[airflow] branch v2-1-test updated: Sane detection of the host/port in entrypoint prod (#17847)

[potiuk]

This is an automated email from the ASF dual-hosted git repository.

potiuk pushed a commit to branch v2-1-test
in repository https://gitbox.apache.org/repos/asf/airflow.git


The following commit(s) were added to refs/heads/v2-1-test by this push:
     new c11d368  Sane detection of the host/port in entrypoint prod (#17847)
c11d368 is described below

commit c11d368b236f27681b85c35dcf5ad2ec10df6193
Author: Jarek Potiuk <ja...@potiuk.com>
AuthorDate: Fri Aug 27 17:22:15 2021 +0200

    Sane detection of the host/port in entrypoint prod (#17847)
    
    The previous regexp parsing was well, not perfect closely following the
    ancient Chinese proverb "If you have problem, introduce regexp - you
    will have two problems".
    
    This PR replaces regexp matching with python urlsplit method.
    
    Fixes: #17828
    (cherry picked from commit 275e0d1b91949ad1a1b916b0ffa27009e0797fea)
---
 scripts/in_container/prod/entrypoint_prod.sh | 25 +++++++------------------
 1 file changed, 7 insertions(+), 18 deletions(-)

diff --git a/scripts/in_container/prod/entrypoint_prod.sh 
b/scripts/in_container/prod/entrypoint_prod.sh
index 674b569..bcf57c0 100755
--- a/scripts/in_container/prod/entrypoint_prod.sh
+++ b/scripts/in_container/prod/entrypoint_prod.sh
@@ -75,6 +75,7 @@ function run_nc() {
     nc -zvvn "${ip}" "${port}"
 }
 
+
 function wait_for_connection {
     # Waits for Connection to the backend specified via URL passed as first 
parameter
     # Detects backend type depending on the URL schema and assigns
@@ -83,24 +84,12 @@ function wait_for_connection {
     # It tries `CONNECTION_CHECK_MAX_COUNT` times and sleeps 
`CONNECTION_CHECK_SLEEP_TIME` between checks
     local connection_url
     connection_url="${1}"
-    local detected_backend=""
-    local detected_host=""
-    local detected_port=""
-
-    # Auto-detect DB parameters
-    # Examples:
-    #  postgres://YourUserName:password@YourHostname:5432/YourDatabaseName
-    #  postgres://YourUserName:password@YourHostname:5432/YourDatabaseName
-    #  postgres://YourUserName:@YourHostname:/YourDatabaseName
-    #  postgres://YourUserName@YourHostname/YourDatabaseName
-    [[ ${connection_url} =~ 
([^:]*)://([^:@]*):?([^@]*)@?([^/:]*):?([0-9]*)/([^\?]*)\??(.*) ]] && \
-        detected_backend=${BASH_REMATCH[1]} &&
-        # Not used USER match
-        # Not used PASSWORD match
-        detected_host=${BASH_REMATCH[4]} &&
-        detected_port=${BASH_REMATCH[5]} &&
-        # Not used SCHEMA match
-        # Not used PARAMS match
+    local detected_backend
+    detected_backend=$(python -c "from urllib.parse import urlsplit; import 
sys; print(urlsplit(sys.argv[1]).scheme)" "${connection_url}")
+    local detected_host
+    detected_host=$(python -c "from urllib.parse import urlsplit; import sys; 
print(urlsplit(sys.argv[1]).hostname)" "${connection_url}")
+    local detected_port
+    detected_port=$(python -c "from urllib.parse import urlsplit; import sys; 
print(urlsplit(sys.argv[1]).port or '')" "${connection_url}")
 
     echo BACKEND="${BACKEND:=${detected_backend}}"
     readonly BACKEND