Nothing
#' @import biocViews
#' @importFrom RBGL transitive.closure
.computeBiocViewsTransitiveClosure = function() {
data("biocViewsVocab", package = "biocViews")
return(RBGL::transitive.closure(biocViewsVocab))
}
#' @importFrom graph edges nodes
.biocChildToParent = function() {
biocViewsTC = .computeBiocViewsTransitiveClosure()
biocViewsTCEdges = edges(biocViewsTC)
df = data.frame(child = unlist(biocViewsTCEdges, use.names = FALSE),
parent = rep(names(biocViewsTCEdges), sapply(biocViewsTCEdges, length)),
stringsAsFactors = FALSE)
# Remove BiocViews, as this is simply the root
df = df[df$parent!='BiocViews',]
return(split(df$parent, df$child))
}
#' Get full Bioconductor software package listing, with details
#'
#' The BiocViews-generated \code{VIEWS} file is available
#' for Bioconductor release and devel repositories. It
#' contains quite a bit more information from the
#' package \code{DESCRIPTION} files than the \code{PACKAGES}
#' file. In particular, it contains \code{biocViews} annotations
#' and URLs for vignettes and developer URLs.
#'
#' Since packages are annotated with the most specific
#' views, the default functionality here is to add parent terms
#' for all views for each package. For example, in the bioCsoft
#' repository, all packages will have at least "Software" added
#' to their biocViews. If one wants to stick to only the most
#' specific terms, set \code{addBiocViewParents} to \code{FALSE}.
#'
#' @param version The requested Bioconductor version. Will
#' default to use the BiocManager defaults (ie., \code{version()}).
#' @param repo The requested biooconductor repository. The default will be the
#' Bioconductor software repository: BioCsoft. Available repos include:
#' "BioCsoft", "BioCann", "BioCexp", "BioCworkflows", and "CRAN". Note
#' that not all repos are available for all versions, particularly older
#' versions (but who would use those, right?).
#'
#' @param addBiocViewParents logical(), whether to add all biocViews
#' parents to biocViews annotations.
#'
#' @return An object of class \code{tbl_df}.
#'
#' @importFrom BiocManager repositories version
#' @importFrom stringr str_split str_replace_all str_remove_all str_squish
#' @importFrom tibble as_tibble
#'
#' @examples
#' bpkgl = biocPkgList()
#' bpkgl
#' unlist(bpkgl[1,'Depends'])
#'
#' # Get a list of all packages that
#' # import "GEOquery"
#' library(dplyr)
#' bpkgl %>%
#' filter(Package=='GEOquery') %>%
#' pull(c('importsMe'))
#'
#' @export
biocPkgList = function(version = BiocManager::version(), repo='BioCsoft',
addBiocViewParents = TRUE) {
if(!is.logical(addBiocViewParents) ||
!(addBiocViewParents %in% c(TRUE,FALSE)))
stop('addBiocViewParents must be a single logical.')
if (!is.character(repo))
stop('repo must be a character vector.')
# nasty hack here, but BiocManager::repositories throws errors for
# repositories not matching the current R version--counterproductive
# for this function.
repos <- BiocManager:::.repositories(site_repository=NA, version = version)
mask <- !repo %in% names(repos)
if (any(mask))
warning(sprintf("repo %s not found in known repositories. Valid values for repo can be shown by running names(BiocManager::repositories()).",
paste(repo[mask], collapse=", ")))
if (all(mask))
stop("No repo was found in known repositories.")
repo <- repo[!mask]
ret <- lapply(as.list(repo),
function(r) {
if (r != "CRAN") {
viewsFileUrl = paste(repos[r], 'VIEWS', sep = '/')
con = url(viewsFileUrl)
ret <- suppressWarnings(try(read.dcf(con), silent = TRUE))
if(methods::is(ret,'try-error'))
stop('No dcf file found for repo ',r,'(version ',version,') at ',
viewsFileUrl,'. Check that the version is available.')
ret = as.data.frame(ret, stringsAsFactors = FALSE)
close(con)
} else {
cranFileUrl <- sprintf("%s/web/packages/packages.rds",
repos[r])
con <- url(cranFileUrl)
ret <- as.data.frame(readRDS(gzcon(con)),
stringsAsFactors=FALSE)
close(con)
rownames(ret) <- NULL
## to increase the overlap of the information from
## CRAN with Bioconductor, replace CRAN column names
## to Bioconductor column names for 'reverse imports',
## 'depends' and 'suggests'
biocrevnames <- c(importsMe="Reverse imports",
dependsOnMe="Reverse depends",
suggestsMe="Reverse suggests")
mt <- match(biocrevnames, colnames(ret))
mask <- !is.na(mt)
colnames(ret)[mt[mask]] <- names(biocrevnames)[mask]
ret$biocViews<- NA_character_
}
# convert comma-delimited text columns into
# list columns
commaCols = c('Depends', 'Suggests', 'dependsOnMe', 'Imports', 'importsMe',
'Enhances', 'vignettes', 'vignetteTitles', 'suggestsMe',
'Maintainer', 'biocViews', 'Archs', 'linksToMe', 'LinkingTo',
'Rfiles')
commaCols = intersect(commaCols, colnames(ret))
for(commaCol in commaCols) {
ret[[commaCol]] = str_split(ret[[commaCol]],'\\s?,\\s?')
}
if(r != "CRAN" && addBiocViewParents) {
child2parentMap = .biocChildToParent()
tmp = lapply(ret$biocViews, function(views) {
return(unique(c(unlist(child2parentMap[views]),views)))
})
ret$biocViews = tmp
}
ret[["Author"]] = ret[["Author"]] %>%
str_replace_all("\n", " ") %>%
str_remove_all("\\[.*?\\]") %>%
str_remove_all("<.*?>") %>%
str_remove_all("\\(.*?\\)") %>%
str_squish() %>%
str_replace_all("\\w* contributions ?\\w*", ", ") %>%
str_replace_all("\\sand\\s", ", ") %>%
str_replace_all(",\\s+,", ",") %>%
str_replace_all(",+", ",")
ret[["Author"]] = lapply(
str_split(ret[["Author"]], ","),
str_squish
)
ret
})
## build a table with the common columns provided by the different
## repositories. if no common columns exist, then prompt an error
commonCols <- Reduce(intersect, lapply(ret, colnames))
if (length(commonCols) == 0)
stop(sprintf("No common columns between repos in %s",
paste(repo, collapse=", ")))
ret <- lapply(ret, function(x, cnames) x[, cnames], commonCols)
ret <- do.call("rbind", ret)
ret = as_tibble(ret)
class(ret) = c("biocPkgList", class(ret))
ret
}
stripVersionString = function(s) sub('\\s?\\(.*\\)\\s?','',s)
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.