diff options
Diffstat (limited to 'src/utils/twtstr.nim')
| -rw-r--r-- | src/utils/twtstr.nim | 37 |
1 files changed, 0 insertions, 37 deletions
diff --git a/src/utils/twtstr.nim b/src/utils/twtstr.nim index b235651f..da9af19a 100644 --- a/src/utils/twtstr.nim +++ b/src/utils/twtstr.nim @@ -7,7 +7,6 @@ import strutils import tables import unicode -import bindings/libunicode import types/opt import utils/charcategory import utils/map @@ -576,42 +575,6 @@ proc passRealloc*(opaque: pointer, p: pointer, size: csize_t): pointer {.cdecl.} = return realloc(p, size) -proc mnormalize*(rs: var seq[Rune], form = UNICODE_NFC) = {.cast(noSideEffect).}: - if rs.len == 0: return - var outbuf: ptr uint32 - let p = cast[ptr uint32](unsafeAddr rs[0]) - let out_len = unicode_normalize(addr outbuf, p, cint(rs.len), form, nil, - passRealloc) - if out_len < 0: - raise newException(Defect, "Unicode normalization failed") - if out_len == 0: - return - rs = cast[seq[Rune]](newSeqUninitialized[uint32](out_len)) - copyMem(addr rs[0], outbuf, out_len * sizeof(uint32)) - dealloc(outbuf) - -#TODO maybe a utf8 normalization procedure? -proc mnormalize*(s: var string) = - if s.isAscii(): - return # no need to normalize ascii - var rs = s.toRunes() - rs.mnormalize() - s = $rs - -func normalize*(rs: seq[Rune], form = UNICODE_NFC): seq[Rune] = {.cast(noSideEffect).}: - if rs.len == 0: return - var outbuf: ptr uint32 - let out_len = unicode_normalize(addr outbuf, - cast[ptr uint32](unsafeAddr rs[0]), - cint(rs.len), form, nil, passRealloc) - if out_len < 0: - raise newException(Defect, "Unicode normalization failed") - if out_len == 0: - return - result = cast[seq[Rune]](newSeqUninitialized[uint32](out_len)) - copyMem(addr result[0], outbuf, out_len * sizeof(uint32)) - dealloc(outbuf) - # https://www.w3.org/TR/xml/#NT-Name const NameStartCharRanges = [ (0xC0, 0xD6), |