Copyright | (c) 2008, 2009 Tom Harper, (c) 2009, 2010 Bryan O'Sullivan, (c) 2009 Duncan Coutts |
---|---|
License | BSD-style |
Maintainer | bos@serpentine.com |
Stability | experimental |
Portability | GHC |
Safe Haskell | None |
Language | Haskell98 |
A module containing private Text
internals. This exposes the
Text
representation and low level construction functions.
Modules which extend the Text
system may need to use this module.
You should not use this module unless you are determined to monkey with the internals, as the functions here do just about nothing to preserve data invariants. You have been warned!
Types
Internally, the Text
type is represented as an array of Word16
UTF-16 code units. The offset and length fields in the constructor
are in these units, not units of Char
.
Invariants that all functions must maintain:
- Since the
Text
type uses UTF-16 internally, it cannot represent characters in the reserved surrogate code point range U+D800 to U+DFFF. To maintain this invariant, thesafe
function mapsChar
values in this range to the replacement character (U+FFFD, '�'). - A leading (or "high") surrogate code unit (0xD800–0xDBFF) must always be followed by a trailing (or "low") surrogate code unit (0xDC00-0xDFFF). A trailing surrogate code unit must always be preceded by a leading surrogate code unit.
A space efficient, packed, unboxed Unicode text type.
Eq Text | |
Data Text | This instance preserves data abstraction at the cost of inefficiency. We omit reflection services for the sake of data abstraction. This instance was created by copying the updated behavior of
The original discussion is archived here: could we get a Data instance for Data.Text.Text? The followup discussion that changed the behavior of |
Ord Text | |
Read Text | |
Show Text | |
IsString Text | |
Monoid Text | |
NFData Text | |
Typeable * Text |
Construction
text :: Array -> Int -> Int -> Text Source
Construct a Text
without invisibly pinning its byte array in
memory if its length has dwindled to zero.
Safety
Map a Char
to a Text
-safe value.
UTF-16 surrogate code points are not included in the set of Unicode
scalar values, but are unfortunately admitted as valid Char
values by Haskell. They cannot be represented in a Text
. This
function remaps those code points to the Unicode replacement
character (U+FFFD, '�'), and leaves other code points
unchanged.
Code that must be here for accessibility
Utilities
firstf :: (a -> c) -> Maybe (a, b) -> Maybe (c, b) Source
Apply a function to the first element of an optional pair.