Lib.rs

tklog is a high-performance structured logging library for Rust [中文]

tklog featuring ease-of-use, efficiency, and a rich feature suite. It supports functionalities such as console logging, file logging, both synchronous and asynchronous logging modes, alongside advanced capabilities like log slicing by time or size and compressed backup of log files.

Features

• Function support includes console logging, file logging, synchronous logging, asynchronous logging.
• Log level settings mirror those of the standard library: trace, debug, info, warn, error, fatal.
• Formatted output with customizable formats that can include log level flags, formatted timestamps, and log file locations.
• Log file slicing by time intervals: hourly, daily, or monthly.
• Log file slicing by specified file size.
• File rolling mechanism that automatically deletes older log files once a maximum backup count is reached to prevent excess logs from accumulating.
• Compression of archived backup log files.
• Supports the official log library standard API

official website

Simple Usage Description

The simplest way to use tklog involves direct macro calls:

use tklog::{trace, debug, error, fatal, info, warn};
fn testlog() {
    trace!("trace>>>>", "aaaaaaaaa", 1, 2, 3, 4);
    debug!("debug>>>>", "bbbbbbbbb", 1, 2, 3, 5);
    info!("info>>>>", "ccccccccc", 1, 2, 3, 5);
    warn!("warn>>>>", "dddddddddd", 1, 2, 3, 6);
    error!("error>>>>", "eeeeeeee", 1, 2, 3, 7);
    fatal!("fatal>>>>", "ffffffff", 1, 2, 3, 8);
}

By default, it will print console log, not files. Execution Result:

[TRACE] 2024-05-26 11:47:22 testlog.rs 27:trace>>>>,aaaaaaaaa,1,2,3,4
[DEBUG] 2024-05-26 11:47:22 testlog.rs 28:debug>>>>,bbbbbbbbb,1,2,3,5
[INFO] 2024-05-26 11:47:22 testlog.rs 29:info>>>>,ccccccccc,1,2,3,5
[WARN] 2024-05-26 11:47:22 testlog.rs 30:warn>>>>,dddddddddd,1,2,3,6
[ERROR] 2024-05-26 11:47:22 testlog.rs 31:error>>>>,eeeeeeee,1,2,3,7
[FATAL] 2024-05-26 11:47:22 testlog.rs 32:fatal>>>>,ffffffff,1,2,3,8

For initialization and customization, tklog furnishes methods to configure options such as console output, log levels, formatting styles, cutting strategies, and custom formatters.

use tklog::{
    sync::Logger,LEVEL, LOG,
    Format, MODE,
};

fn log_init() {
    LOG.set_console(true)       // Enables console logging
        .set_level(LEVEL::Info)  // Sets the log level; default is Debug
        .set_format(Format::LevelFlag | Format::Time | Format::ShortFileName)  // Defines structured log output with chosen details
        .set_cutmode_by_size("tklogsize.txt", 1<<20, 10, true)  // Cuts logs by file size (1 MB), keeps 10 backups, compresses backups
        .set_formatter("{level}{time} {file}:{message}\n");   // Customizes log output format; default is "{level}{time} {file}:{message}"
}

This illustrates global, singleton-style logging setup. Additionally, tklog facilitates custom multi-instance logging configurations, useful in systems requiring distinct logging structures across different components.

Multi-Instance Logging

tklog also accommodates multiple instances for scenarios that require distinct logging configurations. Each instance can possess its unique settings for console output, log level, file rotation, and a custom formatter.

use tklog::{
    debugs, errors, fatals, infos,
    sync::Logger,LEVEL, LOG,
    traces, warns, Format, MODE,
};
fn testmutlilog() {
    let mut log = Logger::new();
    log.set_console(true)
        .set_level(LEVEL::Debug) //Set the log level to Debug
        .set_cutmode_by_time("tklogs.log", MODE::DAY, 10, true)   //Split log files daily, keep up to 10 backups, and compress them
        .set_formatter("{message} | {time} {file}{level}\n");  //Customize the log structure's output format and additional content
    let mut logger = Arc::clone(&Arc::new(Mutex::new(log)));
    let log = logger.borrow_mut();
    traces!(log, "traces>>>>", "AAAAAAAAA", 1, 2, 3, 4);
    debugs!(log, "debugs>>>>", "BBBBBBBBB", 1, 2, 3, 5);
    infos!(log, "infos>>>>", "CCCCCCCCC", 1, 2, 3, 5);
    warns!(log, "warns>>>>", "DDDDDDDDDD", 1, 2, 3, 6);
    errors!(log, "errors>>>>", "EEEEEEEE", 1, 2, 3, 7);
    fatals!(log, "fatals>>>>", "FFFFFFFF", 1, 2, 3, 8);
    thread::sleep(Duration::from_secs(1))
}

