path: root/man/man7/string_copying.7

.\" Copyright 2022 Alejandro Colomar <alx@kernel.org>
.\"
.\" SPDX-License-Identifier: BSD-3-Clause
.\"
.TH string_copying 7 (date) "Linux man-pages (unreleased)"
.\" ----- NAME :: -----------------------------------------------------/
.SH NAME
stpcpy,
strcpy, strcat,
stpecpy,
strtcpy,
strlcpy, strlcat,
stpncpy,
strncpy,
strncat
\- copying strings and character sequences
.\" ----- SYNOPSIS :: -------------------------------------------------/
.SH SYNOPSIS
.\" ----- SYNOPSIS :: (Null-terminated) strings -----------------------/
.SS Strings
.nf
// Chain-copy a string.
.BI "char *stpcpy(char *restrict " dst ", const char *restrict " src );
.P
// Copy/catenate a string.
.BI "char *strcpy(char *restrict " dst ", const char *restrict " src );
.BI "char *strcat(char *restrict " dst ", const char *restrict " src );
.P
// Chain-copy a string with truncation.
.BI "char *stpecpy(char *" dst ", char " end "[0], const char *restrict " src );
.P
// Copy/catenate a string with truncation.
.BI "ssize_t strtcpy(char " dst "[restrict ." dsize "], \
const char *restrict " src ,
.BI "               size_t " dsize );
.BI "size_t strlcpy(char " dst "[restrict ." dsize "], \
const char *restrict " src ,
.BI "               size_t " dsize );
.BI "size_t strlcat(char " dst "[restrict ." dsize "], \
const char *restrict " src ,
.BI "               size_t " dsize );
.fi
.\" ----- SYNOPSIS :: Null-padded character sequences --------/
.SS Null-padded character sequences
.nf
// Fill a fixed-size buffer with characters from a string
// and pad with null bytes.
.BI "char *strncpy(char " dst "[restrict ." dsize "], \
const char *restrict " src ,
.BI "               size_t " dsize );
.BI "char *stpncpy(char " dst "[restrict ." dsize "], \
const char *restrict " src ,
.BI "               size_t " dsize );
.P
// Chain-copy a null-padded character sequence into a character sequence.
.I mempcpy(dst, src, strnlen(src, NITEMS(src)));
.P
// Chain-copy a null-padded character sequence into a string.
.I stpcpy(mempcpy(dst, src, strnlen(src, NITEMS(src))), \[dq]\[dq]);
.P
// Catenate a null-padded character sequence into a string.
.BI "char *strncat(char *restrict " dst ", const char " src "[restrict ." ssize ],
.BI "               size_t " ssize );
.fi
.\" ----- SYNOPSIS :: Known-length character sequences --------------------/
.SS Known-length character sequences
.nf
// Chain-copy a known-length character sequence.
.BI "void *mempcpy(void " dst "[restrict ." len "], \
const void " src "[restrict ." len ],
.BI "               size_t " len );
.P
// Chain-copy a known-length character sequence into a string.
.I stpcpy(mempcpy(dst, src, len), \[dq]\[dq]);
.fi
.SH DESCRIPTION
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: -----------------/
.SS Terms (and abbreviations)
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: string (str) ----/
.TP
.IR "string " ( str )
is a sequence of zero or more non-null characters followed by a null character.
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: null-padded character seq
.TP
.I character sequence
is a sequence of zero or more non-null characters.
A program should never use a character sequence where a string is required.
However, with appropriate care,
a string can be used in the place of a character sequence.
.RS
.TP
.I null-padded character sequence
Character sequences can be contained in fixed-size buffers,
which contain padding null bytes after the character sequence,
to fill the rest of the buffer
without affecting the character sequence;
however, those padding null bytes are not part of the character sequence.
Don't confuse null-padded with null-terminated:
null-padded means 0 or more padding null bytes,
while null-terminated means exactly 1 terminating null character.
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: known-length character sequence
.TP
.I known-length character sequence
Character sequence delimited by its length.
