lassik
/
encoding-declaration
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Skip non-ASCII bytes altogether 

This permits UTF-16, with or without byte order mark
This commit is contained in:
Lassi Kortela 2019-05-12 17:21:48 +03:00
parent 0a9c678e64
commit aac4a4bdec
1 changed files with 12 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
encoding-reader.scm
View File

@ -10,17 +10,19 @@
(lambda (port) (read-bytevector 1000 port))))) (lambda (port) (read-bytevector 1000 port)))))
(if (eof-object? bytes) (make-bytevector 0) bytes))) (if (eof-object? bytes) (make-bytevector 0) bytes)))
(i 0)) (i 0))
(define (peek-next-ascii-byte)
(if (not (< i (bytevector-length bytes)))
(eof-object)
(let ((next-byte (bytevector-u8-ref bytes i)))
(if (<= 1 next-byte 126)
next-byte
(begin (set! i (+ i 1))
(peek-next-ascii-byte))))))
(define (read-char? k) (define (read-char? k)
(let* ((remain? (< i (bytevector-length bytes))) (let* ((next-byte (peek-next-ascii-byte))
(next-byte (if remain? (next-char (if (eof-object? next-byte)
(bytevector-u8-ref bytes i) next-byte
(eof-object))) (integer->char next-byte)))
(next-char (cond ((eof-object? next-byte)
next-byte)
((<= 1 next-byte 126)
(integer->char next-byte))
(else
next-byte)))
(consume? (cond ((procedure? k) (k next-char)) (consume? (cond ((procedure? k) (k next-char))
((char? k) (eqv? k next-char)) ((char? k) (eqv? k next-char))
(else #f)))) (else #f))))