mirror of
git://git.code.sf.net/p/zsh/code
synced 2024-11-19 13:33:52 +01:00
3076 lines
82 KiB
C
3076 lines
82 KiB
C
/*
|
|
* pattern.c - pattern matching
|
|
*
|
|
* This file is part of zsh, the Z shell.
|
|
*
|
|
* Copyright (c) 1999 Peter Stephenson
|
|
* All rights reserved.
|
|
*
|
|
* Permission is hereby granted, without written agreement and without
|
|
* license or royalty fees, to use, copy, modify, and distribute this
|
|
* software and to distribute modified versions of this software for any
|
|
* purpose, provided that the above copyright notice and the following
|
|
* two paragraphs appear in all copies of this software.
|
|
*
|
|
* In no event shall Peter Stephenson or the Zsh Development Group be liable
|
|
* to any party for direct, indirect, special, incidental, or consequential
|
|
* damages arising out of the use of this software and its documentation,
|
|
* even if Peter Stephenson and the Zsh Development Group have been advised of
|
|
* the possibility of such damage.
|
|
*
|
|
* Peter Stephenson and the Zsh Development Group specifically disclaim any
|
|
* warranties, including, but not limited to, the implied warranties of
|
|
* merchantability and fitness for a particular purpose. The software
|
|
* provided hereunder is on an "as is" basis, and Peter Stephenson and the
|
|
* Zsh Development Group have no obligation to provide maintenance,
|
|
* support, updates, enhancements, or modifications.
|
|
*
|
|
* Pattern matching code derived from the regexp library by Henry
|
|
* Spencer, which has the following copyright.
|
|
*
|
|
* Copyright (c) 1986 by University of Toronto.
|
|
* Written by Henry Spencer. Not derived from licensed software.
|
|
*
|
|
* Permission is granted to anyone to use this software for any
|
|
* purpose on any computer system, and to redistribute it freely,
|
|
* subject to the following restrictions:
|
|
*
|
|
* 1. The author is not responsible for the consequences of use of
|
|
* this software, no matter how awful, even if they arise
|
|
* from defects in it.
|
|
*
|
|
* 2. The origin of this software must not be misrepresented, either
|
|
* by explicit claim or by omission.
|
|
*
|
|
* 3. Altered versions must be plainly marked as such, and must not
|
|
* be misrepresented as being the original software.
|
|
*
|
|
* Eagle-eyed readers will notice this is an altered version. Incredibly
|
|
* sharp-eyed readers might even find bits that weren't altered.
|
|
*
|
|
*
|
|
* And I experienced a sense that, like certain regular
|
|
* expressions, seemed to match the day from beginning to end, so
|
|
* that I did not need to identify the parenthesised subexpression
|
|
* that told of dawn, nor the group of characters that indicated
|
|
* the moment when my grandfather returned home with news of
|
|
* Swann's departure for Paris; and the whole length of the month
|
|
* of May, as if matched by a closure, fitted into the buffer of my
|
|
* life with no sign of overflowing, turning the days, like a
|
|
* procession of insects that could consist of this or that
|
|
* species, into a random and unstructured repetition of different
|
|
* sequences, anchored from the first day of the month to the last
|
|
* in the same fashion as the weeks when I knew I would not see
|
|
* Gilberte and would search in vain for any occurrences of the
|
|
* string in the avenue of hawthorns by Tansonville, without my
|
|
* having to delimit explicitly the start or finish of the pattern.
|
|
*
|
|
* M. Proust, "In Search of Lost Files",
|
|
* bk I, "The Walk by Bourne's Place".
|
|
*/
|
|
|
|
#include "zsh.mdh"
|
|
|
|
/*
|
|
* The following union is used mostly for alignment purposes.
|
|
* Normal nodes are longs, while certain nodes take a char * as an argument;
|
|
* here we make sure that they both work out to the same length.
|
|
* The compiled regexp we construct consists of upats stuck together;
|
|
* anything else to be added (strings, numbers) is stuck after and
|
|
* then aligned to a whole number of upat units.
|
|
*
|
|
* Note also that offsets are in terms of the sizes of these things.
|
|
*/
|
|
union upat {
|
|
long l;
|
|
unsigned char *p;
|
|
};
|
|
|
|
typedef union upat *Upat;
|
|
|
|
#include "pattern.pro"
|
|
|
|
/* Number of active parenthesized expressions allowed in backreferencing */
|
|
#define NSUBEXP 9
|
|
|
|
/* definition number opnd? meaning */
|
|
#define P_END 0x00 /* no End of program. */
|
|
#define P_EXCSYNC 0x01 /* no Test if following exclude already failed */
|
|
#define P_EXCEND 0x02 /* no Test if exclude matched orig branch */
|
|
#define P_BACK 0x03 /* no Match "", "next" ptr points backward. */
|
|
#define P_EXACTLY 0x04 /* lstr Match this string. */
|
|
#define P_NOTHING 0x05 /* no Match empty string. */
|
|
#define P_ONEHASH 0x06 /* node Match this (simple) thing 0 or more times. */
|
|
#define P_TWOHASH 0x07 /* node Match this (simple) thing 1 or more times. */
|
|
#define P_GFLAGS 0x08 /* long Match nothing and set globbing flags */
|
|
#define P_ISSTART 0x09 /* no Match start of string. */
|
|
#define P_ISEND 0x0a /* no Match end of string. */
|
|
/* numbered so we can test bit 5 for a branch */
|
|
#define P_BRANCH 0x20 /* node Match this alternative, or the next... */
|
|
#define P_WBRANCH 0x21 /* uc* node P_BRANCH, but match at least 1 char */
|
|
/* excludes are also branches, but have bit 4 set, too */
|
|
#define P_EXCLUDE 0x30 /* uc* node Exclude this from previous branch */
|
|
#define P_EXCLUDP 0x31 /* uc* node Exclude, using full file path so far */
|
|
/* numbered so we can test bit 6 so as not to match initial '.' */
|
|
#define P_ANY 0x40 /* no Match any one character. */
|
|
#define P_ANYOF 0x41 /* str Match any character in this string. */
|
|
#define P_ANYBUT 0x42 /* str Match any character not in this string. */
|
|
#define P_STAR 0x43 /* no Match any set of characters. */
|
|
#define P_NUMRNG 0x44 /* zr, zr Match a numeric range. */
|
|
#define P_NUMFROM 0x45 /* zr Match a number >= X */
|
|
#define P_NUMTO 0x46 /* zr Match a number <= X */
|
|
#define P_NUMANY 0x47 /* no Match any set of decimal digits */
|
|
/* spaces left for P_OPEN+n,... for backreferences */
|
|
#define P_OPEN 0x80 /* no Mark this point in input as start of n. */
|
|
#define P_CLOSE 0x90 /* no Analogous to OPEN. */
|
|
/*
|
|
* no no argument
|
|
* zr the range type zrange_t: may be zlong or unsigned long
|
|
* char a single char
|
|
* uc* a pointer to unsigned char, used at run time and initialised
|
|
* to NULL.
|
|
* str null-terminated, metafied string
|
|
* lstr length as long then string, not null-terminated, unmetafied.
|
|
*/
|
|
|
|
/*
|
|
* Notes on usage:
|
|
* P_WBRANCH: This works like a branch and is used in complex closures,
|
|
* to ensure we don't succeed on a zero-length match of the pattern,
|
|
* since that would cause an infinite loop. We do this by recording
|
|
* the positions where we have already tried to match. See the
|
|
* P_WBRANCH test in patmatch().
|
|
*
|
|
* P_ANY, P_ANYOF: the operand is a null terminated
|
|
* string. Normal characters match as expected. Characters
|
|
* in the range Meta+PP_ALPHA..Meta+PP_UNKNWN do the appropriate
|
|
* Posix range tests. This relies on imeta returning true for these
|
|
* characters. We treat unknown POSIX ranges as never matching.
|
|
* PP_RANGE means the next two (possibly metafied) characters form
|
|
* the limits of a range to test; it's too much like hard work to
|
|
* expand the range.
|
|
*
|
|
* P_EXCLUDE, P_EXCSYNC, PEXCEND: P_EXCLUDE appears in the pattern like
|
|
* P_BRANCH, but applies to the immediately preceding branch. The code in
|
|
* the corresponding branch is followed by a P_EXCSYNC, which simply
|
|
* acts as a marker that a P_EXCLUDE comes next. The P_EXCLUDE
|
|
* has a pointer to char embeded in it, which works
|
|
* like P_WBRANCH: if we get to the P_EXCSYNC, and we already matched
|
|
* up to the same position, fail. Thus we are forced to backtrack
|
|
* on closures in the P_BRANCH if the first attempt was excluded.
|
|
* Corresponding to P_EXCSYNC in the original branch, there is a
|
|
* P_EXCEND in the exclusion. If we get to this point, and we did
|
|
* *not* match in the original branch, the exclusion itself fails,
|
|
* otherwise it succeeds since we know the tail already matches,
|
|
* so P_EXCEND is the end of the exclusion test.
|
|
* The whole sorry mess looks like this, where the upper lines
|
|
* show the linkage of the branches, and the lower shows the linkage
|
|
* of their pattern arguments.
|
|
*
|
|
* --------------------- ----------------------
|
|
* ^ v ^ v
|
|
* ( <BRANCH>:apat-><EXCSYNC> <EXCLUDE>:excpat-><EXCEND> ) tail
|
|
* ^
|
|
* | |
|
|
* --------------------------------------
|
|
*
|
|
* P_EXCLUDP: this behaves exactly like P_EXCLUDE, with the sole exception
|
|
* that we prepend the path so far to the exclude pattern. This is
|
|
* for top level file globs, e.g. ** / *.c~*foo.c
|
|
* ^ I had to leave this space
|
|
* P_NUM*: zl is a zlong if that is 64-bit, else an unsigned long.
|
|
*/
|
|
#define PP_ALPHA 1
|
|
#define PP_ALNUM 2
|
|
#define PP_ASCII 3
|
|
#define PP_BLANK 4
|
|
#define PP_CNTRL 5
|
|
#define PP_DIGIT 6
|
|
#define PP_GRAPH 7
|
|
#define PP_LOWER 8
|
|
#define PP_PRINT 9
|
|
#define PP_PUNCT 10
|
|
#define PP_SPACE 11
|
|
#define PP_UPPER 12
|
|
#define PP_XDIGIT 13
|
|
#define PP_IDENT 14
|
|
#define PP_IFS 15
|
|
#define PP_IFSSPACE 16
|
|
#define PP_WORD 17
|
|
#define PP_UNKWN 18
|
|
#define PP_RANGE 19
|
|
|
|
#define P_OP(p) ((p)->l & 0xff)
|
|
#define P_NEXT(p) ((p)->l >> 8)
|
|
#define P_OPERAND(p) ((p) + 1)
|
|
#define P_ISBRANCH(p) ((p)->l & 0x20)
|
|
#define P_ISEXCLUDE(p) (((p)->l & 0x30) == 0x30)
|
|
#define P_NOTDOT(p) ((p)->l & 0x40)
|
|
|
|
/* Specific to lstr type, i.e. P_EXACTLY. */
|
|
#define P_LS_LEN(p) ((p)[1].l) /* can be used as lvalue */
|
|
#define P_LS_STR(p) ((char *)((p) + 2))
|
|
|
|
/* Flags needed when pattern is executed */
|
|
#define P_SIMPLE 0x01 /* Simple enough to be #/## operand. */
|
|
#define P_HSTART 0x02 /* Starts with # or ##'d pattern. */
|
|
#define P_PURESTR 0x04 /* Can be matched with a strcmp */
|
|
|
|
/*
|
|
* Increment pointer which may be on a Meta (x is a pointer variable),
|
|
* returning the incremented value (i.e. like pre-increment).
|
|
*
|
|
* In future the following will need to refer to metafied multibyte
|
|
* characters. References to invidual characters are not turned
|
|
* into a macro when the characters is metafied (c.f. CHARREF()
|
|
* below then the character is not metafied) and will need tracking
|
|
* down.
|
|
*/
|
|
#define METAINC(x) ((x) += (*(x) == Meta) ? 2 : 1)
|
|
/*
|
|
* Return unmetafied char from string (x is any char *)
|
|
*/
|
|
#define UNMETA(x) (*(x) == Meta ? (x)[1] ^ 32 : *(x))
|
|
|
|
#if defined(ZSH_64_BIT_TYPE) || defined(LONG_IS_64_BIT)
|
|
typedef zlong zrange_t;
|
|
#define ZRANGE_T_IS_SIGNED (1)
|
|
#else
|
|
typedef unsigned long zrange_t;
|
|
#endif
|
|
|
|
/*
|
|
* Characters which terminate a pattern segment. We actually use
|
|
* a pointer patendseg which skips the first character if we are not
|
|
* parsing a file pattern.
|
|
* Note that the size of this and the next array are hard-wired
|
|
* via the definitions.
|
|
*/
|
|
|
|
static char endseg[] = {
|
|
'/', /* file only */
|
|
'\0', Bar, Outpar, /* all patterns */
|
|
Tilde /* extended glob only */
|
|
};
|
|
|
|
#define PATENDSEGLEN_NORM 4
|
|
#define PATENDSEGLEN_EXT 5
|
|
|
|
/* Characters which terminate a simple string */
|
|
|
|
static char endstr[] = {
|
|
'/', /* file only */
|
|
'\0', Bar, Outpar, Quest, Star, Inbrack, Inpar, Inang,
|
|
/* all patterns */
|
|
Tilde, Hat, Pound /* extended glob only */
|
|
};
|
|
|
|
#define PATENDSTRLEN_NORM 9
|
|
#define PATENDSTRLEN_EXT 12
|
|
|
|
|
|
/* Default size for pattern buffer */
|
|
#define P_DEF_ALLOC 256
|
|
|
|
/* Flags used in compilation */
|
|
static char *patstart, *patparse; /* input pointers */
|
|
static int patnpar; /* () count */
|
|
static char *patcode; /* point of code emission */
|
|
static long patsize; /* size of code */
|
|
static char *patout; /* start of code emission string */
|
|
static long patalloc; /* size allocated for same */
|
|
static char *patendseg; /* characters ending segment */
|
|
static int patendseglen; /* length of same */
|
|
static char *patendstr; /* characters ending plain string */
|
|
static int patendstrlen; /* length of sameo */
|
|
|
|
/* Flags used in both compilation and execution */
|
|
static int patflags; /* flags passed down to patcompile */
|
|
static int patglobflags; /* globbing flags & approx */
|
|
|
|
/* Add n more characters, ensuring there is enough space. */
|
|
|
|
enum {
|
|
PA_NOALIGN = 1,
|
|
PA_UNMETA = 2
|
|
};
|
|
|
|
/**/
|
|
static void
|
|
patadd(char *add, int ch, long n, int paflags)
|
|
{
|
|
/* Make sure everything gets aligned unless we get PA_NOALIGN. */
|
|
long newpatsize = patsize + n;
|
|
if (!(paflags & PA_NOALIGN))
|
|
newpatsize = (newpatsize + sizeof(union upat) - 1) &
|
|
~(sizeof(union upat) - 1);
|
|
if (patalloc < newpatsize) {
|
|
long newpatalloc =
|
|
2*(newpatsize > patalloc ? newpatsize : patalloc);
|
|
patout = (char *)zrealloc((char *)patout, newpatalloc);
|
|
patcode = patout + patsize;
|
|
patalloc = newpatalloc;
|
|
}
|
|
patsize = newpatsize;
|
|
if (add) {
|
|
if (paflags & PA_UNMETA) {
|
|
/*
|
|
* Unmetafy and untokenize the string as we go.
