Note

For anyone who is using Elm 0.19 and blocked by this library, consider using hecrj/html-parser for now. It seems still WIP but passed the same test cases. I think that means the most difficult part of the HTML spec should be already covered (e.g. <li> does not need a closing tag). Also, it uses the official parser elm/parser which I was planning to use in the newer version. So I think contributing to hecrj/html-parser is the fastest path for everyone getting happy :)

elm-html-parser

Parse HTML in Elm! (DEMO)

Parse

import HtmlParser as HtmlParser exposing (..)

parse "text" == [ Text "text" ]

parse "<h1>Hello<br>World</h1>"
  == [ Element "h1" [] [ Text "Hello", Element "br" [] [], Text "World" ] ]

parse """<a href="http://example.com">Example</a>"""
  == [ Element "a" [("href", "http://example.com")] [ Text "Example" ] ]

Query

import HtmlParser exposing (..)
import HtmlParser.Util exposing (..)

table = """
  <table border=0 cellpadding=0 cellspacing=0 width=216 style='border-collapse:
   collapse;width:162pt'>
  <!--StartFragment-->
   <col width=72 span=3 style='width:54pt'>
   <tr height=18 style='height:13.5pt'>
    <td height=18 align=right width=72 style='height:13.5pt;width:54pt'>1</td>
    <td align=right width=72 style='width:54pt'>2</td>
    <td align=right width=72 style='width:54pt'>3</td>
   </tr>
   <tr height=18 style='height:13.5pt'>
    <td height=18 class=xl69 align=right style='height:13.5pt'>2</td>
    <td class=xl66 align=right>3</td>
    <td align=right>4</td>
   </tr>
  <!--EndFragment-->
  </table>
"""

( parse table
  |> getElementsByTagName "tr"
  |> mapElements
    (\_ _ innerTr ->
      innerTr
        |> mapElements (\_ _ innerTd -> textContent innerTd)
        |> String.join "\t"
        |> String.trim
    )
  |> String.join "\n"
) == "1\t2\t3\n2\t3\t4"

LICENSE

BSD3