博客

Migrating to a Logger with Sensei

November 30, 2020
Alan Richardson

Migrating to a Logger with Sensei

This post describes creating a recipe to migrate from System.out.println to using a Java Logger.

‍

‍

When I'm hacking out code, rather than using TDD, and I make mistakes, I have the bad habit of printing a line to  System.out, and I've wanted to wean myself off that habit.

I made a bunch of mistakes when writing the code below :

   private String getCountdownString() {
        String output = "";
        String prefix="";
        for(int countdown = 10; countdown > 0; countdown-- ){
            output = output + prefix + countdown;
            System.out.println(output);
            prefix=", ";
        }
        System.out.println(output);
        return output;
    }


Initially, I wrote countdown++ and the loop didn't finish.

And I used countdown > 1 so didn't get the output I wanted.

In the end, I littered my code with System.out.println to help me debug. And the experience reinforced that I need to learn how to use a logger as my default approach.

Research

Fortunately, I read through the Sensei documentation and decided to use the "Getting Started" guide to help me create a recipe to convert from System.out.println and encourage me to use a logger:

  • java.util.logging.Logger

Creating a Recipe

The first thing I do is click on the println then alt+enter to create a new recipe.

System Out Print Create A New Recipe


I create it with the following details:


Name: Logger: use logger instead of println
Description: use logger instead of println - remember to stop using System.out.println
Level: Error


And I'll start by matching methodcall with the name println


search:
  methodcall:
    name: "println"

And the preview shows me all the matches in my code.


Recipe Settings Methodcall Println


I can see that all the matches in my code are for System.out.println but I don't trust that long term this will be the only match. I want to match a more qualified statement I want to change.

I expand the matcher to search for a methodcall on a field named out in the class System.


search:
  methodcall:
    name: "println"
    "on":
      field:
        in:
          class:
            name: "System"
        name: "out"



I could, if I wanted, fully qualify the System name to java.lang.System

Recipe Settings Methodcall name system

Amending the Code to log


Next, I want to create the QuickFix.

First I want to amend the line of code that logs the output:


availableFixes:
- name: "use Logger"
  actions:
  - rewrite:
      to: "logger.log(Level.INFO, {{{ arguments.0 }}})"



I don't have to remember the mustache template format. I used the Show Variables in the GUI to show me the argument and double-clicked on it. Then the GUI filled in the appropriate matching mustache template.



When I try it out, I can see that I still have to alt+enter to import the Level enum. But if I amend my QuickFix to have a fully qualified item then Sensei will add the import for me e.g.

  • It will replace System.out.println(output); with

logger.log(Level.INFO, output);

  • And add an import for the enum:

import java.util.logging.Level;

  • If I rewrite to:
logger.log(java.util.logging.Level.INFO, {{{ arguments.0 }}})

And this will work, but I will still have to remember the syntax to instantiate the logger in the first place.

Syntax To Instantiate The Logger In The First Place


Amending the code to add the logger field

I can amend my QuickFix to create the field for me as well.

I will code the logger first, and then add that to my recipe so I never have to code it again.

Logger logger2 = Logger.getLogger(SysOutTest.class.getName());


I tend to write the example code I want to see generated first because then I can use IntelliJ code completion and syntax checking to make sure I get it right. As a side-effect, it will then be in the code preview when I edit the recipe to add the QuickFix lines that will create that code.

And when I write the example code, I want to use a different field name (here I'm using logger2) because Sensei is clever enough not to add a duplicate field, so I have to fool it by using a different name.

So I'll amend the recipe to create this code by adding a field called logger.

availableFixes:

- name: "use Logger"
  actions:
  - rewrite:
      to: "logger.log(java.util.logging.Level.INFO, {{{ arguments.0 }}})"
  - addField:
      field: "java.util.logging.Logger logger = Logger.getLogger({{{ containingClass.name\
        \ }}}.class.getName())"
      target: "parentClass"


Note that I changed SysOutTest to be a mustache variable so that it picks up the name of any class I use this recipe in. And again, I didn't remember the mustache syntax, I used the GUI Show Variables to find the replacement I needed.

By fully qualifying the Logger to java.util.logging.Logger, Sensei will add the import and write the line of code that I want i.e.

