基于文件名日期归档文件并100天后删除归档，脚本故障求助

Hey there, let's work through getting your script back on track. It sounds like you're dealing with a few common pitfalls when handling filename patterns, archiving logic, and auto-deletion rules. Let's break this down step by step.

First, Let's Cover the Most Likely Problems #

Chances are your script is failing for one of these reasons:

• Your regex isn't correctly parsing the YYYYMMDD date from that specific filename format
• The script isn't creating the 2018 archive directory properly (or lacks permissions to do so)
• The auto-deletion logic is targeting the wrong files, or the date calculation is off
• You're hitting permission issues when trying to move or delete files

Here's a Corrected Bash Script That Meets Your Needs #

I've written a script that handles your exact filename structure, archives 2018 files, and sets up 100-day auto-deletion, with comments to explain each part:

#!/bin/bash

# Update these paths to match your system
SOURCE_CSV_DIR="/home/your-user/csv-files"
ARCHIVE_ROOT="/home/your-user/csv-archives"

# 1. Create 2018 archive directory if it doesn't exist
mkdir -p "${ARCHIVE_ROOT}/2018"

# Loop through all CSV files in the source directory
for csv_file in "${SOURCE_CSV_DIR}"/*.csv; do
    # Skip if there are no matching CSV files
    [ -e "$csv_file" ] || continue

    # Extract the YYYYMMDD date from the filename using regex
    # Filename format: 11 random letters_6 random alnum_YYYYMMDD.random_num_random_num.csv
    if [[ "$csv_file" =~ ^[a-zA-Z]{11}_[a-zA-Z0-9]{6}_([0-9]{8})\.[0-9]+_[0-9]+\.csv$ ]]; then
        extracted_date="${BASH_REMATCH[1]}"
        
        # Check if the date belongs to 2018
        if [[ "$extracted_date" == "2018"* ]]; then
            echo "Archiving: $csv_file -> ${ARCHIVE_ROOT}/2018/"
            mv "$csv_file" "${ARCHIVE_ROOT}/2018/"
        fi
    else
        echo "Skipping non-matching file: $csv_file"
    fi
done

# 2. Delete archived CSV files older than 100 days
echo "Cleaning up archived files older than 100 days..."
find "${ARCHIVE_ROOT}" -type f -name "*.csv" -mtime +100 -delete

# Optional: Print status to verify
echo "Done! Here's what's left in the 2018 archive:"
ls -lh "${ARCHIVE_ROOT}/2018/"

How to Debug Your Original Script #

If you want to fix your existing script instead of using this example, try these checks:

1. Test your regex: Run this quick test to make sure your pattern captures the date correctly:

   test_filename="abcdefghijk_123abc_20180515.789_101.csv"
   if [[ "$test_filename" =~ ^[a-zA-Z]{11}_[a-zA-Z0-9]{6}_([0-9]{8})\.[0-9]+_[0-9]+\.csv$ ]]; then
       echo "Captured date: ${BASH_REMATCH[1]}"
   else
       echo "Pattern didn't match!"
   fi
   

2. Check directory permissions: Run ls -ld "${SOURCE_CSV_DIR}" "${ARCHIVE_ROOT}/2018" to make sure your user has read/write access to both locations.
3. Validate the auto-deletion rule: Test the find command without the -delete flag first to see which files it would target:

   find "${ARCHIVE_ROOT}" -type f -name "*.csv" -mtime +100
   

4. Add debug echoes: Insert echo "Processing $csv_file" or echo "Date extracted: $extracted_date" into your original script to see where it's failing.

Key Fixes in the Example Script #

• Precise regex matching: The pattern exactly matches your filename structure, so only valid files get processed
• Safe directory creation: mkdir -p ensures the archive directory exists (even if parent folders don't)
• Guard against empty loops: The [ -e "$csv_file" ] check prevents errors when there are no CSV files in the source directory
• Explicit date filtering: We explicitly check if the extracted date starts with 2018 to avoid archiving the wrong files

内容的提问来源于stack exchange，提问作者are