|
|
* The Meta characters in add aren't counted in n.
|
|
*/
|
|
while (n--) {
|
|
if (itok(*add))
|
|
*patcode++ = ztokens[*add++ - Pound];
|
|
else if (*add == Meta) {
|
|
add++;
|
|
*patcode++ = *add++ ^ 32;
|
|
} else {
|
|
*patcode++ = *add++;
|
|
}
|
|
}
|
|
} else {
|
|
while (n--)
|
|
*patcode++ = *add++;
|
|
}
|
|
} else
|
|
*patcode++ = ch;
|
|
patcode = patout + patsize;
|
|
}
|
|
|
|
static long rn_offs;
|
|
/* operates on pointers to union upat, returns a pointer */
|
|
#define PATNEXT(p) ((rn_offs = P_NEXT(p)) ? \
|
|
(P_OP(p) == P_BACK) ? \
|
|
((p)-rn_offs) : ((p)+rn_offs) : NULL)
|
|
|
|
/* Called before parsing a set of file matchs to initialize flags */
|
|
|
|
/**/
|
|
void
|
|
patcompstart(void)
|
|
{
|
|
if (isset(CASEGLOB))
|
|
patglobflags = 0;
|
|
else
|
|
patglobflags = GF_IGNCASE;
|
|
}
|
|
|
|
/*
|
|
* Top level pattern compilation subroutine
|
|
* exp is a null-terminated, metafied string.
|
|
* inflags is an or of some PAT_* flags.
|
|
* endexp, if non-null, is set to a pointer to the end of the
|
|
* part of exp which was compiled. This is used when
|
|
* compiling patterns for directories which must be
|
|
* matched recursively.
|
|
*/
|
|
|
|
/**/
|
|
mod_export Patprog
|
|
patcompile(char *exp, int inflags, char **endexp)
|
|
{
|
|
int flags = 0;
|
|
long len = 0;
|
|
long startoff;
|
|
Upat pscan;
|
|
char *lng, *strp = NULL;
|
|
Patprog p;
|
|
|
|
startoff = sizeof(struct patprog);
|
|
/* Ensure alignment of start of program string */
|
|
startoff = (startoff + sizeof(union upat) - 1) & ~(sizeof(union upat) - 1);
|
|
|
|
/* Allocate reasonable sized chunk if none, reduce size if too big */
|
|
if (patalloc != P_DEF_ALLOC)
|
|
patout = (char *)zrealloc(patout, patalloc = P_DEF_ALLOC);
|
|
patcode = patout + startoff;
|
|
patsize = patcode - patout;
|
|
patstart = patparse = exp;
|
|
/*
|
|
* Note global patnpar numbers parentheses 1..9, while patnpar
|
|
* in struct is actual count of parentheses.
|
|
*/
|
|
patnpar = 1;
|
|
patflags = inflags & ~(PAT_PURES|PAT_HAS_EXCLUDP);
|
|
|
|
patendseg = endseg;
|
|
patendseglen = isset(EXTENDEDGLOB) ? PATENDSEGLEN_EXT : PATENDSEGLEN_NORM;
|
|
patendstr = endstr;
|
|
patendstrlen = isset(EXTENDEDGLOB) ? PATENDSTRLEN_EXT : PATENDSTRLEN_NORM;
|
|
|
|
if (!(patflags & PAT_FILE)) {
|
|
patendseg++;
|
|
patendstr++;
|
|
patendseglen--;
|
|
patendstrlen--;
|
|
remnulargs(patparse);
|
|
patglobflags = 0;
|
|
}
|
|
/*
|
|
* Have to be set now, since they get updated during compilation.
|
|
*/
|
|
((Patprog)patout)->globflags = patglobflags;
|
|
|
|
if (!(patflags & PAT_ANY)) {
|
|
/* Look for a really pure string, with no tokens at all. */
|
|
if (!patglobflags
|
|
#ifdef __CYGWIN__
|
|
/*
|
|
* If the OS treats files case-insensitively and we
|
|
* are looking at files, we don't need to use pattern
|
|
* matching to find the file.
|
|
*/
|
|
|| (!(patglobflags & ~GF_IGNCASE) && (patflags & PAT_FILE))
|
|
#endif
|
|
)
|
|
{
|
|
/*
|
|
* Waah! I wish I understood this.
|
|
* Empty metafied strings have an initial Nularg.
|
|
* This never corresponds to a real character in
|
|
* a glob pattern or string, so skip it.
|
|
*/
|
|
if (*exp == Nularg)
|
|
exp++;
|
|
for (strp = exp; *strp &&
|
|
(!(patflags & PAT_FILE) || *strp != '/') && !itok(*strp);
|
|
strp++)
|
|
;
|
|
}
|
|
if (!strp || (*strp && *strp != '/')) {
|
|
/* No, do normal compilation. */
|
|
strp = NULL;
|
|
if (patcompswitch(0, &flags) == 0)
|
|
return NULL;
|
|
} else {
|
|
/*
|
|
* Yes, copy the string, and skip compilation altogether.
|
|
* Null terminate for the benefit of globbing.
|
|
* Leave metafied both for globbing and for our own
|
|
* efficiency.
|
|
*/
|
|
patparse = strp;
|
|
len = strp - exp;
|
|
patadd(exp, 0, len + 1, 0);
|
|
patout[startoff + len] = '\0';
|
|
patflags |= PAT_PURES;
|
|
}
|
|
}
|
|
|
|
/* end of compilation: safe to use pointers */
|
|
p = (Patprog)patout;
|
|
p->startoff = startoff;
|
|
p->patstartch = '\0';
|
|
p->globend = patglobflags;
|
|
p->flags = patflags;
|
|
p->mustoff = 0;
|
|
p->size = patsize;
|
|
p->patmlen = len;
|
|
p->patnpar = patnpar-1;
|
|
|
|
if (!strp) {
|
|
pscan = (Upat)(patout + startoff);
|
|
|
|
if (!(patflags & PAT_ANY) && P_OP(PATNEXT(pscan)) == P_END) {
|
|
/* only one top level choice */
|
|
pscan = P_OPERAND(pscan);
|
|
|
|
if (flags & P_PURESTR) {
|
|
/*
|
|
* The pattern can be matched with a simple strncmp/strcmp.
|
|
* Careful in case we've overwritten the node for the next ptr.
|
|
*/
|
|
char *dst = patout + startoff;
|
|
Upat next;
|
|
p->flags |= PAT_PURES;
|
|
for (; pscan; pscan = next) {
|
|
next = PATNEXT(pscan);
|
|
if (P_OP(pscan) == P_EXACTLY) {
|
|
char *opnd = P_LS_STR(pscan), *mtest;
|
|
long oplen = P_LS_LEN(pscan), ilen;
|
|
int nmeta = 0;
|
|
/*
|
|
* Unfortunately we unmetafied the string
|
|
* and we need to put any metacharacters
|
|
* back now we know it's a pure string.
|
|
* This shouldn't happen too often, it's
|
|
* just that there are some cases such
|
|
* as . and .. in files where we really
|
|
* need a pure string even if there are
|
|
* pattern characters flying around.
|
|
*/
|
|
for (mtest = opnd, ilen = oplen; ilen;
|
|
mtest++, ilen--)
|
|
if (imeta(*mtest))
|
|
nmeta++;
|
|
if (nmeta) {
|
|
char *oldpatout = patout;
|
|
patadd(NULL, 0, nmeta, 0);
|
|
/*
|
|
* Yuk.
|
|
*/
|
|
p = (Patprog)patout;
|
|
opnd = patout + (opnd - oldpatout);
|
|
dst = patout + startoff;
|
|
}
|
|
|
|
while (oplen--) {
|
|
if (imeta(*opnd)) {
|
|
*dst++ = Meta;
|
|
*dst++ = *opnd ^ 32;
|
|
} else {
|
|
*dst++ = *opnd++;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
p->size = dst - patout;
|
|
/* patmlen is really strlen. We don't need a null. */
|
|
p->patmlen = p->size - startoff;
|
|
} else {
|
|
/* starting point info */
|
|
if (P_OP(pscan) == P_EXACTLY && !p->globflags &&
|
|
P_LS_LEN(pscan))
|
|
p->patstartch = *P_LS_STR(pscan);
|
|
/*
|
|
* Find the longest literal string in something expensive.
|
|
* This is itself not all that cheap if we have
|
|
* case-insensitive matching or approximation, so don't.
|
|
*/
|
|
if ((flags & P_HSTART) && !p->globflags) {
|
|
lng = NULL;
|
|
len = 0;
|
|
for (; pscan; pscan = PATNEXT(pscan))
|
|
if (P_OP(pscan) == P_EXACTLY &&
|
|
P_LS_LEN(pscan) >= len) {
|
|
lng = P_LS_STR(pscan);
|
|
len = P_LS_LEN(pscan);
|
|
}
|
|
if (lng) {
|
|
p->mustoff = lng - patout;
|
|
p->patmlen = len;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
/*
|
|
* The pattern was compiled in a fixed buffer: unless told otherwise,
|
|
* we stick the compiled pattern on the heap. This is necessary
|
|
* for files where we will often be compiling multiple segments at once.
|
|
* But if we get the ZDUP flag we always put it in zalloc()ed memory.
|
|
*/
|
|
if (patflags & PAT_ZDUP) {
|
|
Patprog newp = (Patprog)zalloc(patsize);
|
|
memcpy((char *)newp, (char *)p, patsize);
|
|
p = newp;
|
|
} else if (!(patflags & PAT_STATIC)) {
|
|
Patprog newp = (Patprog)zhalloc(patsize);
|
|
memcpy((char *)newp, (char *)p, patsize);
|
|
p = newp;
|
|
}
|
|
|
|
if (endexp)
|
|
*endexp = patparse;
|
|
return p;
|
|
}
|
|
|
|
/*
|
|
* Main body or parenthesized subexpression in pattern
|
|
* Parenthesis (and any ksh_glob gubbins) will have been removed.
|
|
*/
|
|
|
|
/**/
|
|
static long
|
|
patcompswitch(int paren, int *flagp)
|
|
{
|
|
long starter, br, ender, excsync = 0;
|
|
int parno = 0;
|
|
int flags, gfchanged = 0, savglobflags = patglobflags;
|
|
Upat ptr;
|
|
|
|
*flagp = 0;
|
|
|
|
if (paren && (patglobflags & GF_BACKREF) && patnpar <= NSUBEXP) {
|
|
/*
|
|
* parenthesized: make an open node.
|
|
* We can only refer to the first nine parentheses.
|
|
* For any others, we just use P_OPEN on its own; there's
|
|
* no gain in arbitrarily limiting the number of parentheses.
|
|
*/
|
|
parno = patnpar++;
|
|
starter = patnode(P_OPEN + parno);
|
|
} else
|
|
starter = 0;
|
|
|
|
br = patnode(P_BRANCH);
|
|
if (!patcompbranch(&flags))
|
|
return 0;
|
|
if (patglobflags != savglobflags)
|
|
gfchanged++;
|
|
if (starter)
|
|
pattail(starter, br);
|
|
else
|
|
starter = br;
|
|
|
|
*flagp |= flags & (P_HSTART|P_PURESTR);
|
|
|
|
while (*patparse == Bar ||
|
|
(isset(EXTENDEDGLOB) && *patparse == Tilde &&
|
|
(patparse[1] == '/' ||
|
|
!memchr(patendseg, patparse[1], patendseglen)))) {
|
|
int tilde = *patparse++ == Tilde;
|
|
long gfnode = 0, newbr;
|
|
|
|
*flagp &= ~P_PURESTR;
|
|
|
|
if (tilde) {
|
|
union upat up;
|
|
/* excsync remembers the P_EXCSYNC node before a chain of
|
|
* exclusions: all point back to this. only the
|
|
* original (non-excluded) branch gets a trailing P_EXCSYNC.
|
|
*/
|
|
if (!excsync) {
|
|
excsync = patnode(P_EXCSYNC);
|
|
patoptail(br, excsync);
|
|
}
|
|
/*
|
|
* By default, approximations are turned off in exclusions:
|
|
* we need to do this here as otherwise the code compiling
|
|
* the exclusion doesn't know if the flags have really
|
|
* changed if the error count gets restored.
|
|
*/
|
|
patglobflags &= ~0xff;
|
|
if (!(patflags & PAT_FILET) || paren) {
|
|
br = patnode(P_EXCLUDE);
|
|
} else {
|
|
/*
|
|
* At top level (paren == 0) in a file glob !(patflags
|
|
* &PAT_FILET) do the exclusion prepending the file path
|
|
* so far. We need to flag this to avoid unnecessarily
|
|
* copying the path.
|
|
*/
|
|
br = patnode(P_EXCLUDP);
|
|
patflags |= PAT_HAS_EXCLUDP;
|
|
}
|
|
up.p = NULL;
|
|
patadd((char *)&up, 0, sizeof(up), 0);
|
|
/* / is not treated as special if we are at top level */
|
|
if (!paren && *patendseg == '/') {
|
|
tilde++;
|
|
patendseg++;
|
|
patendseglen--;
|
|
patendstr++;
|
|
patendstrlen--;
|
|
}
|
|
} else {
|
|
excsync = 0;
|
|
br = patnode(P_BRANCH);
|
|
/*
|
|
* The position of the following statements means globflags
|
|
* set in the main branch carry over to the exclusion.
|
|
*/
|
|
if (!paren) {
|
|
patglobflags = 0;
|
|
if (((Patprog)patout)->globflags) {
|
|
/*
|
|
* If at top level, we need to reinitialize flags to zero,
|
|
* since (#i)foo|bar only applies to foo and we stuck
|
|
* the #i into the global flags.
