haskal
/
utop
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

UTop_lexer: support combined glyph for more kind of token, Symbol, Ident, Blanks, Literals...

This commit is contained in:
ZAN DoYe 2019-05-12 00:25:06 +08:00
parent dd1f834985
commit 99f28d9f17
1 changed files with 18 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
src/lib/uTop_lexer.mll
View File

@ -173,9 +173,24 @@ and symbol syntax context idx acc = parse
{ let loc = lexeme_loc idx lexbuf in { let loc = lexeme_loc idx lexbuf in
let tok = Symbol (lexeme lexbuf) in let tok = Symbol (lexeme lexbuf) in
tokens syntax context loc.idx2 ((tok, loc) :: acc) lexbuf } tokens syntax context loc.idx2 ((tok, loc) :: acc) lexbuf }
| uchar | uchar as uchar
{ let loc = mkloc idx (idx + 1) (lexeme_start lexbuf) (lexeme_end lexbuf) in { let uChar= Zed_utf8.unsafe_extract uchar 0 in
tokens syntax context loc.idx2 ((Error, loc) :: acc) lexbuf } if Zed_char.is_combining_mark uChar then
let tok, loc= List.hd acc
and tl= List.tl acc in
let tok= match tok with
| Symbol str-> Symbol (str ^ (lexeme lexbuf))
| Lident str-> Lident (str ^ (lexeme lexbuf))
| Uident str-> Uident (str ^ (lexeme lexbuf))
| Constant str-> Constant (str ^ (lexeme lexbuf))
| _-> tok
in
let loc= { loc with ofs2= lexeme_end lexbuf } in
tokens syntax context loc.idx2 ((tok, loc) :: tl) lexbuf
else
let loc = mkloc idx (idx + 1) (lexeme_start lexbuf) (lexeme_end lexbuf) in
tokens syntax context loc.idx2 ((Error, loc) :: acc) lexbuf
}
and camlp4_toplevel syntax context idx acc = parse and camlp4_toplevel syntax context idx acc = parse
| '<' (':' ident)? ('@' lident)? '<' | '<' (':' ident)? ('@' lident)? '<'