Add curation rates stats to JSON output

Refs pombase/pombase-chado#651

src/bin/pombase-chado-json.rs

@@ -2,6 +2,7 @@ extern crate getopts;
 
 use deadpool_postgres::{Pool, Manager};
 use pombase::bio::pdb_reader::read_pdb_data;
+use pombase::db::ChadoQueries;
 
 use std::str::FromStr;
 
@@ -131,6 +132,8 @@ async fn main() -> Result<(), Box<dyn Error>> {
 
     let raw = Raw::new(&mut client).await?;
 
+    let chado_queries = ChadoQueries::new(&mut client).await?;
+
     let interpro_data = parse_interpro(&config, &interpro_json);
     let pfam_data = maybe_pfam_json.map(|pfam_json| parse_pfam(&pfam_json));
     let rnacentral_data =
@@ -151,6 +154,7 @@ async fn main() -> Result<(), Box<dyn Error>> {
     let web_data_build = WebDataBuild::new(&raw, &interpro_data, &pfam_data,
                                            &rnacentral_data, &gene_history,
                                            &pdb_entry_map, &pdb_ref_entry_map,
+                                           &chado_queries,
                                            &config);
     let web_data = web_data_build.get_web_data();
 

src/pombase/data_types.rs

@@ -53,6 +53,7 @@ use std::cmp::Ordering;
 use std::hash::{Hash, Hasher};
 
 use crate::api::search_types::SolrMatchHighlight;
+use crate::db::chado_queries::CommunityResponseRate;
 use crate::web::config::*;
 use crate::types::*;
 use crate::interpro::InterProMatch;
@@ -2595,6 +2596,7 @@ pub struct DetailedStats {
     pub ltp_genes_per_pub_per_year_range: StatsFloatTable,
     pub ltp_annotations_per_pub_per_year_range: StatsFloatTable,
     pub htp_annotations_per_pub_per_year_range: StatsFloatTable,
+    pub community_response_rates: Vec<CommunityResponseRate>,
 }
 
 #[derive(Serialize, Deserialize, Clone, Debug)]

src/pombase/db/chado_queries.rs

@@ -0,0 +1,68 @@
+extern crate tokio_postgres;
+
+use self::tokio_postgres::Client;
+
+#[derive(Serialize, Deserialize, Clone, Debug)]
+pub struct CommunityResponseRate {
+  pub year: i32,
+  pub submitted: i64,
+  pub sent_sessions: i64,
+  pub response_rate: f32,
+}
+
+#[derive(Serialize, Deserialize, Clone, Debug)]
+pub struct ChadoQueries {
+  pub community_response_rates: Vec<CommunityResponseRate>,
+}
+
+const RESPONSE_RATE_SQL: &str = "
+WITH counts as (SELECT year,
+
+  (SELECT COUNT (*)
+   FROM pombase_publication_curation_summary
+   WHERE canto_curator_role = 'community'
+   AND (canto_annotation_status = 'NEEDS_APPROVAL' OR canto_annotation_status = 'APPROVAL_IN_PROGRESS' OR canto_annotation_status = 'APPROVED')
+     AND (canto_session_submitted_date IS NOT NULL
+          AND canto_session_submitted_date <= (YEAR || '-12-30')::date)) AS submitted,
+
+  (SELECT COUNT (*)
+   FROM pombase_publication_curation_summary
+   WHERE canto_curator_role = 'community'
+   AND (canto_approved_date is not null OR canto_first_sent_to_curator_year IS NOT NULL
+     AND canto_first_sent_to_curator_year <= YEAR)) AS sent_sessions
+
+FROM generate_series(2013,
+                       (SELECT extract(YEAR
+                                       FROM CURRENT_DATE))::integer) AS YEAR)
+SELECT year, submitted, sent_sessions, trunc(100.0*submitted/sent_sessions,1)::real as response_rate from counts;
+";
+
+async fn get_community_response_rates(conn: &mut Client)
+  -> Result<Vec<CommunityResponseRate>, tokio_postgres::Error>
+{
+  let mut ret = vec![];
+
+  let result = conn.query(RESPONSE_RATE_SQL, &[]).await?;
+
+  for row in &result {
+    let el = CommunityResponseRate {
+      year: row.get(0),
+      submitted: row.get(1),
+      sent_sessions: row.get(2),
+      response_rate: row.get(3),
+    };
+
+    ret.push(el);
+  }
+
+  Ok(ret)
+}
+
+impl ChadoQueries {
+  pub async fn new(conn: &mut Client) -> Result<ChadoQueries, tokio_postgres::Error> {
+    let community_response_rates = get_community_response_rates(conn).await?;
+    Ok(ChadoQueries {
+      community_response_rates,
+    })
+  }
+}

