/*- * Copyright (c) 1991, 1993 * The Regents of the University of California. All rights reserved. * * This code is derived from software contributed to Berkeley by * Kenneth Almquist. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. */ #ifndef lint #if 0 static char sccsid[] = "@(#)parser.c 8.7 (Berkeley) 5/16/95"; #endif #endif /* not lint */ #include /* __FBSDID("$FreeBSD: src/bin/sh/parser.c,v 1.51.2.1 2005/03/03 03:43:20 obrien Exp $"); */ #include #include #include "shell.h" #include "parser.h" #include "nodes.h" #include "expand.h" /* defines rmescapes() */ #include "syntax.h" #include "options.h" #include "input.h" #include "output.h" #include "var.h" #include "error.h" #include "memalloc.h" #include "mystring.h" #include "alias.h" #include "show.h" #include "eval.h" #if !defined(NO_HISTORY) && !defined(EDITLINE) #include "myhistedit.h" #endif /* * Shell command parser. */ #define EOFMARKLEN 79 #define PROMPTLEN 128 /* values returned by readtoken */ #include "token.h" struct heredoc { struct heredoc *next; /* next here document in list */ union node *here; /* redirection node */ char *eofmark; /* string indicating end of input */ int striptabs; /* if set, strip leading tabs */ }; STATIC struct heredoc *heredoclist; /* list of here documents to read */ STATIC int parsebackquote; /* nonzero if we are inside backquotes */ STATIC int doprompt; /* if set, prompt the user */ STATIC int needprompt; /* true if interactive and at start of line */ STATIC int lasttoken; /* last token read */ MKINIT int tokpushback; /* last token pushed back */ STATIC char *wordtext; /* text of last word returned by readtoken */ MKINIT int checkkwd; /* 1 == check for kwds, 2 == also eat newlines */ STATIC struct nodelist *backquotelist; STATIC union node *redirnode; STATIC struct heredoc *heredoc; STATIC int quoteflag; /* set if (part of) last token was quoted */ STATIC int startlinno; /* line # where last token started */ /* XXX When 'noaliases' is set to one, no alias expansion takes place. */ static int noaliases = 0; #define GDB_HACK 1 /* avoid local declarations which gdb can't handle */ #ifdef GDB_HACK static const char argvars[5] = {CTLVAR, VSNORMAL|VSQUOTE, '@', '=', '\0'}; static const char types[] = "}-+?="; #endif STATIC union node *list(int); STATIC union node *andor(void); STATIC union node *pipeline(void); STATIC union node *command(void); STATIC union node *simplecmd(union node **, union node *); STATIC union node *makename(void); STATIC void parsefname(void); STATIC void parseheredoc(void); STATIC int peektoken(void); STATIC int readtoken(void); STATIC int xxreadtoken(void); STATIC int readtoken1(int, char const *, char *, int); STATIC int noexpand(char *); STATIC void synexpect(int); STATIC void synerror(char *); STATIC void setprompt(int); /* * Read and parse a command. Returns NEOF on end of file. (NULL is a * valid parse tree indicating a blank line.) */ union node * parsecmd(int interact) { int t; extern int exitstatus; tokpushback = 0; doprompt = interact; if (doprompt) setprompt(exitstatus == 0 || (vpse.flags & VUNSET) ? 1 : -1); else setprompt(0); needprompt = 0; t = readtoken(); if (t == TEOF) return NEOF; if (t == TNL) return NULL; tokpushback++; return list(1); } STATIC union node * list(int nlflag) { union node *n1, *n2, *n3; int tok; checkkwd = 2; if (nlflag == 0 && tokendlist[peektoken()]) return NULL; n1 = NULL; for (;;) { n2 = andor(); tok = readtoken(); if (tok == TBACKGND) { if (n2->type == NCMD || n2->type == NPIPE) { n2->ncmd.backgnd = 1; } else if (n2->type == NREDIR) { n2->type = NBACKGND; } else { n3 = (union node *)stalloc(sizeof (struct nredir)); n3->type = NBACKGND; n3->nredir.n = n2; n3->nredir.redirect = NULL; n2 = n3; } } if (n1 == NULL) { n1 = n2; } else { n3 = (union node *)stalloc(sizeof (struct nbinary)); n3->type = NSEMI; n3->nbinary.ch1 = n1; n3->nbinary.ch2 = n2; n1 = n3; } switch (tok) { case TBACKGND: case TSEMI: tok = readtoken(); /* FALLTHROUGH */ case TNL: if (tok == TNL) { parseheredoc(); if (nlflag) return n1; } else { tokpushback++; } checkkwd = 2; if (tokendlist[peektoken()]) return n1; break; case TEOF: if (heredoclist) parseheredoc(); else pungetc(); /* push back EOF on input */ return n1; default: if (nlflag) synexpect(-1); tokpushback++; return n1; } } } STATIC union node * andor(void) { union node *n1, *n2, *n3; int t; n1 = pipeline(); for (;;) { if ((t = readtoken()) == TAND) { t = NAND; } else if (t == TOR) { t = NOR; } else { tokpushback++; return n1; } n2 = pipeline(); n3 = (union node *)stalloc(sizeof (struct nbinary)); n3->type = t; n3->nbinary.ch1 = n1; n3->nbinary.ch2 = n2; n1 = n3; } } STATIC union node * pipeline(void) { union node *n1, *n2, *pipenode; struct nodelist *lp, *prev; int negate; negate = 0; TRACE(("pipeline: entered\n")); while (readtoken() == TNOT) negate = !negate; tokpushback++; n1 = command(); if (readtoken() == TPIPE) { pipenode = (union node *)stalloc(sizeof (struct npipe)); pipenode->type = NPIPE; pipenode->npipe.backgnd = 0; lp = (struct nodelist *)stalloc(sizeof (struct nodelist)); pipenode->npipe.cmdlist = lp; lp->n = n1; do { prev = lp; lp = (struct nodelist *)stalloc(sizeof (struct nodelist)); lp->n = command(); prev->next = lp; } while (readtoken() == TPIPE); lp->next = NULL; n1 = pipenode; } tokpushback++; if (negate) { n2 = (union node *)stalloc(sizeof (struct nnot)); n2->type = NNOT; n2->nnot.com = n1; return n2; } else return n1; } STATIC union node * command(void) { union node *n1, *n2; union node *ap, **app; union node *cp, **cpp; union node *redir, **rpp; int t, negate = 0; checkkwd = 2; redir = NULL; n1 = NULL; rpp = &redir; /* Check for redirection which may precede command */ while (readtoken() == TREDIR) { *rpp = n2 = redirnode; rpp = &n2->nfile.next; parsefname(); } tokpushback++; while (readtoken() == TNOT) { TRACE(("command: TNOT recognized\n")); negate = !negate; } tokpushback++; switch (readtoken()) { case TIF: n1 = (union node *)stalloc(sizeof (struct nif)); n1->type = NIF; if ((n1->nif.test = list(0)) == NULL) synexpect(-1); if (readtoken() != TTHEN) synexpect(TTHEN); n1->nif.ifpart = list(0); n2 = n1; while (readtoken() == TELIF) { n2->nif.elsepart = (union node *)stalloc(sizeof (struct nif)); n2 = n2->nif.elsepart; n2->type = NIF; if ((n2->nif.test = list(0)) == NULL) synexpect(-1); if (readtoken() != TTHEN) synexpect(TTHEN); n2->nif.ifpart = list(0); } if (lasttoken == TELSE) n2->nif.elsepart = list(0); else { n2->nif.elsepart = NULL; tokpushback++; } if (readtoken() != TFI) synexpect(TFI); checkkwd = 1; break; case TWHILE: case TUNTIL: { int got; n1 = (union node *)stalloc(sizeof (struct nbinary)); n1->type = (lasttoken == TWHILE)? NWHILE : NUNTIL; if ((n1->nbinary.ch1 = list(0)) == NULL) synexpect(-1); if ((got=readtoken()) != TDO) { TRACE(("expecting DO got %s %s\n", tokname[got], got == TWORD ? wordtext : "")); synexpect(TDO); } n1->nbinary.ch2 = list(0); if (readtoken() != TDONE) synexpect(TDONE); checkkwd = 1; break; } case TFOR: if (readtoken() != TWORD || quoteflag || ! goodname(wordtext)) synerror("Bad for loop variable"); n1 = (union node *)stalloc(sizeof (struct nfor)); n1->type = NFOR; n1->nfor.var = wordtext; if (readtoken() == TWORD && ! quoteflag && equal(wordtext, "in")) { app = ≈ while (readtoken() == TWORD) { n2 = (union node *)stalloc(sizeof (struct narg)); n2->type = NARG; n2->narg.text = wordtext; n2->narg.backquote = backquotelist; *app = n2; app = &n2->narg.next; } *app = NULL; n1->nfor.args = ap; if (lasttoken != TNL && lasttoken != TSEMI) synexpect(-1); } else { #ifndef GDB_HACK static const char argvars[5] = {CTLVAR, VSNORMAL|VSQUOTE, '@', '=', '\0'}; #endif n2 = (union node *)stalloc(sizeof (struct narg)); n2->type = NARG; n2->narg.text = (char *)argvars; n2->narg.backquote = NULL; n2->narg.next = NULL; n1->nfor.args = n2; /* * Newline or semicolon here is optional (but note * that the original Bourne shell only allowed NL). */ if (lasttoken != TNL && lasttoken != TSEMI) tokpushback++; } checkkwd = 2; if ((t = readtoken()) == TDO) t = TDONE; else if (t == TBEGIN) t = TEND; else synexpect(-1); n1->nfor.body = list(0); if (readtoken() != t) synexpect(t); checkkwd = 1; break; case TCASE: n1 = (union node *)stalloc(sizeof (struct ncase)); n1->type = NCASE; if (readtoken() != TWORD) synexpect(TWORD); n1->ncase.expr = n2 = (union node *)stalloc(sizeof (struct narg)); n2->type = NARG; n2->narg.text = wordtext; n2->narg.backquote = backquotelist; n2->narg.next = NULL; while (readtoken() == TNL); if (lasttoken != TWORD || ! equal(wordtext, "in")) synerror("expecting \"in\""); cpp = &n1->ncase.cases; noaliases = 1; /* turn off alias expansion */ checkkwd = 2, readtoken(); while (lasttoken != TESAC) { *cpp = cp = (union node *)stalloc(sizeof (struct nclist)); cp->type = NCLIST; app = &cp->nclist.pattern; if (lasttoken == TLP) readtoken(); for (;;) { *app = ap = (union node *)stalloc(sizeof (struct narg)); ap->type = NARG; ap->narg.text = wordtext; ap->narg.backquote = backquotelist; if (checkkwd = 2, readtoken() != TPIPE) break; app = &ap->narg.next; readtoken(); } ap->narg.next = NULL; if (lasttoken != TRP) noaliases = 0, synexpect(TRP); cp->nclist.body = list(0); checkkwd = 2; if ((t = readtoken()) != TESAC) { if (t != TENDCASE) noaliases = 0, synexpect(TENDCASE); else checkkwd = 2, readtoken(); } cpp = &cp->nclist.next; } noaliases = 0; /* reset alias expansion */ *cpp = NULL; checkkwd = 1; break; case TLP: n1 = (union node *)stalloc(sizeof (struct nredir)); n1->type = NSUBSHELL; n1->nredir.n = list(0); n1->nredir.redirect = NULL; if (readtoken() != TRP) synexpect(TRP); checkkwd = 1; break; case TBEGIN: n1 = list(0); if (readtoken() != TEND) synexpect(TEND); checkkwd = 1; break; /* Handle an empty command like other simple commands. */ case TSEMI: case TAND: case TOR: /* * An empty command before a ; doesn't make much sense, and * should certainly be disallowed in the case of `if ;'. */ if (!redir) synexpect(-1); case TNL: case TEOF: case TWORD: case TRP: tokpushback++; n1 = simplecmd(rpp, redir); goto checkneg; default: synexpect(-1); } /* Now check for redirection which may follow command */ while (readtoken() == TREDIR) { *rpp = n2 = redirnode; rpp = &n2->nfile.next; parsefname(); } tokpushback++; *rpp = NULL; if (redir) { if (n1->type != NSUBSHELL) { n2 = (union node *)stalloc(sizeof (struct nredir)); n2->type = NREDIR; n2->nredir.n = n1; n1 = n2; } n1->nredir.redirect = redir; } checkneg: if (negate) { n2 = (union node *)stalloc(sizeof (struct nnot)); n2->type = NNOT; n2->nnot.com = n1; return n2; } else return n1; } STATIC union node * simplecmd(union node **rpp, union node *redir) { union node *args, **app; union node **orig_rpp = rpp; union node *n = NULL, *n2; int negate = 0; /* If we don't have any redirections already, then we must reset */ /* rpp to be the address of the local redir variable. */ if (redir == 0) rpp = &redir; args = NULL; app = &args; /* * We save the incoming value, because we need this for shell * functions. There can not be a redirect or an argument between * the function name and the open parenthesis. */ orig_rpp = rpp; while (readtoken() == TNOT) { TRACE(("command: TNOT recognized\n")); negate = !negate; } tokpushback++; for (;;) { if (readtoken() == TWORD) { n = (union