From 752e23eb4b8b262aa21b315ff340b21cdb1f32eb Mon Sep 17 00:00:00 2001
From: Stefano Zacchiroli <zack@upsilon.cc>
Date: Tue, 7 Jul 2015 11:07:19 +0200
Subject: [PATCH] SQL: move stuff not strictly related to ghlister to
 crawler.sql

---
 sql/crawler.sql | 78 +++++++++++++++++++++++++++++++++++++++++++++++++
 sql/pimp_db.sql | 78 -------------------------------------------------
 2 files changed, 78 insertions(+), 78 deletions(-)
 create mode 100644 sql/crawler.sql

diff --git a/sql/crawler.sql b/sql/crawler.sql
new file mode 100644
index 0000000..491f382
--- /dev/null
+++ b/sql/crawler.sql
@@ -0,0 +1,78 @@
+
+-- -- return a random sample of repos, containing %percent repositories
+-- create or replace function repos_random_sample_array(percent real)
+-- returns setof repos as $$
+-- declare
+--     samples integer;
+--     repo repos%rowtype;
+--     ids integer[];
+-- begin
+--     select floor(count(*) / 100 * percent) into samples from repos;
+--     ids := array(select id from repos order by id);
+--     for i in 1 .. samples loop
+--      select * into repo
+--          from repos
+--          where id = ids[round(random() * samples)];
+--      return next repo;
+--     end loop;
+--     return;
+-- end
+-- $$
+-- language plpgsql;
+
+-- return a random sample of repositories
+create or replace function repos_random_sample(percent real)
+returns setof repos as $$
+declare
+    sample_size integer;
+begin
+    select floor(count(*) / 100 * percent) into sample_size from repos;
+    return query
+        select * from repos
+        order by random()
+        limit sample_size;
+    return;
+end
+$$
+language plpgsql;
+
+-- -- return a random sample of repositories
+-- create or replace function random_sample_sequence(percent real)
+-- returns setof repos as $$
+-- declare
+--     sample_size integer;
+--     seq_size integer;
+--     min_id integer;
+--     max_id integer;
+-- begin
+--     select floor(count(*) / 100 * percent) into sample_size from repos;
+--     select min(id) into min_id from repos;
+--     select max(id) into max_id from repos;
+--     seq_size := sample_size * 3;  -- IDs are sparse, generate a larger sequence
+--                                   -- to have enough of them
+--     return query
+--         select * from repos
+--         where id in
+--             (select floor(random() * (max_id - min_id + 1))::integer
+--                     + min_id
+--              from generate_series(1, seq_size))
+--         order by random() limit sample_size;
+--     return;
+-- end
+-- $$
+-- language plpgsql;
+
+create or replace function repos_well_known()
+returns setof repos as $$
+begin
+    return query
+        select * from repos
+	where full_name like 'apache/%'
+	   or full_name like 'eclipse/%'
+	   or full_name like 'mozilla/%'
+	   or full_name = 'torvalds/linux'
+	   or full_name = 'gcc-mirror/gcc';
+    return;
+end
+$$
+language plpgsql;
diff --git a/sql/pimp_db.sql b/sql/pimp_db.sql
index f038bf7..9e59e1c 100644
--- a/sql/pimp_db.sql
+++ b/sql/pimp_db.sql
@@ -34,81 +34,3 @@ CREATE VIEW repo_creations AS
          (yesterday.ts = (SELECT max(ts)
                           FROM repos_history
                           WHERE ts < today.ts));
-
--- -- return a random sample of repos, containing %percent repositories
--- create or replace function repos_random_sample_array(percent real)
--- returns setof repos as $$
--- declare
---     samples integer;
---     repo repos%rowtype;
---     ids integer[];
--- begin
---     select floor(count(*) / 100 * percent) into samples from repos;
---     ids := array(select id from repos order by id);
---     for i in 1 .. samples loop
---      select * into repo
---          from repos
---          where id = ids[round(random() * samples)];
---      return next repo;
---     end loop;
---     return;
--- end
--- $$
--- language plpgsql;
-
--- return a random sample of repositories
-create or replace function repos_random_sample(percent real)
-returns setof repos as $$
-declare
-    sample_size integer;
-begin
-    select floor(count(*) / 100 * percent) into sample_size from repos;
-    return query
-        select * from repos
-        order by random()
-        limit sample_size;
-    return;
-end
-$$
-language plpgsql;
-
--- -- return a random sample of repositories
--- create or replace function random_sample_sequence(percent real)
--- returns setof repos as $$
--- declare
---     sample_size integer;
---     seq_size integer;
---     min_id integer;
---     max_id integer;
--- begin
---     select floor(count(*) / 100 * percent) into sample_size from repos;
---     select min(id) into min_id from repos;
---     select max(id) into max_id from repos;
---     seq_size := sample_size * 3;  -- IDs are sparse, generate a larger sequence
---                                   -- to have enough of them
---     return query
---         select * from repos
---         where id in
---             (select floor(random() * (max_id - min_id + 1))::integer
---                     + min_id
---              from generate_series(1, seq_size))
---         order by random() limit sample_size;
---     return;
--- end
--- $$
--- language plpgsql;
-
-create or replace function repos_well_known()
-returns setof repos as $$
-begin
-    return query
-        select * from repos
-	where full_name like 'apache/%'
-	   or full_name like 'eclipse/%'
-	   or full_name like 'mozilla/%'
-	   or full_name = 'torvalds/linux'
-	   or full_name = 'gcc-mirror/gcc';
-    return;
-end
-$$
-language plpgsql;