Delete download orphans the arr cannot match instead of holding them forever — past the age gate an unnameable entry will never import, and clearing it lets a monitored title search for a copy the arr can parse

2026-07-27 17:38:49 -04:00
parent 74ea1a5218
commit 3084546b32
1 changed files with 43 additions and 7 deletions
@@ -33,10 +33,15 @@
 #                movie file → delete (the arr already refused it as not-an-upgrade)
 #   IMPORTABLE — parse API matches it but the library is missing episodes / the movie
 #                → trigger DownloadedEpisodesScan/DownloadedMoviesScan on the folder and
-#                leave it; whatever imports gets swept as REDUNDANT next week, whatever
+#                leave it; whatever imports gets swept as REDUNDANT next run, whatever
 #                the arr rejects (XEM-blocked, season-span files) stays HELD for a human
-#   HELD       — parse API can't match it (series/movie not in the arr) → report only,
-#                deciding whether to add the series or bin the files is a human call
+#   UNMATCHED  — parse API can't match it (series/movie not in the arr) → delete. Past the
+#                age gate an entry the arr cannot even name is not going to import: if the
+#                title is in the library and monitored, clearing it lets the arr search a
+#                copy it can actually parse; if it is not in the library, nothing is
+#                tracking it and it is dead weight either way. Guarded — see safeguard 6
+#   HELD       — IMPORTABLE entries the arr keeps refusing (XEM-blocked, season-span
+#                files), and everything skipped by a guard → report only, human call
 #
 # The queue fetch is a hard gate: if it fails, the whole arr is skipped — with no queue
 # there is no way to tell tracked from orphaned, and guessing means deleting active imports.
@@ -49,8 +54,14 @@
 #   2. Download dir must exist and live under /mnt/ — refuses to walk anything else
 #   3. Queue fetch must succeed (see above)
 #   4. Age gate — nothing under DOWNLOAD_ORPHAN_AGE days is touched
-#   5. Deletion only for JUNK and parse-verified REDUNDANT — never for HELD/IMPORTABLE
-#   6. Run total over DOWNLOAD_ORPHAN_MAX_DELETE_GB aborts the delete pass and notifies —
+#   5. Deletion only for JUNK, parse-verified REDUNDANT, and UNMATCHED — never for
+#      IMPORTABLE or anything a guard has held
+#   6. UNMATCHED deletes require the arr to report a non-empty library. An empty or
+#      restoring database answers every parse with "no match", which would condemn the
+#      entire download dir; the library is queried directly rather than inferred from the
+#      run's own matches, since a small batch that is legitimately all-unmatched is normal
+#      once daily runs have caught up and would otherwise read as a broken database
+#   7. Run total over DOWNLOAD_ORPHAN_MAX_DELETE_GB aborts the delete pass and notifies —
 #      a queue fetch that returned partial data would classify live downloads as orphans,
 #      and an abnormally large delete total is the visible symptom of exactly that
 #      (--i-know-what-im-doing overrides, e.g. for a first run against a known backlog)
@@ -157,10 +168,12 @@ for arr in sonarr radarr; do
        queue_endpoint="queue?pageSize=1000&includeUnknownSeriesItems=true"
        exts_var="SONARR_EXTENSIONS"
        scan_command="DownloadedEpisodesScan"
+        library_endpoint="series"
    else
        queue_endpoint="queue?pageSize=1000&includeUnknownMovieItems=true"
        exts_var="RADARR_EXTENSIONS"
        scan_command="DownloadedMoviesScan"
+        library_endpoint="movie"
    fi

    QUEUE_JSON=$(arr_api "$arr_url" "$arr_key" "v3" "$queue_endpoint" "${arr^}") || {
@@ -179,6 +192,8 @@ for arr in sonarr radarr; do
    DELETE_SIZES=()
    DELETE_LABELS=()
    SCAN_PATHS=()
+    UNMATCHED_PATHS=()
+    UNMATCHED_SIZES=()
    arr_tracked=0; arr_recent=0; arr_held=0; arr_delete_mb=0

    while IFS= read -r entry; do
@@ -231,8 +246,8 @@ for arr in sonarr radarr; do
        fi

        if [[ "$matched" != true ]]; then
-            echo "  $ICON_WARN HELD (no match in ${arr^}): $base (${size_mb}M)"
-            arr_held=$((arr_held + 1))
+            UNMATCHED_PATHS+=("$entry")
+            UNMATCHED_SIZES+=("$size_mb")
        elif (( missing == 0 )); then
            DELETE_PATHS+=("$entry")
            DELETE_SIZES+=("$size_mb")
@@ -244,6 +259,27 @@ for arr in sonarr radarr; do
        fi
    done < <(find "$dl_dir" -mindepth 1 -maxdepth 1 2>/dev/null)

+    # An arr with an empty or still-restoring database answers every parse with "no match",
+    # which would turn the whole download dir into UNMATCHED and delete it. Confirm the
+    # library actually holds titles before trusting a no-match to mean what it says. This
+    # has to be asked of the arr directly — inferring health from the run's own matches
+    # fails on a small batch that is legitimately all-unmatched, which is the normal case
+    # once daily runs have caught up.
+    if (( ${#UNMATCHED_PATHS[@]} > 0 )); then
+        library_count=$(arr_api "$arr_url" "$arr_key" "v3" "$library_endpoint" "${arr^}" | jq 'length' 2>/dev/null)
+        if [[ ! "$library_count" =~ ^[0-9]+$ ]] || (( library_count == 0 )); then
+            warn "  ${arr^}: library reports ${library_count:-no} titles — cannot trust 'no match', holding ${#UNMATCHED_PATHS[@]} unmatched"
+            arr_held=$((arr_held + ${#UNMATCHED_PATHS[@]}))
+        else
+            for i in "${!UNMATCHED_PATHS[@]}"; do
+                DELETE_PATHS+=("${UNMATCHED_PATHS[$i]}")
+                DELETE_SIZES+=("${UNMATCHED_SIZES[$i]}")
+                DELETE_LABELS+=("UNMATCHED")
+                arr_delete_mb=$((arr_delete_mb + UNMATCHED_SIZES[i]))
+            done
+        fi
+    fi
+
    if (( arr_delete_mb / 1024 > DOWNLOAD_ORPHAN_MAX_DELETE_GB )) && [[ "$I_KNOW" != true ]]; then
        error "${arr^}: delete total $((arr_delete_mb / 1024))G exceeds cap of ${DOWNLOAD_ORPHAN_MAX_DELETE_GB}G — aborting delete pass"
        notify "${arr^} download orphan delete total $((arr_delete_mb / 1024))G exceeds ${DOWNLOAD_ORPHAN_MAX_DELETE_GB}G cap on $(hostname) — possible partial queue data, nothing deleted. Re-run with --i-know-what-im-doing if legitimate." \