如何在AWK中生成JSON对象？自定义实现需求问询

Got it, let's tackle this AWK JSON generation task step by step. I'll cover two scenarios: directly processing your input file, and generating JSON from the pre-built multi-dimensional array you already have.

Solution 1: Process the Input File Directly #

This script will skip irrelevant lines, extract shareholder data, map it to your desired JSON schema, handle escaped quotes, and output the JSON objects as requested.

BEGIN {
    # Define dynamic JSON schema mapping: input field → JSON key
    schema["imie"] = "Name"
    schema["nazwisko"] = "Surname"
    schema["pesel"] = "ID"
    schema["funkcja"] = "Role"
    
    in_shareholders = 0  # Flag to track if we're in the shareholder data block
    entry_count = 0       # Count of key-value pairs collected for the current entry
}

# Trigger entry into shareholder data mode
/dane wspólników/ {
    in_shareholders = 1
    next
}

# Exit shareholder data mode when hitting unrelated sections
/czwarta linia|reprezentanci/ {
    in_shareholders = 0
    delete current_entry  # Clear incomplete data if any
    entry_count = 0
    next
}

# Process lines while in shareholder data mode
in_shareholders {
    key = $1
    # Extract full value (handles cases like Ja"nD with quotes/embedded spaces)
    value = substr($0, index($0, $2))
    
    # Escape double quotes in values to valid JSON
    gsub(/"/, "\\\"", value)
    
    # Map input key to JSON key if it exists in our schema
    if (schema[key] != "") {
        current_entry[schema[key]] = value
        entry_count++
    }
    
    # When we have a full set of 4 fields, output the JSON object
    if (entry_count == 4) {
        printf "{ "
        first_field = 1
        for (json_key in current_entry) {
            if (!first_field) printf ", "
            printf "\"%s\": \"%s\"", json_key, current_entry[json_key]
            first_field = 0
        }
        printf " } "
        
        # Reset for next entry
        delete current_entry
        entry_count = 0
    }
}

How to use:
Run it with your input file like this:

awk -f generate_json.awk your_input_file.txt

Solution 2: Generate JSON from Your Pre-built Multi-dimensional Array #

If you already have the cleaned multi-dimensional array, this simpler script will map it to your schema and output the required JSON:

BEGIN {
    # Define schema order (matches your array's field positions)
    json_keys[1] = "Name"
    json_keys[2] = "Surname"
    json_keys[3] = "ID"
    json_keys[4] = "Role"
    
    # Populate with your existing array data (adjust indices as needed)
    data[1][1] = "JanA"
    data[1][2] = "NowakA"
    data[1][3] = "1234A"
    data[1][4] = "PrezesA"
    
    data[2][1] = "JanD"
    data[2][2] = "NowakD"
    data[2][3] = "12341D"
    data[2][4] = "PrezesD"
    
    data[3][1] = "JanC"
    data[3][2] = "NowakC"
    data[3][3] = "12342C"
    data[3][4] = "PrezesC"
    
    # Loop through the array to generate JSON
    for (i = 1; i <= length(data); i++) {
        printf "{ "
        first_field = 1
        for (j = 1; j <= 4; j++) {
            if (!first_field) printf ", "
            # Optional: Remove quotes for numeric IDs if needed
            # if (data[i][j] ~ /^[0-9]+$/) {
            #     printf "\"%s\": %s", json_keys[j], data[i][j]
            # } else {
                printf "\"%s\": \"%s\"", json_keys[j], data[i][j]
            # }
            first_field = 0
        }
        printf " } "
    }
    printf "\n"
}

Optional tweak:
If some IDs are pure numbers and you want them unquoted in JSON, uncomment the numeric check section in the script.

内容的提问来源于stack exchange，提问作者djsowa