Execution Result

debugs>>>>,BBBBBBBBB,1,2,3,5 | 2024-05-26 14:13:25 testlog.rs 70[DEBUG]
infos>>>>,CCCCCCCCC,1,2,3,5 | 2024-05-26 14:13:25 testlog.rs 71[INFO]
warns>>>>,DDDDDDDDDD,1,2,3,6 | 2024-05-26 14:13:25 testlog.rs 72[WARN]
errors>>>>,EEEEEEEE,1,2,3,7 | 2024-05-26 14:13:25 testlog.rs 73[ERROR]
fatals>>>>,FFFFFFFF,1,2,3,8 | 2024-05-26 14:13:25 testlog.rs 74[FATAL]

Note: The structured log output above conforms to the format specified by "{message} | {time} {file}{level}\n". The formatter includes identifiers like {message}, {time}, {file}, {level}, and any additional text or separators outside these placeholders.

Detailed Usage Guide

1. Log Levels: Trace < Debug < Info < Warn < Error < Fatal.

Example:

   LOG.set_level(LEVEL::Info) //Sets the log level to Info

2. Console Logging: Enable or disable via .set_console(bool).

   LOG.set_console(false) //Disables console logging (default is true)

3. Log Formats:

Format::Nano ： No formatting
Format::Date  ： Outputs date (e.g., 2024-05-26)
Format::Time  ： Outputs time to seconds (e.g., 14:13:25)
Format::Microseconds ：Outputs time with microseconds (e.g., 18:09:17.462245)
Format::LongFileName ：Full file path with line number (e.g., tests/testlog.rs 25)
Format::ShortFileName ： Abbreviated file path with line number (e.g., testlog.rs 25)
Format::LevelFlag ： Log level marker (e.g., [Debug]).

For custom formats:

LOG.set_format(Format::LevelFlag | Format::Time | Format::ShortFileName)

4. Custom Format Strings:

Default is "{level}{time} {file}:{message}\n".

• {level}: Log level indicator, e.g., [Debug].

• {time}: Logged timestamp.

• {file}: Filename and line number.

• {message}: Log content.

Example:

   LOG.set_formatter("{message} | {time} {file}{level}\n")

Reminder: Text outside the {level}, {time}, {file}, and {message} tags is output verbatim, including delimiters, spaces, and newlines.

5. Time-Based Log File Rotation:

Modes: MODE::HOUR, MODE::DAY, MODE::MONTH.

Use .set_cutmode_by_time() with:

• File path
• Time mode
• Maximum backup count
• Compression option

Example:

   let mut log = Logger::new(); 
   log.set_cutmode_by_time("/usr/local/tklogs.log", MODE::DAY, 0, false);

This configures the log to be stored at /usr/local/tklogs.log, rotated daily, with no limit on backups, and without compressing daily logs.

Backup Naming Conventions:

• Daily:
  • tklogs_20240521.log
  • tklogs_20240522.log
• Hourly:
  • tklogs_2024052110.log
  • tklogs_2024052211.log
• Monthly:
  • tklogs_202403.log
  • tklogs_202404.log

6. Size-Based Log File Rotation:

Utilize .set_cutmode_by_size() with the following parameters:

• File path
• Roll size
• Max backups
• Compress backups

Example:

let mut log = Logger::new(); 
log.set_cutmode_by_size("tklogs.log", 100<<20, 10, true);

Here, tklogs.log denotes the path, with files rolling at 100 MB each, retaining 10 backups, and compressing them.

