SRFI-207: String-notated bytevectors
This egg allows bytevectors (AKA u8vectors) to be treated as string-like objects, known as a bytestrings. It provides both reader support for bytestring literals (which are bytevectors, but are written in a convenient string-like notation) and a substantial library modelled on srfi-13 for bytestring processing. Procedures for converting between hexadecimal or base64-encoded strings and bytestrings are also included.
In addition to the procedure specified by SRFI 207, this egg includes forms based on procedures from (chicken base) and srfi-152. To use these extensions, import (srfi 207 extensions).
To use bytestring literals in compiled code, compile with -X srfi-207.
This page includes excerpts from the SRFI document, but is primarily intended to document the forms exported by the egg. For a full description of the SRFI, see the SRFI document.
Most of the procedures of this SRFI begin with bytestring- in order to distinguish them from other bytevector procedures. This does not mean that they accept or return a separate bytestring type: bytestrings and bytevectors are exactly the same type.
The following names are used for the arguments:
|obj||Any Scheme object.|
|pred||A predicate that accepts zero or more arguments.|
|list||A Scheme list.|
|start, end||Exact integers specifying a half-open interval of indexes for a sub-bytevector. When omitted, start defaults to 0 and end to the length of the corresponding bytevector argument. It is an error unless 0 ≤ start ≤ end ≤ (bytevector-length bytevector).|
It is an error (unless otherwise noted) if the procedures are passed arguments that do not have the type implied by the argument names.
The basic form of a string-notated bytevector is:
The contents of a string-notated bytevector can be ASCII characters, hexadecimal sequences, or various mnemonic sequences. In general, the syntax closely follows the syntax for string literals given in R7RS §6.7; the main exception is that all characters must be ASCII. Unicode codepoints above U+007f must be expressed by hex sequences. Since bytestrings are just bytevectors, though, they can't contain any element numerically greater than 255 (#xff).
Within the content of a string-notated bytevector:
- The sequence \" represents the integer 34;
- The sequence \\ represents the integer 92;
- The following mnemonic sequences represent the corresponding integers:
- The sequence \x followed by zero or more 0 characters, followed by one or two hexadecimal digits, followed by ; represents the integer specified by the hexadecimal digits;
- The sequence \ followed by zero or more intraline whitespace characters, followed by a newline, followed by zero or more further intraline whitespace characters, is ignored and corresponds to no entry in the resulting bytevector;
- Any other printable ASCII character represents the character number of that character in the ASCII/Unicode code chart; and
- It is an error to use any other character or sequence beginning with \ within a string-notated bytevector.
When the Scheme reader encounters a string-notated bytevector, it produces a datum as if that bytevector had been written out in full. That is, #u8"A" is exactly equivalent to #u8(65).
- bytestring arg …procedure
Converts the args into a sequence of small integers and returns them as a bytevector as follows:
- If arg is an exact integer in the range 0–255 inclusive, it is added to the result.
- If arg is an ASCII character (that is, its codepoint is in the range 0–127 inclusive), it is converted to its codepoint and added to the result.
- If arg is a bytevector, its elements are added to the result.
- If arg is a string of ASCII characters, it is converted to a sequence of codepoints which are added to the result.
Otherwise, an error satisfying bytestring-error? is signaled.
(bytestring "lor" #\r #x65 #u8(#x6d)) ⇒ #u8"lorem" (bytestring "η" #\space #u8(#x65 #x71 #x75 #x69 #x76)) ⇒ ; error
- make-bytestring listprocedure
If the elements of list are suitable arguments for bytestring, returns the bytevector that would be the result of applying bytestring to list. Otherwise, an error satisfying bytestring-error? is signaled.
- make-bytestring! bytevector at listprocedure
If the elements of list are suitable arguments for bytestring, writes the bytes of the bytevector that would be the result of calling make-bytestring into bytevector starting at index at.
(define bstring (make-bytevector 10 #x20)) (make-bytestring! bstring 2 '(#\s #\c "he" #u8(#x6d #x65)) bstring ⇒ #u8" scheme "
- bytevector->hex-string bytevectorprocedure
- hex-string->bytevector stringprocedure
Converts between a bytevector and a string containing pairs of hexadecimal digits. If string is not pairs of hexadecimal digits, an error satisfying bytestring-error? is raised.
