R/regex.R

In tdhock/plotMakefile: Parse and plot Makefiles

### Parse several occurances of pattern from each of several strings
### using (named) capturing regular expressions, returning a list of
### matrices (with column names).
str_match_all_perl <- function(string,pattern){
  stopifnot(is.character(string))
  stopifnot(is.character(pattern))
  stopifnot(length(pattern)==1)
  parsed <- gregexpr(pattern,string,perl=TRUE)
  lapply(seq_along(parsed),function(i){
    r <- parsed[[i]]
    starts <- attr(r,"capture.start")
    if(r[1]==-1)return(matrix(nrow=0,ncol=1+ncol(starts)))
    names <- attr(r,"capture.names")
    lengths <- attr(r,"capture.length")
    full <- substring(string[i],r,r+attr(r,"match.length")-1)
    subs <- substring(string[i],starts,starts+lengths-1)
    m <- matrix(c(full,subs),ncol=length(names)+1)
    colnames(m) <- c("",names)
    m
  })
}