It may be a slice of a larger character sequence,
or even of a string.
.RE
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: length (len) ----/
.TP
.IR "length " ( len )
is the number of non-null characters in a string or character sequence.
It is the return value of
.I strlen(str)
and of
.IR "strnlen(buf, size)" .
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: size ------------/
.TP
.I size
refers to the entire buffer
where the string or character sequence is contained.
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: end -------------/
.TP
.I end
is the name of a pointer to one past the last element of a buffer.
It is equivalent to
.IR &str[size] .
It is used as a sentinel value,
to be able to truncate strings or character sequences
instead of overrunning the containing buffer.
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: copy ------------/
.TP
.I copy
This term is used when
the writing starts at the first element pointed to by
.IR dst .
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: catenate --------/
.TP
.I catenate
This term is used when
a function first finds the terminating null character in
.IR dst ,
and then starts writing at that position.
.\" ----- DESCRIPTION :: Terms (and abbreviations) :: chain -----------/
.TP
.I chain
This term is used when
it's the programmer who provides
a pointer to the terminating null character in the string
.I dst
(or one after the last character in a character sequence),
and the function starts writing at that location.
The function returns
a pointer to the new location of the terminating null character
(or one after the last character in a character sequence)
after the call,
so that the programmer can use it to chain such calls.
.\" ----- DESCRIPTION :: Copy, catenate, and chain-copy ---------------/
.SS Copy, catenate, and chain-copy
Originally,
there was a distinction between functions that copy and those that catenate.
However, newer functions that copy while allowing chaining
cover both use cases with a single API.
They are also algorithmically faster,
since they don't need to search for
the terminating null character of the existing string.
However, functions that catenate have a much simpler use,
so if performance is not important,
it can make sense to use them for improving readability.
.P
The pointer returned by functions that allow chaining
is a byproduct of the copy operation,
so it has no performance costs.
Functions that return such a pointer,
and thus can be chained,
have names of the form
.RB * stp *(),
since it's common to name the pointer just
.IR p .
.P
Chain-copying functions that truncate
should accept a pointer to the end of the destination buffer,
and have names of the form
.RB * stpe *().
This allows not having to recalculate the remaining size after each call.
.\" ----- DESCRIPTION :: Truncate or not? -----------------------------/
.SS Truncate or not?
The first thing to note is that programmers should be careful with buffers,
so they always have the correct size,
and truncation is not necessary.
.P
In most cases,
truncation is not desired,
and it is simpler to just do the copy.
Simpler code is safer code.
Programming against programming mistakes by adding more code
just adds more points where mistakes can be made.
.P
Nowadays,
compilers can detect most programmer errors with features like
compiler warnings,
static analyzers, and
.B \%_FORTIFY_SOURCE
(see
.BR ftm (7)).
Keeping the code simple
helps these overflow-detection features be more precise.
.P
When validating user input,
code should normally not truncate,
but instead fail and prevent the copy at all.
.P
In some cases,
however,
it makes sense to truncate.
.P
Functions that truncate:
.IP \[bu] 3
.BR stpecpy ()
.IP \[bu]
.BR strtcpy ()
.IP \[bu]
.BR strlcpy (3bsd)
and
.BR strlcat (3bsd)
are similar, but have important performance problems; see BUGS.
.IP \[bu]
.BR stpncpy (3)
and
.BR strncpy (3)
also truncate, but they don't write strings,
but rather null-padded character sequences.
.\" ----- DESCRIPTION :: Null-padded character sequences --------------/
.SS Null-padded character sequences
For historic reasons,
some standard APIs and file formats,
such as
.BR utmpx (5)
and
.BR tar (1),
use null-padded character sequences in fixed-size buffers.
To interface with them,
specialized functions need to be used.
.P
To copy bytes from strings into these buffers, use
.BR strncpy (3)
or
.BR stpncpy (3).
