mail duplicates one-off script

3 files changed, 145 insertions, 86 deletions

diff --git a/find-mail-duplicates.sh b/find-mail-duplicates.sh
new file mode 100755
index 0000000..c8ae097
--- /dev/null
+++ b/find-mail-duplicates.sh
@@ -0,0 +1,60 @@
+#!/usr/bin/env bash
+# find-mail-duplicates.sh
+# Usage: find-mail-duplicates.sh /path/to/maildir
+
+set -e
+
+# Check if a maildir path was provided
+if [ $# -lt 1 ]; then
+    echo "Usage: $0 /path/to/maildir"
+    echo "Example: $0 ~/Mail/ben@bensima.com/Trash"
+    exit 1
+fi
+
+MAILDIR="$1"
+
+# Validate the maildir
+if [ ! -d "$MAILDIR" ]; then
+    echo "Error: Maildir '$MAILDIR' does not exist or is not a directory"
+    exit 1
+fi
+
+
+# Find all UIDs in the maildir
+TMP_UID_FILE=$(mktemp)
+find "$MAILDIR" -type f -name "*,U=*" | sed -E 's/.*,U=([0-9]+)[,:].*/\1/' | sort > "$TMP_UID_FILE"
+
+# Find duplicated UIDs
+DUPLICATE_UIDS=$(uniq -d "$TMP_UID_FILE")
+
+# Count duplicates
+DUPLICATE_COUNT=$(echo "$DUPLICATE_UIDS" | wc -l)
+if [ -z "$DUPLICATE_UIDS" ]; then
+    echo "No duplicate UIDs found in $MAILDIR"
+    rm "$TMP_UID_FILE"
+    exit 0
+fi
+
+
+# Process each duplicate UID
+for UID_NUM in $DUPLICATE_UIDS; do
+    echo "$MAILDIR $UID_NUM"
+    # echo "--------------------------------------------"
+    # echo "Duplicate UID: $UID_NUM"
+
+    # # Find messages with this UID
+    # DUPLICATE_FILES=$(find "$MAILDIR" -type f -name "*,U=$UID_NUM[,:]*")
+    # FILE_COUNT=$(echo "$DUPLICATE_FILES" | wc -l)
+
+    # echo "Found $FILE_COUNT messages with UID $UID_NUM"
+
+    # # List the duplicates
+    # for f in $DUPLICATE_FILES; do
+    #     echo "  $f"
+    # done
+
+    # echo "To fix these duplicates, run:"
+    # echo "  fix-mail-duplicates.sh \"$MAILDIR\" $UID_NUM"
+done
+
+rm "$TMP_UID_FILE"
diff --git a/fix-duplicate-mail-uids.sh b/fix-duplicate-mail-uids.sh
deleted file mode 100755
index 5543fe9..0000000
--- a/fix-duplicate-mail-uids.sh
+++ /dev/null
@@ -1,86 +0,0 @@
-#!/usr/bin/env bash
-#
-# fix-duplicate-uids.sh - Automatically fix duplicate UIDs in maildir
-#
-# this is needed because sometimes iOS mail and mbsync download or move the same
-# file, and i end up with UID collisions. if i just delete the UID from the
-# filename, then mu will generate a new one that doesn't collide.
-
-# Configuration - edit these paths as needed
-MAIL_DIR="$HOME/Mail"
-
-# Color output helpers
-GREEN='\033[0;32m'
-YELLOW='\033[1;33m'
-NC='\033[0m' # No Color
-
-# Create secure temporary files
-TEMP_FILE=$(mktemp)
-trap 'rm -f "$TEMP_FILE"' EXIT
-
-# Process each mailbox separately
-find "$MAIL_DIR" -type d -name cur | while read -r mailbox_cur; do
-    # Get the actual mailbox path (parent of cur)
-    mailbox=$(dirname "$mailbox_cur")
-    mailbox_name=$(basename "$mailbox")
-
-    echo -e "${YELLOW}Checking mailbox: ${mailbox}${NC}"
-
-    # Find duplicate UIDs in this specific mailbox
-    find "$mailbox_cur" -type f -name "*,U=*:*" \
-      | sed -E 's/.*,U=([0-9]+).*/\1/' \
-      | sort \
-      | uniq -c \
-      | awk '$1 > 1 {print $2}' \
-      > "$TEMP_FILE"
-
-    if [ ! -s "$TEMP_FILE" ]; then
-        echo -e "  ${GREEN}No duplicate UIDs found.${NC}"
-        continue
-    fi
-
-    echo -e "  ${YELLOW}Found $(wc -l < "$TEMP_FILE") duplicate UID(s).${NC}"
-
-    # Process each duplicate UID
-    while read -r uid; do
-        echo -e "  ${YELLOW}Fixing duplicate UID: ${uid}${NC}"
-
-        # Find all files with this UID in the current mailbox
-        files=$(find "$mailbox_cur" -type f -name "*,U=${uid}:*" | sort)
-
-        # Find the oldest file with this UID (by creation time)
-        oldest_file=""
-        oldest_time=9999999999
-
-        while read -r file; do
-            # Extract the timestamp from the filename (typical maildir format)
-            timestamp=$(basename "$file" | cut -d. -f1)
