From 17a3ea880402338420699e03bcb24181e4ff3924 Mon Sep 17 00:00:00 2001
From: Rutger Broekhoff
Date: Thu, 2 May 2024 20:27:40 +0200
Subject: Initial commit

Based on dc4ba6a
---
 script/synckv6/oeuf-synckv6.sh | 43 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)
 create mode 100755 script/synckv6/oeuf-synckv6.sh

(limited to 'script/synckv6/oeuf-synckv6.sh')

diff --git a/script/synckv6/oeuf-synckv6.sh b/script/synckv6/oeuf-synckv6.sh
new file mode 100755
index 0000000..6b24347
--- /dev/null
+++ b/script/synckv6/oeuf-synckv6.sh
@@ -0,0 +1,43 @@
+#!/usr/bin/env bash
+
+set -eu
+set -o pipefail
+
+export AWS_ACCESS_KEY_ID="$S3_ACCESS_KEY_ID"
+export AWS_SECRET_ACCESS_KEY="$S3_SECRET_ACCESS_KEY"
+
+set +x
+all_files=()
+declare -A metafiles
+
+while IFS=' ' read -r size filename; do
+	if [[ "$filename" == *.parquet.meta.json ]]; then
+		metafiles["$filename"]=1
+	else
+		all_files+=($filename)
+	fi
+done < <(rclone ls \
+	--s3-provider "$S3_PROVIDER" \
+	--s3-region "$S3_REGION" \
+	--s3-endpoint "$S3_ENDPOINT" \
+	--s3-env-auth \
+	:s3:$S3_BUCKET)
+
+files=()
+for filename in "${all_files[@]}"; do
+	if [[ -v metafiles["$filename.meta.json"] ]]; then
+		files+=($filename)
+	fi
+done
+
+echo "Found ${#files[@]} relevant KV6 Parquet files"
+echo "Synching this directory with these files"
+
+printf "%s\n" "${files[@]}" | rclone copy \
+ 	--s3-provider "$S3_PROVIDER" \
+ 	--s3-region "$S3_REGION" \
+ 	--s3-endpoint "$S3_ENDPOINT" \
+ 	--s3-env-auth \
+	--progress \
+ 	--files-from - \
+ 	:s3:$S3_BUCKET ./
-- 
cgit v1.2.3