From f6f7450903bdbe36bd5fff146b942e34ad221557 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@robocracy.org>
Date: Fri, 25 Jan 2019 18:41:33 -0800
Subject: transform and import fixes/tweaks

---
 extra/elasticsearch/fatcat_schema.json | 109 +++++++++++++++++++++++++++++++++
 1 file changed, 109 insertions(+)
 create mode 100644 extra/elasticsearch/fatcat_schema.json

(limited to 'extra/elasticsearch/fatcat_schema.json')

diff --git a/extra/elasticsearch/fatcat_schema.json b/extra/elasticsearch/fatcat_schema.json
new file mode 100644
index 00000000..05583330
--- /dev/null
+++ b/extra/elasticsearch/fatcat_schema.json
@@ -0,0 +1,109 @@
+{
+"settings": {
+    "index": {
+        "analysis": {
+            "analyzer": {
+                "default": {
+                    "type": "custom",
+                    "tokenizer": "standard",
+                    "filter": [ "lowercase", "asciifolding" ]
+                },
+                "textIcu": {
+                    "type": "custom",
+                    "tokenizer": "icu_tokenizer",
+                    "char_filter": [ "icu_normalizer" ],
+                    "filter": [ "icu_folding" ]
+                },
+                "textIcuSearch": {
+                    "type": "custom",
+                    "tokenizer": "icu_tokenizer",
+                    "char_filter": [ "icu_normalizer" ],
+                    "filter": [ "icu_folding" ]
+                }
+            }
+        }
+    }
+},
+"mappings": {
+    "release": {
+        "properties": {
+            "ident":          { "type": "keyword" },
+            "state":          { "type": "keyword" },
+            "revision":       { "type": "keyword" },
+            "title":          { "type": "text", "index": true, "analyzer": "textIcu", "search_analyzer":"textIcuSearch" },
+            "original_title": { "type": "text", "index": true, "analyzer": "textIcu", "search_analyzer":"textIcuSearch" },
+            "release_date":   { "type": "date" },
+            "release_year":   { "type": "integer" },
+            "release_type":   { "type": "keyword" },
+            "release_status": { "type": "keyword" },
+            "language":       { "type": "keyword" },
+            "doi":            { "type": "keyword" },
+            "pmid":           { "type": "keyword" },
+            "pmcid":          { "type": "keyword" },
+            "isbn13":         { "type": "keyword" },
+            "wikidata_qid":   { "type": "keyword" },
+            "core_id":        { "type": "keyword" },
+            "axiv_id":        { "type": "keyword" },
+            "jstor_id":       { "type": "keyword" },
+            "license":        { "type": "keyword" },
+            "publisher":                { "type": "text", "index": true, "analyzer": "textIcu", "search_analyzer":"textIcuSearch" },
+            "container_name":           { "type": "text", "index": true, "analyzer": "textIcu", "search_analyzer":"textIcuSearch" },
+            "container_issnl":          { "type": "keyword" },
+            "container_type":           { "type": "keyword" },
+            "contrib_count":        { "type": "integer" },
+            "contrib_names":        { "type": "text", "index": true, "analyzer": "textIcu", "search_analyzer":"textIcuSearch" },
+            "ref_count":            { "type": "integer" },
+            "file_count":           { "type": "integer" },
+            "fileset_count":        { "type": "integer" },
+            "webcapture_count":     { "type": "integer" },
+            "any_abstract":         { "type": "boolean" },
+
+            "best_pdf_url":         { "type": "keyword" },
+            "ia_pdf_url":           { "type": "keyword" },
+            "is_oa":                { "type": "boolean" },
+            "is_longtail_oa":       { "type": "boolean" },
+            "is_preserved":         { "type": "boolean" },
+            "in_kbart":             { "type": "boolean" },
+            "in_jstor":             { "type": "boolean" },
+            "in_dweb":              { "type": "boolean" },
+            "in_web":               { "type": "boolean" },
+            "in_ia":                { "type": "boolean" },
+            "in_ia_sim":            { "type": "boolean" },
+            "in_shadows":           { "type": "boolean" },
+
+            "author":         { "type": "alias", "path": "contrib_names" },
+            "journal":        { "type": "alias", "path": "container_name" },
+            "date":           { "type": "alias", "path": "release_date" },
+            "year":           { "type": "alias", "path": "release_year" },
+            "issn":           { "type": "alias", "path": "container_issnl" },
+            "oa":             { "type": "alias", "path": "is_oa" },
+            "longtail":       { "type": "alias", "path": "is_longtail_oa" },
+            "lang":           { "type": "alias", "path": "language" },
+            "file_pdf_url":   { "type": "alias", "path": "best_pdf_url" },
+            "is_kept":        { "type": "alias", "path": "in_kbart" }
+        }
+    },
+    "changelog": {
+        "properties": {
+            "index":            { "type": "integer" },
+            "editgorup_id":     { "type": "keyword" },
+            "timestamp":        { "type": "date" },
+            "username":         { "type": "keyword" },
+            "is_bot":           { "type": "boolean" },
+            "is_admin":         { "type": "boolean" },
+            "agent":            { "type": "keyword" },
+            "containers":       { "type": "integer" },
+            "creators":         { "type": "integer" },
+            "files":            { "type": "integer" },
+            "filessets":        { "type": "integer" },
+            "webcaptures":      { "type": "integer" },
+            "releases":         { "type": "integer" },
+            "works":            { "type": "integer" },
+            "created":          { "type": "integer" },
+            "updated":          { "type": "integer" },
+            "deleted":          { "type": "integer" },
+            "total":            { "type": "integer" }
+        }
+    }
+}
+}
-- 
cgit v1.2.3