Semi-Structured Data

Modules
	Locking
	Tree Creation
	Subtree Operations
	Path Syntax
	Value Access
	Value Modification
	Pointer Access (Lock Required)
	Array Operations
	Advanced Operations

Detailed Description

The Semi-Structured Data (SSD) module provides a flexible tree structure for storing hierarchical data similar to JSON. It supports hashtables (objects), arrays, and single values as nodes. The tree is fully thread-safe with reader-writer locks and supports complex operations like path-based traversal, cloning, and grafting subtrees.

Key Features:

• Three node types: hashtables (key-value), arrays (indexed), and single values
• Thread-safe operations with reader-writer locking
• Path-based access with automatic traversal (e.g., "bucket/paints[2]/yellow")
• Type-safe value storage using the stvar system
• Support for custom node types through factory functions
• Deep cloning and subtree grafting operations

Basic Usage:

// Create a hashtable tree
SSDNode *root = ssdCreateHashtable();
 
// Set values using paths
ssdSet(root, _SL("user/name"), true, stvar(string, _SL("Alice")));
ssdSet(root, _SL("user/age"), true, stvar(int32, 30));
 
// Get values
int32 age = ssdVal(int32, root, _SL("user/age"), 0);
 
// Use within locked transactions for thread safety
ssdLockedTransaction(root) {
    strref name = ssdStrRef(root, _SL("user/name"));
    // ... use name while lock is held
}
 
// Clean up
objRelease(&root);

Path Syntax: Paths use '/' separators and array indices in brackets:

• bucket/paints[2]/yellow - Navigate through hashtable keys and array indices
• [1]/mango - Access array at root level

Locking: Most operations automatically manage locks, but pointer-returning functions like ssdPtr() and ssdStrRef() require explicit lock management via ssdLockedTransaction() or manual locking with ssdLockRead()/ssdLockWrite().