-" Last Change: 2010 Aug 12
+" Last Change: 2010 Sep 09
"
" Additional contributors:
"
@@ -13,6 +13,276 @@
let s:cpo_sav = &cpo
set cpo-=C
-func! tohtml#Convert2HTML(line1, line2)
+" Automatically find charsets from all encodings supported natively by Vim. With
+" the 8bit- and 2byte- prefixes, Vim can actually support more encodings than
+" this. Let the user specify these however since they won't be supported on
+" every system. TODO: how? g:html_charsets and g:html_encodings?
+"
+" Note, not all of Vim's supported encodings have a charset to use.
+"
+" Names in this list are from:
+" http://www.iana.org/assignments/character-sets
+" g:tohtml#encoding_to_charset: {{{
+let g:tohtml#encoding_to_charset = {
+ \ 'latin1' : 'ISO-8859-1',
+ \ 'iso-8859-2' : 'ISO-8859-2',
+ \ 'iso-8859-3' : 'ISO-8859-3',
+ \ 'iso-8859-4' : 'ISO-8859-4',
+ \ 'iso-8859-5' : 'ISO-8859-5',
+ \ 'iso-8859-6' : 'ISO-8859-6',
+ \ 'iso-8859-7' : 'ISO-8859-7',
+ \ 'iso-8859-8' : 'ISO-8859-8',
+ \ 'iso-8859-9' : 'ISO-8859-9',
+ \ 'iso-8859-10' : 'ISO-8859-10',
+ \ 'iso-8859-13' : 'ISO-8859-13',
+ \ 'iso-8859-14' : 'ISO-8859-14',
+ \ 'iso-8859-15' : 'ISO-8859-15',
+ \ 'koi8-r' : 'KOI8-R',
+ \ 'koi8-u' : 'KOI8-U',
+ \ 'macroman' : 'macintosh',
+ \ 'cp437' : 'IBM437',
+ \ 'cp775' : 'IBM775',
+ \ 'cp850' : 'IBM850',
+ \ 'cp852' : 'IBM852',
+ \ 'cp855' : 'IBM855',
+ \ 'cp857' : 'IBM857',
+ \ 'cp860' : 'IBM860',
+ \ 'cp861' : 'IBM861',
+ \ 'cp862' : 'IBM862',
+ \ 'cp863' : 'IBM863',
+ \ 'cp865' : 'IBM865',
+ \ 'cp866' : 'IBM866',
+ \ 'cp869' : 'IBM869',
+ \ 'cp874' : 'IBM-Thai',
+ \ 'cp1250' : 'windows-1250',
+ \ 'cp1251' : 'windows-1251',
+ \ 'cp1253' : 'windows-1253',
+ \ 'cp1254' : 'windows-1254',
+ \ 'cp1255' : 'windows-1255',
+ \ 'cp1256' : 'windows-1256',
+ \ 'cp1257' : 'windows-1257',
+ \ 'cp1258' : 'windows-1258',
+ \ 'euc-jp' : 'EUC-JP',
+ \ 'sjis' : 'Shift_JIS',
+ \ 'cp932' : 'Shift_JIS',
+ \ 'cp949' : 'KS_C_5601-1987',
+ \ 'euc-kr' : 'EUC-KR',
+ \ 'cp936' : 'GBK',
+ \ 'euc-cn' : 'GB_2312-80',
+ \ 'big5' : 'Big5',
+ \ 'cp950' : 'Big5',
+ \ 'utf-8' : 'UTF-8',
+ \ 'ucs-2' : 'ISO-10646-UCS-2',
+ \ 'ucs-2le' : 'ISO-10646-UCS-2',
+ \ 'utf-16' : 'UTF-16',
+ \ 'utf-16le' : 'UTF-16LE',
+ \ 'ucs-4' : 'UTF-32',
+ \ 'ucs-4le' : 'UTF-32LE',
+ \ }
+lockvar g:tohtml#encoding_to_charset
+" }}}
+
+" Only automatically find encodings supported natively by Vim, let the user
+" specify the encoding if it's not natively supported. This function is only
+" used when the user specifies the charset, they better know what they are
+" doing!