Logger logger = Logger.getLogger(SysOutTest.class.getName());


Fully Qualifying The Logger To java.util.logging.Logger


One useful thing about this recipe is that because it will only add the logger field once, I can use this on any existing code where I have used `System.out.println` and use Sensei to change all the occurrences in my code file at the same time.


Use Logger Fix All problems In The File



Next Steps

Once I get used to this, I'll eventually train myself away from using System.out.println.

I can use Sensei to help me proactively write code by creating a second recipe which helps me create a logger.

e.g. I can match on a class, where there is no field called logger, and add one.

If I create a recipe of level Information


Name: Logger: add logger

Description: Add logger to class


To match on a class without a logger field:

search:
  class:
    without:
      child:
        field:
          name: "logger"


And then I will reuse part of the QuickFix we saw earlier:


availableFixes:

  - name: "Add Logger"
    actions:
      - addField:
          field: "java.util.logging.Logger logger = Logger.getLogger({{{ containingClass.name\
        \ }}}.class.getName())"
          target: "self"


Note the difference in target here compared to the first QuickFix. This uses self because our Search matched the class. The first QuickFix uses parentClass because we matched code within the class itself.

Summary

This represents one of the key flows associated with using Sensei to help improve your personal programming skillset:

  • create a recipe to help with your immediate 'best practice'
  • once you know how to use that best practice... create a recipe to make your workflow faster


---


You can install Sensei from within IntelliJ using "Preferences \ Plugins" (Mac) or "Settings \ Plugins" (Windows)  then just search for "sensei secure code".


The source code and recipes for this can be found in the `sensei-blog-examples` repository in the Secure Code Warrior GitHub account, in the `pojoexamples` module.




标语

在 AI 驱动的开发成果发布前实现有效治理

评估 AI 辅助风险,在提交代码时强制执行安全编码策略,并加速整个软件开发生命周期的安全交付。

预约演示
资源库

探索更多博客

获取有关安全编码、人工智能治理和软件风险管理的专家内容。

浏览全部
案例研究
筛选标签
这是 div 块中的一段文本。

Supercharged Security Awareness: How Tournaments are Inspiring Developers at Erste Group

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

了解更多
案例研究
筛选标签
这是 div 块中的一段文本。

Security as culture: How Blue Prism cultivates world-class secure developers

Learn how Blue Prism, the global leader in intelligent automation for the enterprise, used Secure Code Warrior's agile learning platform to create a security-first culture with their developers, achieve their business goals, and ship secure code at speed

了解更多
案例研究
筛选标签
这是 div 块中的一段文本。

统一的安全文化:Sage 如何打造其安全倡导者项目

了解 Sage 如何通过灵活且以关系为导向的方法提升安全性,成功培养了 200 多名安全倡导者,并实现了显著的风险降低。

了解更多
博客
筛选标签
这是 div 块中的一段文本。

2026年网络安全意识月:理解人工智能安全对企业防御至关重要

目前,大多数顶级人工智能公司已公开呼吁全球放缓人工智能的发展速度,以便安全防护措施能够跟上步伐。然而事实是,严重的安全问题已经显现。随着人工智能实施带来的经济成本和风险不断累积,无数安全领导者和高管团队正处于紧迫的境地。

了解更多
博客
筛选标签
这是 div 块中的一段文本。

赋能要素 8:打造项目品牌

成功的安全编码项目不仅需要优质的内容。赋能要素 8 展示了如何通过品牌化将参与过程转变为一种令人向往、以地位为驱动的体验。

了解更多
博客
筛选标签
这是 div 块中的一段文本。

每一位员工现在都处于 AI 网络安全的最前线

企业技术格局正以一种鲜有人预料到的速度发生剧变,直到为时已晚。我们已正式跨越了从人工编写代码和基础辅助编程,进入“智能体开发生命周期”(ADLC)的时代。虽然自主 AI 智能体有望在多个职能领域实现前所未有的效率,但它们也带来了全新的安全和监管风险。

了解更多

在产品发布前确保 AI 驱动的开发安全

洞察开发风险,执行安全策略,并在整个软件开发生命周期中预防漏洞。

预约演示
未找到相关项目。