aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* refactor citation lookups; add high-level helperBryan Newbold2021-01-191-5/+32
|
* parse: handle pre-quoted strings betterBryan Newbold2021-01-191-1/+10
|
* citation querying helpers (GROBID, fuzzycat, etc)Bryan Newbold2021-01-191-0/+175
|
* query parsing: sniff for citations; more corner casesBryan Newbold2021-01-191-7/+106
|
* grobid2json: add namespace override option for some parsingBryan Newbold2021-01-191-3/+3
| | | | | | | This is to make the reference parsing code re-usable with the simpler processCitations response, which is an XML fragment with no namespace. Should have no impact on existing code paths.
* pipenv: fuzzycatBryan Newbold2021-01-192-13/+102
|
* initial crude query parsing routinesBryan Newbold2021-01-191-0/+71
|
* initial notes on crude query parsingBryan Newbold2021-01-191-0/+18
|
* notes on re-indexing under focalBryan Newbold2021-01-191-0/+31
|
* fix indexing bug (false-y publisher_type?)Bryan Newbold2021-01-181-0/+2
|
* lint: fix small bugs and type annotationsBryan Newbold2021-01-183-5/+6
|
* settings: default to local search/indexingBryan Newbold2021-01-181-1/+1
|
* worker: switch to ES helper for bulk indexingBryan Newbold2021-01-181-8/+10
| | | | | This seems to resolve the problems with index workers failing after a couple hundred docs.
* add missing dotfiles (due to gitignore oops)Bryan Newbold2021-01-184-0/+28
|
* improve 'oa' tag calculationBryan Newbold2021-01-161-4/+4
|
* small corrections to schema/transformBryan Newbold2021-01-162-3/+8
|
* about: another small copy-editing tweakBryan Newbold2021-01-151-2/+2
|
* html: actually remove redundant CSSBryan Newbold2021-01-151-1/+1
|
* make fmtBryan Newbold2021-01-152-14/+19
|
* api: fix /search test, and mypy error on implementationBryan Newbold2021-01-152-12/+22
|
* html: publisher is title text for container nameBryan Newbold2021-01-151-1/+1
|
* html: some CSS refactoring; make journal row a macroBryan Newbold2021-01-152-42/+72
|
* html: improvements to reactive search filters and main serp columnBryan Newbold2021-01-153-23/+31
|
* html: date hover text for yearsBryan Newbold2021-01-141-1/+1
|
* about: small copy editing changesBryan Newbold2021-01-141-13/+6
|
* search: parse and embed a copy of ScholarDoc object in resultsBryan Newbold2021-01-141-1/+6
| | | | Maybe should refactor this to simply replace the object? Hrm.
* crude bibtex and citation formatting, as a demoBryan Newbold2021-01-141-0/+49
|
* schema: make fulltext body optional (eg, for search results)Bryan Newbold2021-01-141-1/+1
|
* serp: refactor ext id display into macro; add dblp,doajBryan Newbold2021-01-141-15/+26
|
* serp: tweak display of tag labelsBryan Newbold2021-01-141-2/+6
|
* api: add type annotations for /search responseBryan Newbold2021-01-141-4/+29
|
* fastapi: /favicon.ico handlerBryan Newbold2021-01-141-1/+5
| | | | | | | HTML responses in the browser should use the indicated path (/static/ia-favicon.ico), but some other responses fall through (like API docs, I think? or unhandled exceptions), so giving up and just handling this route. haproxy should cache if it gets lots of hits.
* fastapi: add basic error/exception handler and pageBryan Newbold2021-01-142-1/+61
|
* html: try goatcounter title redaction againBryan Newbold2021-01-141-1/+1
| | | | | If the value is actually null, goatcounter uses the default, so setting an empty string. Waiting to confirm if this actually works.
* settings: working 'make serve-qa', but only from inside cluster IP spaceBryan Newbold2021-01-141-0/+1
|
* html: resolve footer float on large screens using CSS calc()Bryan Newbold2021-01-141-1/+1
|
* html: fix 'rounded' header barBryan Newbold2021-01-141-2/+3
|
* goatcounter: supress 'title', for privacyBryan Newbold2021-01-141-0/+7
|
* help: fix hanging sentence in 'known issues' sectionBryan Newbold2021-01-141-1/+1
|
* pipenv: lock minio S3 library to <7.0.0Bryan Newbold2021-01-142-11/+20
| | | | | | | | | | | In this upstream commit: https://github.com/minio/minio-py/commit/b81883a98e6f8a09e2903609caabbf0956dd0ec9 The API for errors changes, which makes it harder for use to catch specific exceptions (such as "NoSuchKey" as a Not Found / 404 error). Instead of refactoring, just going to pin the library. We should probably remove this library for a non-implementation-specific S3 client at some point; minio seems simpler than, eg, boto3, but there is probably something ever simpler out there.
* add mocks to work pipeline testBryan Newbold2021-01-141-1/+63
|
* settings: more test configs to prevent actual network attemptsBryan Newbold2021-01-141-0/+3
|
* pipenv: add responses mock dep (for HTTP requests)Bryan Newbold2021-01-142-8/+56
|
* update logo PNGsBryan Newbold2021-01-143-0/+2
| | | | | | | | | | | | | | | | | | | These are legit (paid) exports from placeit.net, to replace the previous demo ones which had watermarks half-removed. The formal PNG download images did not have the same "glow" effect on the red "scholar" text, so I extracted from the browser base64 PNG as before. Used gimp to "crop to selection", then cropped out the lowest "start" background and saved. For small, resized to 200px, retaining ratio. Also ran both files through pngquant to reduce size; I can't visually notice any artifacts or anything from this. It is noticable that placeit.net changed the blue font and has less "depth" to the 3D effect, which also makes the logo appear lighter overall. Very noticble when flipping back and forth, don't think it is a big deal otherwise.
* pytest: suppress warnings for some urllib3 thingsBryan Newbold2021-01-141-0/+2
|
* add support for new identifiers and size_bytes schema fieldsBryan Newbold2021-01-142-4/+16
|
* schema: add more alias helpersBryan Newbold2021-01-141-2/+5
|
* schema: remove unused mag,s2 identifiers; add doaj,dblp,oaiBryan Newbold2021-01-141-4/+10
|
* pipenv: bump elasticsearch to 7.8+ (for async) and fatcat client to 0.3.3+Bryan Newbold2021-01-142-41/+194
|
* serp: expand abstracts using <details>Bryan Newbold2021-01-142-1/+15
| | | | This isn't very polished, but a start, and uses no javascript.