+"
+" Names in this list are from:
+" http://www.iana.org/assignments/character-sets
+" g:tohtml#charset_to_encoding: {{{
+let g:tohtml#charset_to_encoding = {
+ \ 'ISO_8859-1:1987' : 'latin1',
+ \ 'iso-ir-100' : 'latin1',
+ \ 'ISO_8859-1' : 'latin1',
+ \ 'ISO-8859-1' : 'latin1',
+ \ 'latin1' : 'latin1',
+ \ 'l1' : 'latin1',
+ \ 'IBM819' : 'latin1',
+ \ 'CP819' : 'latin1',
+ \ 'csISOLatin1' : 'latin1',
+ \ 'ISO_8859-2:1987' : 'iso-8859-2',
+ \ 'iso-ir-101' : 'iso-8859-2',
+ \ 'ISO_8859-2' : 'iso-8859-2',
+ \ 'ISO-8859-2' : 'iso-8859-2',
+ \ 'latin2' : 'iso-8859-2',
+ \ 'l2' : 'iso-8859-2',
+ \ 'csISOLatin2' : 'iso-8859-2',
+ \ 'ISO_8859-3:1988' : 'iso-8859-3',
+ \ 'iso-ir-109' : 'iso-8859-3',
+ \ 'ISO_8859-3' : 'iso-8859-3',
+ \ 'ISO-8859-3' : 'iso-8859-3',
+ \ 'latin3' : 'iso-8859-3',
+ \ 'l3' : 'iso-8859-3',
+ \ 'csISOLatin3' : 'iso-8859-3',
+ \ 'ISO_8859-4:1988' : 'iso-8859-4',
+ \ 'iso-ir-110' : 'iso-8859-4',
+ \ 'ISO_8859-4' : 'iso-8859-4',
+ \ 'ISO-8859-4' : 'iso-8859-4',
+ \ 'latin4' : 'iso-8859-4',
+ \ 'l4' : 'iso-8859-4',
+ \ 'csISOLatin4' : 'iso-8859-4',
+ \ 'ISO_8859-5:1988' : 'iso-8859-5',
+ \ 'iso-ir-144' : 'iso-8859-5',
+ \ 'ISO_8859-5' : 'iso-8859-5',
+ \ 'ISO-8859-5' : 'iso-8859-5',
+ \ 'cyrillic' : 'iso-8859-5',
+ \ 'csISOLatinCyrillic' : 'iso-8859-5',
+ \ 'ISO_8859-6:1987' : 'iso-8859-6',
+ \ 'iso-ir-127' : 'iso-8859-6',
+ \ 'ISO_8859-6' : 'iso-8859-6',
+ \ 'ISO-8859-6' : 'iso-8859-6',
+ \ 'ECMA-114' : 'iso-8859-6',
+ \ 'ASMO-708' : 'iso-8859-6',
+ \ 'arabic' : 'iso-8859-6',
+ \ 'csISOLatinArabic' : 'iso-8859-6',
+ \ 'ISO_8859-7:1987' : 'iso-8859-7',
+ \ 'iso-ir-126' : 'iso-8859-7',
+ \ 'ISO_8859-7' : 'iso-8859-7',
+ \ 'ISO-8859-7' : 'iso-8859-7',
+ \ 'ELOT_928' : 'iso-8859-7',
+ \ 'ECMA-118' : 'iso-8859-7',
+ \ 'greek' : 'iso-8859-7',
+ \ 'greek8' : 'iso-8859-7',
+ \ 'csISOLatinGreek' : 'iso-8859-7',
+ \ 'ISO_8859-8:1988' : 'iso-8859-8',
+ \ 'iso-ir-138' : 'iso-8859-8',
+ \ 'ISO_8859-8' : 'iso-8859-8',
+ \ 'ISO-8859-8' : 'iso-8859-8',
+ \ 'hebrew' : 'iso-8859-8',
+ \ 'csISOLatinHebrew' : 'iso-8859-8',
+ \ 'ISO_8859-9:1989' : 'iso-8859-9',
+ \ 'iso-ir-148' : 'iso-8859-9',
+ \ 'ISO_8859-9' : 'iso-8859-9',
+ \ 'ISO-8859-9' : 'iso-8859-9',
+ \ 'latin5' : 'iso-8859-9',
+ \ 'l5' : 'iso-8859-9',
+ \ 'csISOLatin5' : 'iso-8859-9',
+ \ 'ISO-8859-10' : 'iso-8859-10',
+ \ 'iso-ir-157' : 'iso-8859-10',
+ \ 'l6' : 'iso-8859-10',
+ \ 'ISO_8859-10:1992' : 'iso-8859-10',
+ \ 'csISOLatin6' : 'iso-8859-10',
+ \ 'latin6' : 'iso-8859-10',
+ \ 'ISO-8859-13' : 'iso-8859-13',
+ \ 'ISO-8859-14' : 'iso-8859-14',
+ \ 'iso-ir-199' : 'iso-8859-14',
+ \ 'ISO_8859-14:1998' : 'iso-8859-14',
+ \ 'ISO_8859-14' : 'iso-8859-14',
+ \ 'latin8' : 'iso-8859-14',
+ \ 'iso-celtic' : 'iso-8859-14',
+ \ 'l8' : 'iso-8859-14',
+ \ 'ISO-8859-15' : 'iso-8859-15',
+ \ 'ISO_8859-15' : 'iso-8859-15',
+ \ 'Latin-9' : 'iso-8859-15',
+ \ 'KOI8-R' : 'koi8-r',
+ \ 'csKOI8R' : 'koi8-r',
+ \ 'KOI8-U' : 'koi8-u',
+ \ 'macintosh' : 'macroman',
+ \ 'mac' : 'macroman',
+ \ 'csMacintosh' : 'macroman',
+ \ 'IBM437' : 'cp437',
+ \ 'cp437' : 'cp437',