.P
To read a null-padded character sequence,
use
.IR "strnlen(src,\ NITEMS(src))" ,
and then you can treat it as a known-length character sequence;
or use
.BR strncat (3)
directly.
.\" ----- DESCRIPTION :: Known-length character sequences -----------------/
.SS Known-length character sequences
The simplest character sequence copying function is
.BR mempcpy (3).
It requires always knowing the length of your character sequences,
for which structures can be used.
It makes the code much faster,
since you always know the length of your character sequences,
and can do the minimal copies and length measurements.
.BR mempcpy (3)
copies character sequences,
so you need to explicitly set the terminating null character
if you need a string.
.P
In programs that make considerable use of strings or character sequences,
and need the best performance,
using overlapping character sequences can make a big difference.
It allows holding subsequences of a larger character sequence,
while not duplicating memory
nor using time to do a copy.
.P
However, this is delicate,
since it requires using character sequences.
C library APIs use strings,
so programs that use character sequences
will have to take care of differentiating strings from character sequences.
.P
To copy a known-length character sequence, use
.BR mempcpy (3).
.P
To copy a known-length character sequence into a string, use
.IR "\%stpcpy(mempcpy(dst,\ src,\ len),\ \[dq]\[dq])" .
.P
A string is also accepted as input,
because
.BR mempcpy (3)
asks for the length,
and a string is composed of a character sequence of the same length
plus a terminating null character.
.\" ----- DESCRIPTION :: String vs character sequence -----------------/
.SS String vs character sequence
Some functions only operate on strings.
Those require that the input
.I src
is a string,
and guarantee an output string
(even when truncation occurs).
Functions that catenate
also require that
.I dst
holds a string before the call.
List of functions:
.IP \[bu] 3
.PD 0
.BR stpcpy (3)
.IP \[bu]
.BR strcpy (3),
.BR strcat (3)
.IP \[bu]
.BR stpecpy ()
.IP \[bu]
.BR strtcpy ()
.IP \[bu]
.BR strlcpy (3bsd),
.BR strlcat (3bsd)
.PD
.P
Other functions require an input string,
but create a character sequence as output.
These functions have confusing names,
and have a long history of misuse.
List of functions:
.IP \[bu] 3
.PD 0
.BR stpncpy (3)
.IP \[bu]
.BR strncpy (3)
.PD
.P
Other functions operate on an input character sequence,
and create an output string.
Functions that catenate
also require that
.I dst
holds a string before the call.
.BR strncat (3)
has an even more misleading name than the functions above.
List of functions:
.IP \[bu] 3
.BR strncat (3)
.P
Other functions operate on an input character sequence
to create an output character sequence.
List of functions:
.IP \[bu] 3
.BR mempcpy (3)
.\" ----- DESCRIPTION :: Functions :: ---------------------------------/
.SS Functions
.\" ----- DESCRIPTION :: Functions :: stpcpy(3) -----------------------/
.TP
.BR stpcpy (3)
Copy the input string into a destination string.
The programmer is responsible for allocating a buffer large enough.
It returns a pointer suitable for chaining.
.\" ----- DESCRIPTION :: Functions :: strcpy(3), strcat(3) ------------/
.TP
.BR strcpy (3)
.TQ
.BR strcat (3)
Copy and catenate the input string into a destination string.
The programmer is responsible for allocating a buffer large enough.
The return value is useless.
.IP
.BR stpcpy (3)
is a faster alternative to these functions.
.\" ----- DESCRIPTION :: Functions :: stpecpy() -----------------------/
.TP
.BR stpecpy ()
Chain-copy the input string into a destination string.
If the destination buffer,
limited by a pointer to its end,
isn't large enough to hold the copy,
the resulting string is truncated
(but it is guaranteed to be null-terminated).
It returns a pointer suitable for chaining.
Truncation needs to be detected only once after the last chained call.
.IP
This function is not provided by any library;
see EXAMPLES for a reference implementation.
.\" ----- DESCRIPTION :: Functions :: strtcpy() -----------------------/
.TP
.BR strtcpy ()
Copy the input string into a destination string.
