Skip to main content

​
Overview

Social Analyzer supports two primary output formats: Pretty (human-readable) and JSON (machine-readable). The format is controlled by the --output or output parameter.

​
Output Format Selection

​
CLI

# Pretty output (default)
social-analyzer --username "johndoe"
social-analyzer --username "johndoe" --output pretty

# JSON output
social-analyzer --username "johndoe" --output json

​
Python API

from social_analyzer import SocialAnalyzer

sa = SocialAnalyzer()

# Pretty output
result = sa.run_as_object(username="johndoe", output="pretty")

# JSON output
result = sa.run_as_object(username="johndoe", output="json")

​
Pretty Output Format

The pretty format provides human-readable, color-coded console output with profile counts and formatted results.

​
Example Output

[init] Detections are updated very often, make sure to get the most up-to-date ones
[init] languages.json & sites.json loaded successfully
[Init] Selected websites: 900
[Checking] youtube
[Checking] twitter
[Checking] reddit
...
[Detected] 15 Profile[s]
[unknown] 3 Profile[s]
[failed] 2 Profile[s]

[Custom output with colored formatting]

​
Features

  • Color-coded status messages
  • Progress indicators
  • Profile counts by category
  • Formatted profile details
  • Real-time status updates

​
JSON Output Format

The JSON format provides structured data suitable for programmatic processing, integration, and storage.

​
Structure Overview

​
detected
array
Array of successfully detected profiles with high confidence
​
unknown
array
Array of profiles that exist but couldn’t be confidently verified
​
failed
array
Array of profiles that failed to load or returned errors

​
Complete JSON Structure

{
  "detected": [
    {
      "link": "https://www.youtube.com/johndoe",
      "rate": "100%",
      "title": "John Doe - YouTube",
      "status": "good",
      "text": "Profile text content...",
      "language": "English",
      "type": "Video",
      "global_rank": 2,
      "rate_percent": 100,
      "extracted": {
        "profiles": ["@johndoe", "johndoe123"],
        "urls": ["https://example.com"],
        "patterns": ["pattern1", "pattern2"]
      },
      "metadata": {
        "names": ["John Doe"],
        "images": ["https://example.com/photo.jpg"],
        "emails": ["[email protected]"]
      }
    }
  ],
  "unknown": [
    {
      "link": "https://www.example.com/johndoe"
    }
  ],
  "failed": [
    {
      "link": "https://www.brokensite.com/johndoe"
    }
  ]
}

​
Profile Object Fields

​
Detected Profile Fields

Detected profiles contain the most comprehensive information:
​
link
string
required
Full URL to the detected profile
"https://www.youtube.com/johndoe"
​
rate
string
required
Confidence rate as percentage string
"100%", "85.5%", "50%"
​
rate_percent
number
Numeric confidence percentage (0-100)
100, 85.5, 50
​
status
string
required
Detection quality indicator
  • good - High confidence match (greater than 80%)
  • maybe - Medium confidence (40-80%)
  • bad - Low confidence (less than 40%)
​
title
string
Page title from the profile
"John Doe (@johndoe) - YouTube"
​
text
string
Extracted text content from the profile page
"Welcome to my channel! Subscribe for daily content..."
​
language
string
Detected page language
"English", "Spanish (Maybe)", "unavailable"
​
type
string
Website category
"Social", "Video", "Music", "Gaming", "Adult"
​
global_rank
number
Alexa/popularity ranking (0 if unranked)
2, 150, 5000, 0
​
extracted
object
Extracted data when --extract flag is used
​
metadata
object
OSINT metadata when --metadata flag is used

​
Unknown Profile Fields

Unknown profiles contain minimal information:
​
link
string
required
Full URL to the profile (existence unverified)

​
Failed Profile Fields

Failed profiles only contain the attempted URL:
​
link
string
required
Full URL that failed to load or returned errors