+ \ '437' : 'cp437',
+ \ 'csPC8CodePage437' : 'cp437',
+ \ 'IBM775' : 'cp775',
+ \ 'cp775' : 'cp775',
+ \ 'csPC775Baltic' : 'cp775',
+ \ 'IBM850' : 'cp850',
+ \ 'cp850' : 'cp850',
+ \ '850' : 'cp850',
+ \ 'csPC850Multilingual' : 'cp850',
+ \ 'IBM852' : 'cp852',
+ \ 'cp852' : 'cp852',
+ \ '852' : 'cp852',
+ \ 'csPCp852' : 'cp852',
+ \ 'IBM855' : 'cp855',
+ \ 'cp855' : 'cp855',
+ \ '855' : 'cp855',
+ \ 'csIBM855' : 'cp855',
+ \ 'IBM857' : 'cp857',
+ \ 'cp857' : 'cp857',
+ \ '857' : 'cp857',
+ \ 'csIBM857' : 'cp857',
+ \ 'IBM860' : 'cp860',
+ \ 'cp860' : 'cp860',
+ \ '860' : 'cp860',
+ \ 'csIBM860' : 'cp860',
+ \ 'IBM861' : 'cp861',
+ \ 'cp861' : 'cp861',
+ \ '861' : 'cp861',
+ \ 'cp-is' : 'cp861',
+ \ 'csIBM861' : 'cp861',
+ \ 'IBM862' : 'cp862',
+ \ 'cp862' : 'cp862',
+ \ '862' : 'cp862',
+ \ 'csPC862LatinHebrew' : 'cp862',
+ \ 'IBM863' : 'cp863',
+ \ 'cp863' : 'cp863',
+ \ '863' : 'cp863',
+ \ 'csIBM863' : 'cp863',
+ \ 'IBM865' : 'cp865',
+ \ 'cp865' : 'cp865',
+ \ '865' : 'cp865',
+ \ 'csIBM865' : 'cp865',
+ \ 'IBM866' : 'cp866',
+ \ 'cp866' : 'cp866',
+ \ '866' : 'cp866',
+ \ 'csIBM866' : 'cp866',
+ \ 'IBM869' : 'cp869',
+ \ 'cp869' : 'cp869',
+ \ '869' : 'cp869',
+ \ 'cp-gr' : 'cp869',
+ \ 'csIBM869' : 'cp869',
+ \ 'windows-1250' : 'cp1250',
+ \ 'windows-1251' : 'cp1251',
+ \ 'windows-1253' : 'cp1253',
+ \ 'windows-1254' : 'cp1254',
+ \ 'windows-1255' : 'cp1255',
+ \ 'windows-1256' : 'cp1256',
+ \ 'windows-1257' : 'cp1257',
+ \ 'windows-1258' : 'cp1258',
+ \ 'Extended_UNIX_Code_Packed_Format_for_Japanese' : 'euc-jp',
+ \ 'csEUCPkdFmtJapanese' : 'euc-jp',
+ \ 'EUC-JP' : 'euc-jp',
+ \ 'Shift_JIS' : 'sjis',
+ \ 'MS_Kanji' : 'sjis',
+ \ 'sjis' : 'sjis',
+ \ 'csShiftJIS' : 'sjis',
+ \ 'IBM-Thai' : 'cp874',
+ \ 'csIBMThai' : 'cp874',
+ \ 'KS_C_5601-1987' : 'cp949',
+ \ 'iso-ir-149' : 'cp949',
+ \ 'KS_C_5601-1989' : 'cp949',
+ \ 'KSC_5601' : 'cp949',
+ \ 'korean' : 'cp949',
+ \ 'csKSC56011987' : 'cp949',
+ \ 'EUC-KR' : 'euc-kr',
+ \ 'csEUCKR' : 'euc-kr',
+ \ 'GBK' : 'cp936',
+ \ 'CP936' : 'cp936',
+ \ 'MS936' : 'cp936',
+ \ 'windows-936' : 'cp936',
+ \ 'GB_2312-80' : 'euc-cn',
+ \ 'iso-ir-58' : 'euc-cn',
+ \ 'chinese' : 'euc-cn',
+ \ 'csISO58GB231280' : 'euc-cn',
+ \ 'Big5' : 'big5',
+ \ 'csBig5' : 'big5',
+ \ 'UTF-8' : 'utf-8',
+ \ 'ISO-10646-UCS-2' : 'ucs-2',
+ \ 'csUnicode' : 'ucs-2',
+ \ 'UTF-16' : 'utf-16',
+ \ 'UTF-16BE' : 'utf-16',
+ \ 'UTF-16LE' : 'utf-16le',
+ \ 'UTF-32' : 'ucs-4',
+ \ 'UTF-32BE' : 'ucs-4',
+ \ 'UTF-32LE' : 'ucs-4le',
+ \ 'ISO-10646-UCS-4' : 'ucs-4',
+ \ 'csUCS4' : 'ucs-4'
+ \ }
+lockvar g:tohtml#charset_to_encoding
+"}}}
+
+func! tohtml#Convert2HTML(line1, line2) "{{{
let s:settings = tohtml#GetUserSettings()
@@ -17,6 +287,6 @@
let s:settings = tohtml#GetUserSettings()
- if !&diff || s:settings.diff_one_file
+ if !&diff || s:settings.diff_one_file "{{{
if a:line2 >= a:line1
let g:html_start_line = a:line1
let g:html_end_line = a:line2
@@ -24,11 +294,11 @@
let g:html_start_line = a:line2
let g:html_end_line = a:line1
endif
- runtime syntax/2html.vim
- else