(bytevector->hex-string #u8"Ford") ⇒ "467f7264" (hex-string->bytevector "5a6170686f64") ⇒ #u8"Zaphod"
- bytevector->base64 bytevector #!optional digitsprocedure
- base64->bytevector string #!optional digitsprocedure
Converts between a bytevector and its base-64 encoding as a string. The 64 digits are represented by the characters 0–9, A–Z, a–z, and the symbols + and /. However, there are different variants of base-64 encoding which use different representations of the 62nd and 63rd digit. If the optional argument digits (a two-character string) is provided, those two characters will be used as the 62nd and 63rd digit instead. Details can be found in RFC 4648. If string is not in base-64 format, an error satisfying bytestring-error? is raised. However, characters that satisfy char-whitespace? are silently ignored.
(bytevector->base64 #u8(1 2 3 4 5 6)) ⇒ "AQIDBAUG" (bytevector->base64 #u8"Arthur Dent") ⇒ "QXJ0aHVyIERlbnQ=" (base64->bytevector "+/ /+") ⇒ #u8(#xfb #xff #xfe)
- bytestring->list bytevector #!optional start endprocedure
Converts all or part of bytevector into a list of the same length containing characters for elements in the range 32 to 127 and exact integers for all other elements.
(bytestring->list #u8(#x41 #x42 1 2) 1 3) ⇒ (#\B 1)
- make-bytestring-generator arg …procedure
Returns a SRFI 158 generator that when invoked will return consecutive bytes of the bytevector that bytestring would create when applied to args, but without creating any bytevectors. The args are validated before any bytes are generated; if they are ill-formed, an error satisfying bytestring-error? is raised.
(generator->list (make-bytestring-generator "lorem")) ⇒ (#x6c #x6f #x72 #x65 #x6d)
- bytestring-pad bytevector len char-or-u8procedure
- bytestring-pad-right bytevector len char-or-u8procedure
Returns a newly allocated bytevector with the contents of bytevector plus sufficient additional bytes at the beginning/end containing char-or-u8 (which can be either an ASCII character or an exact integer in the range 0–255) such that the length of the result is at least len.
(bytestring-pad #u8"Zaphod" 10 #\_) ⇒ #u8"____Zaphod" (bytestring-pad-right #u8(#x80 #x7f) 8 0) ⇒ #u8(#x80 #x7f 0 0 0 0 0 0)
- bytestring-trim bytevector predprocedure
- bytestring-trim-right bytevector predprocedure
- bytestring-trim-both bytevector predprocedure
Returns a newly allocated bytevector with the contents of bytevector, except that consecutive bytes at the beginning / the end / both the beginning and the end that satisfy pred are not included.
(bytestring-trim #u8" Trillian" (lambda (b) (= b #x20))) ⇒ #u8"Trillian" (bytestring-trim-both #u8(0 0 #x80 #x7f 0 0 0) zero?) ⇒ #u8(#x80 #x7f)
- (bytestring-replace bytevector₁ bytevector₂ start₁ end₁ [start₂ end₂])procedure
Returns a newly allocated bytevector with the contents of bytevector₁, except that the bytes indexed by start₁ and end₁ are not included but are replaced by the bytes of bytevector₂ indexed by start₂ and end₂.
(bytestring-replace #u8"Vogon torture" #u8"poetry" 6 13) ⇒ #u8"Vogon poetry"
- bytestring<? bytevector₁ bytevector₂ bytevector₃ …procedure
- bytestring>? bytevector₁ bytevector₂ bytevector₃ …procedure
- bytestring<=? bytevector₁ bytevector₂ bytevector₃ …procedure
- bytestring>=? bytevector₁ bytevector₂ bytevector₃ …procedure
Returns #t if the bytevectors are monotonically less than / greater than / less than or equal to / greater than or equal. Comparisons are lexicographical: shorter bytevectors compare before longer ones, all elements being equal.
Note: u8vector= from srfi-160 rounds out this family. For binary comparison only, use bytevector=? from r6rs-bytevectors or plain old equal?.
(The ability to compare more than two bytevectors is an extension to SRFI 207.)
(bytestring<? #u8"Heart Of Gold" #u8"Heart of Gold") ⇒ #t (bytestring<=? #u8(#x81 #x95) #u8(#x80 #xa0)) ⇒ #f (bytestring>? #u8(1 2 3) #u8(1 2)) ⇒ #t
- (bytestring-index bytevector pred [start end])procedure
- (bytestring-index-right bytevector pred [start end])procedure
Searches bytevector from start to end / from end to start for the first byte that satisfies pred, and returns the index into bytevector containing that byte. In either direction, start is inclusive and end is exclusive. If there are no such bytes, returns #f.
