Scope Logging¶

How RAII Scope Logging Works¶

siddiqsoft::ScopeTrace uses the C++ RAII pattern to measure scope execution duration and emit trace logs upon scope entry and exit.

#include <siddiqsoft/ScopeTrace.hpp>

void compute()
{
    // Defaults scope name to plain function name ("compute")
    siddiqsoft::ScopeTrace scope;

    // Explicit nested scope ("compute-stage1")
    auto sub = scope.nest("stage1");

    sub.info("Processing stage 1 payload...");
} // Destructors automatically log completion status and elapsed duration at LogLevel::debug severity upon scope exit

Source Location & Function Name Extraction¶

Using std::source_location::current(), ScopeTrace captures caller details automatically upon construction: - Source file path (m_location.file_name()) - Line number (m_location.line()) - Enclosing function signature (m_location.function_name())

ScopeTrace also includes internal protected helper routines (extract_func_name() / function_name()) to extract clean plain function names matching __func__ from full signature strings. When ScopeTrace is declared in the global scope (outside of any enclosing function), extract_func_name() evaluates to "GLOBAL".

Nesting Depth & ISO 8601 Timestamps¶

ScopeTrace tracks nested scope depth strictly by parentage (child.depth() = parent.depth() + 1), producing formatted visual indentation for hierarchical log trees. Each output line is prefixed with an ISO 8601 UTC timestamp:

Console Output (Hierarchical Indentation)

  Creating NEW SCOPE compute:6
2026-08-15T20:49:27.060564Z  [info  ] compute - Outer computation started
  Creating NEW SCOPE compute/stage1:4
2026-08-15T20:49:27.060600Z  [info  ]   compute/stage1 - Processing stage 1 payload...
2026-08-15T20:49:27.061250Z  [debug ]   compute/stage1 - COMPLETED: time:650us
2026-08-15T20:49:27.061300Z  [debug ] compute - COMPLETED: time:1200us

Dynamic Log Level Filtering¶

Each ScopeTrace instance maintains a configurable log level threshold (LogLevel / trace_level, defaulted to LogLevel::critical upon construction). You can set the threshold at scope creation or dynamically update it via set_level():

// Set logging threshold to debug at construction
siddiqsoft::ScopeTrace scope("compute", siddiqsoft::LogLevel::debug);

// Dynamically change logging threshold
scope.set_level(siddiqsoft::LogLevel::trace);

Filtering Rules¶

Always Logged: Messages logged with critical, exception, or error level are always output regardless of m_log_level.
LogLevel::trace: Enables all diagnostic logs (warning, info, debug, trace). Recommended for high-volume I/O operations.
LogLevel::debug: Enables warning, info, and debug logs (excludes trace).
LogLevel::info: Enables warning and info logs (excludes debug and trace).
LogLevel::warning: Enables warning logs only.

Log Level Colors & Output Styling¶

ScopeTrace applies distinct ANSI color escape codes to each log level tag to make terminal output visually scannable:

Log Level / Method	Tag Label	Color Output	ANSI Escape Code	Visual Terminal Preview
`trace_level::critical`	`[crit ]`	Red	`\033[0;31m`	[crit ] System memory exhaustion
`trace_level::exception`	`[except]`	Red / Bold Red	`\033[0;31m`	[except] std::runtime_error - Timeout
`trace_level::error`	`[error ]`	Orange	`\033[38;5;208m`	[error ] Failed to connect to db host
`trace_level::warning`	`[warn ]`	Dark Yellow / Gold	`\033[38;5;136m`	[warn ] Cache capacity reached 92%
`trace_level::info`	`[info ]`	Default / Neutral	`\033[0m`	[info ] Processing batch item 42
`trace_level::debug`	`[debug ]`	Light Gray	`\033[38;5;250m`	[debug ] Worker thread pool depth: 4
`trace_level::trace`	`[trace ]`	Dark Blue	`\033[38;5;19m`	[trace ] RX buffer dump: 0x41 0x42 0x43
Scope Exit	`COMPLETED`	Green Time	`\033[0;32m`	COMPLETED: time:450us

[!TIP] Best Practice: High-Frequency I/O Operations Should Use trace (scope.trace(...))

High-volume diagnostic logging — such as socket reads/writes, raw payload packet dumps, file stream buffer transfers, HTTP payload tracing, or inner loop iterations — should always use trace_level::trace (scope.trace(...)).

When m_log_level is set to LogLevel::debug or info, all scope.trace(...) statements are bypassed via a zero-allocation threshold check (level <= m_log_level), eliminating string formatting and stream output overhead during standard debugging while keeping log buffers clean.

In-Scope Logging Sinks¶

Within an active scope, you can output formatted contextual messages to std::cerr with depth indentation and ANSI color coding:

scope.trace("..."): Trace diagnostic details (light gray, active when m_log_level >= trace). Ideal for I/O packet/buffer dumps.
scope.debug("..."): Debug diagnostic details (light gray, active when m_log_level >= debug).
scope.info("..."): Informational messages (active when m_log_level >= info).
scope.warn("..."): Warning messages (colored dark yellow, active when m_log_level >= warning).
scope.err("..."): Error messages (colored orange, always logged).
scope.err_throw<EX>("..."): Unified error logging & exception throwing shortcut (colored orange, throws EX(formatted_msg), always logged).
scope.exp(e): Exception handler shortcut. Logs exception type (typeid(e).name()) in bold red and e.what() (always logged).
scope.exp(e, "..."): Exception handler shortcut with context. Logs exception type, italicized e.what(), and custom formatted contextual details (always logged).

Unified Error Throwing & Catching Patterns¶

ScopeTrace simplifies error flow by combining console logging with exception throwing (err_throw()) and exception handling (exp()):

1. Throw Site (`err_throw`)¶

Instead of separate log and throw statements, err_throw<EX>(fmt, args...) logs the formatted error message with timestamps, depth indentation, and exception type details, then immediately throws EX:

void validate_input(int value)
{
    siddiqsoft::ScopeTrace scope;

    if (value < 0) {
        // Logs orange error message and throws std::invalid_argument
        scope.err_throw<std::invalid_argument>("Value must be non-negative, got: {}", value);
    }
}

2. Catch Site (`exp`)¶

Inside catch blocks, exp() logs caught exception details with zero boilerplate:

try {
    validate_input(-5);
}
catch (const std::exception& e) {
    // Basic exception log: outputs typeid name and e.what()
    scope.exp(e);

    // Exception log with custom formatted contextual details:
    scope.exp(e, "Validation failed during request ID: {}", request_id);
}