+ runtime syntax/2html.vim "}}}
+ else "{{{
let win_list = []
let buf_list = []
windo | if &diff | call add(win_list, winbufnr(0)) | endif
let s:settings.whole_filler = 1
let g:html_diff_win_num = 0
for window in win_list
@@ -29,7 +299,8 @@
let win_list = []
let buf_list = []
windo | if &diff | call add(win_list, winbufnr(0)) | endif
let s:settings.whole_filler = 1
let g:html_diff_win_num = 0
for window in win_list
+ " switch to the next buffer to convert
exe ":" . bufwinnr(window) . "wincmd w"
@@ -35,4 +306,18 @@
exe ":" . bufwinnr(window) . "wincmd w"
+
+ " figure out whether current charset and encoding will work, if not
+ " default to UTF-8
+ if !exists('g:html_use_encoding') &&
+ \ (&l:fileencoding!='' && &l:fileencoding!=s:settings.vim_encoding ||
+ \ &l:fileencoding=='' && &encoding!=s:settings.vim_encoding)
+ echohl WarningMsg
+ echomsg "TOhtml: mismatched file encodings in Diff buffers, using UTF-8"
+ echohl None
+ let s:settings.vim_encoding = 'utf-8'
+ let s:settings.encoding = 'UTF-8'
+ endif
+
+ " set up for diff-mode conversion
let g:html_start_line = 1
let g:html_end_line = line('$')
let g:html_diff_win_num += 1
@@ -36,4 +321,6 @@
let g:html_start_line = 1
let g:html_end_line = line('$')
let g:html_diff_win_num += 1
+
+ " convert this file
runtime syntax/2html.vim
@@ -39,5 +326,7 @@
runtime syntax/2html.vim
+
+ " remember the HTML buffer for later combination
call add(buf_list, bufnr('%'))
endfor
unlet g:html_diff_win_num
call tohtml#Diff2HTML(win_list, buf_list)
@@ -40,9 +329,9 @@
call add(buf_list, bufnr('%'))
endfor
unlet g:html_diff_win_num
call tohtml#Diff2HTML(win_list, buf_list)
- endif
+ endif "}}}
unlet g:html_start_line
unlet g:html_end_line
unlet s:settings
@@ -45,6 +334,6 @@
unlet g:html_start_line
unlet g:html_end_line
unlet s:settings
-endfunc
+endfunc "}}}
@@ -50,5 +339,5 @@
-func! tohtml#Diff2HTML(win_list, buf_list)
+func! tohtml#Diff2HTML(win_list, buf_list) "{{{
let xml_line = ""
let tag_close = '>'
@@ -87,7 +376,7 @@
call add(html, '')
" include encoding as close to the top as possible, but only if not already
- " contained in XML information (to avoid haggling over content type)
+ " contained in XML information
if s:settings.encoding != "" && !s:settings.use_xhtml
call add(html, "')
1
@@ -151,5 +441,5 @@
let insert_index += 1
endif
endfor
- endif
+ endif " }}}
@@ -155,4 +445,6 @@
+ " everything new will get added before the diff styles so diff highlight
+ " properly overrides normal highlight
if diff_style_start != 0
let insert_index = diff_style_start
endif
@@ -203,6 +495,20 @@
" just in case some user autocmd creates content in the new buffer, make sure
" it is empty before proceeding
%d
+
+ " set the fileencoding to match the charset we'll be using
+ let &l:fileencoding=s:settings.vim_encoding
+
+ " According to http://www.w3.org/TR/html4/charset.html#doc-char-set, the byte
+ " order mark is highly recommend on the web when using multibyte encodings. But,
+ " it is not a good idea to include it on UTF-8 files. Otherwise, let Vim
+ " determine when it is actually inserted.
