view src/testdir/test_regexp_utf8.vim @ 18478:94223687df0e

Added tag v8.1.2233 for changeset e93cab5d0f0f27fad7882f1f412927df055b090d
author Bram Moolenaar <Bram@vim.org>
date Tue, 29 Oct 2019 04:30:05 +0100
parents f4ce361bb1e5
children bb87c5c1e29b
line wrap: on
line source

" Tests for regexp in utf8 encoding

func s:equivalence_test()
  let str = "AÀÁÂÃÄÅĀĂĄǍǞǠẢ BḂḆ CÇĆĈĊČ DĎĐḊḎḐ EÈÉÊËĒĔĖĘĚẺẼ FḞ GĜĞĠĢǤǦǴḠ HĤĦḢḦḨ IÌÍÎÏĨĪĬĮİǏỈ JĴ KĶǨḰḴ LĹĻĽĿŁḺ MḾṀ NÑŃŅŇṄṈ OÒÓÔÕÖØŌŎŐƠǑǪǬỎ PṔṖ Q RŔŖŘṘṞ SŚŜŞŠṠ TŢŤŦṪṮ UÙÚÛÜŨŪŬŮŰŲƯǓỦ VṼ WŴẀẂẄẆ XẊẌ YÝŶŸẎỲỶỸ ZŹŻŽƵẐẔ aàáâãäåāăąǎǟǡả bḃḇ cçćĉċč dďđḋḏḑ eèéêëēĕėęěẻẽ fḟ gĝğġģǥǧǵḡ hĥħḣḧḩẖ iìíîïĩīĭįǐỉ jĵǰ kķǩḱḵ lĺļľŀłḻ mḿṁ nñńņňʼnṅṉ oòóôõöøōŏőơǒǫǭỏ pṕṗ q rŕŗřṙṟ sśŝşšṡ tţťŧṫṯẗ uùúûüũūŭůűųưǔủ vṽ wŵẁẃẅẇẘ xẋẍ yýÿŷẏẙỳỷỹ zźżžƶẑẕ"
  let groups = split(str)
  for group1 in groups
      for c in split(group1, '\zs')
	" next statement confirms that equivalence class matches every
	" character in group
        call assert_match('^[[=' . c . '=]]*$', group1)
        for group2 in groups
          if group2 != group1
	    " next statement converts that equivalence class doesn't match
	    " character in any other group
            call assert_equal(-1, match(group2, '[[=' . c . '=]]'))
          endif
        endfor
      endfor
  endfor
endfunc

func Test_equivalence_re1()
  set re=1
  call s:equivalence_test()
  set re=0
endfunc

func Test_equivalence_re2()
  set re=2
  call s:equivalence_test()
  set re=0
endfunc

func s:classes_test()
  set isprint=@,161-255
  call assert_equal('Motörhead', matchstr('Motörhead', '[[:print:]]\+'))

  let alnumchars = ''
  let alphachars = ''
  let backspacechar = ''
  let blankchars = ''
  let cntrlchars = ''
  let digitchars = ''
  let escapechar = ''
  let graphchars = ''
  let lowerchars = ''
  let printchars = ''
  let punctchars = ''
  let returnchar = ''
  let spacechars = ''
  let tabchar = ''
  let upperchars = ''
  let xdigitchars = ''
  let identchars = ''
  let identchars1 = ''
  let kwordchars = ''
  let kwordchars1 = ''
  let fnamechars = ''
  let fnamechars1 = ''
  let i = 1
  while i <= 255
    let c = nr2char(i)
    if c =~ '[[:alpha:]]'
      let alphachars .= c
    endif
    if c =~ '[[:alnum:]]'
      let alnumchars .= c
    endif
    if c =~ '[[:backspace:]]'
      let backspacechar .= c
    endif
    if c =~ '[[:blank:]]'
      let blankchars .= c
    endif
    if c =~ '[[:cntrl:]]'
      let cntrlchars .= c
    endif
    if c =~ '[[:digit:]]'
      let digitchars .= c
    endif
    if c =~ '[[:escape:]]'
      let escapechar .= c
    endif
    if c =~ '[[:graph:]]'
      let graphchars .= c
    endif
    if c =~ '[[:lower:]]'
      let lowerchars .= c
    endif
    if c =~ '[[:print:]]'
      let printchars .= c
    endif
    if c =~ '[[:punct:]]'
      let punctchars .= c
    endif
    if c =~ '[[:return:]]'
      let returnchar .= c
    endif
    if c =~ '[[:space:]]'
      let spacechars .= c
    endif
    if c =~ '[[:tab:]]'
      let tabchar .= c
    endif
    if c =~ '[[:upper:]]'
      let upperchars .= c
    endif
    if c =~ '[[:xdigit:]]'
      let xdigitchars .= c
    endif
    if c =~ '[[:ident:]]'
      let identchars .= c
    endif
    if c =~ '\i'
      let identchars1 .= c
    endif
    if c =~ '[[:keyword:]]'
      let kwordchars .= c
    endif
    if c =~ '\k'
      let kwordchars1 .= c
    endif
    if c =~ '[[:fname:]]'
      let fnamechars .= c
    endif
    if c =~ '\f'
      let fnamechars1 .= c
    endif
    let i += 1
  endwhile

