Shell脚本正则表达式无法限制带重音字符问题排查

Ah, I see the issue here! Your regex ^[a-zA-Z0-9]+$ isn't blocking accented strings like Léo because of bash's locale-dependent regex behavior. Let me break this down and show you how to fix it.

Why Your Current Regex Fails #

When your system uses a UTF-8 locale (like en_US.UTF-8, which is common these days), bash expands the [a-zA-Z] character class to include all Unicode letter characters—not just the ASCII A-Z/a-z range. That means accented characters like é get treated as valid letters, so Léo slips past your check.

Solution 1: Force ASCII Mode with Locale #

The simplest fix is to temporarily switch bash to the C locale (pure ASCII mode) when defining and using your regex. This ensures [a-zA-Z0-9] only matches ASCII letters and numbers, blocking any accented or non-ASCII characters.

Here's how to update your script:

# Switch to ASCII locale to enforce strict character matching
LC_ALL=C
regex="^[a-zA-Z0-9]+$"

while read -r name; do
    if [[ "$name" =~ $regex ]]; then
        echo "Valid name: $name"
    else
        echo "Invalid name (contains accented/non-ASCII characters): $name"
    fi
done

Solution 2: Use Explicit ASCII Range or POSIX Class #

You can also use an explicit ASCII range or POSIX character class (again, paired with LC_ALL=C for consistency):

LC_ALL=C
# Option 1: Explicit ASCII letter/number range
regex="^[0-9A-Za-z]+$"

# Option 2: POSIX alnum class (only matches ASCII chars in C locale)
regex="^[[:alnum:]]+$"

Test It Out #

Let's verify this works as expected:

# Test with accented string
echo "Léo" | ./your_script.sh
# Output: Invalid name (contains accented/non-ASCII characters): Léo

# Test with valid ASCII string
echo "Leo123" | ./your_script.sh
# Output: Valid name: Leo123

内容的提问来源于stack exchange，提问作者danaso