+ if s:settings.vim_encoding == 'utf-8'
+ setlocal nobomb
+ else
+ setlocal bomb
+ endif
+
call append(0, html)
if len(style) > 0
@@ -210,7 +516,7 @@
let style_start = search('^')-1
" Insert javascript to toggle matching folds open and closed in all windows,
- " if dynamic folding is active.
+ " if dynamic folding is active. {{{
if s:settings.dynamic_folds
call append(style_start, [
\ ""
\ ])
- endif
+ endif "}}}
" Insert styles from all the generated html documents and additional styles
" for the table-based layout of the side-by-side diff. The diff should take
" up the full browser window (but not more), and be static in size,
" horizontally scrollable when the lines are too long. Otherwise, the diff
@@ -238,9 +544,9 @@
" Insert styles from all the generated html documents and additional styles
" for the table-based layout of the side-by-side diff. The diff should take
" up the full browser window (but not more), and be static in size,
" horizontally scrollable when the lines are too long. Otherwise, the diff
- " is pretty useless for really long lines.
+ " is pretty useless for really long lines. {{{
if s:settings.use_css
call append(style_start,
\ [''
- \ ])
- endif
+ \])
+ endif "}}}
endif
let &paste = s:old_paste
let &magic = s:old_magic
@@ -257,8 +563,8 @@
endif
let &paste = s:old_paste
let &magic = s:old_magic
-endfunc
+endfunc "}}}
" Gets a single user option and sets it in the passed-in Dict, or gives it the
" default value if the option doesn't actually exist.
@@ -262,9 +568,9 @@
" Gets a single user option and sets it in the passed-in Dict, or gives it the
" default value if the option doesn't actually exist.
-func! tohtml#GetOption(settings, option, default)
+func! tohtml#GetOption(settings, option, default) "{{{
if exists('g:html_'.a:option)
let a:settings[a:option] = g:html_{a:option}
else
let a:settings[a:option] = a:default
endif
@@ -266,10 +572,10 @@
if exists('g:html_'.a:option)
let a:settings[a:option] = g:html_{a:option}
else
let a:settings[a:option] = a:default
endif
-endfunc
+endfunc "}}}
" returns a Dict containing the values of all user options for 2html, including
" default values for those not given an explicit value by the user. Discards the
" html_ prefix of the option for nicer looking code.
@@ -272,8 +578,8 @@
" returns a Dict containing the values of all user options for 2html, including
" default values for those not given an explicit value by the user. Discards the
" html_ prefix of the option for nicer looking code.
-func! tohtml#GetUserSettings()
+func! tohtml#GetUserSettings() "{{{
if exists('s:settings')
" just restore the known options if we've already retrieved them
return s:settings
@@ -289,7 +595,7 @@
let g:html_use_xhtml = g:use_xhtml
endif
- " get current option settings with appropriate defaults
+ " get current option settings with appropriate defaults {{{
call tohtml#GetOption(user_settings, 'no_progress', !has("statusline") )
call tohtml#GetOption(user_settings, 'diff_one_file', 0 )
call tohtml#GetOption(user_settings, 'number_lines', &number )
@@ -302,4 +608,5 @@
call tohtml#GetOption(user_settings, 'no_pre', 0 )
call tohtml#GetOption(user_settings, 'whole_filler', 0 )
call tohtml#GetOption(user_settings, 'use_xhtml', 0 )
+ " }}}
@@ -305,5 +612,5 @@
- " override those settings that need it
+ " override those settings that need it {{{
" hover opening implies dynamic folding
if user_settings.hover_unfold
@@ -330,5 +637,5 @@
" aren't allowed inside a block
if !user_settings.use_css
let user_settings.no_pre = 1
- endif
+ endif "}}}
@@ -334,4 +641,5 @@
- " Figure out proper MIME charset from the 'encoding' option.
