-*todo.txt* For Vim version 7.0aa. Last change: 2005 Jun 06
+*todo.txt* For Vim version 7.0aa. Last change: 2005 Jun 07
VIM REFERENCE MANUAL by Bram Moolenaar
*known-bugs*
-------------------- Known bugs and current work -----------------------
-Patch in if_cscope.c also in 6.3? (Froloff)
- Sergey says it's OK.
-
Add extra list of file locations. Can be used with:
:ltag list of matching tags, like :tselect
- Add SPELLCHECKER, with support for many languages.
- Use "engspchk" from Charles Campbell for ideas (commands, rare words).
- - Should quickly return if there is no word with the character.
- Use array with flags, indicating if there is a word starting with this
- byte. Quickly skip bytes where no word can start.
- Spell checking code todo's:
- - Also allow replacing a word list. Need some mechanism to tell
- wether a spell file adds or replaces one found later in
- 'runtimepath'.
- Is "-" to be considered a word character? "last-minute".
No, in Dutch it can be added optionally. Then make English
dictionaries consistent.
- - Implement user and project word lists. Commands to add words and to
- mark words as wrong.
- - In .aff use RAR to define affix name for rare word.
- - In .aff use HUH to define affix name for keep-case word.
- 's morgens/= does not match 'S morgens
- Implement compound words?
- When @Spell and @NoSpell are both used only do spell checking for
@Spell items, not where they both appear. Useful for Perl pod.
author: Kevin Hendricks <kevin.hendricks@sympatico.ca>
- More complicated: Regions with different languages? E.g. comments in
English, strings in German (po file).
- - Commands required:
- add word to private dict: wrong and OK (in popup menu for evim)
- :spell good <word> zg
- :spell wrong <word> zw
- - Update option window for 'verbosefile', 'spell' and 'spelllang'.
+ - Update option window for 'verbosefile', 'spell', 'spellfile' and
+ 'spelllang'.
- Distribution: Need wordlists for many languages; "language pack"
Put them on the ftp site, ready to download. Include README for
copyrights.
cluster" but change the contains list directly for matching syntax
items.
- Install spell files with src/main.aap.
- Alternatives using ispell or aspell:
+ - Alternate Dutch word list at www.nederlandsewoorden.nl (use script to
+ obtain).
- REFACTORING: The main() function is very long. Move parts to separate
functions, especially loops. Ideas from Walter Briscoe (2003 Apr 3, 2004
--- /dev/null
+" Vim indent file
+" Language: Pascal
+" Maintainer: Neil Carter <n.carter@swansea.ac.uk>
+" Created: 2004 Jul 13
+" Last Change: 2005 Jun 07
+" TODO: Reduce indentation on line after a statement that flowed across
+" two lines (e.g. parameter list closed on second line). Also, increase
+" indent of a becomes-statement that flows onto second line.
+
+" Only load this indent file when no other was loaded.
+if exists("b:did_indent")
+ finish
+endif
+let b:did_indent = 1
+
+setlocal indentexpr=GetPascalIndent(v:lnum)
+" Appending an & to an option sets it to its default value.
+setlocal indentkeys&
+setlocal indentkeys+=~end;,=~const,=~type,=~var,=~begin,=~repeat,=~until,=~for
+setlocal indentkeys+=~program,=~function,=~procedure,=~object,=~private
+setlocal indentkeys+=~record,=~if,=~else,=~case
+
+if exists("*GetPascalIndent")
+ finish
+endif
+
+
+function s:GetPrevLineNum( line_num )
+
+ " Skip over comments and conditional directives
+ let SKIP_LINES = '^\s*\((\*\)\|\(\*\ \)\|\(\*)\)\|\({\$\)'
+
+ let nline = a:line_num
+ while nline > 0
+ let nline = prevnonblank(nline-1)
+ if getline(nline) !~? SKIP_LINES
+ break
+ endif
+ endwhile
+
+" call input( "nline = ".nline )
+
+ return nline
+
+endfunction
+
+
+function! GetPascalIndent( line_num )
+ if a:line_num == 0
+ return 0
+ endif
+
+ " If in the middle of a three-part comment
+ if getline( a:line_num ) =~ '^\s*\*\ '
+ return indent( a:line_num )
+ endif
+
+ " We have to subtract one to start on the line before the current
+ " one. Otherwise, prevnonblank() returns the current line!