-            if [[ $timestamp =~ ^[0-9]+$ ]]; then
-                if [ "$timestamp" -lt "$oldest_time" ]; then
-                    oldest_time=$timestamp
-                    oldest_file=$file
-                fi
-            else
-                # If we can't extract timestamp from filename, use the file's mtime
-                file_time=$(stat -c %Y "$file")
-                if [ "$file_time" -lt "$oldest_time" ]; then
-                    oldest_time=$file_time
-                    oldest_file=$file
-                fi
-            fi
-        done <<< "$files"
-
-        echo "    Keeping original: $(basename "$oldest_file")"
-
-        # Rename all other files with this UID (removing the UID part)
-        while read -r file; do
-            if [ "$file" != "$oldest_file" ]; then
-                new_file=$(echo "$file" | sed -E 's/(.*),U=[0-9]+:(.*)/\1\2/')
-                echo "    Renaming: $(basename "$file") → $(basename "$new_file")"
-                mv "$file" "$new_file"
-            fi
-        done <<< "$files"
-    done < "$TEMP_FILE"
-done
-
-echo -e "${GREEN}Fix complete! Run 'mbsync -a' to resync with new UIDs.${NC}"
diff --git a/fix-mail-duplicates.sh b/fix-mail-duplicates.sh
new file mode 100755
index 0000000..7803649
--- /dev/null
+++ b/fix-mail-duplicates.sh
@@ -0,0 +1,85 @@
+#!/usr/bin/env bash
+# fix-mail-duplicates.sh
+# Usage: fix-mail-duplicates.sh /path/to/maildir 682
+
+set -e
+
+# Check if arguments were provided
+if [ $# -ne 2 ]; then
+    echo "Usage: $0 /path/to/maildir UID_NUMBER"
+    echo "Example: $0 ~/Mail/ben@bensima.com/Archive 682"
+    exit 1
+fi
+
+MAILDIR="$1"
+UID_NUM="$2"
+
+# Validate the arguments
+if [ ! -d "$MAILDIR" ]; then
+    echo "Error: Maildir '$MAILDIR' does not exist or is not a directory"
+    exit 1
+fi
+
+if ! [[ "$UID_NUM" =~ ^[0-9]+$ ]]; then
+    echo "Error: UID must be a number, got '$UID_NUM'"
+    exit 1
+fi
+
+echo "Looking for duplicate UID $UID_NUM in $MAILDIR..."
+
+# Find all files with the EXACT UID (using word boundaries)
+DUPLICATE_FILES=$(find "$MAILDIR" -type f -name "*,U=$UID_NUM[,:]*")
+FILE_COUNT=$(echo "$DUPLICATE_FILES" | wc -l)
+
+if [ "$FILE_COUNT" -le 1 ]; then
+    echo "No duplicates found. Nothing to fix."
+    exit 0
+fi
+
+echo "Found $FILE_COUNT files with the same UID $UID_NUM"
+
+# Get timestamps for each file to determine which is newer
+declare -A FILE_TIMES
+for f in $DUPLICATE_FILES; do
+    # Get the message Date: header if possible, otherwise use file time
+    if command -v formail >/dev/null 2>&1; then
+        DATE_HDR=$(formail -xDate: < "$f" 2>/dev/null)
+        if [ -n "$DATE_HDR" ]; then
+            # Convert to epoch seconds if possible
+            MSG_TIME=$(date -d "$DATE_HDR" +%s 2>/dev/null)
+            if [ $? -eq 0 ]; then
+                FILE_TIMES["$f"]="$MSG_TIME"
+                continue
+            fi
+        fi
+    fi
+
+    # Fallback to file modification time
+    FILE_TIMES["$f"]=$(stat -c %Y "$f")
+done
+
+# Find the oldest file (we'll keep its UID)
+OLDEST_FILE=""
+OLDEST_TIME=9999999999
+for f in "${!FILE_TIMES[@]}"; do
+    if [ "${FILE_TIMES[$f]}" -lt "$OLDEST_TIME" ]; then
+        OLDEST_TIME="${FILE_TIMES[$f]}"
+        OLDEST_FILE="$f"
+    fi
+done
+
+echo "Keeping original UID on oldest file: $OLDEST_FILE"
+
+# Rename all newer files by removing the UID part
+for f in $DUPLICATE_FILES; do
+    if [ "$f" != "$OLDEST_FILE" ]; then
+        # More precise sed pattern to avoid partial UID matches
+        NEW_NAME=$(echo "$f" | sed "s/,U=$UID_NUM\([,:]\)/, \1/" | sed "s/:2,/,/")
+        echo "Renaming newer duplicate:"
+        echo "  From: $f"
+        echo "  To:   $NEW_NAME"
+        mv "$f" "$NEW_NAME"
+    fi
+done
+
+echo "Fixed UID collision for UID $UID_NUM. Please run mbsync again."