If the destination buffer isn't large enough to hold the copy,
the resulting string is truncated
(but it is guaranteed to be null-terminated).
It returns the length of the string,
or \-1 if it truncated.
.IP
This function is not provided by any library;
see EXAMPLES for a reference implementation.
.\" ----- DESCRIPTION :: Functions :: strlcpy(3bsd), strlcat(3bsd) ----/
.TP
.BR strlcpy (3bsd)
.TQ
.BR strlcat (3bsd)
Copy and catenate the input string into a destination string.
If the destination buffer,
limited by its size,
isn't large enough to hold the copy,
the resulting string is truncated
(but it is guaranteed to be null-terminated).
They return the length of the total string they tried to create.
.IP
Check BUGS before using these functions.
.IP
.BR strtcpy ()
and
.BR stpecpy ()
are better alternatives to these functions.
.\" ----- DESCRIPTION :: Functions :: stpncpy(3) ----------------------/
.TP
.BR stpncpy (3)
Copy the input string into
a destination null-padded character sequence in a fixed-size buffer.
If the destination buffer,
limited by its size,
isn't large enough to hold the copy,
the resulting character sequence is truncated.
Since it creates a character sequence,
it doesn't need to write a terminating null character.
It's impossible to distinguish truncation by the result of the call,
from a character sequence that just fits the destination buffer;
truncation should be detected by
comparing the length of the input string
with the size of the destination buffer.
.\" ----- DESCRIPTION :: Functions :: strncpy(3) ----------------------/
.TP
.BR strncpy (3)
This function is identical to
.BR stpncpy (3)
except for the useless return value.
.IP
.BR stpncpy (3)
is a more useful alternative to this function.
.\" ----- DESCRIPTION :: Functions :: strncat(3) ----------------------/
.TP
.BR strncat (3)
Catenate the input character sequence,
contained in a null-padded fixed-size buffer,
into a destination string.
The programmer is responsible for allocating a buffer large enough.
The return value is useless.
.IP
Do not confuse this function with
.BR strncpy (3);
they are not related at all.
.IP
.I \%stpcpy(mempcpy(dst,\ src,\ strnlen(src,\ NITEMS(src))),\ \[dq]\[dq])
is a faster alternative to this function.
.\" ----- DESCRIPTION :: Functions :: mempcpy(3) ----------------------/
.TP
.BR mempcpy (3)
Copy the input character sequence,
limited by its length,
into a destination character sequence.
The programmer is responsible for allocating a buffer large enough.
It returns a pointer suitable for chaining.
.\" ----- RETURN VALUE :: ---------------------------------------------/
.SH RETURN VALUE
.TP
.BR stpcpy (3)
A pointer to the terminating null character in the destination string.
.TP
.BR stpecpy ()
A pointer to the terminating null character in the destination string,
on success.
On error,
NULL is returned,
and
.I errno
is set to indicate the error.
.TP
.BR mempcpy (3)
.TQ
.BR stpncpy (3)
A pointer to one after the last character
in the destination character sequence.
.TP
.BR strtcpy ()
The length of the string,
on success.
On error,
\-1 is returned,
and
.I errno
is set to indicate the error.
.TP
.BR strlcpy (3bsd)
.TQ
.BR strlcat (3bsd)
The length of the total string that they tried to create
(as if truncation didn't occur).
.TP
.BR strcpy (3)
.TQ
.BR strcat (3)
.TQ
.BR strncpy (3)
.TQ
.BR strncat (3)
The
.I dst
pointer,
which is useless.
.\" ----- ERRORS ------------------------------------------------------/
.SH ERRORS
Most of these functions don't set
.IR errno .
.TP
.BR stpecpy ()
.TQ
.BR strtcpy ()
.RS
.TP
.B ENOBUFS
.I dsize
was
.BR 0 .
.TP
.B E2BIG
The string has been truncated.
