Awk-如何改进正则表达式？

我有一个文件：

@Book{gjn2011ske, 
  author =   {Grzegorz J. Nalepa},
  title =    {Semantic Knowledge Engineering. A Rule-Based Approach},
  publisher =    {Wydawnictwa AGH},
  year =     2011,
  address =  {Krak\'ow}
}

@article{gjn2010jucs,
  Author =   {Grzegorz J. Nalepa},
  Journal =  {Journal of Universal Computer Science},
  Number =   7,
  Pages =    {1006-1023},
  Title =    {Collective Knowledge Engineering with Semantic Wikis},
  Volume =   16,
  Year =     2010
}

我想改进只删除第一行的正则表达式注意：无法更改记录分隔符

RS=“}\n”

我试过：

awk 'BEGIN{ RS="}\n" } {gsub(/@.*,/,"") ; print }' file

我想打印结果：

  author =   {Grzegorz J. Nalepa},
  title =    {Semantic Knowledge Engineering. A Rule-Based Approach},
  publisher =    {Wydawnictwa AGH},
  year =     2011,
  address =  {Krak\'ow}

  Author =   {Grzegorz J. Nalepa},
  Journal =  {Journal of Universal Computer Science},
  Number =   7,
  Pages =    {1006-1023},
  Title =    {Collective Knowledge Engineering with Semantic Wikis},
  Volume =   16,
  Year =     2010

谢谢你的帮助

编辑：

我提议的解决办法：

awk 'BEGIN{ RS="}\n" }{sub(",","@"); sub(/@.*@/,""); print }' file 

---

使用指定的

RS

设置很难实现您想要的结果（因为

地址={Krak\ow}

有一个额外的记录结尾）。我宁愿选择：

awk '$0 !~ "^@" && $0 !~ "^} *$" { print }' FILE 

看到了

编辑我不知道为什么它必须与一个regexp解决方案，你能解释一下吗

无论如何，还有一个（）解决方案使用regexp，但不是您所期望的：

awk 'BEGIN{ RS="}\n" }
{
  split($0,a,"\n")
  for (e=1;e<=length(a);e++) {
      if (a[e] ~ "{" && a[e] !~ "}") {
          sub("$","}",a[e])
      }
      if (a[e] ~ "=") { print a[e] }
  }
  printf("\n")
}' INPUTFILE

---

一种不使用正则表达式的方法。将字段分隔符设置为换行符，现在寄存器的每个键都将是一个字段。这样，遍历每个字段并打印那些不以

@

开头的字段：

awk '
    BEGIN { 
        RS="}\n"; 
        FS=OFS="\n"; 
    } 
    { 
        for (i=1; i<=NF; i++) { 
            if ( substr($i, 1, 1) != "@" ) { 
                printf "%s%s", $i, (i == NF) ? RS : OFS; 
            } 
        } 
    }
' file

---

我将使用

GNU sed

来执行此操作：

sed '/^@/,/^}$/ { //d }' file.txt

结果:

  author =   {Grzegorz J. Nalepa},
  title =    {Semantic Knowledge Engineering. A Rule-Based Approach},
  publisher =    {Wydawnictwa AGH},
  year =     2011,
  address =  {Krak\'ow}

  Author =   {Grzegorz J. Nalepa},
  Journal =  {Journal of Universal Computer Science},
  Number =   7,
  Pages =    {1006-1023},
  Title =    {Collective Knowledge Engineering with Semantic Wikis},
  Volume =   16,
  Year =     2010

请注意，您可以使用

-i

标志进行适当的更改（即覆盖文件内容），也可以使用

-s

标志对多个文件进行更改。例如：

sed -s -i '/^@/,/^}$/ { //d }' *.txt

测试如下：

> awk '{if($0!~/@/&&$0!~/^}/)print}' temp
  author =       {Grzegorz J. Nalepa},
  title =        {Semantic Knowledge Engineering. A Rule-Based Approach},
  publisher =    {Wydawnictwa AGH},
  year =         2011,
  address =      {Krak\'ow}

  Author =       {Grzegorz J. Nalepa},
  Journal =      {Journal of Universal Computer Science},
  Number =       7,
  Pages =        {1006-1023},
  Title =        {Collective Knowledge Engineering with Semantic Wikis},
  Volume =       16,
  Year =         2010
>

---

谢谢你的解决方案。您的示例将“}”保留在最后一行的末尾。请参阅我的编辑和建议的解决方案。感谢您提供的解决方案，但等待仍然存在，例如，正则表达式。请参阅“我的编辑”和“建议的解决方案”。@Tedee12345:无法更改

awk

的记录分隔符会产生比解决问题更多的问题。围绕这些问题编写代码从来都不是一个好主意。你应该考虑发布为什么你认为保持代码> RS=“}\n”是个好主意。如果是，请提供更多样本数据。祝你好运。谢谢你提供的解决方案，但是等待仍然存在，例如，正则表达式。请参阅我的编辑和建议的解决方案。再次感谢您的回复。你提出的第一个解决方案很适合我。这个答案与20小时前佐尔特的答案几乎相同。你应该考虑像我一样投票，谢谢你的解决方案的例子。

sed -s -i '/^@/,/^}$/ { //d }' *.txt

awk '{if($0!~/@/&&$0!~/^}/)print}' temp

> awk '{if($0!~/@/&&$0!~/^}/)print}' temp
  author =       {Grzegorz J. Nalepa},
  title =        {Semantic Knowledge Engineering. A Rule-Based Approach},
  publisher =    {Wydawnictwa AGH},
  year =         2011,
  address =      {Krak\'ow}

  Author =       {Grzegorz J. Nalepa},
  Journal =      {Journal of Universal Computer Science},
  Number =       7,
  Pages =        {1006-1023},
  Title =        {Collective Knowledge Engineering with Semantic Wikis},
  Volume =       16,
  Year =         2010
>