- if exists("g:html_use_encoding")
+ if exists("g:html_use_encoding") "{{{
+ " user specified the desired MIME charset, figure out proper
+ " 'fileencoding' from it or warn the user if we cannot
let user_settings.encoding = g:html_use_encoding
@@ -337,2 +645,10 @@
let user_settings.encoding = g:html_use_encoding
+ let user_settings.vim_encoding = tohtml#EncodingFromCharset(g:html_use_encoding)
+ if user_settings.vim_encoding == ''
+ echohl WarningMsg
+ echomsg "TOhtml: file encoding for"
+ \ g:html_use_encoding
+ \ "unknown, please set 'fileencoding'"
+ echohl None
+ endif
else
@@ -338,5 +654,6 @@
else
- let vim_encoding = &encoding
- if vim_encoding =~ '^8bit\|^2byte'
- let vim_encoding = substitute(vim_encoding, '^8bit-\|^2byte-', '', '')
+ " Figure out proper MIME charset from 'fileencoding' if possible
+ if &l:fileencoding != ''
+ let user_settings.vim_encoding = &l:fileencoding
+ call tohtml#CharsetFromEncoding(user_settings)
endif
@@ -342,21 +659,12 @@
endif
- if vim_encoding == 'latin1'
- let user_settings.encoding = 'iso-8859-1'
- elseif vim_encoding =~ "^cp12"
- let user_settings.encoding = substitute(vim_encoding, 'cp', 'windows-', '')
- elseif vim_encoding == 'sjis' || vim_encoding == 'cp932'
- let user_settings.encoding = 'Shift_JIS'
- elseif vim_encoding == 'big5' || vim_encoding == 'cp950'
- let user_settings.encoding = "Big5"
- elseif vim_encoding == 'euc-cn'
- let user_settings.encoding = 'GB_2312-80'
- elseif vim_encoding == 'euc-tw'
- let user_settings.encoding = ""
- elseif vim_encoding =~ '^euc\|^iso\|^koi'
- let user_settings.encoding = substitute(vim_encoding, '.*', '\U\0', '')
- elseif vim_encoding == 'cp949'
- let user_settings.encoding = 'KS_C_5601-1987'
- elseif vim_encoding == 'cp936'
- let user_settings.encoding = 'GBK'
- elseif vim_encoding =~ '^ucs\|^utf'
+
+ " else from 'encoding' if possible
+ if &l:fileencoding == '' || user_settings.encoding == ''
+ let user_settings.vim_encoding = &encoding
+ call tohtml#CharsetFromEncoding(user_settings)
+ endif
+
+ " else default to UTF-8 and warn user
+ if user_settings.encoding == ''
+ let user_settings.vim_encoding = 'utf-8'
let user_settings.encoding = 'UTF-8'
@@ -362,4 +670,5 @@
let user_settings.encoding = 'UTF-8'
- else
- let user_settings.encoding = ""
+ echohl WarningMsg
+ echomsg "TOhtml: couldn't determine MIME charset, using UTF-8"
+ echohl None
endif
@@ -365,7 +674,7 @@
endif
- endif
+ endif "}}}
" TODO: font
return user_settings
endif
@@ -367,11 +676,45 @@
" TODO: font
return user_settings
endif
-endfunc
+endfunc "}}}
+
+" get the proper HTML charset name from a Vim encoding option.
+function! tohtml#CharsetFromEncoding(settings) "{{{
+ let l:vim_encoding = a:settings.vim_encoding
+ if exists('g:html_charset_override') && has_key(g:html_charset_override, l:vim_encoding)
+ let a:settings.encoding = g:html_charset_override[l:vim_encoding]
+ " since the user has overridden the default charset for this encoding, we
+ " may need to convert to a different encoding to make the two match
+ let a:settings.vim_encoding = tohtml#EncodingFromCharset(a:settings.encoding)
+ else
+ if l:vim_encoding =~ '^8bit\|^2byte'
+ " 8bit- and 2byte- prefixes are to indicate encodings available on the
+ " system that Vim will convert with iconv(), look up just the encoding name,
+ " not Vim's prefix.
+ let l:vim_encoding = substitute(l:vim_encoding, '^8bit-\|^2byte-', '', '')
+ endif
+ if has_key(g:tohtml#encoding_to_charset, l:vim_encoding)
+ let a:settings.encoding = g:tohtml#encoding_to_charset[l:vim_encoding]
+ else
+ let a:settings.encoding = ""
+ endif
+ endif
+endfun "}}}
+
+" Get the proper Vim encoding option setting from an HTML charset name.
+function! tohtml#EncodingFromCharset(encoding) "{{{
+ if exists('g:html_encoding_override') && has_key(g:html_encoding_override, a:encoding)
+ return g:html_encoding_override[a:encoding]
+ elseif has_key(g:tohtml#charset_to_encoding, a:encoding)
+ return g:tohtml#charset_to_encoding[a:encoding]
+ else
+ return ""
+ endif
+endfun "}}}
let &cpo = s:cpo_sav
unlet s:cpo_sav
" Make sure any patches will probably use consistent indent
@@ -373,6 +716,6 @@
let &cpo = s:cpo_sav
unlet s:cpo_sav
" Make sure any patches will probably use consistent indent
-" vim: ts=8 sw=2 sts=2 noet
+" vim: ts=8 sw=2 sts=2 noet fdm=marker
diff --git a/runtime/doc/syntax.txt b/runtime/doc/syntax.txt
--- a/runtime/doc/syntax.txt
+++ b/runtime/doc/syntax.txt
@@ -475,14 +475,32 @@
This will use
at the end of each line and use " " for repeated
spaces.