(bytestring-index #u8(#x65 #x72 #x83 #x6f) (lambda (b) (> b #x7f))) ⇒ 2 (bytestring-index #u8"Beeblebrox" (lambda (b) (> b #x7f))) ⇒ #f (bytestring-index-right #u8"Zaphod" odd?) ⇒ 4
- bytestring-break bytevector predprocedure
- bytestring-span bytevector predprocedure
Returns two values, a bytevector containing the maximal sequence of characters (searching from the beginning of bytevector to the end) that do not satisfy / do satisfy pred, and another bytevector containing the remaining characters.
(bytestring-break #u8(#x50 #x4b 0 0 #x1 #x5) zero?) ⇒ #u8(#x50 #x4b) #u8(0 0 #x1 #x5) (bytestring-span #u8"ABCDefg" (lambda (b) (and (> b 40) (< b 91)))) ⇒ #u8"ABCD" #u8"efg"
Joining and splitting
- bytestring-join bytevector-list delimiter #!optional grammarprocedure
Pastes the bytevectors in bytevector-list together using the delimiter, which can be anything suitable as an argument to bytestring. The grammar argument is a symbol that determines how the delimiter is used, and defaults to infix. It is an error for grammar to be any symbol other than these four:
- infix means an infix or separator grammar: inserts the delimiter between list elements. An empty list will produce an empty bytevector.
- strict-infix means the same as infix if the list is non-empty, but will signal an error satisfying bytestring-error? if given an empty list.
- suffix means a suffix or terminator grammar: inserts the delimiter after every list element.
- prefix means a prefix grammar: inserts the delimiter before every list element.
(bytestring-join '(#u8"Heart" #u8"of" #u8"Gold") #x20) ⇒ #u8"Heart of Gold" (bytestring-join '(#u8(#xef #xbb) #u8(#xbf)) 0 'prefix) ⇒ #u8(0 #xef #xbb 0 #xbf) (bytestring-join '() 0 'strict-infix) ⇒ ; error
- bytestring-split bytevector delimiter #!optional grammarprocedure
Divides the elements of bytevector and returns a list of newly allocated bytevectors using the delimiter (an ASCII character or exact integer in the range 0–255 inclusive). Delimiter bytes are not included in the result bytevectors.
The grammar argument is used to control how bytevector is divided. It has the same default and meaning as in bytestring-join, except that infix and strict-infix mean the same thing. That is, if grammar is prefix or suffix, then ignore any delimiter in the first or last position of bytevector respectively.
(bytestring-split #u8"Beeblebrox" #x62) ⇒ (#u8"Bee" #u8"le" #u8"rox") (bytestring-split #u8(1 0 2 0) 0 'suffix) ⇒ (#u8(1) #u8(2))
- read-textual-bytestring prefix #!optional portprocedure
Reads a string in the external format described in this SRFI from port and return it as a bytevector. If the prefix argument is false, this procedure assumes that "#u8" has already been read from port. If port is omitted, it defaults to the value of (current-input-port). If the characters read are not in the external format, an error satisfying bytestring-error? is raised.
(call-with-port (open-input-string "#u8\"AB\\xad;\\xf0;\\x0d;CD\"") (lambda (port) (read-textual-bytestring #t port))) ⇒ #u8(#x41 #x42 #xad #xf0 #x0d #x43 #x44)
- write-textual-bytestring bytevector #!optional portprocedure
Writes bytevector in the external format described in this SRFI to port. Bytes representing non-graphical ASCII characters are unencoded: all other bytes are encoded with a single letter if possible, otherwise with a \x escape. If port is omitted, it defaults to the value of (current-output-port).
(call-with-port (open-output-string) (lambda (port) (write-textual-bytestring #u8(#x9 #x41 #x72 #x74 #x68 #x75 #x72 #xa) port) (get-output-string port))) ⇒ "#u8\"\\tArthur\\n\""
- write-binary-bytestring port arg …procedure
Outputs each arg to the binary output port port using the same interpretations as bytestring, but without creating any bytevectors. The args are validated before any bytes are written to port; if they are ill-formed, an error satisfying bytestring-error? is raised.
