Shell脚本if语句语法错误排查及日志旧信息删除方法咨询

Hey there! Let’s break down your two shell script problems and solve them step by step.

1. Fixing the if/elif/else Syntax Error ("Unexpected token 'else'") #

That error almost always traces back to a formatting misstep in your conditional blocks. Let’s go through the most common culprits and how to fix them:

Common Causes & Quick Fixes #

• Missing spaces around brackets: When using the [ test command, you must add spaces after [ and before ].
  ❌ Wrong: if[$count -gt 5]
  ✅ Correct: if [ $count -gt 5 ]
• Missing semicolon for inline then: If you write if and then on the same line, a semicolon is required before then.
  ❌ Wrong: if [ $count -gt 5 ] then
  ✅ Correct: if [ $count -gt 5 ]; then
• Unmatched parentheses: If you’re using extended tests with [[, make sure you close it with ]]—don’t mix [ and [[ in the same condition.
• Typos: Double-check that you wrote elif (not elif or else if) and that else has no extra characters before it.

Example of a Corrected Block #

Here’s a properly formatted if/elif/else structure that won’t throw syntax errors:

#!/bin/bash
count=$(wc -l < input.txt)

if [ $count -gt 100 ]; then
    echo "Cleaning up excess entries..."
    # Your cleanup logic here
elif [ $count -lt 10 ]; then
    echo "Not enough entries to clean."
else
    echo "Entry count is within normal range."
fi

Double-check line 30 of your count.sh—it’s almost certainly one of these formatting issues causing the error.

2. Deleting Old Status Entries with date #

To remove the oldest status information from output.txt, you’ll need to identify timestamped lines, compare them to a cutoff date, and keep only newer entries. Here are two practical approaches:

Approach 1: Keep the Latest N Entries #

If your status lines are ordered by timestamp (newest last), use tail to retain only the most recent entries:

# Keep the last 500 entries (adjust the number to your needs)
tail -n 500 output.txt > temp.txt && mv temp.txt output.txt

If entries aren’t ordered, sort them first by timestamp, then keep the latest N:

# Assuming timestamp is in the first two fields (e.g., 2024-05-20 14:30:00)
sort -k1,2r output.txt | head -n 500 | sort -k1,2 > temp.txt && mv temp.txt output.txt

• sort -k1,2r: Sorts lines in reverse order by date + time.
• head -n 500: Grabs the top 500 newest entries.
• sort -k1,2: Re-sorts them back to chronological order.

Approach 2: Delete Entries Older Than X Days #

If you want to remove entries older than a specific timeframe (e.g., 7 days ago), use date to generate a cutoff timestamp and filter lines with awk:

#!/bin/bash
# Calculate cutoff timestamp (7 days ago, in seconds since epoch)
cutoff=$(date -d "7 days ago" +%s)

# Filter entries newer than the cutoff (adjust $1 $2 to match your timestamp fields)
awk -v cutoff="$cutoff" '{
    timestamp_str = $1 " " $2
    cmd = "date -d \"" timestamp_str "\" +%s"
    cmd | getline entry_timestamp
    close(cmd)
    if (entry_timestamp > cutoff) print $0
}' output.txt > temp.txt && mv temp.txt output.txt

If your timestamps use ISO 8601 format (e.g., 2024-05-20), skip timestamp conversion and compare strings directly:

cutoff_date=$(date -d "7 days ago" +%Y-%m-%d)
awk -v cutoff="$cutoff_date" '$1 >= cutoff' output.txt > temp.txt && mv temp.txt output.txt

内容的提问来源于stack exchange，提问作者Desert