node *)stalloc(sizeof (struct narg)); n->type = NARG; n->narg.text = wordtext; n->narg.backquote = backquotelist; *app = n; app = &n->narg.next; } else if (lasttoken == TREDIR) { *rpp = n = redirnode; rpp = &n->nfile.next; parsefname(); /* read name of redirection file */ } else if (lasttoken == TLP && app == &args->narg.next && rpp == orig_rpp) { /* We have a function */ if (readtoken() != TRP) synexpect(TRP); #ifdef notdef if (! goodname(n->narg.text)) synerror("Bad function name"); #endif n->type = NDEFUN; n->narg.next = command(); goto checkneg; } else { tokpushback++; break; } } *app = NULL; *rpp = NULL; n = (union node *)stalloc(sizeof (struct ncmd)); n->type = NCMD; n->ncmd.backgnd = 0; n->ncmd.args = args; n->ncmd.redirect = redir; checkneg: if (negate) { n2 = (union node *)stalloc(sizeof (struct nnot)); n2->type = NNOT; n2->nnot.com = n; return n2; } else return n; } STATIC union node * makename(void) { union node *n; n = (union node *)stalloc(sizeof (struct narg)); n->type = NARG; n->narg.next = NULL; n->narg.text = wordtext; n->narg.backquote = backquotelist; return n; } void fixredir(union node *n, const char *text, int err) { TRACE(("Fix redir %s %d\n", text, err)); if (!err) n->ndup.vname = NULL; if (is_digit(text[0]) && text[1] == '\0') n->ndup.dupfd = digit_val(text[0]); else if (text[0] == '-' && text[1] == '\0') n->ndup.dupfd = -1; else { if (err) synerror("Bad fd number"); else n->ndup.vname = makename(); } } STATIC void parsefname(void) { union node *n = redirnode; if (readtoken() != TWORD) synexpect(-1); if (n->type == NHERE) { struct heredoc *here = heredoc; struct heredoc *p; int i; if (quoteflag == 0) n->type = NXHERE; TRACE(("Here document %d\n", n->type)); if (here->striptabs) { while (*wordtext == '\t') wordtext++; } if (! noexpand(wordtext) || (i = strlen(wordtext)) == 0 || i > EOFMARKLEN) synerror("Illegal eof marker for << redirection"); rmescapes(wordtext); here->eofmark = wordtext; here->next = NULL; if (heredoclist == NULL) heredoclist = here; else { for (p = heredoclist ; p->next ; p = p->next); p->next = here; } } else if (n->type == NTOFD || n->type == NFROMFD) { fixredir(n, wordtext, 0); } else { n->nfile.fname = makename(); } } /* * Input any here documents. */ STATIC void parseheredoc(void) { struct heredoc *here; union node *n; while (heredoclist) { here = heredoclist; heredoclist = here->next; if (needprompt) { setprompt(2); needprompt = 0; } readtoken1(pgetc(), here->here->type == NHERE? SQSYNTAX : DQSYNTAX, here->eofmark, here->striptabs); n = (union node *)stalloc(sizeof (struct narg)); n->narg.type = NARG; n->narg.next = NULL; n->narg.text = wordtext; n->narg.backquote = backquotelist; here->here->nhere.doc = n; } } STATIC int peektoken(void) { int t; t = readtoken(); tokpushback++; return (t); } STATIC int readtoken(void) { int t; int savecheckkwd = checkkwd; struct alias *ap; #if DEBUG int alreadyseen = tokpushback; #endif top: t = xxreadtoken(); if (checkkwd) { /* * eat newlines */ if (checkkwd == 2) { checkkwd = 0; while (t == TNL) { parseheredoc(); t = xxreadtoken(); } } else checkkwd = 0; /* * check for keywords and aliases */ if (t == TWORD && !quoteflag) { const char * const *pp; for (pp = parsekwd; *pp; pp++) { if (**pp == *wordtext && equal(*pp, wordtext)) { lasttoken = t = pp - parsekwd + KWDOFFSET; TRACE(("keyword %s recognized\n", tokname[t])); goto out; } } if (noaliases == 0 && (ap = lookupalias(wordtext, 1)) != NULL) { pushstring(ap->val, strlen(ap->val), ap); checkkwd = savecheckkwd; goto top; } } out: checkkwd = (t == TNOT) ? savecheckkwd : 0; } #if DEBUG if (!alreadyseen) TRACE(("token %s %s\n", tokname[t], t == TWORD ? wordtext : "")); else TRACE(("reread token %s %s\n", tokname[t], t == TWORD ? wordtext : "")); #endif