summaryrefslogtreecommitdiff
path: root/ref_queries
diff options
context:
space:
mode:
Diffstat (limited to 'ref_queries')
-rw-r--r--ref_queries/update_roster_match_relationship.sql21
1 files changed, 21 insertions, 0 deletions
diff --git a/ref_queries/update_roster_match_relationship.sql b/ref_queries/update_roster_match_relationship.sql
new file mode 100644
index 0000000..a9c8cdc
--- /dev/null
+++ b/ref_queries/update_roster_match_relationship.sql
@@ -0,0 +1,21 @@
+-- i am thinking how fast slow this will be and if we should do this in DB.
+-- maybe we need a script to run a cursor and update records - dont know...
+-- a few runs on 35K records was closing my server connection on my shitty laptop
+
+-- this will update all roster with match_apiId
+update "roster"
+set "match_apiId" = match."apiId"
+from match
+where (roster."apiId" = roster_1 or roster."apiId" = roster_2)
+
+
+-- this is same as before however should only pick empty records
+update "roster"
+set "match_apiId" = match."apiId"
+from match
+where roster."match_apiId" is not null and (roster."apiId" = roster_1 or roster."apiId" = roster_2)
+
+
+
+-- this should always return a count of 2 for each match_apiId :)
+select "match_apiId", count(*) from roster group by "match_apiId"