如何在Shell中解析文本文件指定段落并构建数组？

Got it, let's figure out how to parse those multi-line user blocks into a shell array—since you already know how to handle single-line key-value pairs with grep and cut, this is just a step up for multi-line content.

First, let's assume your users.txt looks something like this (filling in the blanks based on your description):

Authorized Admins:
admin1
admin2
admin3
Regular Users:
user1
user2

1. Use awk to extract the target block (cleanest method) #

awk is built for handling line ranges, so it's perfect here. This command will grab all lines between the Authorized Admins: header and the next section header (assuming section headers end with a colon):

# Populate the admins array
admins=($(awk '/Authorized Admins:/ {flag=1; next} /:$/ {flag=0} flag' users.txt))

# Verify the array contents
echo "Total admins: ${#admins[@]}"
for admin in "${admins[@]}"; do
    echo "- $admin"
done

Here's the breakdown of the awk logic:

• /Authorized Admins:/ {flag=1; next}: When we hit the target header, flip a flag to "active" and skip the header line itself
• /:$/ {flag=0}: When we hit a line ending with : (the next section header), flip the flag to "inactive"
• flag: Only print lines where the flag is active

2. Use grep with context (if sections are separated by blank lines) #

If your users.txt uses blank lines to split sections, like this:

Authorized Admins:
admin1
admin2

Regular Users:
userA

You can use grep's -A flag to pull lines after the header, then filter out the unwanted content:

admins=($(grep -A 100 "Authorized Admins:" users.txt | grep -vE "Authorized Admins:|^$|Regular Users:"))

• -A 100: Grab the header plus the next 100 lines (adjust the number if you have more admins)
• grep -vE ...: Exclude the header line, blank lines, and the next section's header

Pro Tip for tricky usernames #

If your usernames might include spaces or special characters, don't use ($(...)) to populate the array—it'll split on spaces. Instead, read lines one by one with a while loop to preserve each username as a single array element:

admins=()
while IFS= read -r line; do
    admins+=("$line")
done < <(awk '/Authorized Admins:/ {flag=1; next} /:$/ {flag=0} flag' users.txt)

内容的提问来源于stack exchange，提问作者Gavin Morton