(call-with-port (open-output-bytevector) (lambda (port) (write-binary-bytestring port #\Z #x61 #x70 "hod") (get-output-bytevector port))) ⇒ #u8"Zaphod"
- bytestring-error? objprocedure
Returns #t if obj is an object signaled by any of the following procedures, in the circumstances described above:
Like R7RS error objects, the bytestring-error objects provided by this implementation encapsulate a message and a collection of irritants. The former is a string; the latter can be any Scheme objects, generally those which caused the error to be signaled.
In this implementation, bytestring errors are conditions of kind (exn bytestring). They have the location, message, and arguments properties. See Module (chicken condition) for more on inspecting conditions.
The following forms are provided by the (srfi 207 extensions) module. They are extensions to the SRFI.
- bytestring-translate bytevector from toprocedure
from and to may be bytes (exact integers), ASCII characters, or lists of bytes/characters. to must contain at least as many elements as from.
Translates each occurrence of from to to in bytevector, returning a newly allocated bytevector. If from and to are lists, then the ith element of from is replaced with the ith element of to.
(Extension based on string-translate from (chicken string)).
(bytestring-translate #u8"Zaphod" #\d #\z) ⇒ #u8"Zaphoz" (bytestring-translate #u8"gargleblaster" '(#\g #\e) '(#\b #\o)) ⇒ #u8"barbloblastor"
- bytestring-substitute bytevector alistprocedure
Each element of alist must be a pair of the form (x . y), where x and y are characters or bytes. Returns a newly allocated bytevector in which each occurrence of each x is replaced with y.
(Extension based on string-translate* from (chicken string)).
(bytestring-substitute #u8"Zaphod" '((#\d . #\z))) ⇒ #u8"Zaphoz" (bytestring-translate #u8"gargleblaster" '((#\g . #\b) (#\e . #\o))) ⇒ #u8"barbloblastor"
- (subbytestring=? bytevector₁ bytevector₂ [start₁ start₂ length])procedure
Compares sub-bytevectors of bytevector₁ and bytevector₂ and returns #t if they are equal and #f otherwise. The spans of length length are compared, starting at start₁ of bytevector₁ and at start₂ of bytevector₂. Both start arguments default to 0, and length defaults to the minimum remaining length between the two bytevectors.
(Extension based on substring=? from (chicken string).)
(subbytestring=? #u8"Vogon poetry" #u8"not Vogon torture!" 0 4 5) ⇒ #t
- bytestring-compare3 bytevector₁ bytevector₂procedure
Returns -1, 0, or 1 if bytevector₁ is lexicographically less than, equal to, or greater than bytevector₂, respectively.
(Extension based on bytestring-compare3 from (chicken string).)
(bytestring-compare3 #u8"Zaphod" #u8"just Zaphod") ⇒ -1 (bytestring-compare3 #u8"gargleblaster" #u8"Vogon") ⇒ 1
- bytestring-chomp bytevector #!optional suffixprocedure
Returns a newly allocated bytevector with the contents of bytevector, except that the bytevector suffix is trimmed from the end if it is present. suffix defaults to #u8"\n".
(Extension based on string-chomp from (chicken string).)
(bytestring-chomp #u8"Vogon, " #u8", ") ⇒ #u8"Vogon"
- bytestring-prefix-length bytevector₁ bytevector₂procedure
- bytestring-suffix-length bytevector₁ bytevector₂procedure
Return the length of the longest common prefix/suffix of bytevector₁ and bytevector₂. For prefixes, this is equivalent to their "mismatch index".
(bytestring-prefix-length #u8"Heart Of Gold" #u8"Heart of Gold") ⇒ 6 (bytestring-suffix-length #u8"Heart Of Gold" #u8"Heart of Gold") ⇒ 6 (bytestring-prefix-length-ci #u8"Heart Of Gold" #u8"Heart of Gold") ⇒ 13
(Extension based on string-prefix-length, etc. from srfi-152.)
- bytestring-prefix? bytevector₁ bytevector₂procedure
- bytestring-suffix? bytevector₁ bytevector₂procedure
Is bytevector₁ a prefix/suffix of bytevector₂?
(Extension based on string-prefix?, etc. from srfi-152.)
- bytestring-segment bytevector kprocedure
Returns a list of bytestrings representing the consecutive subvectors of bytevector of length k. The last bytevector may be shorter than k.