|
|
* We could have done it so that they only got set in the
|
|
* first branch, but it's quite convenient having any
|
|
* global flags set in the header and not buried in the
|
|
* pattern. (Or maybe it isn't and we should
|
|
* forget this bit and always stick in an explicit GFLAGS
|
|
* statement instead of using the header.)
|
|
* Also, this can't happen for file globs where there are
|
|
* no top-level |'s.
|
|
*
|
|
* No gfchanged, as nothing to follow branch at top
|
|
* level.
|
|
*/
|
|
union upat up;
|
|
gfnode = patnode(P_GFLAGS);
|
|
up.l = patglobflags;
|
|
patadd((char *)&up, 0, sizeof(union upat), 0);
|
|
}
|
|
} else {
|
|
patglobflags = savglobflags;
|
|
}
|
|
}
|
|
newbr = patcompbranch(&flags);
|
|
if (tilde == 2) {
|
|
/* restore special treatment of / */
|
|
patendseg--;
|
|
patendseglen++;
|
|
patendstr--;
|
|
patendstrlen++;
|
|
}
|
|
if (!newbr)
|
|
return 0;
|
|
if (gfnode)
|
|
pattail(gfnode, newbr);
|
|
if (!tilde && patglobflags != savglobflags)
|
|
gfchanged++;
|
|
pattail(starter, br);
|
|
if (excsync)
|
|
patoptail(br, patnode(P_EXCEND));
|
|
*flagp |= flags & P_HSTART;
|
|
}
|
|
|
|
/*
|
|
* Make a closing node, hooking it to the end.
|
|
* Note that we can't optimize P_NOTHING out here, since another
|
|
* branch at that point would indicate the current choices continue,
|
|
* which they don't.
|
|
*/
|
|
ender = patnode(paren ? parno ? P_CLOSE+parno : P_NOTHING : P_END);
|
|
pattail(starter, ender);
|
|
|
|
/*
|
|
* Hook the tails of the branches to the closing node,
|
|
* except for exclusions which terminate where they are.
|
|
*/
|
|
for (ptr = (Upat)patout + starter; ptr; ptr = PATNEXT(ptr))
|
|
if (!P_ISEXCLUDE(ptr))
|
|
patoptail(ptr-(Upat)patout, ender);
|
|
|
|
/* check for proper termination */
|
|
if ((paren && *patparse++ != Outpar) ||
|
|
(!paren && *patparse &&
|
|
!((patflags & PAT_FILE) && *patparse == '/')))
|
|
return 0;
|
|
|
|
if (paren && gfchanged) {
|
|
/*
|
|
* Restore old values of flags when leaving parentheses.
|
|
* gfchanged detects a change in any branch (except exclusions
|
|
* which are separate), since we need to emit this even if
|
|
* a later branch happened to put the flags back.
|
|
*/
|
|
pattail(ender, patnode(P_GFLAGS));
|
|
patglobflags = savglobflags;
|
|
patadd((char *)&savglobflags, 0, sizeof(long), 0);
|
|
}
|
|
|
|
return starter;
|
|
}
|
|
|
|
/*
|
|
* Compile something ended by Bar, Outpar, Tilde, or end of string.
|
|
* Note the BRANCH or EXCLUDE tag must already have been omitted:
|
|
* this returns the position of the operand of that.
|
|
*/
|
|
|
|
/**/
|
|
static long
|
|
patcompbranch(int *flagp)
|
|
{
|
|
long chain, latest = 0, starter;
|
|
int flags = 0;
|
|
|
|
*flagp = P_PURESTR;
|
|
|
|
starter = chain = 0;
|
|
while (!memchr(patendseg, *patparse, patendseglen) ||
|
|
(*patparse == Tilde && patparse[1] != '/' &&
|
|
memchr(patendseg, patparse[1], patendseglen))) {
|
|
if (isset(EXTENDEDGLOB) &&
|
|
((!isset(SHGLOB) &&
|
|
(*patparse == Inpar && patparse[1] == Pound)) ||
|
|
(isset(KSHGLOB) && *patparse == '@' && patparse[1] == Inpar &&
|
|
patparse[2] == Pound))) {
|
|
/* Globbing flags. */
|
|
char *pp1 = patparse;
|
|
int oldglobflags = patglobflags, ignore;
|
|
long assert;
|
|
patparse += (*patparse == '@') ? 3 : 2;
|
|
if (!patgetglobflags(&patparse, &assert, &ignore))
|
|
return 0;
|
|
if (!ignore) {
|
|
if (assert) {
|
|
/*
|
|
* Start/end assertion looking like flags, but
|
|
* actually handled as a normal node
|
|
*/
|
|
latest = patnode(assert);
|
|
flags = 0;
|
|
} else {
|
|
if (pp1 == patstart) {
|
|
/* Right at start of pattern, the simplest case.
|
|
* Put them into the flags and don't emit anything.
|
|
*/
|
|
((Patprog)patout)->globflags = patglobflags;
|
|
continue;
|
|
} else if (!*patparse) {
|
|
/* Right at the end, so just leave the flags for
|
|
* the next Patprog in the chain to pick up.
|
|
*/
|
|
break;
|
|
}
|
|
/*
|
|
* Otherwise, we have to stick them in as a pattern
|
|
* matching nothing.
|
|
*/
|
|
if (oldglobflags != patglobflags) {
|
|
/* Flags changed */
|
|
union upat up;
|
|
latest = patnode(P_GFLAGS);
|
|
up.l = patglobflags;
|
|
patadd((char *)&up, 0, sizeof(union upat), 0);
|
|
} else {
|
|
/* No effect. */
|
|
continue;
|
|
}
|
|
}
|
|
} else if (!*patparse)
|
|
break;
|
|
else
|
|
continue;
|
|
} else if (isset(EXTENDEDGLOB) && *patparse == Hat) {
|
|
/*
|
|
* ^pat: anything but pat. For proper backtracking,
|
|
* etc., we turn this into (*~pat), except without the
|
|
* parentheses.
|
|
*/
|
|
patparse++;
|
|
latest = patcompnot(0, &flags);
|
|
} else
|
|
latest = patcomppiece(&flags);
|
|
if (!latest)
|
|
return 0;
|
|
if (!starter)
|
|
starter = latest;
|
|
if (!(flags & P_PURESTR))
|
|
*flagp &= ~P_PURESTR;
|
|
if (!chain)
|
|
*flagp |= flags & P_HSTART;
|
|
else
|
|
pattail(chain, latest);
|
|
chain = latest;
|
|
}
|
|
/* check if there was nothing in the loop, i.e. () */
|
|
if (!chain)
|
|
starter = patnode(P_NOTHING);
|
|
|
|
return starter;
|
|
}
|
|
|
|
/* get glob flags, return 1 for success, 0 for failure */
|
|
|
|
/**/
|
|
int
|
|
patgetglobflags(char **strp, long *assertp, int *ignore)
|
|
{
|
|
char *nptr, *ptr = *strp;
|
|
zlong ret;
|
|
|
|
*assertp = 0;
|
|
*ignore = 1;
|
|
/* (#X): assumes we are still positioned on the first X */
|
|
for (; *ptr && *ptr != Outpar; ptr++) {
|
|
if (*ptr == 'q') {
|
|
/* Glob qualifiers, ignored in pattern code */
|
|
while (*ptr && *ptr != Outpar)
|
|
ptr++;
|
|
break;
|
|
} else {
|
|
*ignore = 0;
|
|
switch (*ptr) {
|
|
case 'a':
|
|
/* Approximate matching, max no. of errors follows */
|
|
ret = zstrtol(++ptr, &nptr, 10);
|
|
/*
|
|
* We can't have more than 254, because we need 255 to
|
|
* mark 254 errors in wbranch and exclude sync strings
|
|
* (hypothetically --- hope no-one tries it).
|
|
*/
|
|
if (ret < 0 || ret > 254 || ptr == nptr)
|
|
return 0;
|
|
patglobflags = (patglobflags & ~0xff) | (ret & 0xff);
|
|
ptr = nptr-1;
|
|
break;
|
|
|
|
case 'l':
|
|
/* Lowercase in pattern matches lower or upper in target */
|
|
patglobflags = (patglobflags & ~GF_IGNCASE) | GF_LCMATCHUC;
|
|
break;
|
|
|
|
case 'i':
|
|
/* Fully case insensitive */
|
|
patglobflags = (patglobflags & ~GF_LCMATCHUC) | GF_IGNCASE;
|
|
break;
|
|
|
|
case 'I':
|
|
/* Restore case sensitivity */
|
|
patglobflags &= ~(GF_LCMATCHUC|GF_IGNCASE);
|
|
break;
|
|
|
|
case 'b':
|
|
/* Make backreferences */
|
|
patglobflags |= GF_BACKREF;
|
|
break;
|
|
|
|
case 'B':
|
|
/* Don't make backreferences */
|
|
patglobflags &= ~GF_BACKREF;
|
|
break;
|
|
|
|
case 'm':
|
|
/* Make references to complete match */
|
|
patglobflags |= GF_MATCHREF;
|
|
break;
|
|
|
|
case 'M':
|
|
/* Don't */
|
|
patglobflags &= ~GF_MATCHREF;
|
|
break;
|
|
|
|
case 's':
|
|
*assertp = P_ISSTART;
|
|
break;
|
|
|
|
case 'e':
|
|
*assertp = P_ISEND;
|
|
break;
|
|
|
|
default:
|
|
return 0;
|
|
}
|
|
}
|
|
}
|
|
if (*ptr != Outpar)
|
|
return 0;
|
|
/* Start/end assertions must appear on their own. */
|
|
if (*assertp && (*strp)[1] != Outpar)
|
|
return 0;
|
|
*strp = ptr + 1;
|
|
return 1;
|
|
}
|
|
|
|
/*
|
|
* compile a chunk such as a literal string or a [...] followed
|
|
* by a possible hash operator
|
|
*/
|
|
|
|
/**/
|
|
static long
|
|
patcomppiece(int *flagp)
|
|
{
|
|
long starter = 0, next, pound, op;
|
|
int flags, flags2, kshchar, len, ch, patch, nmeta;
|
|
union upat up;
|
|
char *nptr, *str0, *ptr, cbuf[2];
|
|
zrange_t from, to;
|
|
|
|
flags = 0;
|
|
str0 = patparse;
|
|
for (;;) {
|
|
/*
|
|
* Check if we have a string. First, we need to make sure
|
|
* the string doesn't introduce a ksh-like parenthesized expression.
|
|
*/
|
|
kshchar = '\0';
|
|
if (isset(KSHGLOB) && *patparse && patparse[1] == Inpar) {
|
|
if (strchr("?*+!@", *patparse))
|
|
kshchar = STOUC(*patparse);
|
|
else if (*patparse == Star || *patparse == Quest)
|
|
kshchar = STOUC(ztokens[*patparse - Pound]);
|
|
}
|
|
|
|
/*
|
|
* End of string (or no string at all) if ksh-type parentheses,
|
|
* or special character, unless that character is a tilde and
|
|
* the character following is an end-of-segment character. Thus
|
|
* tildes are not special if there is nothing following to
|
|
* be excluded.
|
|
*/
|
|
if (kshchar || (memchr(patendstr, *patparse, patendstrlen) &&
|
|
(*patparse != Tilde ||
|
|
patparse[1] == '/' ||
|
|
!memchr(patendseg, patparse[1], patendseglen))))
|
|
break;
|
|
|
|
METAINC(patparse);
|
|
}
|
|
|
|
if (patparse > str0) {
|
|
long slen = patparse - str0;
|
|
int morelen;
|
|
|
|
/* Ordinary string: cancel kshchar lookahead */
|
|
kshchar = '\0';
|
|
/*
|
|
* Assume it matches a simple string until we find otherwise.
|
|
*/
|
|
flags |= P_PURESTR;
|
|
DPUTS(patparse == str0, "BUG: matched nothing in patcomppiece.");
|
|
/* more than one character matched? */
|
|
morelen = str0 + (*str0 == Meta ? 2 : 1) < patparse;
|
|
/*
|
|
* If we have more than one character, a following hash only
|
|
* applies to the last, so decrement.
|
|
*/
|
|
if (isset(EXTENDEDGLOB) && *patparse == Pound && morelen)
|
|
patparse -= (patparse > str0 + 1 && patparse[-2] == Meta) ? 2 : 1;
|
|
/*
|
|
* If len is 1, we can't have an active # following, so doesn't
|
|
* matter that we don't make X in `XX#' simple.
|
|
*/
|
|
if (!morelen)
|
|
flags |= P_SIMPLE;
|
|
starter = patnode(P_EXACTLY);
|
|
|
|
/* Get length of string without metafication. */
|
|
nmeta = 0;
|
|
/* inherited from domatch, but why, exactly? */
|
|
if (*str0 == Nularg)
|
|
str0++;
|
|
for (ptr = str0; ptr < patparse; ptr++) {
|
|
if (*ptr == Meta) {
|
|
nmeta++;
|
|
ptr++;
|
|
}
|
|
}
|
|
slen = (patparse - str0) - nmeta;
|
|
/* First add length, which is a long */
|
|
patadd((char *)&slen, 0, sizeof(long), 0);
|
|
/*
|
|
* Then the string, not null terminated.
|
|
* Unmetafy and untokenize; pass the final length,
|
|
* which is what we need to allocate, i.e. not including
|
|
* a count for each Meta in the string.
|
|
*/
|
|
patadd(str0, 0, slen, PA_UNMETA);
|
|
nptr = P_LS_STR((Upat)patout + starter);
|
|
/*
|
|
* It's much simpler to turn off pure string mode for
|
|
* any case-insensitive or approximate matching; usually,
|
|
* that is correct, or they wouldn't have been turned on.
|
|
* However, we need to make sure we match a "." or ".."
|
|
* in a file name as a pure string. There's a minor bug
|
|
* that this will also apply to something like
|
|
* ..(#a1).. (i.e. the (#a1) has no effect), but if you're
|
|
* going to write funny patterns, you get no sympathy from me.
