JSON Output

Export query results as JSON objects in NDJSON (newline-delimited JSON) format.

Usage

loq -o:JSON "SELECT * FROM data.csv"
loq -o:NDJSON "SELECT * FROM data.csv"

Both -o:JSON and -o:NDJSON produce the same output.

Output Format

Each row is output as a JSON object on its own line (NDJSON format):

{"name":"Alice","age":32,"city":"New York"}
{"name":"Bob","age":28,"city":"San Francisco"}
{"name":"Carol","age":35,"city":"Chicago"}

This format is:

• Streaming-friendly: Process line by line
• Memory-efficient: No need to load entire array
• Compatible: Works with jq, grep, etc.

Output to File

# To file
loq -o:JSON --ofile:output.json "SELECT * FROM data.csv"

# To stdout (default)
loq -o:JSON "SELECT * FROM data.csv"

Type Handling

SQL Type	JSON Type
Integer	Number
Float	Number
Boolean	Boolean
String	String
NULL	null
DateTime	String (ISO 8601)

Examples

{"id":42,"price":19.99,"active":true,"name":"Widget","deleted":null}
{"timestamp":"2024-01-15T14:30:00"}

Examples

Basic Export

loq -o:JSON "SELECT * FROM users.csv"

Filtered Export

loq -o:JSON --ofile:errors.json \
          "SELECT timestamp, level, message FROM logs.csv WHERE level = 'error'"

Aggregated Export

loq -o:JSON "SELECT category, COUNT(*) AS count, SUM(amount) AS total
                   FROM sales.csv
                   GROUP BY category"

Output:

{"category":"Electronics","count":150,"total":45000.50}
{"category":"Clothing","count":89,"total":12500.00}
{"category":"Books","count":210,"total":8400.75}

With Column Aliases

loq -o:JSON "SELECT
                       name AS customerName,
                       email AS contactEmail,
                       total AS orderTotal
                   FROM orders.csv"

Processing with jq

jq is a powerful JSON processor.

Filter

loq -o:JSON "SELECT * FROM users.csv" | jq 'select(.age > 30)'

Extract Field

loq -o:JSON "SELECT * FROM users.csv" | jq -r '.name'

Transform

loq -o:JSON "SELECT * FROM users.csv" | jq '{user: .name, years: .age}'

Collect into Array

loq -o:JSON "SELECT * FROM users.csv" | jq -s '.'

Output:

[
  {"name":"Alice","age":32},
  {"name":"Bob","age":28}
]

Statistics

loq -o:JSON "SELECT * FROM sales.csv" | jq -s 'map(.amount) | add'

API Integration

POST to API

loq -o:JSON "SELECT * FROM data.csv" | \
  curl -X POST -H "Content-Type: application/json" \
       --data-binary @- https://api.example.com/import

Save and Upload

loq -o:JSON --ofile:data.json "SELECT * FROM data.csv"
curl -X POST -H "Content-Type: application/json" \
     -d @data.json https://api.example.com/import

Streaming Processing

Process large files line by line:

Python

import json
import subprocess

proc = subprocess.Popen(
    ['loq', '-o:JSON', 'SELECT * FROM huge.csv'],
    stdout=subprocess.PIPE,
    text=True
)

for line in proc.stdout:
    record = json.loads(line)
    # Process record
    print(record['name'])

Node.js

const { spawn } = require('child_process');
const readline = require('readline');

const proc = spawn('loq', ['-o:JSON', 'SELECT * FROM huge.csv']);
const rl = readline.createInterface({ input: proc.stdout });

rl.on('line', (line) => {
    const record = JSON.parse(line);
    console.log(record.name);
});

Bash

loq -o:JSON "SELECT * FROM data.csv" | while read line; do
    name=$(echo "$line" | jq -r '.name')
    echo "Processing: $name"
done

Converting to JSON Array

If you need a JSON array instead of NDJSON:

Using jq

loq -o:JSON "SELECT * FROM users.csv" | jq -s '.'

Using Shell

echo '['
loq -o:JSON "SELECT * FROM users.csv" | sed 's/$/,/' | sed '$ s/,$//'
echo ']'

Nested JSON

loq outputs flat JSON objects. For nested structures, use jq:

# Create nested structure
loq -o:JSON "SELECT id, name, email, city, country FROM users.csv" | \
  jq '{id, name, contact: {email}, location: {city, country}}'

Best Practices

Use Meaningful Aliases

SELECT
    user_id AS userId,
    first_name AS firstName,
    last_name AS lastName
FROM users.csv

Handle NULLs

NULL values are output as JSON null:

{"name":"Alice","email":null}

Check for nulls in processing:

loq -o:JSON "SELECT * FROM users.csv" | jq 'select(.email != null)'

Large Files

NDJSON is efficient for large files:

# Process 1M rows without loading all into memory
loq -o:JSON "SELECT * FROM huge.csv" | process_stream

Troubleshooting

Invalid JSON Characters

Special characters are automatically escaped:

{"text":"Line 1\nLine 2","quote":"He said \"hello\""}

Unicode

JSON output uses UTF-8 encoding:

{"name":"日本語","emoji":"😀"}

Numeric Precision

Large integers maintain precision:

{"id":9007199254740993}

See Also

• Output Formats Overview
• CSV Output
• JSON Input