-The current value of 'encoding' is used to specify the charset of the HTML
-file. This only works for those values of 'encoding' that have an equivalent
-HTML charset name. To overrule this set g:html_use_encoding to the name of
-the charset to be used: >
- :let g:html_use_encoding = "foobar"
-To omit the line that specifies the charset, set g:html_use_encoding to an
-empty string: >
+For diff mode on a single file (with g:html_diff_one_file) a sequence of more
+than 3 filler lines is displayed as three lines with the middle line
+mentioning the total number of inserted lines. If you prefer to see all the
+inserted lines as with the side-by-side diff, use: >
+ :let g:html_whole_filler = 1
+And to go back to displaying up to three lines again: >
+ :unlet g:html_whole_filler
+<
+TOhtml uses the current value of 'fileencoding' if set, or 'encoding' if not,
+to determine the charset and 'fileencoding' of the HTML file. By default,
+this only works for encodings supported natively by Vim (mentioned
+specifically by name in |encoding-names|) which also appear in the IANA
+registry: http://www.iana.org/assignments/character-sets. Note that not all
+names in the IANA registry are actually widely supported, but TOhtml will
+choose them anyway unless you override with a different charset.
+
+To overrule all automatic charset detection, set g:html_use_encoding to the
+name of the charset to be used. TOhtml will try to determine the appropriate
+'fileencoding' setting from the charset, but you may need to set it manually
+if TOhtml cannot determine the encoding. It is probably a very good idea to
+set this variable to something widely supported, like UTF-8, for anything you
+will be serving up on a webserver: >
+ :let g:html_use_encoding = "UTF-8"
+You can also use this option to omit the line that specifies the charset
+entirely, by setting g:html_use_encoding to an empty string: >
:let g:html_use_encoding = ""
To go back to the automatic mechanism, delete the g:html_use_encoding
variable: >
:unlet g:html_use_encoding
@@ -485,5 +503,14 @@
:let g:html_use_encoding = ""
To go back to the automatic mechanism, delete the g:html_use_encoding
variable: >
:unlet g:html_use_encoding
+
+If you specify a charset with g:html_use_encoding for which TOhtml cannot
+automatically detect the corresponding 'fileencoding' setting, you can use
+g:html_encoding_override to allow TOhtml to detect the correct encoding.
+This is a dictionary of charset-encoding pairs that will replace existing
+pairs automatically detected by TOhtml, or supplement with new pairs. For
+example, to allow TOhtml to detect the HTML charset "windows-1252" properly as
+the encoding "8bit-cp1252", use: >
+ :let g:html_encoding_override = {'windows-1252': '8bit-cp1252'}
<
@@ -489,13 +516,13 @@
<
-For diff mode a sequence of more than 3 filler lines is displayed as three
-lines with the middle line mentioning the total number of inserted lines. If
-you prefer to see all the inserted lines use: >
- :let g:html_whole_filler = 1
-And to go back to displaying up to three lines again: >
- :unlet g:html_whole_filler
-<
+The g:html_charset_override is similar, it allows TOhtml to detect the charset
+for any 'fileencoding' or 'encoding' which is not detected automatically. You
+can also use it to override specific existing encoding-charset pairs. For
+example, many user agents do not handle UTF-32 very well, so you could use the
+following to convert such documents to UTF-8 instead: >
+ :let g:html_charset_override = {'ucs-4': 'UTF-8'}
+
*convert-to-XML* *convert-to-XHTML*
An alternative is to have the script generate XHTML (XML compliant HTML). To
do this set the "html_use_xhtml" variable: >
:let g:html_use_xhtml = 1
@@ -497,10 +524,10 @@
*convert-to-XML* *convert-to-XHTML*
An alternative is to have the script generate XHTML (XML compliant HTML). To
do this set the "html_use_xhtml" variable: >
:let g:html_use_xhtml = 1
-Any of these options can be enabled or disabled by setting them explicitly to
-the desired value, or restored to their default by removing the variable using
-|:unlet|.
+Any of the on/off options can be enabled or disabled by setting them
+explicitly to the desired value, or restored to their default by removing the
+variable using |:unlet|.
Remarks:
@@ -505,7 +532,4 @@
Remarks:
-- This only works in a version with GUI support. If the GUI is not actually
- running (possible for X11) it still works, but not very well (the colors
- may be wrong).
- Some truly ancient browsers may not show the background colors.
- From most browsers you can also print the file (in color)!
@@ -510,5 +534,8 @@
- Some truly ancient browsers may not show the background colors.
- From most browsers you can also print the file (in color)!
+- This version of TOhtml may work with older versions of Vim, but some
+ features such as conceal support will not function, and the colors may be
+ incorrect for an old Vim without GUI support compiled in.
Here is an example how to run the script over all .c and .h files from a
Unix shell: >
diff --git a/runtime/plugin/tohtml.vim b/runtime/plugin/tohtml.vim
--- a/runtime/plugin/tohtml.vim
+++ b/runtime/plugin/tohtml.vim
@@ -1,8 +1,8 @@
" Vim plugin for converting a syntax highlighted file to HTML.