return (t); } /* * Read the next input token. * If the token is a word, we set backquotelist to the list of cmds in * backquotes. We set quoteflag to true if any part of the word was * quoted. * If the token is TREDIR, then we set redirnode to a structure containing * the redirection. * In all cases, the variable startlinno is set to the number of the line * on which the token starts. * * [Change comment: here documents and internal procedures] * [Readtoken shouldn't have any arguments. Perhaps we should make the * word parsing code into a separate routine. In this case, readtoken * doesn't need to have any internal procedures, but parseword does. * We could also make parseoperator in essence the main routine, and * have parseword (readtoken1?) handle both words and redirection.] */ #define RETURN(token) return lasttoken = token STATIC int xxreadtoken(void) { int c; if (tokpushback) { tokpushback = 0; return lasttoken; } if (needprompt) { setprompt(2); needprompt = 0; } startlinno = plinno; for (;;) { /* until token or start of word found */ c = pgetc_macro(); if (c == ' ' || c == '\t') continue; /* quick check for white space first */ switch (c) { case ' ': case '\t': continue; case '#': while ((c = pgetc()) != '\n' && c != PEOF); pungetc(); continue; case '\\': if (pgetc() == '\n') { startlinno = ++plinno; if (doprompt) setprompt(2); else setprompt(0); continue; } pungetc(); goto breakloop; case '\n': plinno++; needprompt = doprompt; RETURN(TNL); case PEOF: RETURN(TEOF); case '&': if (pgetc() == '&') RETURN(TAND); pungetc(); RETURN(TBACKGND); case '|': if (pgetc() == '|') RETURN(TOR); pungetc(); RETURN(TPIPE); case ';': if (pgetc() == ';') RETURN(TENDCASE); pungetc(); RETURN(TSEMI); case '(': RETURN(TLP); case ')': RETURN(TRP); default: goto breakloop; } } breakloop: return readtoken1(c, BASESYNTAX, (char *)NULL, 0); #undef RETURN } /* * If eofmark is NULL, read a word or a redirection symbol. If eofmark * is not NULL, read a here document. In the latter case, eofmark is the * word which marks the end of the document and striptabs is true if * leading tabs should be stripped from the document. The argument firstc * is the first character of the input token or document. * * Because C does not have internal subroutines, I have simulated them * using goto's to implement the subroutine linkage. The following macros * will run code that appears at the end of readtoken1. */ #define CHECKEND() {goto checkend; checkend_return:;} #define PARSEREDIR() {goto parseredir; parseredir_return:;} #define PARSESUB() {goto parsesub; parsesub_return:;} #define PARSEBACKQOLD() {oldstyle = 1; goto parsebackq; parsebackq_oldreturn:;} #define PARSEBACKQNEW() {oldstyle = 0; goto parsebackq; parsebackq_newreturn:;} #define PARSEARITH() {goto parsearith; parsearith_return:;} STATIC int readtoken1(int firstc, char const *syntax, char *eofmark, int striptabs) { int c = firstc; char *out; int len; char line[EOFMARKLEN + 1]; struct nodelist *bqlist; int quotef; int dblquote; int varnest; /* levels of variables expansion */ int arinest; /* levels of arithmetic expansion */ int parenlevel; /* levels of parens in arithmetic */ int oldstyle; char const *prevsyntax; /* syntax before arithmetic */ int synentry; #if __GNUC__ /* Avoid longjmp clobbering */ (void) &out; (void) "ef; (void) &dblquote; (void) &varnest; (void) &arinest; (void) &parenlevel; (void) &oldstyle; (void) &prevsyntax; (void) &syntax; (void) &synentry; #endif startlinno = plinno; dblquote = 0; if (syntax == DQSYNTAX) dblquote = 1; quotef = 0; bqlist = NULL; varnest = 0; arinest = 0; parenlevel = 0; STARTSTACKSTR(out); loop: { /* for each line, until end of word */ CHECKEND(); /* set c to PEOF if at end of here document */ for (;;) { /* until end