Backup File Naming Convention:

tklogs_1.log.gz
tklogs_2.log.gz
tklogs_3.log.gz

Log Printing Methods:

• Global Singleton:

  • trace!, debug!, info!, warn!, error!, fatal!

• Multiple Instances:

  • traces!, debugs!, infos!, warns!, errors!, fatals!

Asynchronous Logging

• Global Singleton Async:

  • async_trace!, async_debug!, async_info!, async_warn!, async_error!, async_fatal!

• Multiple Instances Async:

  • async_traces!, async_debugs!, async_infos!, async_warns!, async_errors!, async_fatals!

Example: Global Asynchronous Usage

use tklog::{
    async_debug, async_error, async_fatal, async_info, async_trace, async_warn, LEVEL, Format, ASYNC_LOG
};

async fn async_log_init() {
    // Configure global singleton
    ASYNC_LOG
        .set_console(false) // Disable console output
        .set_level(LEVEL::Trace) // Set log level to Trace
        .set_format(Format::LevelFlag | Format::Time | Format::ShortFileName) // Define structured logging output
        .set_cutmode_by_size("tklog_async.txt", 10000, 10, false) // Rotate log files by size, every 10,000 bytes, with 10 backups
        .await;
}

#[tokio::test]
async fn testlog() {
    async_log_init().await;
    async_trace!("trace>>>>", "aaaaaaa", 1, 2, 3);
    async_debug!("debug>>>>", "aaaaaaa", 1, 2, 3);
    async_info!("info>>>>", "bbbbbbbbb", 1, 2, 3);
    async_warn!("warn>>>>", "cccccccccc", 1, 2, 3);
    async_error!("error>>>>", "ddddddddddddd", 1, 2, 3);
    async_fatal!("fatal>>>>", "eeeeeeeeeeeeee", 1, 2, 3);
    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
}

Execution Result:

[TRACE] 20:03:32 testasynclog.rs 20:trace>>>>,aaaaaaa,1,2,3
[DEBUG] 20:03:32 testasynclog.rs 21:debug>>>>,aaaaaaa,1,2,3
[INFO] 20:03:32 testasynclog.rs 22:info>>>>,bbbbbbbbb,1,2,3
[WARN] 20:03:32 testasynclog.rs 23:warn>>>>,cccccccccc,1,2,3
[ERROR] 20:03:32 testasynclog.rs 24:error>>>>,ddddddddddddd,1,2,3
[FATAL] 20:03:32 testasynclog.rs 25:fatal>>>>,eeeeeeeeeeeeee,1,2,3

Multiple Instance Asynchronous

use std::sync::Arc;

use tklog::{
    async_debugs, async_errors, async_fatals, async_infos, async_traces, async_warns, LEVEL, Format, ASYNC_LOG, MODE
};

#[tokio::test]
async fn testmultilogs() {
    let mut log = tklog::Async::Logger::new();
    log.set_console(false)
        .set_level(LEVEL::Debug)
        .set_cutmode_by_time("tklogasync.log", MODE::DAY, 10, true)
        .await
        .set_formatter("{message} | {time} {file}{level}\n");

    let mut logger = Arc::clone(&Arc::new(Mutex::new(log)));
    let log = logger.borrow_mut();
    async_traces!(log, "async_traces>>>>", "AAAAAAAAAA", 1, 2, 3);
    async_debugs!(log, "async_debugs>>>>", "BBBBBBBBBB", 1, 2, 3);
    async_infos!(log, "async_infos>>>>", "CCCCCCCCCC", 1, 2, 3);
    async_warns!(log, "async_warns>>>>", "DDDDDDDDDD", 1, 2, 3);
    async_errors!(log, "async_errors>>>>", "EEEEEEEEEEE", 1, 2, 3);
    async_fatals!(log, "async_fatals>>>>", "FFFFFFFFFFFF", 1, 2, 3);
    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
}

Execution Result:

async_debugs>>>>,BBBBBBBBBB,1,2,3 | 2024-05-26 20:10:24 testasynclog.rs 45[DEBUG]
async_infos>>>>,CCCCCCCCCC,1,2,3 | 2024-05-26 20:10:24 testasynclog.rs 46[INFO]
async_warns>>>>,DDDDDDDDDD,1,2,3 | 2024-05-26 20:10:24 testasynclog.rs 47[WARN]
async_errors>>>>,EEEEEEEEEEE,1,2,3 | 2024-05-26 20:10:24 testasynclog.rs 48[ERROR]
async_fatals>>>>,FFFFFFFFFFFF,1,2,3 | 2024-05-26 20:10:24 testasynclog.rs 49[FATAL]

