Coerce bigBed columns based on autoSql types (#16)

Co-authored-by: Claude Opus 4.5 <[email protected]>

Author: jayhesselberth

.Rbuildignore

@@ -11,3 +11,4 @@
 ^CRAN-SUBMISSION$
 ^[\.]?air\.toml$
 ^\.vscode$
+^\.clang-format$

.clang-format

@@ -0,0 +1,6 @@
+BasedOnStyle: Google
+IndentWidth: 2
+ColumnLimit: 100
+AllowShortFunctionsOnASingleLine: Empty
+AllowShortIfStatementsOnASingleLine: false
+AllowShortLoopsOnASingleLine: false

R/read.r

@@ -64,11 +64,16 @@ as_granges <- function(x) {
 
 #' Read data from bigBed files.
 #'
+#' Columns are automatically typed based on the autoSql schema embedded
+#' in the bigBed file. Integer types (`uint`, `int`) become R integers,
+#' floating point types (`float`, `double`) become R doubles, and all
+#' other types (including array types like `int[blockCount]`) remain
+#' as character strings.
+#'
 #' @param bbfile filename for bigBed file
 #' @param chrom read data for specific chromosome
 #' @param start start position for data
 #' @param end end position for data
-#' @param convert convert bigBed values to individual columns
 #'
 #' @return \code{tibble}
 #'
@@ -87,8 +92,7 @@ read_bigbed <- function(
   bbfile,
   chrom = NULL,
   start = NULL,
-  end = NULL,
-  convert = TRUE
+  end = NULL
 ) {
   if (!file.exists(bbfile)) {
     stop("File does not exist: ", bbfile)
@@ -99,42 +103,5 @@ read_bigbed <- function(
   }
 
   res <- read_bigbed_cpp(bbfile, chrom, start, end)
-
-  if (!convert) {
-    return(as_tibble(res))
-  }
-
-  vals <- do.call(rbind, strsplit(res[["value"]], "\t"))
-  # merge chrom, start, end with new values
-  res_new <- cbind(res[, 1:3], vals)
-
-  fnames <- bigbed_sql_fields(bbfile)
-  # drop chrom, start, end
-  fnames <- fnames[-(1:3)]
-
-  colnames(res_new) <- c("chrom", "start", "end", fnames)
-  return(as_tibble(res_new))
-}
-
-#' @examples
-#' bb <- system.file("extdata", "test.bb", package = "cpp11bigwig")
-#' bigbed_sql_fields(bb)
-#'
-#' @noRd
-bigbed_sql_fields <- function(bbfile) {
-  res <- bigbed_sql_cpp(bbfile)
-
-  # parse the autoSql
-  lines <- unlist(strsplit(res, "\n"))
-  fields <- lines[grep(";", lines)]
-
-  unlist(
-    lapply(
-      fields,
-      function(line) {
-        field <- sub("^\\s*\\S+\\s+(\\S+);.*", "\\1", line)
-        field
-      }
-    )
-  )
+  as_tibble(res)
 }

README.md

@@ -53,9 +53,9 @@ bb <- system.file("extdata", "test.bb", package = "cpp11bigwig")
 read_bigbed(bb)
 #> # A tibble: 3 × 12
 #>   chrom  start    end name  score strand thickStart thickEnd reserved blockCount
-#>   <chr>  <int>  <int> <chr> <chr> <chr>  <chr>      <chr>    <chr>    <chr>     
-#> 1 chr1  4.80e6 4.84e6 test… 1     +      4797973    4836816  1        9         
-#> 2 chr10 4.85e6 4.88e6 diff… 1     +      4848118    4880877  1        6         
-#> 3 chr20 5.07e6 5.15e6 negs… 1     -      5073253    5152630  1        14        
+#>   <chr>  <int>  <int> <chr> <int> <chr>       <int>    <int>    <int>      <int>
+#> 1 chr1  4.80e6 4.84e6 test…     1 +         4797973  4836816        1          9
+#> 2 chr10 4.85e6 4.88e6 diff…     1 +         4848118  4880877        1          6
+#> 3 chr20 5.07e6 5.15e6 negs…     1 -         5073253  5152630        1         14
 #> # ℹ 2 more variables: blockSizes <chr>, chromStarts <chr>
 ```