gawk中的字符串数字(STRNUM)介绍

(编辑:jimmy 日期: 2024/11/16 浏览:2)

几天前在论坛看到一个帖子 http://bbs.chinaunix.net/thread-4184581-1-1.html 这是个挺有意思的问题。
我自己做了几个试验,发现00e1有这个问题,但是00a1、00b1、00c1等其它字符串却没有这个问题。

前天,怀着忐忑的心情给 bug-gawk@gnu.org 发了一封邮件:

[bug report] gawk doesn’t work correctly when assign special value to variable
To: bug-gawk@gnu.org

1. Symptom:
gawk doesn’s work correctly when assign special value (like 00e1) variable.

2. Steps to repeat this issue:

1) Download and compile the latest gawk 4.1.3
http://ftp.gnu.org/gnu/gawk/
[root]# gawk --version | head -2
GNU Awk 4.1.3, API: 1.1
Copyright (C) 1989, 1991-2015 Free Software Foundation.

2) Prepare a test file
cat > 1.txt << EOF
00e1 00e1
00e2 00e4
00e3 00e1_01
EOF

3) Execute following command to see the result
[root]# gawk -v var="00e1" '$2==var' 1.txt
00e1 00e1
00e2 00e4

The second line should not be there. Because "00e4" is NOT equal to "00e1" obviously.
Is it a bug or by design?

3. Workaround:
gawk -v var="00e1" '$2==""var' 1.txt
gawk -v var="^00e1___FCKpd___0quot; '$2~var' 1.txt

没想到,昨天就收到了来自 Andrew J. Schorr 大神的回复,效率还真是高呢,32个赞:

On Thu, Aug 06, 2015 at 03:55:50PM +0800, Shell_HAT wrote:
> The second line should not be there. Because "00e4" is NOT equal to "00e1" obviously.
> Is it a bug or by design?

This is a tricky area. Please check the manual for the discussion of "String Type versus Numeric Type":
http://www.gnu.org/software/gawk/manual/html_node/Variable-Typing.html

I think the command-line assignment results in a variable of type STRNUM.

> 3. Workaround:
> gawk -v var="00e1" '$2==""var' 1.txt
> gawk -v var="^00e1___FCKpd___1quot; '$2~var' 1.txt

These examples may also be helpful:

bash-4.2$ gawk 'BEGIN {var = "00e1"} $2 == var' 1.txt
00e1 00e1
bash-4.2$ gawk 'BEGIN {var = 00e1} $2 == var' 1.txt
00e1 00e1
00e2 00e4

Regards,
Andy

看完那篇文档,似有所悟。但是还是没搞清楚为啥00a1、00b1、00c1等字符串没问题呢?
喝完一杯咖啡之后,恍然大悟:科学计数法!泥煤的~~~

一句话新闻

微软与英特尔等合作伙伴联合定义“AI PC”:键盘需配有Copilot物理按键
几个月来,英特尔、微软、AMD和其它厂商都在共同推动“AI PC”的想法,朝着更多的AI功能迈进。在近日,英特尔在台北举行的开发者活动中,也宣布了关于AI PC加速计划、新的PC开发者计划和独立硬件供应商计划。
在此次发布会上,英特尔还发布了全新的全新的酷睿Ultra Meteor Lake NUC开发套件,以及联合微软等合作伙伴联合定义“AI PC”的定义标准。