|
|
*/
|
|
if (patglobflags) {
|
|
if (!(patflags & PAT_FILE))
|
|
flags &= ~P_PURESTR;
|
|
else if (!(nptr[0] == '.' &&
|
|
(slen == 1 || (nptr[1] == '.' && slen == 2))))
|
|
flags &= ~P_PURESTR;
|
|
}
|
|
} else {
|
|
if (kshchar)
|
|
patparse++;
|
|
|
|
patch = *patparse;
|
|
METAINC(patparse);
|
|
switch(patch) {
|
|
case Quest:
|
|
flags |= P_SIMPLE;
|
|
starter = patnode(P_ANY);
|
|
break;
|
|
case Star:
|
|
/* kshchar is used as a sign that we can't have #'s. */
|
|
kshchar = -1;
|
|
starter = patnode(P_STAR);
|
|
break;
|
|
case Inbrack:
|
|
flags |= P_SIMPLE;
|
|
if (*patparse == Hat || *patparse == '^' || *patparse == '!') {
|
|
patparse++;
|
|
starter = patnode(P_ANYBUT);
|
|
} else
|
|
starter = patnode(P_ANYOF);
|
|
if (*patparse == Outbrack) {
|
|
patparse++;
|
|
patadd(NULL, ']', 1, PA_NOALIGN);
|
|
}
|
|
while (*patparse && *patparse != Outbrack) {
|
|
/* Meta is not a token */
|
|
if (*patparse == Inbrack && patparse[1] == ':' &&
|
|
(nptr = strchr(patparse+2, ':')) &&
|
|
nptr[1] == Outbrack) {
|
|
/* Posix range. */
|
|
patparse += 2;
|
|
len = nptr - patparse;
|
|
if (!strncmp(patparse, "alpha", len))
|
|
ch = PP_ALPHA;
|
|
else if (!strncmp(patparse, "alnum", len))
|
|
ch = PP_ALNUM;
|
|
else if (!strncmp(patparse, "ascii", len))
|
|
ch = PP_ASCII;
|
|
else if (!strncmp(patparse, "blank", len))
|
|
ch = PP_BLANK;
|
|
else if (!strncmp(patparse, "cntrl", len))
|
|
ch = PP_CNTRL;
|
|
else if (!strncmp(patparse, "digit", len))
|
|
ch = PP_DIGIT;
|
|
else if (!strncmp(patparse, "graph", len))
|
|
ch = PP_GRAPH;
|
|
else if (!strncmp(patparse, "lower", len))
|
|
ch = PP_LOWER;
|
|
else if (!strncmp(patparse, "print", len))
|
|
ch = PP_PRINT;
|
|
else if (!strncmp(patparse, "punct", len))
|
|
ch = PP_PUNCT;
|
|
else if (!strncmp(patparse, "space", len))
|
|
ch = PP_SPACE;
|
|
else if (!strncmp(patparse, "upper", len))
|
|
ch = PP_UPPER;
|
|
else if (!strncmp(patparse, "xdigit", len))
|
|
ch = PP_XDIGIT;
|
|
else if (!strncmp(patparse, "IDENT", len))
|
|
ch = PP_IDENT;
|
|
else if (!strncmp(patparse, "IFS", len))
|
|
ch = PP_IFS;
|
|
else if (!strncmp(patparse, "IFSSPACE", len))
|
|
ch = PP_IFSSPACE;
|
|
else if (!strncmp(patparse, "WORD", len))
|
|
ch = PP_WORD;
|
|
else
|
|
ch = PP_UNKWN;
|
|
patparse = nptr + 2;
|
|
if (ch != PP_UNKWN)
|
|
patadd(NULL, STOUC(Meta+ch), 1, PA_NOALIGN);
|
|
continue;
|
|
}
|
|
if (itok(*patparse)) {
|
|
cbuf[0] = ztokens[*patparse - Pound];
|
|
} else if (*patparse == Meta) {
|
|
cbuf[0] = Meta;
|
|
cbuf[1] = *++patparse;
|
|
} else
|
|
cbuf[0] = *patparse;
|
|
patparse++;
|
|
|
|
if (*patparse == '-' && patparse[1] != Outbrack) {
|
|
patadd(NULL, STOUC(Meta+PP_RANGE), 1, PA_NOALIGN);
|
|
patadd(cbuf, 0, (cbuf[0] == Meta) ? 2 : 1, PA_NOALIGN);
|
|
if (itok(*++patparse)) {
|
|
patadd(0, STOUC(ztokens[*patparse - Pound]), 1,
|
|
PA_NOALIGN);
|
|
} else
|
|
patadd(patparse, 0, (*patparse == Meta) ? 2 : 1,
|
|
PA_NOALIGN);
|
|
METAINC(patparse);
|
|
} else
|
|
patadd(cbuf, 0, (cbuf[0] == Meta) ? 2 : 1, PA_NOALIGN);
|
|
}
|
|
if (*patparse != Outbrack)
|
|
return 0;
|
|
patparse++;
|
|
/* terminate null string and fix alignment */
|
|
patadd(NULL, 0, 1, 0);
|
|
break;
|
|
case Inpar:
|
|
/* is this how to treat parentheses in SHGLOB? */
|
|
if (isset(SHGLOB) && !kshchar)
|
|
return 0;
|
|
if (kshchar == '!') {
|
|
/* This is nasty, we should really either handle all
|
|
* kshglobbing below or here. But most of the
|
|
* others look like non-ksh patterns, while this one
|
|
* doesn't, so we handle it here and leave the rest.
|
|
* We treat it like an extendedglob ^, except that
|
|
* it goes into parentheses.
|
|
*
|
|
* If we did do kshglob here, we could support
|
|
* the old behaviour that things like !(foo)##
|
|
* work, but it makes the code more complicated at
|
|
* the expense of allowing the user to do things
|
|
* they shouldn't.
|
|
*/
|
|
if (!(starter = patcompnot(1, &flags2)))
|
|
return 0;
|
|
} else if (!(starter = patcompswitch(1, &flags2)))
|
|
return 0;
|
|
flags |= flags2 & P_HSTART;
|
|
break;
|
|
case Inang:
|
|
/* Numeric glob */
|
|
len = 0; /* beginning present 1, end present 2 */
|
|
if (idigit(*patparse)) {
|
|
from = (zrange_t) zstrtol((char *)patparse,
|
|
(char **)&nptr, 10);
|
|
patparse = nptr;
|
|
len |= 1;
|
|
}
|
|
DPUTS(*patparse != '-', "BUG: - missing from numeric glob");
|
|
patparse++;
|
|
if (idigit(*patparse)) {
|
|
to = (zrange_t) zstrtol((char *)patparse,
|
|
(char **)&nptr, 10);
|
|
patparse = nptr;
|
|
len |= 2;
|
|
}
|
|
if (*patparse != Outang)
|
|
return 0;
|
|
patparse++;
|
|
switch(len) {
|
|
case 3:
|
|
starter = patnode(P_NUMRNG);
|
|
patadd((char *)&from, 0, sizeof(from), 0);
|
|
patadd((char *)&to, 0, sizeof(to), 0);
|
|
break;
|
|
case 2:
|
|
starter = patnode(P_NUMTO);
|
|
patadd((char *)&to, 0, sizeof(to), 0);
|
|
break;
|
|
case 1:
|
|
starter = patnode(P_NUMFROM);
|
|
patadd((char *)&from, 0, sizeof(from), 0);
|
|
break;
|
|
case 0:
|
|
starter = patnode(P_NUMANY);
|
|
break;
|
|
}
|
|
/* This can't be simple, because it isn't.
|
|
* Mention in manual that matching digits with [...]
|
|
* is more efficient.
|
|
*/
|
|
break;
|
|
case Pound:
|
|
DPUTS(!isset(EXTENDEDGLOB), "BUG: # not treated as string");
|
|
/*
|
|
* A hash here is an error; it should follow something
|
|
* repeatable.
|
|
*/
|
|
return 0;
|
|
break;
|
|
#ifdef DEBUG
|
|
default:
|
|
dputs("BUG: character not handled in patcomppiece");
|
|
return 0;
|
|
break;
|
|
#endif
|
|
}
|
|
}
|
|
|
|
if (!(pound = (*patparse == Pound && isset(EXTENDEDGLOB))) &&
|
|
(kshchar <= 0 || kshchar == '@' || kshchar == '!')) {
|
|
*flagp = flags;
|
|
return starter;
|
|
}
|
|
|
|
/* too much at once doesn't currently work */
|
|
if (kshchar && pound)
|
|
return 0;
|
|
|
|
if (kshchar == '*') {
|
|
op = P_ONEHASH;
|
|
*flagp = P_HSTART;
|
|
} else if (kshchar == '+') {
|
|
op = P_TWOHASH;
|
|
*flagp = P_HSTART;
|
|
} else if (kshchar == '?') {
|
|
op = 0;
|
|
*flagp = 0;
|
|
} else if (*++patparse == Pound) {
|
|
op = P_TWOHASH;
|
|
patparse++;
|
|
*flagp = P_HSTART;
|
|
} else {
|
|
op = P_ONEHASH;
|
|
*flagp = P_HSTART;
|
|
}
|
|
|
|
/*
|
|
* Note optimizations with pointers into P_NOTHING branches: some
|
|
* should logically point to next node after current piece.
|
|
*
|
|
* Backtracking is also encoded in a slightly obscure way: the
|
|
* code emitted ensures we test the non-empty branch of complex
|
|
* patterns before the empty branch on each repetition. Hence
|
|
* each time we fail on a non-empty branch, we try the empty branch,
|
|
* which is equivalent to backtracking.
|
|
*/
|
|
if ((flags & P_SIMPLE) && (op == P_ONEHASH || op == P_TWOHASH) &&
|
|
P_OP((Upat)patout+starter) == P_ANY) {
|
|
/* Optimize ?# to *. Silly thing to do, since who would use
|
|
* use ?# ? But it makes the later code shorter.
|
|
*/
|
|
Upat uptr = (Upat)patout + starter;
|
|
if (op == P_TWOHASH) {
|
|
/* ?## becomes ?* */
|
|
uptr->l = (uptr->l & ~0xff) | P_ANY;
|
|
pattail(starter, patnode(P_STAR));
|
|
} else {
|
|
uptr->l = (uptr->l & ~0xff) | P_STAR;
|
|
}
|
|
} else if ((flags & P_SIMPLE) && op && !(patglobflags & 0xff)) {
|
|
/* Don't simplify if we need to look for approximations. */
|
|
patinsert(op, starter, NULL, 0);
|
|
} else if (op == P_ONEHASH) {
|
|
/* Emit x# as (x&|), where & means "self". */
|
|
up.p = NULL;
|
|
patinsert(P_WBRANCH, starter, (char *)&up, sizeof(up));
|
|
/* Either x */
|
|
patoptail(starter, patnode(P_BACK)); /* and loop */
|
|
patoptail(starter, starter); /* back */
|
|
pattail(starter, patnode(P_BRANCH)); /* or */
|
|
pattail(starter, patnode(P_NOTHING)); /* null. */
|
|
} else if (op == P_TWOHASH) {
|
|
/* Emit x## as x(&|) where & means "self". */
|
|
next = patnode(P_WBRANCH); /* Either */
|
|
up.p = NULL;
|
|
patadd((char *)&up, 0, sizeof(up), 0);
|
|
pattail(starter, next);
|
|
pattail(patnode(P_BACK), starter); /* loop back */
|
|
pattail(next, patnode(P_BRANCH)); /* or */
|
|
pattail(starter, patnode(P_NOTHING)); /* null. */
|
|
} else if (kshchar == '?') {
|
|
/* Emit ?(x) as (x|) */
|
|
patinsert(P_BRANCH, starter, NULL, 0); /* Either x */
|
|
pattail(starter, patnode(P_BRANCH)); /* or */
|
|
next = patnode(P_NOTHING); /* null */
|
|
pattail(starter, next);
|
|
patoptail(starter, next);
|
|
}
|
|
if (*patparse == Pound)
|
|
return 0;
|
|
|
|
return starter;
|
|
}
|
|
|
|
/*
|
|
* Turn a ^foo (paren = 0) or !(foo) (paren = 1) into *~foo with
|
|
* parentheses if necessary. As you see, that's really quite easy.
|
|
*/
|
|
|
|
/**/
|
|
static long
|
|
patcompnot(int paren, int *flagsp)
|
|
{
|
|
union upat up;
|
|
long excsync, br, excl, n, starter;
|
|
int dummy;
|
|
|
|
/* Here, we're matching a star at the start. */
|
|
*flagsp = P_HSTART;
|
|
|
|
starter = patnode(P_BRANCH);
|
|
br = patnode(P_STAR);
|
|
excsync = patnode(P_EXCSYNC);
|
|
pattail(br, excsync);
|
|
pattail(starter, excl = patnode(P_EXCLUDE));
|
|
up.p = NULL;
|
|
patadd((char *)&up, 0, sizeof(up), 0);
|
|
if (!(br = (paren ? patcompswitch(1, &dummy) : patcompbranch(&dummy))))
|
|
return 0;
|
|
pattail(br, patnode(P_EXCEND));
|
|
n = patnode(P_NOTHING); /* just so much easier */
|
|
pattail(excsync, n);
|
|
pattail(excl, n);
|
|
|
|
return starter;
|
|
}
|
|
|
|
/* Emit a node */
|
|
|
|
/**/
|
|
static long
|
|
patnode(long op)
|
|
{
|
|
long starter = (Upat)patcode - (Upat)patout;
|
|
union upat up;
|
|
|
|
up.l = op;
|
|
patadd((char *)&up, 0, sizeof(union upat), 0);
|
|
return starter;
|
|
}
|
|
|
|
/*
|
|
* insert an operator in front of an already emitted operand:
|
|
* we relocate the operand. there had better be nothing else after.
|
|
*/
|
|
|
|
/**/
|
|
static void
|
|
patinsert(long op, int opnd, char *xtra, int sz)
|
|
{
|
|
char *src, *dst, *opdst;
|
|
union upat buf, *lptr;
|
|
|
|
buf.l = 0;
|
|
patadd((char *)&buf, 0, sizeof(buf), 0);
|
|
if (sz)
|
|
patadd(xtra, 0, sz, 0);
|
|
src = patcode - sizeof(union upat) - sz;
|
|
dst = patcode;
|
|
opdst = patout + opnd * sizeof(union upat);
|
|
while (src > opdst)
|
|
*--dst = *--src;
|
|
|
|
/* A cast can't be an lvalue */
|
|
lptr = (Upat)opdst;
|
|
lptr->l = op;
|
|
opdst += sizeof(union upat);
|
|
while (sz--)
|
|
*opdst++ = *xtra++;
|
|
}
|
|
|
|
/* set the 'next' pointer at the end of a node chain */
|
|
|
|
/**/
|
|
static void
|
|
pattail(long p, long val)
|
|
{
|
|
Upat scan, temp;
|
|
long offset;
|
|
|
|
scan = (Upat)patout + p;
|
|
for (;;) {
|
|
if (!(temp = PATNEXT(scan)))
|
|
break;
|
|
scan = temp;
|
|
}
|
|
|
|
offset = (P_OP(scan) == P_BACK)
|
|
? (scan - (Upat)patout) - val : val - (scan - (Upat)patout);
|
|
|
|
scan->l |= offset << 8;
|
|
}
|
|
|
|
/* do pattail, but on operand of first argument; nop if operandless */
|
|
|
|
/**/
|
|
static void patoptail(long p, long val)
|
|
{
|
|
Upat ptr = (Upat)patout + p;
|
|
int op = P_OP(ptr);
|
|
if (!p || !P_ISBRANCH(ptr))
|
|
return;
|
|
if (op == P_BRANCH)
|
|
pattail(P_OPERAND(p), val);
|
|
else
|
|
pattail(P_OPERAND(p) + 1, val);
|
|
}
|
|
|
|
|
|
/*
|
|
* Run a pattern.
