Tidy or "Pretty Print" HTML/XHTML Documents

Pass in HTML content as either plain or raw text or parsed objects (either with the XML or xml2 packages) or as an httr response object along with an options list that specifies how the content will be tidied and get back tidied content of the same object type as passed in to the function.

# S3 method for response
tidy_html(content, options = list(TidyXhtmlOut = TRUE),
  verbose = FALSE)

tidy_html(content, options = list(TidyXhtmlOut = TRUE), verbose = FALSE)

# S3 method for default
tidy_html(content, options = list(TidyXhtmlOut = TRUE),
  verbose = FALSE)

# S3 method for character
tidy_html(content, options = list(TidyXhtmlOut = TRUE),
  verbose = FALSE)

# S3 method for raw
tidy_html(content, options = list(TidyXhtmlOut = TRUE),
  verbose = FALSE)

# S3 method for xml_document
tidy_html(content, options = list(TidyXhtmlOut = TRUE),
  verbose = FALSE)

# S3 method for HTMLInternalDocument
tidy_html(content, options = list(TidyXhtmlOut
  = TRUE), verbose = FALSE)

# S3 method for connection
tidy_html(content, options = list(TidyXhtmlOut = TRUE),
  verbose = FALSE)

Arguments

content
accepts a character vector, raw vector or parsed content from the xml2 or XML packages.
options
named list of options
verbose
output document errors? (default: FALSE)

Value

Tidied HTML/XHTML content. The object type will be the same as that of the input type except when it is a connection, then a character vector will be returned.

Details

The default option TixyXhtmlOut will convert the input content to XHTML.

Currently supported options:

  • Ones taking a logical value: TidyAltText, TidyBodyOnly, TidyBreakBeforeBR, TidyCoerceEndTags, TidyDropEmptyElems, TidyDropEmptyParas, TidyFixBackslash, TidyFixComments, TidyGDocClean, TidyHideComments, TidyHtmlOut, TidyIndentContent, TidyJoinClasses, TidyJoinStyles, TidyLogicalEmphasis, TidyMakeBare, TidyMakeClean, TidyMark, TidyOmitOptionalTags, TidyReplaceColor, TidyUpperCaseAttrs, TidyUpperCaseTags, TidyWord2000, TidyXhtmlOut
  • Ones taking a character value: TidyDoctype, TidyInlineTags, TidyBlockTags, TidyEmptyTags, TidyPreTags
  • Ones taking an integer value: TidyIndentSpaces, TidyTabSize, TidyWrapLen

File https://github.com/hrbrmstr/htmltidy/issues if there are other libtidy options you'd like supported.

It is likely that the most used options will be:

  • TidyXhtmlOut (logical),
  • TidyHtmlOut (logical) and
  • TidyDocType which should be one of "omit", "html5", "auto", "strict" or "loose".

You can clean up Microsoft Word (2000) and Google Docs HTML via logical settings for TidyWord2000 and TidyGDocClean, respectively.

It may also be advantageous to remove all comments with TidyHideComments.

Note

If document parsing errors are severe enough, tidy_html() will not be able to clean the document and will display the errors (this output can be captured with sink() or capture.output()) along with a warning and return a "best effort" cleaned version of the document.

References

http://api.html-tidy.org/tidy/quickref_5.1.25.html & https://github.com/htacg/tidy-html5/blob/master/include/tidyenum.h for definitions of the options supported above and https://www.w3.org/People/Raggett/tidy/ for an explanation of what "tidy" HTML is and some canonical examples of what it can do.

Examples

opts <- list(
  TidyDocType="html5",
  TidyMakeClean=TRUE,
  TidyHideComments=TRUE,
  TidyIndentContent=TRUE,
  TidyWrapLen=200
)

txt <- paste0(
  c("<html><head><style>p { color: red; }</style><body><!-- ===== body ====== -->",
"<p>Test</p></body><!--Default Zone --> <!--Default Zone End--></html>"),
  collapse="")

cat(tidy_html(txt, option=opts))
#> <!DOCTYPE html>
#> <html>
#>   <head>
#>     <meta name="generator" content="HTML Tidy for HTML5 for R version 5.0.0">
#>     <style>
#>     p { color: red; }
#>     </style>
#>     <title></title>
#>   </head>
#>   <body>
#>     <p>
#>       Test
#>     </p>
#>   </body>
#> </html>
#> 

library(httr)
res <- GET("http://rud.is/test/untidy.html")

# look at the original, un-tidy source
cat(content(res, as="text", encoding="UTF-8"))
#> <head>
#> <style>
#> body { font-family: sans-serif; }
#> </style>
#> </head>
#> <body>
#> <b>This is <b>some <i>really </i> poorly formatted HTML</b>
#> 
#> as is this <span id="sp">portion<div>
#> 

# see the tidied version
cat(tidy_html(content(res, as="text", encoding="UTF-8"),
              list(TidyDocType="html5", TidyWrapLen=200)))
#> <!DOCTYPE html>
#> <html>
#> <head>
#> <meta name="generator" content="HTML Tidy for HTML5 for R version 5.0.0">
#> <style>
#> body { font-family: sans-serif; }
#> </style>
#> <title></title>
#> </head>
#> <body>
#> <b>This is some <i>really</i> poorly formatted HTML as is this <span id="sp">portion</span></b>
#> <div><span id="sp"></span></div>
#> </body>
#> </html>
#> 

# but, you could also just do:
cat(tidy_html(url("http://rud.is/test/untidy.html")))
#> <!DOCTYPE html>
#> <html xmlns="http://www.w3.org/1999/xhtml">
#> <head>
#> <meta name="generator" content=
#> "HTML Tidy for HTML5 for R version 5.0.0" />
#> <style>
#> <![CDATA[
#> body { font-family: sans-serif; }
#> ]]>
#> </style>
#> <title></title>
#> </head>
#> <body>
#> <b>This is some <i>really</i> poorly formatted HTMLas is this
#> <span id="sp">portion</span></b>
#> <div><span id="sp"></span></div>
#> </body>
#> </html>
#>