commands, both as documentation and as a way to get argparse to print available commands in a more readable format.
* tweaks to file ingest importerBryan Newbold2019-12-031-0/+6
| | | | | - allow overriding source filter whitelist (common case for CLI use) - fix editgroup description env variable pass-through
* have ingest-file-results importer operate as crawl-botBryan Newbold2019-11-151-1/+1
| | | | As opposed to sandcrawler-bot
* better ingest-file-results import nameBryan Newbold2019-11-151-1/+1
|
* ingest file result importerBryan Newbold2019-11-151-0/+34
|
* small fixes to confluent-kafka importers/workersBryan Newbold2019-09-201-1/+1
| | | | | | | | - decrease default changelog pipeline to 5.0sec - fix missing KafkaException harvester imports - more confluent-kafka tweaks - updates to kafka consumer configs - bump elastic updates consumergroup (again)
* convert importers to confluent-kafka libraryBryan Newbold2019-09-201-2/+3
|
* start chocula importerBryan Newbold2019-09-031-0/+14
|
* support extids in matched importerBryan Newbold2019-06-201-0/+4
|
* faster LargeFile XML importer for PubMedBryan Newbold2019-05-291-1/+1
|
* make pubmed ref lookups configurableBryan Newbold2019-05-221-1/+8
|
* creative importer for bulk JSTOR importsBryan Newbold2019-05-221-0/+18
|
* pubmed importer command and tweaksBryan Newbold2019-05-221-0/+25
|
* arxiv importer robustification and CLI implBryan Newbold2019-05-211-0/+21
|
* JALC bulk file importerBryan Newbold2019-05-211-0/+21
|
* fix default mimetype (impacted pre-1923 files)Bryan Newbold2019-05-151-1/+5
|
* editgroup description overrideBryan Newbold2019-04-221-1/+11
|
* minor arabesque tweaksBryan Newbold2019-04-181-12/+22
|
* arabesque importer using crawl-bot credsBryan Newbold2019-04-181-1/+1
|
* arabesque import tweaksBryan Newbold2019-04-181-0/+4
|
* early version of arabesque importerBryan Newbold2019-04-121-0/+28
|
* importer for CDL/DASH dat pilot dweb datasetsBryan Newbold2019-03-191-1/+29
|
* new importer: wayback_staticBryan Newbold2019-03-191-0/+48
|
* reduce default import batch size to 50Bryan Newbold2019-01-291-1/+1
|
* batch size as a general import paramBryan Newbold2019-01-281-13/+4
|
* add missing bezerk-mode flag to GROBID importBryan Newbold2019-01-281-3/+8
|
* fix typo in crossref importerBryan Newbold2019-01-281-1/+1
|
* update journal meta import/transformBryan Newbold2019-01-251-3/+3
|
* more import script fixesBryan Newbold2019-01-231-1/+4
|
* update importer scriptBryan Newbold2019-01-231-33/+24
|
* pubmed+datacite tokens; no journal,grobid,matched tokensBryan Newbold2019-01-221-2/+2
|
* issn => journal_metadata in several placesBryan Newbold2019-01-171-9/+9
|
* start refactoring API object passingBryan Newbold2019-01-081-13/+36
|
* crossref importer checks for existing DOIsBryan Newbold2018-11-211-3/+7
|
* correct kafka topic namesBryan Newbold2018-11-201-1/+1
|
* start supporting kafka importersBryan Newbold2018-11-191-3/+17
| | | | A nice feature would be some/any log output as to progress.
* bunch of pylint cleanupBryan Newbold2018-11-151-1/+1
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-25 04:04:28 +0000