read_xml: Read HTML or XML.

Description Usage Arguments Value Setting the "user agent" header Examples

View source: R/xml_parse.R

Description

Read HTML or XML.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
read_xml(x, encoding = "", ..., as_html = FALSE, options = "NOBLANKS")

read_html(x, encoding = "", ..., options = c("RECOVER", "NOERROR", "NOBLANKS"))

## S3 method for class 'character'
read_xml(x, encoding = "", ..., as_html = FALSE, options = "NOBLANKS")

## S3 method for class 'raw'
read_xml(
  x,
  encoding = "",
  base_url = "",
  ...,
  as_html = FALSE,
  options = "NOBLANKS"
)

## S3 method for class 'connection'
read_xml(
  x,
  encoding = "",
  n = 64 * 1024,
  verbose = FALSE,
  ...,
  base_url = "",
  as_html = FALSE,
  options = "NOBLANKS"
)

Arguments

x

A string, a connection, or a raw vector.

A string can be either a path, a url or literal xml. Urls will be converted into connections either using base::url or, if installed, curl::curl. Local paths ending in .gz, .bz2, .xz, .zip will be automatically uncompressed.

If a connection, the complete connection is read into a raw vector before being parsed.

encoding

Specify a default encoding for the document. Unless otherwise specified XML documents are assumed to be in UTF-8 or UTF-16. If the document is not UTF-8/16, and lacks an explicit encoding directive, this allows you to supply a default.

...

Additional arguments passed on to methods.

as_html

Optionally parse an xml file as if it's html.

options

Set parsing options for the libxml2 parser. Zero or more of \Sexpr[results=rd]{xml2:::describe_options(xml2:::xml_parse_options())}

base_url

When loading from a connection, raw vector or literal html/xml, this allows you to specify a base url for the document. Base urls are used to turn relative urls into absolute urls.

n

If file is a connection, the number of bytes to read per iteration. Defaults to 64kb.

verbose

When reading from a slow connection, this prints some output on every iteration so you know its working.

Value

An XML document. HTML is normalised to valid XML - this may not be exactly the same transformation performed by the browser, but it's a reasonable approximation.

Setting the "user agent" header

When performing web scraping tasks it is both good practice — and often required — to set the user agent request header to a specific value. Sometimes this value is assigned to emulate a browser in order to have content render in a certain way (e.g. Mozilla/5.0 (Windows NT 5.1; rv:52.0) Gecko/20100101 Firefox/52.0 to emulate more recent Windows browsers). Most often, this value should be set to provide the web resource owner information on who you are and the intent of your actions like this Google scraping bot user agent identifier: Googlebot/2.1 (+http://www.google.com/bot.html).

You can set the HTTP user agent for URL-based requests using httr::set_config() and httr::user_agent():

httr::set_config(httr::user_agent("me@example.com; +https://example.com/info.html"))

httr::set_config() changes the configuration globally, httr::with_config() can be used to change configuration temporarily.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
# Literal xml/html is useful for small examples
read_xml("<foo><bar /></foo>")
read_html("<html><title>Hi<title></html>")
read_html("<html><title>Hi")

# From a local path
read_html(system.file("extdata", "r-project.html", package = "xml2"))

## Not run: 
# From a url
cd <- read_xml(xml2_example("cd_catalog.xml"))
me <- read_html("http://had.co.nz")

## End(Not run)

xml2 documentation built on Dec. 11, 2021, 9:26 a.m.