Regex 使用grep提取两个子字符串之间的字符串

以下内容几乎是我想要的，但不是全部：

echo 'pie(x=x,labels=c("5.3 %","2.8 %","1.5 %","5.2 %","2.9 %","2.3 %","7.9 %","16.9 %","19.5 %","27.5 %","8.3 %"),main="ChIP",col=c("#445FA2","#EB9D86","#799F7A","#6C527F","#5FA1C1","#E8BB77","#A8C5EF","#FDCDB9","#C6E6B5","#F1D5EE","#B4E1F6"),clockwise=TRUE,border=FALSE,radius=0.9,cex=0.8,init.angle=90,density=100)'|grep -oP '(?<=").*?(?= %)'

请告知如何获得“更清洁”的输出，例如：

5.3
2.8
1.5
5.2
2.9
2.3
7.9
16.9
19.5
27.5
8.3

---

最好，请基于上述grep-oP解决方案，但当然也欢迎其他解决方案

如果您只需要0-9和。可以使用的字符：

grep -oP '(?<=")[\d\.]*?(?= %)'

而不是

grep -oP '(?<=").*?(?= %)'

注意：Perl中的：\d匹配0-9

---

. 在Perl中，匹配任何字符，因此您可以将其转义为文字句点字符

简单地将数字与空格匹配就可以获得所需的输出

 grep -oP '[0-9]+.[0-9]+ '

如果小数点和%之间没有空格，也可以这样做

grep -oP '[0-9]+.[0-9]+.*?(?=%)'

扩展为：

…|grep-Eo[0-9]+\[0-9]+

如果您只是查找数字点数字或使用：

…|grep-Eo[0-9]+\[0-9]+|[0-9]+

对于所有可能有点介于之间的数字。

GNU awk解决方案：

s='pie(x=x,labels=c("5.3 %","2.8 %","1.5 %","5.2 %","2.9 %","2.3 %","7.9 %","16.9 %","19.5 %","27.5 %","8.3 %"),main="ChIP",col=c("#445FA2","#EB9D86","#799F7A","#6C527F","#5FA1C1","#E8BB77","#A8C5EF","#FDCDB9","#C6E6B5","#F1D5EE","#B4E1F6"),clockwise=TRUE,border=FALSE,radius=0.9,cex=0.8,init.angle=90,density=100)'

awk -v FPAT='"[0-9]+\\.[0-9]+ %"' '{ for(i=1;i<=NF;i++) { gsub(/[^.0-9]+/,"",$i); print $i } }' <<<"$s"

---

谢谢为了更一般地捕捉数字中也有字符混音的情况，例如：echo'piex=x，labels=cabc5ui.3%，def2.8%，1.5ghi%，5.2%，2.9%，2.3%，7.9%，16.9%，19.5%，27.5%，8.3%，main=ChIP，col=c445FA2，EB9D86799F7A，6C527F，5FA1C1，E8BB77，A8C5EF，FDCDB9，C6E6B5，F1D5EE，B4E1F6，顺时针=真，边框=假，半径=0.9，cex=0.8，初始角度=90，密度=100’| grep-oP’？非常欢迎你。请注意，可以使用\w替换Perl中的a-zA-Z0-9。由于grep的P选项使用Perl regexp，您可以通过键入：perldoc perlre了解更多信息，尽管我的grep手册页警告说-P是实验性的。非常感谢！例如：grep'pie'EZH2|u 2libsMerged|u posnegMerged|u atPRC2MajorStitchedRegions.R | sed-n 2p | grep-oP |？

s='pie(x=x,labels=c("5.3 %","2.8 %","1.5 %","5.2 %","2.9 %","2.3 %","7.9 %","16.9 %","19.5 %","27.5 %","8.3 %"),main="ChIP",col=c("#445FA2","#EB9D86","#799F7A","#6C527F","#5FA1C1","#E8BB77","#A8C5EF","#FDCDB9","#C6E6B5","#F1D5EE","#B4E1F6"),clockwise=TRUE,border=FALSE,radius=0.9,cex=0.8,init.angle=90,density=100)'

awk -v FPAT='"[0-9]+\\.[0-9]+ %"' '{ for(i=1;i<=NF;i++) { gsub(/[^.0-9]+/,"",$i); print $i } }' <<<"$s"

5.3
2.8
1.5
5.2
2.9
2.3
7.9
16.9
19.5
27.5
8.3