JSON Manipulation Patterns

Provides comprehensive guidance on JSON file manipulation patterns, simple path query syntax, transformation strategies, validation approaches, and safe editing workflows.

Simple Path Query Syntax

Basic Path Expressions

Property access:

• name - Direct property
• user.name - Nested property (dot notation)
• config.server.host - Deep nesting

Array access:

• items[0] - First element
• items[1] - Second element
• users[5].name - Property of array element

Wildcards:

• users[*] - All array elements
• users[*].email - Property from all array elements
• *.id - All id properties at current level
• data.*.value - All value properties in data object

Complex Paths

Mixed notation:

users[0].orders[*].items[0].price

Price of first item in all orders of first user.

Wildcard mapping:

products[*].tags[*]

All tags from all products (flattened).

Nested wildcards:

categories[*].items[*].name

Names of all items in all categories.

What's NOT Supported

For complex filtering logic, use TypeScript transform scripts instead:

Filtering by condition (NOT supported in paths):

// Use transform script instead of path queries
export default (data: any) => {
  return data.users.filter(u => u.age > 18);
};

Recursive descent (NOT supported):

// NOT: $..email (find all emails anywhere)
// Instead: Use specific paths or transform scripts

Array slicing (NOT supported):

// NOT: items[0:5] (first 5 items)
// Instead: Use transform script with slice()

Transformation Patterns

Data Filtering

Remove inactive items:

export default (data: any) => {
  data.items = data.items.filter((item: any) => item.active);
  return data;
};

Filter by date:

export default (data: any) => {
  const cutoff = new Date('2024-01-01');
  data.records = data.records.filter((r: any) =>
    new Date(r.date) >= cutoff
  );
  return data;
};

Data Mapping

Transform structure:

export default (data: any) => ({
  version: "2.0",
  items: data.items.map((item: any) => ({
    id: item.itemId,
    name: item.itemName,
    price: item.cost * 1.1, // Add 10% markup
  }))
});

Normalize data:

export default (data: any) => {
  data.users = data.users.map((user: any) => ({
    ...user,
    email: user.email.toLowerCase(),
    name: user.name.trim(),
  }));
  return data;
};

Data Aggregation

Add computed fields:

export default (data: any) => {
  data.summary = {
    totalItems: data.items.length,
    totalValue: data.items.reduce((sum: number, item: any) =>
      sum + item.price, 0
    ),
    averagePrice: data.items.reduce((sum: number, item: any) =>
      sum + item.price, 0) / data.items.length,
  };
  return data;
};

Group by category:

export default (data: any) => {
  const grouped = data.items.reduce((acc: any, item: any) => {
    const category = item.category || 'uncategorized';
    if (!acc[category]) acc[category] = [];
    acc[category].push(item);
    return acc;
  }, {});
  return { categories: grouped };
};

Data Migration

Version upgrade:

export default (data: any) => {
  // Migrate from v1 to v2 format
  if (data.version === "1.0") {
    return {
      version: "2.0",
      config: {
        ...data.settings,
        newField: "default",
      },
      data: data.items,
    };
  }
  return data;
};

Rename properties:

export default (data: any) => {
  return {
    ...data,
    items: data.items.map((item: any) => ({
      id: item.itemId,      // Rename itemId -> id
      title: item.name,     // Rename name -> title
      cost: item.price,     // Rename price -> cost
    }))
  };
};

Schema Validation Strategies

Design JSON Schemas

Basic structure:

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "type": "object",
  "properties": {
    "name": { "type": "string" },
    "age": { "type": "number", "minimum": 0 }
  },
  "required": ["name"]
}

Array validation:

{
  "type": "object",
  "properties": {
    "items": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "id": { "type": "string" },
          "value": { "type": "number" }
        },
        "required": ["id", "value"]
      },
      "minItems": 1
    }
  }
}

Enum and patterns:

{
  "type": "object",
  "properties": {
    "status": {
      "type": "string",
      "enum": ["active", "inactive", "pending"]
    },
    "email": {
      "type": "string",
      "format": "email"
    },
    "phone": {
      "type": "string",
      "pattern": "^\\d{3}-\\d{3}-\\d{4}$"
    }
  }
}

Validation Workflow

1. Define schema for your data structure
2. Validate before edits to catch issues early
3. Validate after edits to ensure correctness
4. Store schemas in version control alongside data

Common Validation Scenarios

Configuration validation:

• Ensure required fields present
• Validate value ranges
• Check format of URLs, emails, etc.