.RE
.\" ----- NOTES :: strscpy(9) -----------------------------------------/
.SH NOTES
The Linux kernel has an internal function for copying strings,
.BR strscpy (9),
which is identical to
.BR strtcpy (),
except that it returns
.B \-E2BIG
instead of \-1
and it doesn't set
.IR errno .
.\" ----- CAVEATS :: --------------------------------------------------/
.SH CAVEATS
Don't mix chain calls to truncating and non-truncating functions.
It is conceptually wrong
unless you know that the first part of a copy will always fit.
Anyway, the performance difference will probably be negligible,
so it will probably be more clear if you use consistent semantics:
either truncating or non-truncating.
Calling a non-truncating function after a truncating one is necessarily wrong.
.\" ----- BUGS :: -----------------------------------------------------/
.SH BUGS
All catenation functions share the same performance problem:
.UR https://www.joelonsoftware.com/\:2001/12/11/\:back\-to\-basics/
Shlemiel the painter
.UE .
As a mitigation,
compilers are able to transform some calls to catenation functions
into normal copy functions,
since
.I strlen(dst)
is usually a byproduct of the previous copy.
.P
.BR strlcpy (3)
and
.BR strlcat (3)
need to read the entire
.I src
string,
even if the destination buffer is small.
This makes them vulnerable to Denial of Service (DoS) attacks
if an attacker can control the length of the
.I src
string.
And if not,
they're still unnecessarily slow.
.\" ----- EXAMPLES :: -------------------------------------------------/
.SH EXAMPLES
The following are examples of correct use of each of these functions.
.\" ----- EXAMPLES :: stpcpy(3) ---------------------------------------/
.TP
.BR stpcpy (3)
.EX
p = buf;
p = stpcpy(p, "Hello ");
p = stpcpy(p, "world");
p = stpcpy(p, "!");
len = p \- buf;
puts(buf);
.EE
.\" ----- EXAMPLES :: strcpy(3), strcat(3) ----------------------------/
.TP
.BR strcpy (3)
.TQ
.BR strcat (3)
.EX
strcpy(buf, "Hello ");
strcat(buf, "world");
strcat(buf, "!");
len = strlen(buf);
puts(buf);
.EE
.\" ----- EXAMPLES :: stpecpy() ---------------------------------------/
.TP
.BR stpecpy ()
.EX
end = buf + NITEMS(buf);
p = buf;
p = stpecpy(p, end, "Hello ");
p = stpecpy(p, end, "world");
p = stpecpy(p, end, "!");
if (p == NULL) {
    len = NITEMS(buf) \- 1;
    goto toolong;
}
len = p \- buf;
puts(buf);
.EE
.\" ----- EXAMPLES :: strtcpy() ---------------------------------------/
.TP
.BR strtcpy ()
.EX
len = strtcpy(buf, "Hello world!", NITEMS(buf));
if (len == \-1)
    goto toolong;
puts(buf);
.EE
.\" ----- EXAMPLES :: strlcpy(3bsd), strlcat(3bsd) --------------------/
.TP
.BR strlcpy (3bsd)
.TQ
.BR strlcat (3bsd)
.EX
if (strlcpy(buf, "Hello ", NITEMS(buf)) >= NITEMS(buf))
    goto toolong;
if (strlcat(buf, "world", NITEMS(buf)) >= NITEMS(buf))
    goto toolong;
len = strlcat(buf, "!", NITEMS(buf));
if (len >= NITEMS(buf))
    goto toolong;
puts(buf);
.EE
.\" ----- EXAMPLES :: stpncpy(3) --------------------------------------/
.TP
.BR stpncpy (3)
.EX