  call assert_equal('ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz', alphachars)
  call assert_equal('0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz', alnumchars)
  call assert_equal("\b", backspacechar)
  call assert_equal("\t ", blankchars)
  call assert_equal("\x01\x02\x03\x04\x05\x06\x07\b\t\n\x0b\f\r\x0e\x0f\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1a\e\x1c\x1d\x1e\x1f\x7f", cntrlchars)
  call assert_equal("0123456789", digitchars)
  call assert_equal("\<Esc>", escapechar)
  call assert_equal('!"#$%&''()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~', graphchars)
  call assert_equal('abcdefghijklmnopqrstuvwxyzµßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ', lowerchars)
  call assert_equal(' !"#$%&''()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ', printchars)
  call assert_equal('!"#$%&''()*+,-./:;<=>?@[\]^_`{|}~', punctchars)
  call assert_equal('ABCDEFGHIJKLMNOPQRSTUVWXYZÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ', upperchars)
  call assert_equal("\r", returnchar)
  call assert_equal("\t\n\x0b\f\r ", spacechars)
  call assert_equal("\t", tabchar)
  call assert_equal('0123456789ABCDEFabcdef', xdigitchars)

  if has('win32')
    let identchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz€‚ƒ„…†‡ˆ‰Š‹ŒŽ‘’“”•–—˜™š›œžŸ ¡¢£¤¥¦§µÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ'
    let kwordchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyzµÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  elseif has('ebcdic')
    let identchars_ok = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz€ŒŽœž¬®µº¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
    let kwordchars_ok = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz€ŒŽœž¬®µº¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  else
    let identchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyzµÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
    let kwordchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyzµÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  endif

  if has('win32')
    let fnamechars_ok = '!#$%+,-./0123456789:=@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]_abcdefghijklmnopqrstuvwxyz{}~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  elseif has('amiga')
    let fnamechars_ok = '$+,-./0123456789:ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  elseif has('vms')
    let fnamechars_ok = '#$%+,-./0123456789:;<>ABCDEFGHIJKLMNOPQRSTUVWXYZ[]_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  elseif has('ebcdic')
    let fnamechars_ok = '#$%+,-./=ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  else
    let fnamechars_ok = '#$%+,-./0123456789=ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
  endif

  call assert_equal(identchars_ok, identchars)
  call assert_equal(kwordchars_ok, kwordchars)
  call assert_equal(fnamechars_ok, fnamechars)

  call assert_equal(identchars1, identchars)
  call assert_equal(kwordchars1, kwordchars)
  call assert_equal(fnamechars1, fnamechars)
endfunc

func Test_classes_re1()
  set re=1
  call s:classes_test()
  set re=0
endfunc

func Test_classes_re2()
  set re=2
  call s:classes_test()
  set re=0
endfunc

func Test_reversed_range()
  for re in range(0, 2)
    exe 'set re=' . re
    call assert_fails('call match("abc def", "[c-a]")', 'E944:')
  endfor
  set re=0
endfunc

func Test_large_class()
  set re=1
  call assert_fails('call match("abc def", "[\u3000-\u4000]")', 'E945:')
  set re=2
  call assert_equal(0, 'abc def' =~# '[\u3000-\u4000]')
  call assert_equal(1, "\u3042" =~# '[\u3000-\u4000]')
  set re=0
endfunc

func Test_optmatch_toolong()
  set re=1
  " Can only handle about 8000 characters.
  let pat = '\\%[' .. repeat('x', 9000) .. ']'
  call assert_fails('call match("abc def", "' .. pat .. '")', 'E339:')
  set re=0
endfunc