Supports the official log library standard API

1. tklog implements the regular use of the official Log interface API
2. Implement the official log library API to be used in asynchronous scenarios

How to enable the official log library API：

tklog enables API support for official logs by calling the uselog() function

Use example

use std::{thread, time::Duration};
use tklog::{Format, LEVEL, LOG};
fn test_synclog() {
    //init  LOG
    LOG.set_console(true)
        .set_level(LEVEL::Debug)
        .set_cutmode_by_size("logsize.log", 10000, 10, true)
        .uselog();  //Enable the official log library
	
	log::trace!("trace>>>>{}{}{}{}{}", "aaaa", 1, 2, 3, 4);
	log::debug!("debug>>>>{}{}",1,2);
    log::info!("info log");
    log::warn!("warn log");
    log::error!("error log");
	thread::sleep(Duration::from_secs(1))
}

Enable the log library API in asynchronous scenarios

use std::{thread, time::Duration};
use tklog::{Format, LEVEL, ASYNC_LOG};
async fn test_synclog() {
    //init ASYNC  LOG 
    ASYNC_LOG.set_console(false)
        .set_cutmode_by_size("asynclogsize.log", 10000, 10, true).await
        .uselog(); //Enable the official log library
	
    log::trace!("trace async log>>>>{}{}{}{}{}", "aaaaaaaaa", 1, 2, 3, 4);
    log::debug!("debug async log>>>>{}{}",1,2);
	log::info!("info async log");
    log::warn!("warn async log");
    log::error!("error async log");
    tokio::time::sleep(tokio::time::Duration::from_secs(3)).await;
}

The module sets log parameters

1. tklog provides set_option and set_mod_option to set the global log parameters of the Logger object and specify the log parameters of the mod
2. In the project, you can use the global LOG object to set log parameters for multiple mod at the same time
3. Different mod can set different log level, log formats, log file, etc
4. The log parameter of mod for ASYNC_LOG is the same as LOG object

set_option example：

tklog::LOG.set_option(LogOption{level:Some(LEVEL::Debug),console: Some(false),format:None,formatter:None,fileoption: Some(Box::new(FileTimeMode::new("day.log",tklog::MODE::DAY,0,true)))});

LogOption instruction

• level level of log
• format format of log
• formatter user-defined log output format
• console console log setting
• fileoption file log setting

set_mod_option File log setting：

tklog::LOG.set_mod_option("testlog::module1",LogOption{level:Some(LEVEL::Debug),console: Some(false),format:None,formatter:None,fileoption: Some(Box::new(FileTimeMode::new("day.log", tklog::MODE::DAY, 0,true)))});

• testlog::module1 is the module name，you can use module_path!() to print out the current module name
• When tklog is used in the module testlog::module1, tklog will use the LogOption object

Complete mod example

mod module1 {
    use std::{thread, time::Duration};
    use tklog::{handle::FileTimeMode, LogOption, LEVEL};
    pub fn testmod() {
        tklog::LOG.set_mod_option("testlog::module1", LogOption { level: Some(LEVEL::Debug), format: None, formatter: None, console: None, fileoption: Some(Box::new(FileTimeMode::new("module1.log", tklog::MODE::DAY, 0, true))) }).uselog();
        tklog::debug!("module1,tklog api,LOG debug log>>", 123);
        tklog::info!("module1,tklog api,LOG info log>>", 456);
        log::debug!("module1,log api,debug log>>{}", 111);
        log::info!("module1,log api,info log>>{}", 222);
        thread::sleep(Duration::from_secs(1))
    }
}

mod module2 {
    use std::{thread, time::Duration};
    use tklog::{handle::FileTimeMode, LogOption, LEVEL};
    pub fn testmod() {
        tklog::LOG.set_mod_option("testlog::module2", LogOption { level: Some(LEVEL::Info), format: None, formatter: None, console: None, fileoption: Some(Box::new(FileTimeMode::new("module2.log", tklog::MODE::DAY, 0, true))) }).uselog();
        tklog::debug!("module2,tklog api,LOG debug log>>", 123);
        tklog::info!("module2,tklog api,LOG info log>>", 456);
        log::debug!("module2,log api,debug log>>{}", 111);
        log::info!("module2,log api,info log>>{}", 222);
        thread::sleep(Duration::from_secs(1))
    }
}