|
|
*/
|
|
static char *patinstart; /* Start of input string */
|
|
static char *patinend; /* End of input string */
|
|
static char *patinput; /* String input pointer */
|
|
static char *patinpath; /* Full path for use with ~ exclusions */
|
|
static int patinlen; /* Length of last successful match.
|
|
* Includes count of Meta characters.
|
|
*/
|
|
|
|
static char *patbeginp[NSUBEXP]; /* Pointer to backref beginnings */
|
|
static char *patendp[NSUBEXP]; /* Pointer to backref ends */
|
|
static int parsfound; /* parentheses (with backrefs) found */
|
|
|
|
static int globdots; /* Glob initial dots? */
|
|
|
|
/*
|
|
* Macros which are currently trivial but are likely to be less
|
|
* so when we handle multibyte characters. They operate on
|
|
* umetafied strings.
|
|
*/
|
|
|
|
/* Get a character from the start point in a string */
|
|
#define CHARREF(x) (STOUC(*x))
|
|
/* Get a pointer to the next character */
|
|
#define CHARNEXT(x) (x+1)
|
|
/* Increment a pointer past the current character. */
|
|
#define CHARINC(x) (x++)
|
|
/* Counter the number of characters between two pointers, largest first */
|
|
#define CHARSUB(x,y) (x-y)
|
|
|
|
/*
|
|
* The following need to be accessed in the globbing scanner for
|
|
* a multi-component file path. See horror story in glob.c.
|
|
*/
|
|
/**/
|
|
int errsfound; /* Total error count so far */
|
|
|
|
/**/
|
|
int forceerrs; /* Forced maximum error count */
|
|
|
|
/**/
|
|
void
|
|
pattrystart(void)
|
|
{
|
|
forceerrs = -1;
|
|
errsfound = 0;
|
|
}
|
|
|
|
/*
|
|
* Test prog against null-terminated, metafied string.
|
|
*/
|
|
|
|
/**/
|
|
mod_export int
|
|
pattry(Patprog prog, char *string)
|
|
{
|
|
return pattryrefs(prog, string, -1, -1, 0, NULL, NULL, NULL);
|
|
}
|
|
|
|
/*
|
|
* Test prog against string of given length, no null termination
|
|
* but still metafied at this point. offset gives an offset
|
|
* to include in reported match indices
|
|
*/
|
|
|
|
/**/
|
|
mod_export int
|
|
pattrylen(Patprog prog, char *string, int len, int unmetalen, int offset)
|
|
{
|
|
return pattryrefs(prog, string, len, unmetalen, offset, NULL, NULL, NULL);
|
|
}
|
|
|
|
/*
|
|
* Test prog against string with given lengths. The input
|
|
* string is metafied; stringlen is the raw string length, and
|
|
* unmetalen the number of characters in the original string (some
|
|
* of which may now be metafied). Either value may be -1
|
|
* to indicate a null-terminated string which will be counted. Note
|
|
* there may be a severe penalty for this if a lot of matching is done
|
|
* on one string.
|
|
*
|
|
* offset is the position in the original string (not seen by
|
|
* the pattern module) at which we are trying to match.
|
|
* This is added in to the positions recorded in patbeginp and patendp
|
|
* when we are looking for substrings. Currently this only happens
|
|
* in the parameter substitution code.
|
|
*
|
|
* Note this is a character offset, i.e. a metafied character
|
|
* counts as 1.
|
|
*
|
|
* The last three arguments are used to report the positions for the
|
|
* backreferences. On entry, *nump should contain the maximum number
|
|
* of positions to report. In this case the match, mbegin, mend
|
|
* arrays are not altered.
|
|
*/
|
|
|
|
/**/
|
|
mod_export int
|
|
pattryrefs(Patprog prog, char *string, int stringlen, int unmetalen,
|
|
int patoffset,
|
|
int *nump, int *begp, int *endp)
|
|
{
|
|
int i, maxnpos = 0, ret, needfullpath, unmetalenp;
|
|
int origlen;
|
|
char **sp, **ep, *tryalloced, *ptr;
|
|
char *progstr = (char *)prog + prog->startoff;
|
|
|
|
if (nump) {
|
|
maxnpos = *nump;
|
|
*nump = 0;
|
|
}
|
|
/* inherited from domatch, but why, exactly? */
|
|
if (*string == Nularg) {
|
|
string++;
|
|
unmetalen--;
|
|
}
|
|
|
|
if (stringlen < 0)
|
|
stringlen = strlen(string);
|
|
origlen = stringlen;
|
|
|
|
patflags = prog->flags;
|
|
/*
|
|
* For a top-level ~-exclusion, we will need the full
|
|
* path to exclude, so copy the path so far and append the
|
|
* current test string.
|
|
*/
|
|
needfullpath = (patflags & PAT_HAS_EXCLUDP) && pathpos;
|
|
|
|
/* Get the length of the full string when unmetafied. */
|
|
if (unmetalen < 0)
|
|
unmetalen = ztrsub(string + stringlen, string);
|
|
if (needfullpath)
|
|
unmetalenp = ztrsub(pathbuf + pathpos, pathbuf);
|
|
else
|
|
unmetalenp = 0;
|
|
|
|
DPUTS(needfullpath && (patflags & (PAT_PURES|PAT_ANY)),
|
|
"rum sort of file exclusion");
|
|
/*
|
|
* Partly for efficiency, and partly for the convenience of
|
|
* globbing, we don't unmetafy pure string patterns, and
|
|
* there's no reason to if the pattern is just a *.
|
|
*/
|
|
if (!(patflags & (PAT_PURES|PAT_ANY))
|
|
&& (needfullpath || unmetalen != stringlen)) {
|
|
/*
|
|
* We need to copy if we need to prepend the path so far
|
|
* (in which case we copy both chunks), or if we have
|
|
* Meta characters.
|
|
*/
|
|
char *dst;
|
|
int icopy, ncopy;
|
|
|
|
dst = tryalloced = zalloc(unmetalen + unmetalenp);
|
|
|
|
if (needfullpath) {
|
|
/* loop twice, copy path buffer first time */
|
|
ptr = pathbuf;
|
|
ncopy = unmetalenp;
|
|
} else {
|
|
/* just loop once, copy string with unmetafication */
|
|
ptr = string;
|
|
ncopy = unmetalen;
|
|
}
|
|
for (icopy = 0; icopy < 2; icopy++) {
|
|
for (i = 0; i < ncopy; i++) {
|
|
if (*ptr == Meta) {
|
|
ptr++;
|
|
*dst++ = *ptr++ ^ 32;
|
|
} else {
|
|
*dst++ = *ptr++;
|
|
}
|
|
}
|
|
if (!needfullpath)
|
|
break;
|
|
/* next time append test string to path so far */
|
|
ptr = string;
|
|
ncopy = unmetalen;
|
|
}
|
|
|
|
if (needfullpath) {
|
|
patinstart = tryalloced + unmetalenp;
|
|
patinpath = tryalloced;
|
|
} else {
|
|
patinstart = tryalloced;
|
|
patinpath = NULL;
|
|
}
|
|
stringlen = unmetalen;
|
|
} else {
|
|
patinstart = string;
|
|
tryalloced = patinpath = NULL;
|
|
}
|
|
|
|
patinend = patinstart + stringlen;
|
|
/*
|
|
* From now on we do not require NULL termination of
|
|
* the test string. There should also be no more references
|
|
* to the variable string.
|
|
*/
|
|
|
|
if (prog->flags & (PAT_PURES|PAT_ANY)) {
|
|
/*
|
|
* Either we are testing against a pure string,
|
|
* or we can match anything at all.
|
|
*/
|
|
int ret;
|
|
if (prog->flags & PAT_ANY) {
|
|
/*
|
|
* Optimisation for a single "*": always matches
|
|
* (except for no_glob_dots, see below).
|
|
*/
|
|
ret = 1;
|
|
} else {
|
|
/*
|
|
* Testing a pure string. See if initial
|
|
* components match.
|
|
*/
|
|
int lendiff = stringlen - prog->patmlen;
|
|
if (lendiff < 0) {
|
|
/* No, the pattern string is too long. */
|
|
ret = 0;
|
|
} else if (!memcmp(progstr, patinstart, prog->patmlen)) {
|
|
/*
|
|
* Initial component matches. Matches either
|
|
* if lengths are the same or we are not anchored
|
|
* to the end of the string.
|
|
*/
|
|
ret = !lendiff || (prog->flags & PAT_NOANCH);
|
|
} else {
|
|
/* No match. */
|
|
ret = 0;
|
|
}
|
|
}
|
|
if (ret) {
|
|
/*
|
|
* For files, we won't match initial "."s unless
|
|
* glob_dots is set.
|
|
*/
|
|
if ((prog->flags & PAT_NOGLD) && *patinstart == '.') {
|
|
ret = 0;
|
|
} else {
|
|
/*
|
|
* Remember the length in case used for ${..#..} etc.
|
|
* In this case, we didn't unmetafy the string.
|
|
*/
|
|
patinlen = (int)prog->patmlen;
|
|
/* if matching files, must update globbing flags */
|
|
patglobflags = prog->globend;
|
|
}
|
|
}
|
|
|
|
if (tryalloced)
|
|
zfree(tryalloced, unmetalen + unmetalenp);
|
|
|
|
return ret;
|
|
} else {
|
|
/*
|
|
* Test for a `must match' string, unless we're scanning for a match
|
|
* in which case we don't need to do this each time.
|
|
*/
|
|
ret = 1;
|
|
if (!(prog->flags & PAT_SCAN) && prog->mustoff)
|
|
{
|
|
char *testptr; /* start pointer into test string */
|
|
char *teststop; /* last point from which we can match */
|
|
char *patptr = (char *)prog + prog->mustoff;
|
|
int patlen = prog->patmlen;
|
|
int found = 0;
|
|
|
|
if (patlen > stringlen) {
|
|
/* Too long, can't match. */
|
|
ret = 0;
|
|
} else {
|
|
teststop = patinend - patlen;
|
|
|
|
for (testptr = patinstart; testptr <= teststop; testptr++)
|
|
{
|
|
if (!memcmp(testptr, patptr, patlen)) {
|
|
found = 1;
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (!found)
|
|
ret = 0;
|
|
}
|
|
}
|
|
if (!ret) {
|
|
if (tryalloced)
|
|
zfree(tryalloced, unmetalen + unmetalenp);
|
|
return 0;
|
|
}
|
|
|
|
patglobflags = prog->globflags;
|
|
if (!(patflags & PAT_FILE)) {
|
|
forceerrs = -1;
|
|
errsfound = 0;
|
|
}
|
|
globdots = !(patflags & PAT_NOGLD);
|
|
parsfound = 0;
|
|
|
|
patinput = patinstart;
|
|
|
|
if (patmatch((Upat)progstr)) {
|
|
/*
|
|
* we were lazy and didn't save the globflags if an exclusion
|
|
* failed, so set it now
|
|
*/
|
|
patglobflags = prog->globend;
|
|
|
|
/*
|
|
* Record length of successful match, including Meta
|
|
* characters. Do it here so that patmatchlen() can return
|
|
* it even if we delete the pattern strings.
|
|
*/
|
|
patinlen = patinput - patinstart;
|
|
/*
|
|
* Optimization: if we didn't find any Meta characters
|
|
* to begin with, we don't need to look for them now.
|
|
*/
|
|
if (unmetalen != origlen) {
|
|
for (ptr = patinstart; ptr < patinput; ptr++)
|
|
if (imeta(*ptr))
|
|
patinlen++;
|
|
}
|
|
|
|
/*
|
|
* Should we clear backreferences and matches on a failed
|
|
* match?
|
|
*/
|
|
if ((patglobflags & GF_MATCHREF) && !(patflags & PAT_FILE)) {
|
|
/*
|
|
* m flag: for global match. This carries no overhead
|
|
* in the pattern matching part.
|
|
*
|
|
* Remember the test pattern is already unmetafied.
|
|
*/
|
|
char *str;
|
|
int mlen = CHARSUB(patinput, patinstart);
|
|
|
|
str = metafy(patinstart, patinput - patinstart, META_DUP);
|
|
setsparam("MATCH", str);
|
|
setiparam("MBEGIN", (zlong)(patoffset + !isset(KSHARRAYS)));
|
|
setiparam("MEND",
|
|
(zlong)(mlen + patoffset +
|
|
!isset(KSHARRAYS) - 1));
|
|
}
|
|
if (prog->patnpar && nump) {
|
|
/*
|
|
* b flag: for backreferences using parentheses. Reported
|
|
* directly.
|
|
*/
|
|
*nump = prog->patnpar;
|
|
|
|
sp = patbeginp;
|
|
ep = patendp;
|
|
|
|
for (i = 0; i < prog->patnpar && i < maxnpos; i++) {
|
|
if (parsfound & (1 << i)) {
|
|
if (begp)
|
|
*begp++ = CHARSUB(*sp, patinstart) + patoffset;
|
|
if (endp)
|
|
*endp++ = CHARSUB(*ep, patinstart) + patoffset
|
|
- 1;
|
|
} else {
|
|
if (begp)
|
|
*begp++ = -1;
|
|
if (endp)
|
|
*endp++ = -1;
|
|
}
|
|
|
|
sp++;
|
|
ep++;
|
|
}
|
|
} else if (prog->patnpar && !(patflags & PAT_FILE)) {
|
|
/*
|
|
* b flag: for backreferences using parentheses.