​
Filtering Output

​
By Status

Control which profile categories are included: 
# Only detected profiles
social-analyzer --username "johndoe" --profiles detected

# Detected and failed
social-analyzer --username "johndoe" --profiles detected,failed

# All categories
social-analyzer --username "johndoe" --profiles all

​
By Confidence

Filter detected profiles by confidence level: 
# Only high confidence
social-analyzer --username "johndoe" --filter good

# Good and maybe
social-analyzer --username "johndoe" --filter good,maybe

# All confidence levels
social-analyzer --username "johndoe" --filter all

​
By Fields

Show only specific fields: 
# Only links and rates
social-analyzer --username "johndoe" --options "link,rate"

# Links, rates, and titles
social-analyzer --username "johndoe" --options "link,rate,title"

# Everything (default)
social-analyzer --username "johndoe" --options ""

​
Simplified Output

The Python CLI supports a simplified mode that outputs only links of 100% confidence profiles: 
python -m social-analyzer --username "johndoe" --simplify

​
Simplified Structure

{
  "detected": [
    {
      "link": "https://www.youtube.com/johndoe"
    },
    {
      "link": "https://twitter.com/johndoe"
    }
  ]
}
Simplified mode automatically:
  • Sets filter to “good”
  • Removes “unknown” and “failed” categories
  • Keeps only 100% confidence matches
  • Shows only the “link” field

​
Web API Response Format

The web API (POST /analyze_string) returns a comprehensive object: 
{
  "username": "johndoe",
  "uuid": "task-uuid-here",
  "info": {
    "items": [],
    "original": "johndoe",
    "corrected": "johndoe",
    "total": 15,
    "checking": "Using johndoe with no lookups"
  },
  "user_info_normal": {
    "data": [
      {
        "link": "https://www.youtube.com/johndoe",
        "rate": "100%",
        "status": "good",
        "title": "John Doe - YouTube"
      }
    ],
    "type": "all"
  },
  "user_info_advanced": {
    "data": [],
    "type": "all"
  },
  "user_info_special": {
    "data": [],
    "type": "all"
  },
  "ages": [],
  "names_origins": [],
  "table": {
    "prefix": [],
    "name": [],
    "number": [],
    "symbol": [],
    "unknown": [],
    "maybe": []
  },
  "common": [],
  "words_info": [],
  "custom_search": [],
  "graph": {
    "graph": {
      "nodes": [],
      "links": []
    }
  },
  "stats": {
    "categories": {},
    "countries": {}
  },
  "logs": "[log content here]"
}

​
Sorting and Organization

Detected profiles are automatically sorted by confidence rate in descending order: 
{
  "detected": [
    {"link": "...", "rate": "100%"},
    {"link": "...", "rate": "95.5%"},
    {"link": "...", "rate": "85%"},
    {"link": "...", "rate": "70%"}
  ]
}

​
Integration Examples

​
Parse JSON in Bash

social-analyzer --username "johndoe" --output json | jq '.detected[].link'

​
Parse JSON in Python

import json
import subprocess

result = subprocess.run(
    ['social-analyzer', '--username', 'johndoe', '--output', 'json', '--silent'],
    capture_output=True,
    text=True
)

data = json.loads(result.stdout)
for profile in data.get('detected', []):
    print(f"{profile['link']} - {profile['rate']}")

​
Parse JSON in Node.js

const { execSync } = require('child_process');

const output = execSync(
  'social-analyzer --username johndoe --output json'
).toString();

const data = JSON.parse(output);
data.detected.forEach(profile => {
  console.log(`${profile.link} - ${profile.rate}`);
});

​
Best Practices

For automation, combine --output json with --silent to get clean JSON without log messages:
social-analyzer --username "johndoe" --output json --silent > results.json
The exact fields returned depend on the flags used (--extract, --metadata, --options, etc.)
When using --trim, long strings will be truncated. Avoid this flag if you need full content.

Build docs developers (and LLMs) love

Get started for freeTalk to us