1：`gsub`

Question

我有刺痛

“ Functionname（'parameter1blue'，'parameter2red'，'14246,14681'，'Simple'，'2018-07-26'）”

应替换为

“ Functionname（'parameter1blue'，'parameter2red'，'14681，XXXXXX'，'Simple'，'2018-07-26'）”

我看过regex函数和其他字符串函数，它们很长/很难编辑任何更改，我需要最简单的方法来对字符串数组进行此操作。

注意：“ XXXXXX”位于最大数字的位置，但它会替换最小数字的值，同时保留升序。

这是一个入门示例代码，其中包含示例数据和所需数据。

#This is the avaliable data
olddata<-data.frame(sqlcode=c("Functionname('parameter1blue','parameter2red','14246,14681','Simple','37748','2018-07-26')",
                              "Functionname('parameter1green','parameter2blue','13027,13559,13914,14246,14681','Simple','24548','2018-07-26')",
                              "Functionname('parameter1white','parameter2red','13587,42254','Complex','36848','2018-07-26')",
                              "Functionname('parameter1green','parameter2green','14246','Simple','37258','2018-07-26')",
                              "Functionname('parameter1red','parameter2white','14246,14681','Complex','37568','2018-07-26')",
                              "Functionname('parameter1blue','parameter2white','13587,42243','Simple','22548','2018-07-26')"),stringsAsFactors = F)

#This is the value which has to be replaced
newval="XXXXXX"

#This is how the new data should look like
#The numbers between the parameter2color and Simple/complex are supposed to be replaced with the newval in a way that the first number between them is replaced with it 
# but placed at the position of the last number
desireddata<-data.frame(sqlcode=c("Functionname('parameter1blue','parameter2red','14681,XXXXXX','Simple','37748','2018-07-26')",
                              "Functionname('parameter1green','parameter2blue','13559,13914,14246,14681,XXXXXX','Simple','24548','2018-07-26')",
                              "Functionname('parameter1white','parameter2red','42254,XXXXXX','Complex','36848','2018-07-26')",
                              "Functionname('parameter1green','parameter2green','XXXXXX','Simple','37258','2018-07-26')",
                              "Functionname('parameter1red','parameter2white','14681,XXXXXX','Complex','37568','2018-07-26')",
                              "Functionname('parameter1blue','parameter2white','42243,XXXXXX','Simple','22548','2018-07-26')"))

Answer 1

好的，新规则，新代码，新测试数据。我将解决方案保留在下面（gsub和`regmatches<-`），但是它们似乎并不遵循规则。这是使用OP中数据的有效代码。

gr1 <- gregexpr("\\(.*\\)", olddata$sqlcode)
args <- strsplit(unlist(regmatches(olddata$sqlcode, gr1)), "','")
arg3 <- sapply(args, `[[`, 3)
arg3new <- sapply(strsplit(arg3, ","), function(a) paste(c(tail(a,n=-1), newval), collapse=","))
regmatches(olddata$sqlcode, gr1) <- sapply(mapply(`[<-`, args, list(3), arg3new, SIMPLIFY=FALSE), paste, collapse="','")
olddata
#                                                                                                           sqlcode
# 1                     Functionname('parameter1blue','parameter2red','14681,XXXXXX','Simple','37748','2018-07-26')
# 2 Functionname('parameter1green','parameter2blue','13559,13914,14246,14681,XXXXXX','Simple','24548','2018-07-26')
# 3                   Functionname('parameter1white','parameter2red','42254,XXXXXX','Complex','36848','2018-07-26')
# 4                        Functionname('parameter1green','parameter2green','XXXXXX','Simple','37258','2018-07-26')
# 5                   Functionname('parameter1red','parameter2white','14681,XXXXXX','Complex','37568','2018-07-26')
# 6                   Functionname('parameter1blue','parameter2white','42243,XXXXXX','Simple','22548','2018-07-26')

此行下方的所有内容都不再需要。

两种方法：第一种（gsub）会更改找到的最大数量的所有实例，这可能/不可能或存在问题；第二个（`regmatches<-`）仅替换which.max返回的最大值，因此它将始终替换最多一个数字。

1：`gsub`

gr <- gregexpr("[0-9]+", olddata$sqlcode)
str( nums <- regmatches(olddata$sqlcode, gr) )
# List of 6
#  $ : chr [1:7] "1" "2" "14246" "14681" ...
#  $ : chr [1:10] "1" "2" "13027" "13559" ...
#  $ : chr [1:7] "1" "2" "13587" "42254" ...
#  $ : chr [1:6] "1" "2" "14246" "2018" ...
#  $ : chr [1:7] "1" "2" "14246" "14681" ...
#  $ : chr [1:7] "1" "2" "13587" "42243" ...
str( inds <- sapply(nums, function(n) which.max(as.integer(n))) )
#  int [1:6] 4 7 4 3 4 4
str( replacethese <- mapply(`[[`, nums, inds) )
#  chr [1:6] "14681" "14681" "42254" "14246" "14681" "42243"