|
|
*/
|
|
int palen = prog->patnpar+1;
|
|
char **matcharr, **mbeginarr, **mendarr;
|
|
char numbuf[DIGBUFSIZE];
|
|
|
|
matcharr = zshcalloc(palen*sizeof(char *));
|
|
mbeginarr = zshcalloc(palen*sizeof(char *));
|
|
mendarr = zshcalloc(palen*sizeof(char *));
|
|
|
|
sp = patbeginp;
|
|
ep = patendp;
|
|
|
|
for (i = 0; i < prog->patnpar; i++) {
|
|
if (parsfound & (1 << i)) {
|
|
matcharr[i] = metafy(*sp, *ep - *sp, META_DUP);
|
|
/*
|
|
* mbegin and mend give indexes into the string
|
|
* in the standard notation, i.e. respecting
|
|
* KSHARRAYS, and with the end index giving
|
|
* the last character, not one beyond.
|
|
* For example, foo=foo; [[ $foo = (f)oo ]] gives
|
|
* (without KSHARRAYS) indexes 1 and 1, which
|
|
* corresponds to indexing as ${foo[1,1]}.
|
|
*/
|
|
sprintf(numbuf, "%ld",
|
|
(long)(CHARSUB(*sp, patinstart) +
|
|
patoffset +
|
|
!isset(KSHARRAYS)));
|
|
mbeginarr[i] = ztrdup(numbuf);
|
|
sprintf(numbuf, "%ld",
|
|
(long)(CHARSUB(*ep, patinstart) +
|
|
patoffset +
|
|
!isset(KSHARRAYS) - 1));
|
|
mendarr[i] = ztrdup(numbuf);
|
|
} else {
|
|
/* Pattern wasn't set: either it was in an
|
|
* unmatched branch, or a hashed parenthesis
|
|
* that didn't match at all.
|
|
*/
|
|
matcharr[i] = ztrdup("");
|
|
mbeginarr[i] = ztrdup("-1");
|
|
mendarr[i] = ztrdup("-1");
|
|
}
|
|
sp++;
|
|
ep++;
|
|
}
|
|
setaparam("match", matcharr);
|
|
setaparam("mbegin", mbeginarr);
|
|
setaparam("mend", mendarr);
|
|
}
|
|
|
|
ret = 1;
|
|
} else
|
|
ret = 0;
|
|
|
|
if (tryalloced)
|
|
zfree(tryalloced, unmetalen + unmetalenp);
|
|
|
|
return ret;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Return length of previous succesful match. This is
|
|
* in metafied bytes, i.e. includes a count of Meta characters.
|
|
* Unusual and futile attempt at modular encapsulation.
|
|
*/
|
|
|
|
/**/
|
|
int
|
|
patmatchlen(void)
|
|
{
|
|
return patinlen;
|
|
}
|
|
|
|
/*
|
|
* Match literal characters with case insensitivity test: the first
|
|
* comes from the input string, the second the current pattern.
|
|
*/
|
|
#define CHARMATCH(chin, chpa) (chin == chpa || \
|
|
((patglobflags & GF_IGNCASE) ? \
|
|
((isupper(chin) ? tolower(chin) : chin) == \
|
|
(isupper(chpa) ? tolower(chpa) : chpa)) : \
|
|
(patglobflags & GF_LCMATCHUC) ? \
|
|
(islower(chpa) && toupper(chpa) == chin) : 0))
|
|
/*
|
|
* The same but caching an expression from the first argument,
|
|
* Requires local charmatch_cache definition.
|
|
*/
|
|
#define CHARMATCH_EXPR(expr, chpa) \
|
|
(charmatch_cache = (expr), CHARMATCH(charmatch_cache, chpa))
|
|
|
|
/*
|
|
* exactpos is used to remember how far down an exact string we have
|
|
* matched, if we are doing approximation and can therefore redo from
|
|
* the same point; we never need to otherwise.
|
|
*
|
|
* exactend is a pointer to the end of the string, which isn't
|
|
* null-terminated.
|
|
*/
|
|
static char *exactpos, *exactend;
|
|
|
|
/*
|
|
* Main matching routine.
|
|
*
|
|
* Testing the tail end of a match is usually done by recursion, but
|
|
* we try to eliminate that in favour of looping for simple cases.
|
|
*/
|
|
|
|
/**/
|
|
static int
|
|
patmatch(Upat prog)
|
|
{
|
|
/* Current and next nodes */
|
|
Upat scan = prog, next, opnd;
|
|
char *start, *save, *chrop, *chrend, *compend;
|
|
int savglobflags, op, no, min, nextch, fail = 0, saverrsfound;
|
|
zrange_t from, to, comp;
|
|
|
|
while (scan) {
|
|
next = PATNEXT(scan);
|
|
|
|
if (!globdots && P_NOTDOT(scan) && patinput == patinstart &&
|
|
patinput < patinend && *patinput == '.')
|
|
return 0;
|
|
|
|
switch (P_OP(scan)) {
|
|
case P_ANY:
|
|
if (patinput == patinend)
|
|
fail = 1;
|
|
else
|
|
CHARINC(patinput);
|
|
break;
|
|
case P_EXACTLY:
|
|
/*
|
|
* acts as nothing if *chrop is null: this is used by
|
|
* approx code.
|
|
*/
|
|
if (exactpos) {
|
|
chrop = exactpos;
|
|
chrend = exactend;
|
|
} else {
|
|
chrop = P_LS_STR(scan);
|
|
chrend = chrop + P_LS_LEN(scan);
|
|
}
|
|
exactpos = NULL;
|
|
while (chrop < chrend && patinput < patinend) {
|
|
int chin = CHARREF(patinput);
|
|
int chpa = CHARREF(chrop);
|
|
if (!CHARMATCH(chin, chpa)) {
|
|
fail = 1;
|
|
break;
|
|
}
|
|
CHARINC(chrop);
|
|
CHARINC(patinput);
|
|
}
|
|
if (chrop < chrend) {
|
|
exactpos = chrop;
|
|
exactend = chrend;
|
|
fail = 1;
|
|
}
|
|
break;
|
|
case P_ANYOF:
|
|
if (patinput == patinend ||
|
|
!patmatchrange((char *)P_OPERAND(scan),
|
|
CHARREF(patinput)))
|
|
fail = 1;
|
|
else
|
|
CHARINC(patinput);
|
|
break;
|
|
case P_ANYBUT:
|
|
if (patinput == patinend ||
|
|
patmatchrange((char *)P_OPERAND(scan),
|
|
CHARREF(patinput)))
|
|
fail = 1;
|
|
else
|
|
CHARINC(patinput);
|
|
break;
|
|
case P_NUMRNG:
|
|
case P_NUMFROM:
|
|
case P_NUMTO:
|
|
/*
|
|
* To do this properly, we really have to treat numbers as
|
|
* closures: that's so things like <1-1000>33 will
|
|
* match 633 (they didn't up to 3.1.6). To avoid making this
|
|
* too inefficient, we see if there's an exact match next:
|
|
* if there is, and it's not a digit, we return 1 after
|
|
* the first attempt.
|
|
*/
|
|
op = P_OP(scan);
|
|
start = (char *)P_OPERAND(scan);
|
|
from = to = 0;
|
|
if (op != P_NUMTO) {
|
|
#ifdef ZSH_64_BIT_TYPE
|
|
/* We can't rely on pointer alignment being good enough. */
|
|
memcpy((char *)&from, start, sizeof(zrange_t));
|
|
#else
|
|
from = *((zrange_t *) start);
|
|
#endif
|
|
start += sizeof(zrange_t);
|
|
}
|
|
if (op != P_NUMFROM) {
|
|
#ifdef ZSH_64_BIT_TYPE
|
|
memcpy((char *)&to, start, sizeof(zrange_t));
|
|
#else
|
|
to = *((zrange_t *) start);
|
|
#endif
|
|
}
|
|
start = compend = patinput;
|
|
comp = 0;
|
|
while (patinput < patinend && idigit(*patinput)) {
|
|
if (comp)
|
|
comp *= 10;
|
|
comp += *patinput - '0';
|
|
patinput++;
|
|
compend++;
|
|
|
|
if (comp & ((zrange_t)1 << (sizeof(comp)*8 -
|
|
#ifdef ZRANGE_T_IS_SIGNED
|
|
2
|
|
#else
|
|
1
|
|
#endif
|
|
))) {
|
|
/*
|
|
* Out of range (allowing for signedness, which
|
|
* we need if we are using zlongs).
|
|
* This is as far as we can go.
|
|
* If we're doing a range "from", skip all the
|
|
* remaining numbers. Otherwise, we can't
|
|
* match beyond the previous point anyway.
|
|
* Leave the pointer to the last calculated
|
|
* position (compend) where it was before.
|
|
*/
|
|
if (op == P_NUMFROM) {
|
|
while (patinput < patinend && idigit(*patinput))
|
|
patinput++;
|
|
}
|
|
}
|
|
}
|
|
save = patinput;
|
|
no = 0;
|
|
while (patinput > start) {
|
|
/* if already too small, no power on earth can save it */
|
|
if (comp < from && patinput <= compend)
|
|
break;
|
|
if ((op == P_NUMFROM || comp <= to) && patmatch(next)) {
|
|
return 1;
|
|
}
|
|
if (!no && P_OP(next) == P_EXACTLY &&
|
|
(!P_LS_LEN(next) ||
|
|
!idigit(STOUC(*P_LS_STR(next)))) &&
|
|
!(patglobflags & 0xff))
|
|
return 0;
|
|
patinput = --save;
|
|
no++;
|
|
/*
|
|
* With a range start and an unrepresentable test
|
|
* number, we just back down the test string without
|
|
* changing the number until we get to a representable
|
|
* one.
|
|
*/
|
|
if (patinput < compend)
|
|
comp /= 10;
|
|
}
|
|
patinput = start;
|
|
fail = 1;
|
|
break;
|
|
case P_NUMANY:
|
|
/* This is <->: any old set of digits, don't bother comparing */
|
|
start = patinput;
|
|
while (patinput < patinend && idigit(CHARREF(patinput)))
|
|
patinput++;
|
|
save = patinput;
|
|
no = 0;
|
|
while (patinput > start) {
|
|
if (patmatch(next))
|
|
return 1;
|
|
if (!no && P_OP(next) == P_EXACTLY &&
|
|
(!P_LS_LEN(next) ||
|
|
!idigit(CHARREF(P_LS_STR(next)))) &&
|
|
!(patglobflags & 0xff))
|
|
return 0;
|
|
patinput = --save;
|
|
no++;
|
|
}
|
|
patinput = start;
|
|
fail = 1;
|
|
break;
|
|
case P_NOTHING:
|
|
break;
|
|
case P_BACK:
|
|
break;
|
|
case P_GFLAGS:
|
|
patglobflags = P_OPERAND(scan)->l;
|
|
break;
|
|
case P_OPEN:
|
|
case P_OPEN+1:
|
|
case P_OPEN+2:
|
|
case P_OPEN+3:
|
|
case P_OPEN+4:
|
|
case P_OPEN+5:
|
|
case P_OPEN+6:
|
|
case P_OPEN+7:
|
|
case P_OPEN+8:
|
|
case P_OPEN+9:
|
|
no = P_OP(scan) - P_OPEN;
|
|
save = patinput;
|
|
|
|
if (patmatch(next)) {
|
|
/*
|
|
* Don't set patbeginp if some later invocation of
|
|
* the same parentheses already has.
|
|
*/
|
|
if (no && !(parsfound & (1 << (no - 1)))) {
|
|
patbeginp[no-1] = save;
|
|
parsfound |= 1 << (no - 1);
|
|
}
|
|
return 1;
|
|
} else
|
|
return 0;
|
|
break;
|
|
case P_CLOSE:
|
|
case P_CLOSE+1:
|
|
case P_CLOSE+2:
|
|
case P_CLOSE+3:
|
|
case P_CLOSE+4:
|
|
case P_CLOSE+5:
|
|
case P_CLOSE+6:
|
|
case P_CLOSE+7:
|
|
case P_CLOSE+8:
|
|
case P_CLOSE+9:
|
|
no = P_OP(scan) - P_CLOSE;
|
|
save = patinput;
|
|
|
|
if (patmatch(next)) {
|
|
DPUTS(!patendp, "patendp not set for backreferencing");
|
|
if (no && !(parsfound & (1 << (no + 15)))) {
|
|
patendp[no-1] = save;
|
|
parsfound |= 1 << (no + 15);
|
|
}
|
|
return 1;
|
|
} else
|
|
return 0;
|
|
break;
|
|
case P_EXCSYNC:
|
|
/* See the P_EXCLUDE code below for where syncptr comes from */
|
|
{
|
|
unsigned char *syncptr;
|
|
Upat after;
|
|
after = P_OPERAND(scan);
|
|
DPUTS(!P_ISEXCLUDE(after),
|
|
"BUG: EXCSYNC not followed by EXCLUDE.");
|
|
DPUTS(!P_OPERAND(after)->p,
|
|
"BUG: EXCSYNC not handled by EXCLUDE");
|
|
syncptr = P_OPERAND(after)->p + (patinput - patinstart);
|
|
/*
|
|
* If we already matched from here, this time we fail.
|
|
* See WBRANCH code for story about error count.
|
|
*/
|
|
if (*syncptr && errsfound + 1 >= *syncptr)
|
|
return 0;
|
|
/*
|
|
* Else record that we (possibly) matched this time.
|
|
* No harm if we don't: then the previous test will just
|
|
* short cut the attempted match that is bound to fail.
|
|
* We never try to exclude something that has already
|
|
* failed anyway.
|
|
*/
|
|
*syncptr = errsfound + 1;
|
|
}
|
|
break;
|
|
case P_EXCEND:
|
|
/*
|
|
* This is followed by a P_EXCSYNC, but only in the P_EXCLUDE
|
|
* branch. Actually, we don't bother following it: all we
|
|
* need to know is that we successfully matched so far up
|
|
* to the end of the asserted pattern; the endpoint
|
|
* in the target string is nulled out.
|
|
*/
|
|
if (!(fail = (patinput < patinend)))
|
|
return 1;
|
|
break;
|
|
case P_BRANCH:
|
|
case P_WBRANCH:
|
|
/* P_EXCLUDE shouldn't occur without a P_BRANCH */
|
|
if (!P_ISBRANCH(next)) {
|
|
/* no choice, avoid recursion */
|
|
DPUTS(P_OP(scan) == P_WBRANCH,
|
|
"BUG: WBRANCH with no alternative.");
|
|
next = P_OPERAND(scan);
|
|
} else {
|
|
do {
|
|
save = patinput;
|
|
savglobflags = patglobflags;
|
|
saverrsfound = errsfound;
|
|
if (P_ISEXCLUDE(next)) {
|
|
/*
|
|
* The strategy is to test the asserted pattern,
|
|
* recording via P_EXCSYNC how far the part to
|
|
* be excluded matched. We then set the
|
|
* length of the test string to that
|
|
* point and see if the exclusion as far as
|
|
* P_EXCEND also matches that string.
