From 7940b4b02492eb8052c33252a7c77188181867e6 Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Mon, 11 Jan 2021 20:30:39 -0800 Subject: elasticsearch: simple new dblp and doaj fields --- extra/elasticsearch/release_schema.json | 3 +++ python/fatcat_tools/transforms/elasticsearch.py | 4 ++++ 2 files changed, 7 insertions(+) diff --git a/extra/elasticsearch/release_schema.json b/extra/elasticsearch/release_schema.json index 666a672f..e1f7a79a 100644 --- a/extra/elasticsearch/release_schema.json +++ b/extra/elasticsearch/release_schema.json @@ -72,6 +72,8 @@ "ark_id": { "type": "keyword", "normalizer": "default", "doc_values": false }, "mag_id": { "type": "keyword", "normalizer": "default", "doc_values": false }, "s2_id": { "type": "keyword", "normalizer": "default", "doc_values": false }, + "dblp_id": { "type": "keyword", "normalizer": "default", "doc_values": false }, + "doaj_id": { "type": "keyword", "normalizer": "default", "doc_values": false }, "license": { "type": "keyword", "normalizer": "default" }, "publisher": { "type": "text", "index": true, "analyzer": "textIcu", "search_analyzer":"textIcuSearch" }, "publisher_type": { "type": "keyword", "normalizer": "default" }, @@ -103,6 +105,7 @@ "is_preserved": { "type": "boolean" }, "in_kbart": { "type": "boolean" }, "in_jstor": { "type": "boolean" }, + "in_doaj": { "type": "boolean" }, "in_dweb": { "type": "boolean" }, "in_web": { "type": "boolean" }, "in_ia": { "type": "boolean" }, diff --git a/python/fatcat_tools/transforms/elasticsearch.py b/python/fatcat_tools/transforms/elasticsearch.py index db860a09..f37aadba 100644 --- a/python/fatcat_tools/transforms/elasticsearch.py +++ b/python/fatcat_tools/transforms/elasticsearch.py @@ -72,6 +72,8 @@ def release_to_elasticsearch(entity: ReleaseEntity, force_bool: bool = True) -> jstor_id = release.ext_ids.jstor, ark_id = release.ext_ids.ark, mag_id = release.ext_ids.mag, + dblp_id = release.ext_ids.dblp, + doaj_id = release.ext_ids.doaj, ) t.update(dict( @@ -84,6 +86,7 @@ def release_to_elasticsearch(entity: ReleaseEntity, force_bool: bool = True) -> in_ia_sim = False, in_kbart = None, in_jstor = False, + in_doaj= bool(release.ext_ids.doaj), in_shadows = False, )) @@ -280,6 +283,7 @@ def _rte_container_helper(container: ContainerEntity, release_year: Optional[int if c_extra.get('doaj'): if c_extra['doaj'].get('as_of'): t['is_oa'] = True + t['in_doaj'] = True if c_extra.get('road'): if c_extra['road'].get('as_of'): t['is_oa'] = True -- cgit v1.2.3 From d31434324cfc1924d4d27b52dcc89f3b3e81ef19 Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Mon, 11 Jan 2021 20:31:43 -0800 Subject: proposals: mark older ES doc as implemented Note that this current group of changes is unrelated. --- proposals/2020_elasticsearch_schemas.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/proposals/2020_elasticsearch_schemas.md b/proposals/2020_elasticsearch_schemas.md index c3e79073..8dbff219 100644 --- a/proposals/2020_elasticsearch_schemas.md +++ b/proposals/2020_elasticsearch_schemas.md @@ -1,5 +1,5 @@ -status: planning +status: implemented This document tracks "easy" elasticsearch schema and behavior changes that could be made while being backwards compatible with the current v0.3 schema and -- cgit v1.2.3