(Extension based on string-segment from srfi-152. See also string-chop from (chicken string).)
(bytestring-segment #u8"Heart of Gold" 3) ⇒ (#u8"Hea" #u8"rt " #u8"of " #u8"Gol" #u8"d")
- (bytestring-contains bytevector₁ bytevector₂ [start₁ end₁ start₂ end₂])procedure
- (bytestring-contains-right bytevector₁ bytevector₂ [start₁ end₁ start₂ end₂])procedure
Does the subvector of bytevector₁ specified by start₁ and end₁ contain the sequence of bytes given by the subvector of bytevector₂ specified by start₂ and end₂?
Returns #f if there is no match. If start₂ = end₂, bytestring-contains returns start₁ but bytestring-contains-right returns end₁. Otherwise returns the index in bytevector₁ for the first character of the first/last match; that index lies within the half-open interval [start₁, end₁), and the match lies entirely within the [start₁, end₁) range of bytevector₁.
(Extension based on string-contains, etc. from srfi-152.)
(bytestring-contains #u8(1 2 3 4 5) #u8(2 3)) ⇒ 1 (bytestring-contains #u8"hitchhiker" #u8"tchh" 2 7) ⇒ 2 (bytestring-contains #u8"hitchhiker" #u8"hacker" 0 10 2) ⇒ #f (bytestring-contains-right #u8"banana" #u8"an") ⇒ 3
- (bytestring-concatenate-reverse bytevector-list [final-bvec end])procedure
With no optional arguments, calling this procedure is equivalent to
(u8vector-concatenate (reverse bytevector-list))
but may be more efficient.
If the optional bytevector argument final-bvec is specified, it is effectively consed onto the beginning of bytevector-list before performing the list-reverse and concatenate operations.
If the optional argument end is given, only the bytes up to but not including end in final-bvec are added to the result.
(Extension based on string-concatenate-reverse from srfi-152.)
(bytestring-concatenate-reverse '(#u8" must be" #u8"Hello, I") #u8" going.XXXX" 7) ⇒ #u8"Hello, I must be going."
- (bytestring-replicate bytevector from to [start end])procedure
This is an "extended substring" procedure that implements replicated copying of a subbytestring.
The subvector of bytevector described by start and end (the whole bytevector, by default) is conceptually replicated both up and down the index space, in both the positive and negative directions, to produce a conceptually infinite bytevector. The subvector from from to to of this bytevector is returned.
- The from/to arguments give a half-open range containing the characters from index from up to, but not including, index to.
- The from/to indexes are not expressed in the index space of bytevector. They refer instead to the replicated index space of the substring defined by bytevector, start, and end.
(Extension based on string-replicate from srfi-152 (AKA xsubstring from srfi-13).)
;; Rotate left. (bytestring-replicate #u8"abcdef" 1 7) ⇒ #u8"bcdefa" ;; Rotate right. (bytestring-replicate #u8"abcdef" -1 5) ⇒ #u8"fabcde" ;; Iterative copy. (bytestring-replicate #u8".oOo" 0 12) ⇒ #u8".oOo.oOo.oOo."
This egg tries to give useful information when things go wrong. Procedure arguments are type-checked. When a type check fails, a condition of kind (exn type assertion) is raised. Bytestring bounds errors are signaled by (exn bounds assertion) conditions. This conforms to the condition protocol used by CHICKEN's internal libraries.
See the Module (chicken condition) page for more information.
About This Egg
The following eggs are required:
In addition, the srfi-133, srfi-158, and srfi-160 eggs are optional dependencies which will be used if present.
by Daphne Preston-Kendal (external notation), John Cowan (procedure design), & Wolfgang Corcoran-Mathe (implementation)
Originally ported to Chicken Scheme 5 by Sergey Goldgaber.
Contact: <wcm at sigwinch dot xyzzy minus the zy>
© 2020 Daphne Preston-Kendal, John Cowan, and Wolfgang Corcoran-Mathe.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice (including the next paragraph) shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
- Ported to Chicken Scheme 5
- Changed maintainer information.
- Simplified dependencies.
- Reader support for bytestring literals, types.
- Simplify, remove hard srfi-160 dependency.
- 1.0 (2022-01-29)
- Extend egg with forms from (chicken string), srfi-152, R7RS.
- 2.0.0 (2022-09-24)
- Reorganize library and move extensions to their own module. Improve checks and follow CHICKEN's condition protocol.