+ let prev_line_num = s:GetPrevLineNum( a:line_num )
+ let prev_line = getline( prev_line_num )
+ let indnt = indent( prev_line_num )
+
+ let this_line = getline( a:line_num )
+
+ " At the start of a block, we have to indent the newly-created line
+ " based on the previous line.
+ " =~ means matches a regular expression
+ " a question mark after =~ means ignore case (# means match case)
+ " const, type, var should always appear at the start of a line, but
+ " begin can appear anywhere in the line.
+ " if one of the following keywords appear in the previous line with
+ " nothing before it but optional whitespace, and nothing after it.
+ " Has to be end of line at end to show this is not a routine
+ " parameter list. Otherwise, you'd end up with cascading vars.
+
+ " These words appear alone on a line (apart from whitespace).
+ if prev_line =~ '^\s*\(const\|var\|begin\|repeat\|private\)$'
+ " Place an & before an option to obtain its value.
+ let indnt = indnt + &shiftwidth
+ endif
+
+ " Words preceded by optional whitespace and followed by anything.
+ if prev_line =~ '^\s*\(for\|if\|else\|case\)'
+ " Place an & before an option to obtain its value.
+ let indnt = indnt + &shiftwidth
+ " if this is a multistatement block then we need to align the
+ " begin with the previous line.
+ if this_line =~ '^\s*begin'
+ let indnt = indnt - &shiftwidth
+ endif
+ endif
+ " These words may have text before them on the line (hence the .*).
+ if prev_line =~ '^.*\s*\<\(object\|record\)\>$'
+ let indnt = indnt + &shiftwidth
+ endif
+ " If we have opened a bracket and the contents spills over one line,
+ " then indent one level beyond the bracket's first line. RE = an
+ " opening bracket followed by any amount of anything other than a
+ " closing bracket and then the end-of-line. If we didn't include the
+ " end of line, this RE would match even closed brackets, since it
+ " would match everything up to the closing bracket.
+ " This test isn't clever enough to handle brackets inside strings or
+ " comments.
+ if prev_line =~ '([^*][^)]*$'
+ let indnt = indnt + &shiftwidth
+ endif
+
+ " If we just closed a bracket that started on a previous line, then
+ " unindent.
+ if prev_line =~ '^[^(]*[^*])'
+ let indnt = indnt - &shiftwidth
+ endif
+
+ " At the end of a block, we have to unindent both the current line
+ " (the 'end;' for instance) and the newly-created line.
+ if this_line =~ '^\s*\(end;\|until\|else\)'
+ let indnt = indnt - &shiftwidth
+ endif
+
+ " Keywords that always appear at the start of a line.
+ " Problem is that function and procedure keywords should be indented
+ " if within a class declaration.
+ if this_line =~ '^\s*\<type\|uses\|$IFDEF\|$ENDIF\|procedure\|function\>'
+ let indnt = 0
+ endif
+ if prev_line =~ '^\s*\<type\|uses\>'
+ let indnt = &shiftwidth
+ endif
+
+ " Put conditional compile directives on first column.
+ if this_line =~ '^\s*{\$'
+ let indnt = 0
+ endif
+
+ return indnt
+endfunction
+
+" TODO: end; should align with the previous (begin/record/object/else).
+" "else begin" is the only case where begin does not appear at the start
+" of the line.
+
+" TODO: Don't align with {$IFDEF}
+
+"Example from vb.vim
+" regular expression match, case insensitive
+"if previous_line =~?
+" start of line, zero or more whitespace
+"'^\s*
+" start of word
+"\<
+"
+"\(
+" begin\|
+" \%(
+" \%(
+" private\|public\|friend
+" \)
+" \s\+
+" \)
+" zero or more of the previous atom
+" \=
+" \%(
+" function\|sub\|property
+" \)
+" \|select\|case\|default\|if
+"\>
+" .\{-}\<then\>\s*$\|else\|elseif\|do\|for\|while\|enum\|with
+"\)
+" end of word
+"\>'
+" let ind = ind + &sw
+"endif
projects / vim / commitdiff