|
|
* We need to keep testing the asserted pattern
|
|
* by backtracking, since the first attempt
|
|
* may be excluded while a later attempt may not.
|
|
* For this we keep a pointer just after
|
|
* the P_EXCLUDE which is tested by the P_EXCSYNC
|
|
* to see if we matched there last time, in which
|
|
* case we fail. If there is nothing to backtrack
|
|
* over, that doesn't matter: we should fail anyway.
|
|
* The pointer also tells us where the asserted
|
|
* pattern matched for use by the exclusion.
|
|
*
|
|
* It's hard to allocate space for this
|
|
* beforehand since we may need to do it
|
|
* recursively.
|
|
*
|
|
* P.S. in case you were wondering, this code
|
|
* is horrible.
|
|
*/
|
|
Upat syncstrp;
|
|
char *origpatinend;
|
|
unsigned char *oldsyncstr;
|
|
char *matchpt = NULL;
|
|
int ret, savglobdots, matchederrs = 0;
|
|
int savparsfound = parsfound;
|
|
DPUTS(P_OP(scan) == P_WBRANCH,
|
|
"BUG: excluded WBRANCH");
|
|
syncstrp = P_OPERAND(next);
|
|
/*
|
|
* Unlike WBRANCH, each test at the same exclude
|
|
* sync point (due to an external loop) is separate,
|
|
* i.e testing (foo~bar)# is no different from
|
|
* (foo~bar)(foo~bar)... from the exclusion point
|
|
* of view, so we use a different sync string.
|
|
*/
|
|
oldsyncstr = syncstrp->p;
|
|
syncstrp->p = (unsigned char *)
|
|
zshcalloc((patinend - patinstart) + 1);
|
|
origpatinend = patinend;
|
|
while ((ret = patmatch(P_OPERAND(scan)))) {
|
|
unsigned char *syncpt;
|
|
char *savpatinstart;
|
|
int savforce = forceerrs;
|
|
int savpatflags = patflags, synclen;
|
|
forceerrs = -1;
|
|
savglobdots = globdots;
|
|
matchederrs = errsfound;
|
|
matchpt = patinput; /* may not be end */
|
|
globdots = 1; /* OK to match . first */
|
|
/* Find the point where the scan
|
|
* matched the part to be excluded: because
|
|
* of backtracking, the one
|
|
* most recently matched will be the first.
|
|
* (Luckily, backtracking is done after all
|
|
* possibilities for approximation have been
|
|
* checked.)
|
|
*/
|
|
for (syncpt = syncstrp->p; !*syncpt; syncpt++)
|
|
;
|
|
synclen = syncpt - syncstrp->p;
|
|
if (patinstart + synclen != patinend) {
|
|
/*
|
|
* Temporarily mark the string as
|
|
* ending at this point.
|
|
*/
|
|
DPUTS(patinstart + synclen > matchpt,
|
|
"BUG: EXCSYNC failed");
|
|
|
|
patinend = patinstart + synclen;
|
|
/*
|
|
* If this isn't really the end of the string,
|
|
* remember this for the (#e) assertion.
|
|
*/
|
|
patflags |= PAT_NOTEND;
|
|
}
|
|
savpatinstart = patinstart;
|
|
next = PATNEXT(scan);
|
|
while (next && P_ISEXCLUDE(next)) {
|
|
patinput = save;
|
|
/*
|
|
* turn off approximations in exclusions:
|
|
* note we keep remaining patglobflags
|
|
* set by asserted branch (or previous
|
|
* excluded branches, for consistency).
|
|
*/
|
|
patglobflags &= ~0xff;
|
|
errsfound = 0;
|
|
opnd = P_OPERAND(next) + 1;
|
|
if (P_OP(next) == P_EXCLUDP && patinpath) {
|
|
/*
|
|
* Top level exclusion with a file,
|
|
* applies to whole path so add the
|
|
* segments already matched.
|
|
* We copied these in front of the
|
|
* test pattern, so patinend doesn't
|
|
* need moving.
|
|
*/
|
|
DPUTS(patinput != patinstart,
|
|
"BUG: not at start excluding path");
|
|
patinput = patinstart = patinpath;
|
|
}
|
|
if (patmatch(opnd)) {
|
|
ret = 0;
|
|
/*
|
|
* Another subtlety: if we exclude the
|
|
* match, any parentheses just found
|
|
* become invalidated.
|
|
*/
|
|
parsfound = savparsfound;
|
|
}
|
|
if (patinpath) {
|
|
patinput = savpatinstart +
|
|
(patinput - patinstart);
|
|
patinstart = savpatinstart;
|
|
}
|
|
if (!ret)
|
|
break;
|
|
next = PATNEXT(next);
|
|
}
|
|
/*
|
|
* Restore original end position.
|
|
*/
|
|
patinend = origpatinend;
|
|
patflags = savpatflags;
|
|
globdots = savglobdots;
|
|
forceerrs = savforce;
|
|
if (ret)
|
|
break;
|
|
patinput = save;
|
|
patglobflags = savglobflags;
|
|
errsfound = saverrsfound;
|
|
}
|
|
zfree((char *)syncstrp->p,
|
|
(patinend - patinstart) + 1);
|
|
syncstrp->p = oldsyncstr;
|
|
if (ret) {
|
|
patinput = matchpt;
|
|
errsfound = matchederrs;
|
|
return 1;
|
|
}
|
|
while ((scan = PATNEXT(scan)) &&
|
|
P_ISEXCLUDE(scan))
|
|
;
|
|
} else {
|
|
int ret = 1, pfree = 0;
|
|
Upat ptrp = NULL;
|
|
unsigned char *ptr;
|
|
if (P_OP(scan) == P_WBRANCH) {
|
|
/*
|
|
* This is where we make sure that we are not
|
|
* repeatedly matching zero-length strings in
|
|
* a closure, which would cause an infinite loop,
|
|
* and also remove exponential behaviour in
|
|
* backtracking nested closures.
|
|
* The P_WBRANCH operator leaves a space for a
|
|
* uchar *, initialized to NULL, which is
|
|
* turned into a string the same length as the
|
|
* target string. Every time we match from a
|
|
* particular point in the target string, we
|
|
* stick a 1 at the corresponding point here.
|
|
* If we come round to the same branch again, and
|
|
* there is already a 1, then the test fails.
|
|
*/
|
|
opnd = P_OPERAND(scan);
|
|
ptrp = opnd++;
|
|
if (!ptrp->p) {
|
|
ptrp->p = (unsigned char *)
|
|
zshcalloc((patinend - patinstart) + 1);
|
|
pfree = 1;
|
|
}
|
|
ptr = ptrp->p + (patinput - patinstart);
|
|
|
|
/*
|
|
* Without approximation, this is just a
|
|
* single bit test. With approximation, we
|
|
* need to know how many errors there were
|
|
* last time we made the test. If errsfound
|
|
* is now smaller than it was, hence we can
|
|
* make more approximations in the remaining
|
|
* code, we continue with the test.
|
|
* (This is why the max number of errors is
|
|
* 254, not 255.)
|
|
*/
|
|
if (*ptr && errsfound + 1 >= *ptr)
|
|
ret = 0;
|
|
*ptr = errsfound + 1;
|
|
} else
|
|
opnd = P_OPERAND(scan);
|
|
if (ret)
|
|
ret = patmatch(opnd);
|
|
if (pfree) {
|
|
zfree((char *)ptrp->p,
|
|
(patinend - patinstart) + 1);
|
|
ptrp->p = NULL;
|
|
}
|
|
if (ret)
|
|
return 1;
|
|
scan = PATNEXT(scan);
|
|
}
|
|
patinput = save;
|
|
patglobflags = savglobflags;
|
|
errsfound = saverrsfound;
|
|
DPUTS(P_OP(scan) == P_WBRANCH,
|
|
"BUG: WBRANCH not first choice.");
|
|
next = PATNEXT(scan);
|
|
} while (scan && P_ISBRANCH(scan));
|
|
return 0;
|
|
}
|
|
break;
|
|
case P_STAR:
|
|
/* Handle specially for speed, although really P_ONEHASH+P_ANY */
|
|
case P_ONEHASH:
|
|
case P_TWOHASH:
|
|
/*
|
|
* This is just simple cases, matching one character.
|
|
* With approximations, we still handle * this way, since
|
|
* no approximation is ever necessary, but other closures
|
|
* are handled by the more complicated branching method
|
|
*/
|
|
op = P_OP(scan);
|
|
/* Note that no counts possibly metafied characters */
|
|
start = patinput;
|
|
if (op == P_STAR) {
|
|
for (no = 0; patinput < patinend; CHARINC(patinput))
|
|
no++;
|
|
/* simple optimization for reasonably common case */
|
|
if (P_OP(next) == P_END)
|
|
return 1;
|
|
} else {
|
|
DPUTS(patglobflags & 0xff,
|
|
"BUG: wrong backtracking with approximation.");
|
|
if (!globdots && P_NOTDOT(P_OPERAND(scan)) &&
|
|
patinput == patinstart && patinput < patinend &&
|
|
CHARREF(patinput) == '.')
|
|
return 0;
|
|
no = patrepeat(P_OPERAND(scan));
|
|
}
|
|
min = (op == P_TWOHASH) ? 1 : 0;
|
|
/*
|
|
* Lookahead to avoid useless matches. This is not possible
|
|
* with approximation.
|
|
*/
|
|
if (P_OP(next) == P_EXACTLY && P_LS_LEN(next) &&
|
|
!(patglobflags & 0xff)) {
|
|
char *nextop = P_LS_STR(next);
|
|
/*
|
|
* If that P_EXACTLY is last (common in simple patterns,
|
|
* such as *.c), then it can be only be matched at one
|
|
* point in the test string, so record that.
|
|
*/
|
|
if (P_OP(PATNEXT(next)) == P_END &&
|
|
!(patflags & PAT_NOANCH)) {
|
|
int ptlen = patinend - patinput;
|
|
int lenmatch = patinend - (min ? CHARNEXT(start) : start);
|
|
/* Are we in the right range? */
|
|
if (P_LS_LEN(next) > lenmatch || P_LS_LEN(next) < ptlen)
|
|
return 0;
|
|
/* Yes, just position appropriately and test. */
|
|
patinput += ptlen - P_LS_LEN(next);
|
|
/*
|
|
* Here we will need to be careful that patinput is not
|
|
* in the middle of a multibyte character.
|
|
*/
|
|
/* Continue loop with P_EXACTLY test. */
|
|
break;
|
|
}
|
|
nextch = CHARREF(nextop);
|
|
} else
|
|
nextch = -1;
|
|
save = patinput;
|
|
savglobflags = patglobflags;
|
|
saverrsfound = errsfound;
|
|
while (no >= min) {
|
|
int charmatch_cache;
|
|
if (nextch < 0 ||
|
|
(patinput < patinend &&
|
|
CHARMATCH_EXPR(CHARREF(patinput), nextch))) {
|
|
if (patmatch(next))
|
|
return 1;
|
|
}
|
|
no--;
|
|
save--;
|
|
/*
|
|
* Here we will need to make sure save is
|
|
* decremented properly to the start of
|
|
* the preceeding multibyte character.
|
|
*/
|
|
patinput = save;
|
|
patglobflags = savglobflags;
|
|
errsfound = saverrsfound;
|
|
}
|
|
/*
|
|
* As with branches, the patmatch(next) stuff for *
|
|
* handles approximation, so we don't need to try
|
|
* anything here.
|
|
*/
|
|
return 0;
|
|
case P_ISSTART:
|
|
if (patinput != patinstart || (patflags & PAT_NOTSTART))
|
|
fail = 1;
|
|
break;
|
|
case P_ISEND:
|
|
if (patinput < patinend || (patflags & PAT_NOTEND))
|
|
fail = 1;
|
|
break;
|
|
case P_END:
|
|
if (!(fail = (patinput < patinend && !(patflags & PAT_NOANCH))))
|
|
return 1;
|
|
break;
|
|
#ifdef DEBUG
|
|
default:
|
|
dputs("BUG: bad operand in patmatch.");
|
|
return 0;
|
|
break;
|
|
#endif
|
|
}
|
|
|
|
if (fail) {
|
|
if (errsfound < (patglobflags & 0xff) &&
|
|
(forceerrs == -1 || errsfound < forceerrs)) {
|
|
/*
|
|
* Approximation code. There are four possibilities
|
|
*
|
|
* 1. omit character from input string
|
|
* 2. transpose characters in input and pattern strings
|
|
* 3. omit character in both input and pattern strings
|
|
* 4. omit character from pattern string.
|
|
*
|
|
* which we try in that order.
|
|
*
|
|
* Of these, 2, 3 and 4 require an exact match string
|
|
* (P_EXACTLY) while 1, 2 and 3 require that we not
|
|
* have reached the end of the input string.
|
|
*
|
|
* Note in each case after making the approximation we
|
|
* need to retry the *same* pattern; this is what
|
|
* requires exactpos, a slightly doleful way of
|
|
* communicating with the exact character matcher.
|
|
*/
|
|
char *savexact = exactpos;
|
|
save = patinput;
|
|
savglobflags = patglobflags;
|
|
saverrsfound = ++errsfound;
|
|
fail = 0;
|
|
|
|
DPUTS(P_OP(scan) != P_EXACTLY && exactpos,
|
|
"BUG: non-exact match has set exactpos");
|
|
|
|
/* Try omitting a character from the input string */
|
|
if (patinput < patinend) {
|
|
CHARINC(patinput);
|
|
/* If we are not on an exact match, then this is
|
|
* our last gasp effort, so we can optimize out
|
|
* the recursive call.
|
|
*/
|
|
if (P_OP(scan) != P_EXACTLY)
|
|
continue;
|
|
if (patmatch(scan))
|
|
return 1;
|
|
}
|
|
|
|
if (P_OP(scan) == P_EXACTLY) {
|
|
char *nextexact = savexact;
|
|
DPUTS(!savexact,
|
|
"BUG: exact match has not set exactpos");
|
|
CHARINC(nextexact);
|
|
|
|
if (save < patinend) {
|
|
char *nextin = save;
|
|
CHARINC(nextin);
|
|
patglobflags = savglobflags;
|
|
errsfound = saverrsfound;
|
|
exactpos = savexact;
|
|
|
|
/*
|
|
* Try swapping two characters in patinput and
|
|
* exactpos
|
|
*/
|
|
if (save < patinend && nextin < patinend &&
|
|
nextexact < exactend) {
|
|
int cin0 = CHARREF(save);
|
|
int cpa0 = CHARREF(exactpos);
|
|
int cin1 = CHARREF(nextin);
|
|
int cpa1 = CHARREF(nextexact);
|
|
|
|
if (CHARMATCH(cin0, cpa1) &&
|
|
CHARMATCH(cin1, cpa0)) {
|
|
patinput = nextin;
|
|
CHARINC(patinput);
|
|
exactpos = nextexact;
|
|
CHARINC(exactpos);
|
|
if (patmatch(scan))
|
|
return 1;
|
|
|
|
patglobflags = savglobflags;
|
|
errsfound = saverrsfound;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Try moving up both strings.
