From 647be363aa582bb9da0349bd9287b820f68805c3 Mon Sep 17 00:00:00 2001
From: Thord Johansson <tpj.productions@gmail.com>
Date: Sun, 30 Apr 2023 16:31:19 +0200
Subject: [PATCH] Better support for URL lists

---
 stash.sh | 53 ++++++++++++++++++++++++++++++++++++++++++-----------
 1 file changed, 42 insertions(+), 11 deletions(-)

diff --git a/stash.sh b/stash.sh
index 7ffd5d7..59ca8b2 100755
--- a/stash.sh
+++ b/stash.sh
@@ -4,24 +4,55 @@ STASH_IMPORT_DIR=""
 STASH_API_KEY=""
 STASH_HOST=""
 STASH_PORT=0
+STASH_YTDLP_FORMAT="%(title)s [%(id)s].%(ext)s"  # https://github.com/yt-dlp/yt-dlp#output-template
+STASH_PRINT_PREFIX="Get: "
+
 source .env
 
 TARGET_DIR=$(readlink -m "$STASH_IMPORT_DIR/$(date +%Y%m)")
 mkdir -p $TARGET_DIR
 
-regex='^http(s?):\/\/[-a-zA-Z0-9@:%._\+~#=]{2,256}\.[a-z]{2,6}\b[-a-zA-Z0-9@:%_\+.~#?&\/\/=]*$'
+download_file() {
+    FILE_URL="$1"
 
-if [[ $1 =~ $regex ]]; then
-    yt-dlp $1 -o "$TARGET_DIR\%(title)s [%(id)s].%(ext)s"
-else
-    if [[ $1 == *.txt ]]; then
-        yt-dlp -o "$TARGET_DIR\%(title)s [%(id)s].%(ext)s" -a $1
-    else
-	echo $1
-	echo $TARGET_DIR
-        rsync "$1" "$TARGET_DIR\\"
+    extensions="(jpg|JPG|jpeg|JPEG|png|PNG|gif|GIF|mp4|MP4)"
+    rgx_file="^.*\.$extensions$"
+    rgx_filename="[A-Za-z0-9_]*.$extensions"
+    rgx_dbu='http(s?)://.*donmai.us.*/posts/'
+    if [[ $FILE_URL =~ $rgx_dbu ]]; then
+	FILE_URL=$(curl -s "$1" | grep -Eo "http(s?)://.*donmai.us.*/original/[A-Za-z0-9/_]*\.(jpg|jpeg|png|gif|mp4)" | grep '__' -m1)
     fi
+
+    if [[ $FILE_URL =~ $rgx_file ]]; then
+	echo $STASH_PRINT_PREFIX $(echo $FILE_URL | grep -Eo "$rgx_filename")
+        curl -sO "$FILE_URL" --output-dir "$2/"
+    else
+        echo $STASH_PRINT_PREFIX $FILE_URL
+        yt-dlp $FILE_URL -o "$2/$3"
+    fi
+}
+
+rgx_url='^http(s?):\/\/[-a-zA-Z0-9@:%._\+~#=]{2,256}\.[a-z]{2,6}\b[-a-zA-Z0-9@:%_\+.~#?&\/\/=]*$'
+
+if [[ $1 =~ $rgx_url ]]; then
+    # Download using yt-dlp
+    download_file "$1" "$TARGET_DIR" "$STASH_YTDLP_FORMAT"
+else
+    isFile=$(file -0 "$1" | cut -d $'\0' -f2)
+    case "$isFile" in
+        (*text*)
+            # Download as multiple URLs from the provided source file
+            echo "Reading list of $(wc -l $1 | awk '{print $1}') URL(s)"
+            while read p; do
+                download_file "$p" "$TARGET_DIR" "$STASH_YTDLP_FORMAT"
+            done <$1
+            ;;
+        (*)
+            rsync "$1" "$TARGET_DIR/"
+            ;;
+    esac
 fi
 
 # Update stash
-curl -X POST -H "ApiKey: $STASH_API_KEY" -H "Content-Type: application/json" --data '{ "query": "mutation { metadataScan (input:{useFileMetadata: false})}" }' $STASH_HOST:$STASH_PORT/graphql
+echo "Updating Stash..."
+curl -S -s -o /dev/null -X POST -H "ApiKey: $STASH_API_KEY" -H "Content-Type: application/json" --data '{ "query": "mutation { metadataScan (input:{useFileMetadata: false})}" }' $STASH_HOST:$STASH_PORT/graphql