#[test]
fn testmod2() {
    module1::testmod();
    module2::testmod();
}

Execution Result:

[DEBUG] 2024-06-19 10:54:07 testlog.rs 54:module1,tklog api,LOG debug log>>,123
[INFO] 2024-06-19 10:54:07 testlog.rs 55:module1,tklog api,LOG info log>>,456
[DEBUG] 2024-06-19 10:54:07 testlog.rs 56:module1,log api,debug log>>111
[INFO] 2024-06-19 10:54:07 testlog.rs 57:module1,log api,info log>>222
[INFO] 2024-06-19 10:54:08 testlog.rs 68:module2,tklog api,LOG info log>>,456
[INFO] 2024-06-19 10:54:08 testlog.rs 70:module2,log api,info log>>222

Example 2: Asynchronous logging

mod module3 {
    use tklog::{handle::FileTimeMode, Format, LogOption, LEVEL};
    pub async fn testmod() {
        tklog::ASYNC_LOG.set_mod_option("testlog::module3", LogOption { level: Some(LEVEL::Debug), format: Some(Format::Date), formatter: None, console: None, fileoption: Some(Box::new(FileTimeMode::new("module3.log", tklog::MODE::DAY, 0, true))) }).await.uselog();
        tklog::async_debug!("async module3,tklog api,LOG debug log>>", 123);
        tklog::async_info!("async module3,tklog api,LOG info log>>", 456);
        log::debug!("async module3,log api,debug log>>{}", 333);
        log::info!("async module3,log api,info log>>{}", 444);
        tokio::time::sleep(tokio::time::Duration::from_secs(1)).await;
    }
}

mod module4 {
    use tklog::{handle::FileTimeMode, Format, LogOption, LEVEL};
    pub async fn testmod() {
        tklog::ASYNC_LOG.set_mod_option("testlog::module4", LogOption { level: Some(LEVEL::Info), format: Some(Format::Date), formatter: None, console: None, fileoption: Some(Box::new(FileTimeMode::new("module4.log", tklog::MODE::DAY, 0, true))) }).await.uselog();
        tklog::async_debug!("async module4,tklog api,LOG debug log>>", 123);
        tklog::async_info!("async module4,tklog api,LOG info log>>", 456);
        log::debug!("async module4,log api,debug log>>{}", 333);
        log::info!("async module4,log api,info log>>{}", 444);
        tokio::time::sleep(tokio::time::Duration::from_secs(1)).await;
    }
}

#[tokio::test]
async fn testmod4() {
    module3::testmod().await;
    module4::testmod().await;
}

Execution Result:

[DEBUG] 2024-06-19 10:59:26 testlog.rs 85:async module3,tklog api,LOG debug log>>,123
[INFO] 2024-06-19 10:59:26 testlog.rs 86:async module3,tklog api,LOG info log>>,456
[DEBUG] 2024-06-19 10:59:26 testlog.rs 87:async module3,log api,debug log>>333
[INFO] 2024-06-19 10:59:26 testlog.rs 88:async module3,log api,info log>>444
[INFO] 2024-06-19 10:59:27 testlog.rs 98:async module4,tklog api,LOG info log>>,456
[INFO] 2024-06-19 10:59:27 testlog.rs 100:async module4,log api,info log>>444

tklog supports multi-instance formatting format! And asynchronous format!