API response validation:

• Verify response structure
• Check data types
• Ensure required fields exist

Data migration validation:

• Validate source data before transform
• Validate transformed data against new schema
• Ensure no data loss

Safe Editing Workflows

Pre-Edit Checklist

1. Backup verification:

   • Auto-backup is enabled by default
   • Backups stored as .bak files
   • Check backup location before destructive operations

2. Dry-run testing:

   • Always use --dry-run true for new operations
   • Review changes before applying
   • Verify path matches expected locations

3. Validation:

   • Validate JSON before editing
   • Validate after editing to ensure correctness
   • Use schemas to enforce constraints

Destructive Operations

Delete operations:

1. Query first: See what will be deleted
2. Dry-run: Preview deletion
3. Confirm: Verify backup created
4. Execute: Apply deletion

Batch operations:

1. Test on single file first
2. Dry-run on all files
3. Review each file's changes
4. Apply to all files
5. Validate all results

Recovery from Mistakes

Restore from backup:

# Backups are `.bak` files in same directory
cp data.json.bak data.json

Revert specific change:

1. Query backup file for original value
2. Set original value in current file
3. Validate result

Best Practices

For simple edits:

• Use set operation with explicit paths
• Avoid wildcards unless necessary
• Validate path with query first

For complex transformations:

• Write TypeScript transform script
• Test on copy of data first
• Use dry-run to preview result
• Keep transform scripts in version control

For batch operations:

• Test on single file first
• Use explicit file lists (avoid * until tested)
• Check each result
• Keep backups until verified

Merge Strategies

Deep Merge

• Use when: Merging configuration files with nested objects
• Behavior: Recursively merges objects, concatenates arrays
• Example: Base config + environment-specific overrides

Shallow Merge

• Use when: Top-level property replacement needed
• Behavior: Only merges first level, later values win
• Example: Simple key-value configurations

Concat

• Use when: Combining array-based files
• Behavior: Flattens all arrays into single array
• Example: Merging multiple data files

Merge Order Matters

Files are merged left-to-right:

file1.json + file2.json + file3.json

• file3 values override file2
• file2 values override file1

Place most general config first, most specific last.

Performance Considerations

Large Files

Query optimization:

• Use specific paths: users[0].name instead of *[*].name
• Access nested data directly: config.server.host instead of wildcards
• For complex filtering, use transform scripts instead of repeated queries

Transform optimization:

• Process in chunks for very large files
• Use streaming for files over 100MB
• Consider splitting into multiple files

Batch operations:

• Process files in parallel when possible
• Use specific file lists, not wildcards
• Monitor memory usage for many files

Common Patterns

Configuration Management

Base + Override pattern:

1. Base config (defaults)
2. Environment config (dev/staging/prod)
3. Local overrides (optional)

Merge order: base → env → local

Validation workflow:

1. Validate base config
2. Merge with overrides
3. Validate merged result
4. Deploy

Data Pipeline

Extract-Transform-Load:

1. Query: Extract relevant data
2. Transform: Apply TypeScript transformation
3. Validate: Check against schema
4. Load: Write to destination

Version Migration

Safe migration:

1. Query current version
2. Backup original
3. Transform to new version
4. Validate against new schema
5. Test with application
6. Deploy

Troubleshooting

Query Returns Nothing

1. Test incrementally: Build path step by step
2. Verify data structure: Query with empty path or root property first
3. Check syntax: Ensure brackets and dots are correct
4. Try wildcards: Use * to see what's available

Edit Doesn't Work

1. Query first: Verify path matches
2. Check data type: Ensure value type matches
3. Review backup: Compare before/after
4. Validate JSON: Ensure file is valid

Transform Fails

1. Check script syntax: Ensure TypeScript is valid
2. Test function: Ensure function exported
3. Handle edge cases: Check for null/undefined
4. Review error message: TypeScript errors are specific

Examples by Use Case

API Response Processing

Extract user emails:

data.users[*].email

Filter active accounts:
Use a transform script for filtering:

export default (data: any) => {
  return data.data.users.filter(u => u.status === "active" && u.verified);
};

Configuration Updates

Update API endpoint:

Operation: set
Path: config.apiEndpoint
Value: "https://api.example.com"

Enable feature flag:
Use a transform script for conditional updates:

export default (data: any) => {
  const feature = data.features.find(f => f.name === "newFeature");
  if (feature) feature.enabled = true;
  return data;
};

Data Migration

Transform user format:

export default (data: any) => ({
  version: "2.0",
  users: data.users.map((u: any) => ({
    userId: u.id,
    profile: {
      name: u.name,
      email: u.email,
    },
    metadata: {
      createdAt: u.created,
      active: u.status === "active",
    }
  }))
});

---

Reference these patterns when helping users with JSON manipulation tasks. Always suggest the safest approach (query → dry-run → validate → execute) for destructive operations.