mapply(function(strings,old) gsub(paste0("\\b", old, "\\b"), newval, strings),
       olddata$sqlcode, replacethese)
# [1] "Functionname('parameter1blue','parameter2red','14246,XXXXXX','Simple','2018-07-26')"                    
# [2] "Functionname('parameter1green','parameter2blue','13027,13559,13914,14246,XXXXXX','Simple','2018-07-26')"
# [3] "Functionname('parameter1white','parameter2red','13587,XXXXXX','Complex','2018-07-26')"                  
# [4] "Functionname('parameter1green','parameter2green','XXXXXX','Simple','2018-07-26')"                       
# [5] "Functionname('parameter1red','parameter2white','14246,XXXXXX','Complex','2018-07-26')"                  
# [6] "Functionname('parameter1blue','parameter2white','13587,XXXXXX','Simple','2018-07-26')"

2：`regmatches<-`

N.B。，此方法在side-effect中运行，方法是就地（在框架内）更改数据；如果这是一个问题，请改为处理数据副本。

从不变的数据开始，唯一的规定是字符串必须为character，而不是factor。（如果将stringsAsFactors=FALSE添加到对data.frame，read.table，read.csv等的呼叫中，则不会有问题。）

olddata$sqlcode <- as.character(olddata$sqlcode)

我们需要一个函数来索引gregexpr的返回值。这很简单，但是因为属性也需要索引，所以看起来有点吵：

index_reg <- function(gr, i) {
  newgr <- gr[i]
  attributes(newgr) <- attributes(gr)
  attr(newgr, "match.length") <- attr(newgr, "match.length")[i]
  newgr
}

有了这个，我们就可以做到：

gr <- gregexpr("[0-9]+", olddata$sqlcode)                  # no change
nums <- regmatches(olddata$sqlcode, gr)                    # no change
inds <- sapply(nums, function(n) which.max(as.integer(n))) # no change
regmatches(olddata$sqlcode, mapply(index_reg, gr, inds, SIMPLIFY=FALSE)) <- newval
olddata # changed in-place, SIDE-EFFECT!
#                                                                                                   sqlcode
# 1                     Functionname('parameter1blue','parameter2red','14246,XXXXXX','Simple','2018-07-26')
# 2 Functionname('parameter1green','parameter2blue','13027,13559,13914,14246,XXXXXX','Simple','2018-07-26')
# 3                   Functionname('parameter1white','parameter2red','13587,XXXXXX','Complex','2018-07-26')
# 4                        Functionname('parameter1green','parameter2green','XXXXXX','Simple','2018-07-26')
# 5                   Functionname('parameter1red','parameter2white','14246,XXXXXX','Complex','2018-07-26')
# 6                   Functionname('parameter1blue','parameter2white','13587,XXXXXX','Simple','2018-07-26')

Answer 2

这是一种stringr的方法，用于替换组中的最后一个数字：

olddata <- data.frame(
  sqlcode = c(
    "Functionname('parameter1blue','parameter2red','14246,14681','Simple','2018-07-26')",
    "Functionname('parameter1green','parameter2blue','13027,13559,13914,14246,14681','Simple','2018-07-26')",
    "Functionname('parameter1white','parameter2red','13587,42254','Complex','2018-07-26')",
    "Functionname('parameter1green','parameter2green','14246','Simple','2018-07-26')",
    "Functionname('parameter1red','parameter2white','14246,14681','Complex','2018-07-26')",
    "Functionname('parameter1blue','parameter2white','13587,42243','Simple','2018-07-26')"
  )
)

library(tidyverse)
desireddata <- olddata %>%
  mutate(sqlcode = str_replace(sqlcode, "\\d{5}(?=','(Simple|Complex))", "XXXXXX"))
desireddata
#>                                                                                                   sqlcode
#> 1                     Functionname('parameter1blue','parameter2red','14246,XXXXXX','Simple','2018-07-26')
#> 2 Functionname('parameter1green','parameter2blue','13027,13559,13914,14246,XXXXXX','Simple','2018-07-26')
#> 3                   Functionname('parameter1white','parameter2red','13587,XXXXXX','Complex','2018-07-26')
#> 4                        Functionname('parameter1green','parameter2green','XXXXXX','Simple','2018-07-26')
#> 5                   Functionname('parameter1red','parameter2white','14246,XXXXXX','Complex','2018-07-26')
#> 6                   Functionname('parameter1blue','parameter2white','13587,XXXXXX','Simple','2018-07-26')

由reprex package（v0.2.0）于2018-08-10创建。

R中字符串内某些动态变化之间的高级替换

2 个答案:

1：`gsub`

2：`regmatches<-`

R中字符串内某些动态变化之间的高级替换

2 个答案:

1：gsub

2：`regmatches<-`

1：`gsub`