a956a81432
In today's html editing, almost nobody writes a "pure" html code (complete, i mean), with body, title, etc... most of people uses CMS's and similar things, which means that the html is just a "piece" of html and not the entire structure, the changes on this branch are meant to ignore those annoying warnings The last line is for structures that includes the "&" character, like in those kind of cases: < a href="http://foo.bar/something.php&value&something&foo" >
102 lines
3.8 KiB
VimL
102 lines
3.8 KiB
VimL
"============================================================================
|
|
"File: tidy.vim
|
|
"Description: Syntax checking plugin for syntastic.vim
|
|
"Maintainer: Martin Grenfell <martin.grenfell at gmail dot com>
|
|
"License: This program is free software. It comes without any warranty,
|
|
" to the extent permitted by applicable law. You can redistribute
|
|
" it and/or modify it under the terms of the Do What The Fuck You
|
|
" Want To Public License, Version 2, as published by Sam Hocevar.
|
|
" See http://sam.zoy.org/wtfpl/COPYING for more details.
|
|
"
|
|
"============================================================================
|
|
" TODO: join this with xhtml.vim for DRY's sake?
|
|
function! s:TidyEncOptByFenc()
|
|
let tidy_opts = {
|
|
\'utf-8' : '-utf8',
|
|
\'ascii' : '-ascii',
|
|
\'latin1' : '-latin1',
|
|
\'iso-2022-jp' : '-iso-2022',
|
|
\'cp1252' : '-win1252',
|
|
\'macroman' : '-mac',
|
|
\'utf-16le' : '-utf16le',
|
|
\'utf-16' : '-utf16',
|
|
\'big5' : '-big5',
|
|
\'sjis' : '-shiftjis',
|
|
\'cp850' : '-ibm858',
|
|
\}
|
|
return get(tidy_opts, &fileencoding, '-utf8')
|
|
endfunction
|
|
|
|
let s:ignore_html_errors = [
|
|
\ "<table> lacks \"summary\" attribute",
|
|
\ "not approved by W3C",
|
|
\ "attribute \"placeholder\"",
|
|
\ "<meta> proprietary attribute \"charset\"",
|
|
\ "<meta> lacks \"content\" attribute",
|
|
\ "inserting \"type\" attribute",
|
|
\ "proprietary attribute \"data-",
|
|
\ "missing <!DOCTYPE> declaration",
|
|
\ "inserting implicit <body>",
|
|
\ "inserting missing 'title' element",
|
|
\ "attribute \"[+",
|
|
\ "unescaped & or unknown entity",
|
|
\ "<input> attribute \"type\" has invalid value \"search\""
|
|
\]
|
|
|
|
function! s:ValidateError(text)
|
|
let valid = 0
|
|
for i in s:ignore_html_errors
|
|
if stridx(a:text, i) != -1
|
|
let valid = 1
|
|
break
|
|
endif
|
|
endfor
|
|
return valid
|
|
endfunction
|
|
|
|
function s:Args()
|
|
let args = s:TidyEncOptByFenc()
|
|
let args .= " --new-blocklevel-tags " . shellescape('section, article, aside, hgroup, header, footer, nav, figure, figcaption')
|
|
let args .= " --new-inline-tags " . shellescape('video, audio, source, embed, mark, progress, meter, time, ruby, rt, rp, canvas, command, details, datalist')
|
|
let args .= " --new-empty-tags " . shellescape('wbr, keygen')
|
|
let args .= " -e"
|
|
endfunction
|
|
|
|
function! SyntaxCheckers_html_tidy_IsAvailable()
|
|
return executable('tidy')
|
|
endfunction
|
|
|
|
function! SyntaxCheckers_html_tidy_GetLocList()
|
|
let makeprg = syntastic#makeprg#build({
|
|
\ 'exe': 'tidy',
|
|
\ 'args': s:Args(),
|
|
\ 'tail': '2>&1',
|
|
\ 'subchecker': 'tidy' })
|
|
let errorformat='%Wline %l column %c - Warning: %m,%Eline %l column %c - Error: %m,%-G%.%#,%-G%.%#'
|
|
|
|
let loclist = SyntasticMake({ 'makeprg': makeprg, 'errorformat': errorformat })
|
|
|
|
" process loclist since we need to add some info and filter out valid HTML5
|
|
" from the errors
|
|
let n = len(loclist) - 1
|
|
let bufnum = bufnr("")
|
|
while n >= 0
|
|
let i = loclist[n]
|
|
" filter out valid HTML5
|
|
if s:ValidateError(i['text']) == 1
|
|
unlet loclist[n]
|
|
else
|
|
"the file name isnt in the output so stick in the buf num manually
|
|
let i['bufnr'] = bufnum
|
|
endif
|
|
let n -= 1
|
|
endwhile
|
|
|
|
return loclist
|
|
endfunction
|
|
|
|
call g:SyntasticRegistry.CreateAndRegisterChecker({
|
|
\ 'filetype': 'html',
|
|
\ 'name': 'tidy'})
|
|
|