" Maintainer: Ben Fritz
-" Last Change: 2010 Aug 12
+" Last Change: 2010 Sep 10
"
" The core of the code is in $VIMRUNTIME/autoload/tohtml.vim and
" $VIMRUNTIME/syntax/2html.vim
"
" TODO:
@@ -4,8 +4,11 @@
"
" The core of the code is in $VIMRUNTIME/autoload/tohtml.vim and
" $VIMRUNTIME/syntax/2html.vim
"
" TODO:
+" * Allow user to specify more encoding/charset pairs to auto-detect, or
+" override some of the built-in ones.
+" * Increase number of supported charset-vim encoding pairs if possible
" * Restore open/closed folds and cursor position after processing each file
" with option not to restore for speed increase
" * Add extra meta info (generation time, etc.)
@@ -13,7 +16,33 @@
" * Implementation detail: add threshold for writing the lines to the html
" buffer before we're done (5000 or so lines should do it)
" * TODO comments for code cleanup scattered throughout
+"
+"
+" Changelog:
+" 7.3_v7b1 (this beta ): Remove use of setwinvar() function which cannot be
+" called in restricted mode (Andy Spencer). Use
+" 'fencoding' instead of 'encoding' to determine by
+" charset, and make sure the 'fenc' of the generated
+" file matches its indicated charset. Add charsets for
+" all of Vim's natively supported encodings.
+" 7.3_v6 (0d3f0e3d289b): Really fix bug with 'nowrapscan', 'magic' and other
+" user settings interfering with diff mode generation,
+" trailing whitespace (e.g. line number column) when
+" using html_no_pre, and bugs when using
+" html_hover_unfold.
+" 7.3_v5 ( unreleased ): Fix bug with 'nowrapscan' and also with out-of-sync
+" folds in diff mode when first line was folded.
+" 7.3_v4 (7e008c174cc3): Bugfixes, especially for xhtml markup, and diff mode.
+" 7.3_v3 (a29075150aee): Refactor option handling and make html_use_css
+" default to true when not set to anything. Use strict
+" doctypes where possible. Rename use_xhtml option to
+" html_use_xhtml for consistency. Use .xhtml extension
+" when using this option. Add meta tag for settings.
+" 7.3_v2 (80229a724a11): Fix syntax highlighting in diff mode to use both the
+" diff colors and the normal syntax colors
+" 7.3_v1 (e7751177126b): Add conceal support and meta tags in output
+" Pre-v1 baseline: Mercurial changeset 3c9324c0800e
if exists('g:loaded_2html_plugin')
finish
endif
@@ -16,8 +45,8 @@
if exists('g:loaded_2html_plugin')
finish
endif
-let g:loaded_2html_plugin = 'vim7.3_v6'
+let g:loaded_2html_plugin = 'vim7.3_v7b1'
" Define the :TOhtml command when:
" - 'compatible' is not set
diff --git a/runtime/syntax/2html.vim b/runtime/syntax/2html.vim
--- a/runtime/syntax/2html.vim
+++ b/runtime/syntax/2html.vim
@@ -1,6 +1,6 @@
" Vim syntax support file
" Maintainer: Ben Fritz
-" Last Change: 2010 Aug 12
+" Last Change: 2010 Sep 04
"
" Additional contributors:
"
@@ -265,6 +265,19 @@
let s:old_magic = &magic
set magic
+" set the fileencoding to match the charset we'll be using
+let &l:fileencoding=s:settings.vim_encoding
+
+" According to http://www.w3.org/TR/html4/charset.html#doc-char-set, the byte
+" order mark is highly recommend on the web when using multibyte encodings. But,
+" it is not a good idea to include it on UTF-8 files. Otherwise, let Vim
+" determine when it is actually inserted.
+if s:settings.vim_encoding == 'utf-8'
+ setlocal nobomb
+else
+ setlocal bomb
+endif
+
let s:lines = []
if s:settings.use_xhtml
@@ -1071,6 +1084,7 @@
let @/ = s:old_search
let &more = s:old_more
exe s:orgwin . "wincmd w"
+let &l:stl = s:origwin_stl
let &l:et = s:old_et
let &l:scrollbind = s:old_bind
exe s:newwin . "wincmd w"
@@ -1074,6 +1088,7 @@
let &l:et = s:old_et
let &l:scrollbind = s:old_bind
exe s:newwin . "wincmd w"
+let &l:stl = s:newwin_stl
exec 'resize' s:old_winheight
let &l:winfixheight = s:old_winfixheight
@@ -1077,8 +1092,6 @@
exec 'resize' s:old_winheight
let &l:winfixheight = s:old_winfixheight
-call setwinvar(s:orgwin,'&stl', s:origwin_stl)
-call setwinvar(s:newwin,'&stl', s:newwin_stl)
let &ls=s:ls
" Save a little bit of memory (worth doing?)