Skip to contents

Find archive within an html-type vector

get_archive.Rd

Find all archive of given type within a vector of html text.

Usage

get_archive(
  origin,
  extension = c("zip", "7z"),
  name = NULL,
  date = NULL,
  version = NULL,
  directory = FALSE,
  local_origin = NULL
)

Arguments

origin

character, the url address where to find archives.

extension

character, vector of acceptable types of archives to be downloaded.

name

character, vector of acceptable names fo archives to be downloaded. See details

date

character, something like a date that should be used as a filter. See details

version

character, something like a version that should be used as a filter.

directory

logical, should directories be found instead of archives. See details.

local_origin

character, the local url address to add to relative links. When set to `NULL`` (default), origin is used.

Value

A character vector of all archives or directory found in origin matching with given constraints.

Details

First, all links are retrieved with get_link_from_html().

If directory is TRUE, only finishing pattern "\" are kept, and this pattern is erased. extension is not used in this case.

When extension is givens, it may contain different possibilities. All are matched at the end of archives' names.

At this point, a check is made to see if remaining links are relative (not starting by "http" or "https") or absolute. All relative links are completed with local_origin. All links are named with base::basename(), those names are used for further selections.

name may contain different possibilities. It is matched at the beginning of archives' names.

version may contain different possibilities. Il will be matched anywhere in archives' names. This may lead an empty character as result.

date may contain either "last", and so anything that can be considered as a date in archives' names ("\ against and the max is taken. If nothing matches, all archives' names are kept. date may also contain anything admissible for create_date(). If so, anything that can be considered as a date in archives' names ("\ "\ date pertain to create_date(date) are kept, possibly nothing.

Examples