src/pombase/db/mod.rs

@@ -1,5 +1,7 @@
 pub mod raw;
 pub mod processed;
+pub mod chado_queries;
 
 pub use raw::Raw;
 pub use processed::Processed;
+pub use chado_queries::ChadoQueries;

src/pombase/web/data.rs

@@ -1575,7 +1575,7 @@ impl WebData {
         let file_name = String::new() + output_dir + "/detailed_stats.json";
         let f = File::create(file_name).expect("Unable to open file");
         let mut writer = BufWriter::new(&f);
-        writer.write_all(s.as_bytes()).expect("Unable to write stats.json");
+        writer.write_all(s.as_bytes()).expect("Unable to write detailed_stats.json");
 
         Ok(())
     }

src/pombase/web/data_build.rs

@@ -13,7 +13,7 @@ use std::collections::{HashMap, HashSet};
 
 use crate::bio::pdb_reader::{PDBGeneEntryMap, PDBRefEntryMap};
 use crate::bio::protein_view::make_protein_view_data_map;
-use crate::db::raw::*;
+use crate::db::{raw::*, ChadoQueries};
 
 use crate::gene_history::GeneHistoryMap;
 use crate::types::*;
@@ -65,6 +65,7 @@ pub struct WebDataBuild<'a> {
     all_gene_history: &'a Option<GeneHistoryMap>,
     pdb_gene_entry_map: &'a Option<PDBGeneEntryMap>,
     pdb_ref_entry_map: &'a Option<PDBRefEntryMap>,
+    chado_queries: &'a ChadoQueries,
     config: &'a Config,
 
     genes: UniquenameGeneMap,
@@ -879,6 +880,7 @@ impl <'a> WebDataBuild<'a> {
                all_gene_history: &'a Option<GeneHistoryMap>,
                pdb_gene_entry_map: &'a Option<PDBGeneEntryMap>,
                pdb_ref_entry_map: &'a Option<PDBRefEntryMap>,
+               chado_queries: &'a ChadoQueries,
                config: &'a Config) -> WebDataBuild<'a>
     {
         WebDataBuild {
@@ -889,6 +891,7 @@ impl <'a> WebDataBuild<'a> {
             all_gene_history,
             pdb_gene_entry_map,
             pdb_ref_entry_map,
+            chado_queries,
             config,
 
             genes: BTreeMap::new(),
@@ -7300,6 +7303,7 @@ phenotypes, so just the first part of this extension will be used:
                  header: annotations_per_year_header,
                  data: htp_annotations_per_year_range,
             },
+            community_response_rates: self.chado_queries.community_response_rates.clone(),
         }
     }
 

tests/test_db_json.rs

@@ -12,6 +12,7 @@ use self::pombase::data_types::*;
 use self::pombase::web::config::*;
 use self::pombase::web::data_build::*;
 use self::pombase::web::data::*;
+use pombase::db::ChadoQueries;
 use pombase::interpro::DomainData;
 
 mod util;
@@ -747,10 +748,13 @@ fn get_test_web_data() -> WebData {
     let rnacentral_data = Some(HashMap::new());
     let pfam_data = Some(HashMap::new());
     let gene_history = None;
+    let chado_queries = ChadoQueries {
+        community_response_rates: vec![],
+    };
 
     let web_data_build = WebDataBuild::new(&raw, &domain_data, &pfam_data,
                                            &rnacentral_data, &gene_history,
-                                           &None, &None, &config);
+                                           &None, &None, &chado_queries, &config);
     web_data_build.get_web_data()
 }