Example：

#[test]
fn testformats() {
	let mut log = Logger::new();
	log.set_console(true)
		.set_level(LEVEL::Debug)
		.set_cutmode_by_time("tklogs.log", MODE::DAY, 10, true);
	let mut logger = Arc::clone(&Arc::new(Mutex::new(log)));
	let log = logger.borrow_mut();

	let v = vec![1, 2, 3];
	tklog::formats!(log, LEVEL::Debug, "Debug>>>{},{}>>>{:?}", 1, 2, v);

	let v2 = vec!['a', 'b'];
	tklog::formats!(log, LEVEL::Info, "Info>>>{},{}>>{:?}", 1, 2, v2);
	tklog::formats!(log, LEVEL::Warn, "Warn>>>{},{}", 1, 2);
	tklog::formats!(log, LEVEL::Error, "Error>>>{},{}", 1, 2);
	tklog::formats!(log, LEVEL::Fatal, "Fatal>>>{},{}", 1, 2);

	thread::sleep(Duration::from_secs(1))
}

Execution Result:

[DEBUG] 2024-06-06 15:54:07 testsynclog.rs 80:Debug>>>1,2>>>[1, 2, 3]
[INFO] 2024-06-06 15:54:07 testsynclog.rs 83:Info>>>1,2>>['a', 'b']
[WARN] 2024-06-06 15:54:07 testsynclog.rs 84:Warn>>>1,2
[ERROR] 2024-06-06 15:54:07 testsynclog.rs 85:Error>>>1,2
[FATAL] 2024-06-06 15:54:07 testsynclog.rs 86:Fatal>>>1,2

asynchronous Example

#[tokio::test]
async fn testformats() {
	let mut log = tklog::Async::Logger::new();
	log.set_console(true)
		.set_level(LEVEL::Debug)
		.set_cutmode_by_time("tklogasyncs.log", MODE::DAY, 10, true)
		.await;
	let mut logger = Arc::clone(&Arc::new(Mutex::new(log)));
	let log = logger.borrow_mut();

	let v = vec![1, 2, 3];
	tklog::async_formats!(log, LEVEL::Debug, "Debug>>>{},{}>>>{:?}", 1, 2, v);

	let v2 = vec!['a', 'b'];
	tklog::async_formats!(log, LEVEL::Info, "Info>>>{},{}>>{:?}", 1, 2, v2);
	tklog::async_formats!(log, LEVEL::Warn, "Warn>>>{},{}", 1, 2);
	tklog::async_formats!(log, LEVEL::Error, "Error>>>{},{}", 1, 2);
	tklog::async_formats!(log, LEVEL::Fatal, "Fatal>>>{},{}", 1, 2);

	tokio::time::sleep(tokio::time::Duration::from_secs(1)).await;
}

Execution Result:

[DEBUG] 2024-06-06 16:09:26 testasynclog.rs 61:Debug>>>1,2>>>[1, 2, 3]
[INFO] 2024-06-06 16:09:26 testasynclog.rs 64:Info>>>1,2>>['a', 'b']
[WARN] 2024-06-06 16:09:26 testasynclog.rs 65:Warn>>>1,2
[ERROR] 2024-06-06 16:09:26 testasynclog.rs 66:Error>>>1,2
[FATAL] 2024-06-06 16:09:26 testasynclog.rs 67:Fatal>>>1,2

Benchmark Test

log_benchmark           time:   [2.9703 µs 2.9977 µs 3.0256 µs]
                        	change: [-95.539% -95.413% -95.268%] (p = 0.00 < 0.05)
                        	Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) high mild
  5 (5.00%) high severe

log_benchmark           time:   [2.9685 µs 3.0198 µs 3.0678 µs]
                        	change: [-3.6839% -1.2170% +1.0120%] (p = 0.34 > 0.05)
                        	No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
  7 (7.00%) high mild

test_debug              time:   [3.3747 µs 3.4599 µs 3.5367 µs]
                               change: [-69.185% -68.009% -66.664%] (p = 0.00 < 0.05)
                               Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
  6 (6.00%) high mild
  3 (3.00%) high severe

test_debug              time:   [3.8377 µs 3.8881 µs 3.9408 µs]
                                change: [-66.044% -65.200% -64.363%] (p = 0.00 < 0.05)
                                Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
  2 (2.00%) high mild

Explanation: The time range gives three data points representing the minimum test execution time (2.9055 microseconds), the value near the average (2.9444microseconds-3.8881microseconds ), and the maximum (3.9408 microseconds).

Conclusion: Log printing function performance: 2 µs /op - 3.9 µs /op (microsecond/time)