|
|
*/
|
|
patinput = nextin;
|
|
exactpos = nextexact;
|
|
if (patmatch(scan))
|
|
return 1;
|
|
|
|
patinput = save;
|
|
patglobflags = savglobflags;
|
|
errsfound = saverrsfound;
|
|
exactpos = savexact;
|
|
}
|
|
|
|
DPUTS(exactpos == exactend, "approximating too far");
|
|
/*
|
|
* Try moving up the exact match pattern.
|
|
* This must be the last attempt, so just loop
|
|
* instead of calling recursively.
|
|
*/
|
|
CHARINC(exactpos);
|
|
continue;
|
|
}
|
|
}
|
|
exactpos = NULL;
|
|
return 0;
|
|
}
|
|
|
|
scan = next;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/**/
|
|
static int
|
|
patmatchrange(char *range, int ch)
|
|
{
|
|
int r1, r2;
|
|
|
|
/*
|
|
* Careful here: unlike other strings, range is a NULL-terminated,
|
|
* metafied string, because we need to treat the Posix and hyphenated
|
|
* ranges specially.
|
|
*/
|
|
for (; *range; range++) {
|
|
if (imeta(STOUC(*range))) {
|
|
switch (STOUC(*range)-STOUC(Meta)) {
|
|
case 0:
|
|
if (STOUC(*++range ^ 32) == ch)
|
|
return 1;
|
|
break;
|
|
case PP_ALPHA:
|
|
if (isalpha(ch))
|
|
return 1;
|
|
break;
|
|
case PP_ALNUM:
|
|
if (isalnum(ch))
|
|
return 1;
|
|
break;
|
|
case PP_ASCII:
|
|
if ((ch & ~0x7f) == 0)
|
|
return 1;
|
|
break;
|
|
case PP_BLANK:
|
|
if (ch == ' ' || ch == '\t')
|
|
return 1;
|
|
break;
|
|
case PP_CNTRL:
|
|
if (iscntrl(ch))
|
|
return 1;
|
|
break;
|
|
case PP_DIGIT:
|
|
if (isdigit(ch))
|
|
return 1;
|
|
break;
|
|
case PP_GRAPH:
|
|
if (isgraph(ch))
|
|
return 1;
|
|
break;
|
|
case PP_LOWER:
|
|
if (islower(ch))
|
|
return 1;
|
|
break;
|
|
case PP_PRINT:
|
|
if (isprint(ch))
|
|
return 1;
|
|
break;
|
|
case PP_PUNCT:
|
|
if (ispunct(ch))
|
|
return 1;
|
|
break;
|
|
case PP_SPACE:
|
|
if (isspace(ch))
|
|
return 1;
|
|
break;
|
|
case PP_UPPER:
|
|
if (isupper(ch))
|
|
return 1;
|
|
break;
|
|
case PP_XDIGIT:
|
|
if (isxdigit(ch))
|
|
return 1;
|
|
break;
|
|
case PP_IDENT:
|
|
if (iident(ch))
|
|
return 1;
|
|
break;
|
|
case PP_IFS:
|
|
if (isep(ch))
|
|
return 1;
|
|
break;
|
|
case PP_IFSSPACE:
|
|
if (iwsep(ch))
|
|
return 1;
|
|
break;
|
|
case PP_WORD:
|
|
/*
|
|
* HERE: when we support multibyte characters,
|
|
* this test needs to be wcsiword().
|
|
*/
|
|
if (iword(ch))
|
|
return 1;
|
|
break;
|
|
case PP_RANGE:
|
|
range++;
|
|
r1 = STOUC(UNMETA(range));
|
|
METAINC(range);
|
|
r2 = STOUC(UNMETA(range));
|
|
if (*range == Meta)
|
|
range++;
|
|
if (r1 <= ch && ch <= r2)
|
|
return 1;
|
|
break;
|
|
case PP_UNKWN:
|
|
DPUTS(1, "BUG: unknown posix range passed through.\n");
|
|
break;
|
|
default:
|
|
DPUTS(1, "BUG: unknown metacharacter in range.");
|
|
break;
|
|
}
|
|
} else if (STOUC(*range) == ch)
|
|
return 1;
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
/* repeatedly match something simple and say how many times */
|
|
|
|
/**/
|
|
static int patrepeat(Upat p)
|
|
{
|
|
int count = 0, tch, charmatch_cache;
|
|
char *scan, *opnd;
|
|
|
|
scan = patinput;
|
|
opnd = (char *)P_OPERAND(p);
|
|
|
|
switch(P_OP(p)) {
|
|
#ifdef DEBUG
|
|
case P_ANY:
|
|
dputs("BUG: ?# did not get optimized to *");
|
|
return 0;
|
|
break;
|
|
#endif
|
|
case P_EXACTLY:
|
|
DPUTS(P_LS_LEN(p) != 1, "closure following more than one character");
|
|
tch = CHARREF(P_LS_STR(p));
|
|
while (scan < patinend &&
|
|
CHARMATCH_EXPR(CHARREF(scan), tch)) {
|
|
count++;
|
|
CHARINC(scan);
|
|
}
|
|
break;
|
|
case P_ANYOF:
|
|
while (scan < patinend && patmatchrange(opnd, CHARREF(scan))) {
|
|
count++;
|
|
CHARINC(scan);
|
|
}
|
|
break;
|
|
case P_ANYBUT:
|
|
while (scan < patinend && !patmatchrange(opnd, CHARREF(scan))) {
|
|
count++;
|
|
CHARINC(scan);
|
|
}
|
|
break;
|
|
#ifdef DEBUG
|
|
default:
|
|
dputs("BUG: something very strange is happening in patrepeat");
|
|
return 0;
|
|
break;
|
|
#endif
|
|
}
|
|
|
|
patinput = scan;
|
|
return count;
|
|
}
|
|
|
|
/* Free a patprog. */
|
|
|
|
/**/
|
|
mod_export void
|
|
freepatprog(Patprog prog)
|
|
{
|
|
if (prog && prog != dummy_patprog1 && prog != dummy_patprog2)
|
|
zfree(prog, prog->size);
|
|
}
|
|
|
|
/**/
|
|
#ifdef ZSH_PAT_DEBUG
|
|
|
|
/* Debugging stuff: print and test a regular expression */
|
|
|
|
/* Dump a regexp onto stdout in vaguely comprehensible form */
|
|
|
|
/**/
|
|
static void
|
|
patdump(Patprog r)
|
|
{
|
|
char *s, *base, op = P_EXACTLY;
|
|
Upat up, codestart, next;
|
|
|
|
base = (char *)r;
|
|
s = base + r->startoff;
|
|
|
|
if (r->flags & PAT_PURES) {
|
|
printf("STRING:%s\n", (char *)s);
|
|
} else {
|
|
codestart = (Upat)s;
|
|
while (op != P_END) {
|
|
up = (Upat)s;
|
|
op = P_OP(up);
|
|
printf("%2d%s", up-codestart, patprop(up));
|
|
next = PATNEXT(up);
|
|
printf("(%d)", next ? next-codestart : 0);
|
|
s += sizeof(union upat);
|
|
if (op == P_EXACTLY) {
|
|
long llen = *(long *)s;
|
|
s += sizeof(long);
|
|
while (llen--) {
|
|
putchar(CHARREF(s));
|
|
CHARINC(s);
|
|
}
|
|
} else if (op == P_ANYOF || op == P_ANYBUT) {
|
|
while (*s != '\0') {
|
|
if (itok(*s)) {
|
|
if (*s == Meta + PP_RANGE) {
|
|
s++;
|
|
printf("<RANGE:%c-", UNMETA(s));
|
|
METAINC(s);
|
|
printf("%c>", UNMETA(s));
|
|
} else {
|
|
printf("<TYPE:%d>", *s - Meta);
|
|
s++;
|
|
continue;
|
|
}
|
|
} else
|
|
putchar(UNMETA(s));
|
|
METAINC(s);
|
|
}
|
|
} else if (op == P_NUMRNG || op == P_NUMFROM || op == P_NUMTO) {
|
|
printf("%lu", (unsigned long)*(zrange_t *)s);
|
|
s += sizeof(zrange_t);
|
|
if (op == P_NUMRNG) {
|
|
printf("-%lu", (unsigned long)*(zrange_t *)s);
|
|
s += sizeof(zrange_t);
|
|
}
|
|
} else if (op == P_GFLAGS) {
|
|
printf("%ld, %ld", (++up)->l & ~0xff, (++up)->l & 0xff);
|
|
s += sizeof(union upat);
|
|
} else if (op == P_WBRANCH || op == P_EXCLUDE ||
|
|
op == P_EXCLUDP) {
|
|
s += sizeof(union upat);
|
|
}
|
|
putchar('\n');
|
|
s = base + (((s - base) + sizeof(union upat) - 1) &
|
|
~(sizeof(union upat) - 1));
|
|
}
|
|
}
|
|
|
|
printf("Total size = %ld\n", r->size);
|
|
if (r->patstartch)
|
|
printf("start `%c' ", r->patstartch);
|
|
if (!(r->flags & PAT_NOANCH))
|
|
printf("EOL-anchor ");
|
|
if (r->patnpar)
|
|
printf("%d active backreferences ", r->patnpar);
|
|
if (r->mustoff)
|
|
printf("must have \"%s\"", (char *)r + r->mustoff);
|
|
printf("\n");
|
|
if (r->globflags) {
|
|
printf("Globbing flags: ");
|
|
if (r->globflags & GF_LCMATCHUC)
|
|
printf("LC matches UC ");
|
|
if (r->globflags & GF_IGNCASE)
|
|
printf("Ignore case");
|
|
printf("\n");
|
|
if (r->globflags & 0xff)
|
|
printf("Max errors = %d\n", r->globflags & 0xff);
|
|
}
|
|
}
|
|
|
|
/**/
|
|
static char *
|
|
patprop(Upat op)
|
|
{
|
|
char *p = NULL;
|
|
static char buf[50];
|
|
|
|
strcpy(buf, ":");
|
|
|
|
switch(P_OP(op)) {
|
|
case P_ANY:
|
|
p = "ANY";
|
|
break;
|
|
case P_ANYOF:
|
|
p = "ANYOF";
|
|
break;
|
|
case P_ANYBUT:
|
|
p = "ANYBUT";
|
|
break;
|
|
case P_BRANCH:
|
|
p = "BRANCH";
|
|
break;
|
|
case P_WBRANCH:
|
|
p = "WBRANCH";
|
|
break;
|
|
case P_EXCLUDE:
|
|
p = "EXCLUDE";
|
|
break;
|
|
case P_EXCLUDP:
|
|
p = "EXCLUDP";
|
|
break;
|
|
case P_EXCSYNC:
|
|
p = "EXCSYNC";
|
|
break;
|
|
case P_EXCEND:
|
|
p = "EXCEND";
|
|
break;
|
|
case P_EXACTLY:
|
|
p = "EXACTLY";
|
|
break;
|
|
case P_GFLAGS:
|
|
p = "GFLAGS";
|
|
break;
|
|
case P_ISSTART:
|
|
p = "ISSTART";
|
|
break;
|
|
case P_ISEND:
|
|
p = "ISEND";
|
|
break;
|
|
case P_NOTHING:
|
|
p = "NOTHING";
|
|
break;
|
|
case P_BACK:
|
|
p = "BACK";
|
|
break;
|
|
case P_END:
|
|
p = "END";
|
|
break;
|
|
case P_OPEN:
|
|
case P_OPEN+1:
|
|
case P_OPEN+2:
|
|
case P_OPEN+3:
|
|
case P_OPEN+4:
|
|
case P_OPEN+5:
|
|
case P_OPEN+6:
|
|
case P_OPEN+7:
|
|
case P_OPEN+8:
|
|
case P_OPEN+9:
|
|
sprintf(buf+strlen(buf), "OPEN%ld", P_OP(op)-P_OPEN);
|
|
p = NULL;
|
|
break;
|
|
case P_CLOSE:
|
|
case P_CLOSE+1:
|
|
case P_CLOSE+2:
|
|
case P_CLOSE+3:
|
|
case P_CLOSE+4:
|
|
case P_CLOSE+5:
|
|
case P_CLOSE+6:
|
|
case P_CLOSE+7:
|
|
case P_CLOSE+8:
|
|
case P_CLOSE+9:
|
|
sprintf(buf+strlen(buf), "CLOSE%ld", P_OP(op)-P_CLOSE);
|
|
p = NULL;
|
|
break;
|
|
case P_STAR:
|
|
p = "STAR";
|
|
break;
|
|
case P_ONEHASH:
|
|
p = "ONEHASH";
|
|
break;
|
|
case P_TWOHASH:
|
|
p = "TWOHASH";
|
|
break;
|
|
case P_NUMRNG:
|
|
p = "NUMRNG";
|
|
break;
|
|
case P_NUMFROM:
|
|
p = "NUMFROM";
|
|
break;
|
|
case P_NUMTO:
|
|
p = "NUMTO";
|
|
break;
|
|
case P_NUMANY:
|
|
p = "NUMANY";
|
|
break;
|
|
default:
|
|
fprintf(stderr, "Bad opcode\n");
|
|
p = NULL;
|
|
break;
|
|
}
|
|
if (p)
|
|
strcat(buf, p);
|
|
return buf;
|
|
}
|
|
|
|
/**/
|
|
int
|
|
bin_patdebug(char *name, char **args, char *ops, int func)
|
|
{
|
|
Patprog prog;
|
|
int ret = 0;
|
|
|
|
tokenize(*args);
|
|
|
|
if (!(prog = patcompile((char *)*args, 0, 0)))
|
|
return 1;
|
|
if (ops['p'] || !args[1]) {
|
|
patdump(prog);
|
|
}
|
|
|
|
while (*++args) {
|
|
if (!pattry(prog, (char *)*args))
|
|
ret++;
|
|
}
|
|
return ret;
|
|
}
|
|
|
|
/**/
|
|
#endif /* ZSH_PAT_DEBUG */
|