tklog | rust logging library

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

ease-of-use, efficiency, a rich feature suite. console logging, file logging, structured data formatting, with capabilities extending to log slicing by time or size, both synchronous and asynchronous logging modes, alongside compressed backup of log files.

Feature

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

Relevant website

Use tklog

[dependencies]
tklog = "0.0.x"   #   "0.0.x" current version

Simple Description of usage

Simplest use: Call macros directly:

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, this prints to the console without writing to a file. A sample output might look like:

[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 provides methods to set options such as console logging, log levels, formats, cutting modes, and formatters.

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

fn log_init() {
        LOG.set_console(true)  //log of console
            .set_level(LEVEL::Info)  //log level，default：Debug
            .set_format(Format::LevelFlag | Format::Time | Format::ShortFileName)  //Structured logs that define the output log information
            .set_cutmode_by_size("tklogsize.txt", 1<<20, 10, true)  //The log file cutting mode is 1 MB file size, 10 backup log files are reserved, and backup logs are compressed
            .set_formatter("{level}{time} {file}:{message}\n")    //defined log output format。default：{level}{time} {file}:{message}

}

The above is an example of global single-instance printing. tklog supports custom multi-instance printing. Multiple instances are generally used in scenarios where the system requires different print structures.

Multi-Instance Logging

tklog also supports multiple instances for scenarios requiring different logging structures. Each instance can have its own settings for console output, log level, file rotation, and 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 by day, reserve a maximum of 10 backups, and compress the backup files
        .set_formatter("{message} | {time} {file}{level}\n")          //customize the input sequence and additional content of log structure information
 
    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 above input structured information is determined by "{message} | {time} {file}{level}\n" formatter. formatter in addition to the key identifier {message} {time} {file} {level}, other contents of the output, such as | , Spaces, newlines and so on.

Detailed Usage Instructions

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

Example:

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

2. Console Logging: Invoke the .set_console(bool) function.

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 down to seconds (e.g., 14:13:25).
Format::Microseconds: Outputs time with microseconds (e.g., 18:09:17.462245).
Format::LongFileName: Full file info with line number (e.g., tests estlog.rs 25).
Format::ShortFileName: Short file info with line number (e.g., testlog.rs 25).
Format::LevelFlag: Log level flag (e.g., [Debug]).

To set a custom format:

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

4. Custom Format Strings:

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

{level}: Log level info, e.g., [Debug].
{time}: Log time.
{file}: File and line number.
{message}: Log content.

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

Note: Text outside {level}, {time}, {file}, and {message} tags is output as is, including separators, spaces, newlines, etc.

5. Time-Based Log File Rotation:
Modes: MODE::HOUR, MODE::DAY, MODE::MONTH.

Use .set_cutmode_by_time() with parameters:

File path
Time mode
Max backup count
Whether to compress backups

Example:

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

This sets the backup location to /usr/local/tklogs.log, rotates daily, keeps unlimited backups, and doesn't compress the daily log file.

Backup file naming convention:

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:
Use .set_cutmode_by_size() with 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 is the path, files roll at 100MB, it keeps 10 backups, and compresses them.

Backup file format:

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() {
    // Setting parameters for global singleton
   ASYNC_LOG
        .set_console(false)   //console
        .set_level(LEVEL::Trace)  //level
        .set_format(Format::LevelFlag | Format::Time | Format::ShortFileName)  //Structured logging defines the output of log information.
        .set_cutmode_by_size("tklog_async.txt", 10000, 10, false).await  //The log file cutting mode is the file size, once every 10,000 bytes, and 10 backup log files are reserved
 }
 
 
#[tokio::test]
async fn testlog() {
    async_log_init().await;  //Parameter 
    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() {
    //create Async::Logger object
    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

tklog implements the regular use of the official Log interface API
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

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
In the project, you can use the global LOG object to set log parameters for multiple mod at the same time
Different mod can set different log level, log formats, log file, etc
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

tklog supports custom log processing functions.

tklog allows the addition of external custom functions through set_custom_handler(), enabling control over the log processing flow and logic.

Example

#[test]
fn test_custom() {
    fn custom_handler(lc: &LogContext) -> bool {
        println!("level >>>>>>>>>>>>>>>>>{:?}", lc.level);
        println!("message >>>>>>>>>>>>>>>>>{:?}", lc.log_body);
        println!("filename >>>>>>>>>>>>>>>>>{:?}", lc.filename);
        println!("line >>>>>>>>>>>>>>>>>{:?}", lc.line);
        println!("modname >>>>>>>>>>>>>>>>>{:?}", lc.modname);
        if lc.level == LEVEL::Debug {
            println!("{}", "debug now");
            return false;
        }
        true
    }

    LOG.set_custom_handler(custom_handler);
    debug!("000000000000000000");
    info!("1111111111111111111");
    thread::sleep(Duration::from_secs(1))
}

Execution Result

---- test_custom stdout ----
level >>>>>>>>>>>>>>>>>Debug
message >>>>>>>>>>>>>>>>>"000000000000000000"
filename >>>>>>>>>>>>>>>>>"tests\\testsynclog.rs"
line >>>>>>>>>>>>>>>>>143
modname >>>>>>>>>>>>>>>>>"testsynclog"
debug now
level >>>>>>>>>>>>>>>>>Info
message >>>>>>>>>>>>>>>>>"1111111111111111111"
filename >>>>>>>>>>>>>>>>>"tests\\testsynclog.rs"
line >>>>>>>>>>>>>>>>>144
modname >>>>>>>>>>>>>>>>>"testsynclog"
[INFO] 2024-08-05 15:39:07 testsynclog.rs 144:1111111111111111111

Explanation

When the function fn custom_handler(lc: &LogContext) -> bool returns true, tklog calls the custom_handler to execute the custom function and then continues with tklog's logging process. When it returns false, tklog does not proceed with its logging process and directly returns. As shown in the example, when the log level is Debug, it returns false, so tklog does not print the Debug log.

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)

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

Feature

Relevant website

Use tklog

Simple Description of usage

Detailed Usage Instructions

Log Printing Methods:

Asynchronous Logging

Example:

Multiple instance asynchronous

Supports the official log library standard API

How to enable the official log library API：

Enable the log library API in asynchronous scenarios

The module sets log parameters

set_option example：

LogOption instruction

set_mod_option File log setting：

Complete mod example

Execution Result:

Example 2: Asynchronous logging

Execution Result:

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

tklog supports custom log processing functions.

Explanation

Benchmark Test

`set_option` example：

`set_mod_option` File log setting：