p = stpncpy(u->ut_user, "alx", NITEMS(u->ut_user));
if (NITEMS(u->ut_user) < strlen("alx"))
    goto toolong;
len = p \- u->ut_user;
fwrite(u->ut_user, 1, len, stdout);
.EE
.\" ----- EXAMPLES :: strncpy(3) --------------------------------------/
.TP
.BR strncpy (3)
.EX
strncpy(u->ut_user, "alx", NITEMS(u->ut_user));
if (NITEMS(u->ut_user) < strlen("alx"))
    goto toolong;
len = strnlen(u->ut_user, NITEMS(u->ut_user));
fwrite(u->ut_user, 1, len, stdout);
.EE
.\" ----- EXAMPLES :: mempcpy(dst, src, strnlen(src, NITEMS(src))) ----/
.TP
.I mempcpy(dst, src, strnlen(src, NITEMS(src)))
.EX
char  buf[NITEMS(u->ut_user)];
p = buf;
p = mempcpy(p, u->ut_user, strnlen(u->ut_user, NITEMS(u->ut_user)));
len = p \- buf;
fwrite(buf, 1, len, stdout);
.EE
.\" ----- EXAMPLES :: stpcpy(mempcpy(dst, src, strnlen(src, NITEMS(src))), "")
.TP
.I stpcpy(mempcpy(dst, src, strnlen(src, NITEMS(src))), \[dq]\[dq])
.EX
char  buf[NITEMS(u->ut_user) + 1];
p = buf;
p = mempcpy(p, u->ut_user, strnlen(u->ut_user, NITEMS(u->ut_user)));
p = stpcpy(p, "");
len = p \- buf;
puts(buf);
.EE
.\" ----- EXAMPLES :: strncat(3) --------------------------------------/
.TP
.BR strncat (3)
.EX
char  buf[NITEMS(u->ut_user) + 1];
strcpy(buf, "");
strncat(buf, u->ut_user, NITEMS(u->ut_user));
len = strlen(buf);
puts(buf);
.EE
.\" ----- EXAMPLES :: mempcpy(3) --------------------------------------/
.TP
.BR mempcpy (3)
.EX
p = buf;
p = mempcpy(p, "Hello ", 6);
p = mempcpy(p, "world", 5);
p = mempcpy(p, "!", 1);
len = p \- buf;
fwrite(buf, 1, len, stdout);
.EE
.\" ----- EXAMPLES :: stpcpy(mempcpy(), "") ---------------------------/
.TP
.I stpcpy(mempcpy(dst, src, len), \[dq]\[dq])
.EX
p = buf;
p = mempcpy(p, "Hello ", 6);
p = mempcpy(p, "world", 5);
p = mempcpy(p, "!", 1);
p = stpcpy(p, "");
len = p \- buf;
puts(buf);
.EE
.\" ----- EXAMPLES :: Implementations :: ------------------------------/
.SS Implementations
Here are reference implementations for functions not provided by libc.
.P
.in +4n
.EX
/* This code is in the public domain. */
\&
.\" ----- EXAMPLES :: Implementations :: stpecpy() --------------------/
char *
.IR stpecpy "(char *dst, char end[0], const char *restrict src)"
{
    size_t  dlen;
\&
    if (dst == NULL)
        return NULL;
\&
    dlen = strtcpy(dst, src, end \- dst);
    return (dlen == \-1) ? NULL : dst + dlen;
}
\&
.\" ----- EXAMPLES :: Implementations :: strtcpy() --------------------/
ssize_t
.IR strtcpy "(char *restrict dst, const char *restrict src, size_t dsize)"
{
    bool    trunc;
    size_t  dlen, slen;
\&
    if (dsize == 0) {
        errno = ENOBUFS;
        return \-1;
    }
\&
    slen = strnlen(src, dsize);
    trunc = (slen == dsize);
    dlen = slen \- trunc;
\&
    stpcpy(mempcpy(dst, src, dlen), "");
    if (trunc)
        errno = E2BIG;
    return trunc ? \-1 : slen;
}
.\" ----- SEE ALSO :: -------------------------------------------------/
.SH SEE ALSO
.BR bzero (3),
.BR memcpy (3),
.BR memccpy (3),
.BR mempcpy (3),
.BR stpcpy (3),
.BR strlcpy (3bsd),
.BR strncat (3),
.BR stpncpy (3),
.BR string (3)