of line or end of word */ CHECKSTRSPACE(3, out); /* permit 3 calls to USTPUTC */ synentry = syntax[c]; switch(synentry) { case CNL: /* '\n' */ if (syntax == BASESYNTAX) goto endword; /* exit outer loop */ USTPUTC(c, out); plinno++; if (doprompt) setprompt(2); else setprompt(0); c = pgetc(); goto loop; /* continue outer loop */ case CWORD: USTPUTC(c, out); break; case CCTL: if (eofmark == NULL || dblquote) USTPUTC(CTLESC, out); USTPUTC(c, out); break; case CBACK: /* backslash */ c = pgetc(); if (c == PEOF) { USTPUTC('\\', out); pungetc(); } else if (c == '\n') { if (doprompt) setprompt(2); else setprompt(0); } else { if (dblquote && c != '\\' && c != '`' && c != '$' && (c != '"' || eofmark != NULL)) USTPUTC('\\', out); if (SQSYNTAX[c] == CCTL) USTPUTC(CTLESC, out); else if (eofmark == NULL) USTPUTC(CTLQUOTEMARK, out); USTPUTC(c, out); quotef++; } break; case CSQUOTE: if (eofmark == NULL) USTPUTC(CTLQUOTEMARK, out); syntax = SQSYNTAX; break; case CDQUOTE: if (eofmark == NULL) USTPUTC(CTLQUOTEMARK, out); syntax = DQSYNTAX; dblquote = 1; break; case CENDQUOTE: if (eofmark != NULL && arinest == 0 && varnest == 0) { USTPUTC(c, out); } else { if (arinest) { syntax = ARISYNTAX; dblquote = 0; } else if (eofmark == NULL) { syntax = BASESYNTAX; dblquote = 0; } quotef++; } break; case CVAR: /* '$' */ PARSESUB(); /* parse substitution */ break; case CENDVAR: /* '}' */ if (varnest > 0) { varnest--; USTPUTC(CTLENDVAR, out); } else { USTPUTC(c, out); } break; case CLP: /* '(' in arithmetic */ parenlevel++; USTPUTC(c, out); break; case CRP: /* ')' in arithmetic */ if (parenlevel > 0) { USTPUTC(c, out); --parenlevel; } else { if (pgetc() == ')') { if (--arinest == 0) { USTPUTC(CTLENDARI, out); syntax = prevsyntax; if (syntax == DQSYNTAX) dblquote = 1; else dblquote = 0; } else USTPUTC(')', out); } else { /* * unbalanced parens * (don't 2nd guess - no error) */ pungetc(); USTPUTC(')', out); } } break; case CBQUOTE: /* '`' */ PARSEBACKQOLD(); break; case CEOF: goto endword; /* exit outer loop */ default: if (varnest == 0) goto endword; /* exit outer loop */ USTPUTC(c, out); } c = pgetc_macro(); } } endword: if (syntax == ARISYNTAX) synerror("Missing '))'"); if (syntax != BASESYNTAX && ! parsebackquote && eofmark == NULL) synerror("Unterminated quoted string"); if (varnest != 0) { startlinno = plinno; synerror("Missing '}'"); } USTPUTC('\0', out); len = out - stackblock(); out = stackblock(); if (eofmark == NULL) { if ((c == '>' || c == '<') && quotef == 0 && len <= 2 && (*out == '\0' || is_digit(*out))) { PARSEREDIR(); return lasttoken = TREDIR; } else { pungetc(); } } quoteflag = quotef; backquotelist = bqlist; grabstackblock(len); wordtext = out; return lasttoken = TWORD; /* end of readtoken routine */ /* * Check to see whether we are at the end of the here document. When this * is called, c is set to the first character of the next input line. If * we are at the end of the here document, this routine sets the c to PEOF. */ checkend: { if (eofmark) { if (striptabs) { while (c == '\t') c = pgetc(); } if (c == *eofmark) { if (pfgets(line, sizeof line) != NULL) { char *p, *q; p = line; for (q = eofmark + 1 ; *q && *p == *q ; p++, q++); if (*p == '\n' && *q == '\0') { c = PEOF; plinno++; needprompt = doprompt; } else { pushstring(line, strlen(line), NULL); } } } } goto checkend_return; } /* * Parse a redirection operator. The variable "out" points to a string * specifying the fd to be redirected. The variable "c" contains the * first character of the redirection operator. */ parseredir: { char fd = *out; union node *np; np = (union node *)stalloc(sizeof (struct nfile)); if (c == '>') { np->nfile.fd = 1; c = pgetc(); if (c == '>') np->type = NAPPEND; else if (c == '&') np->type = NTOFD; else if (c == '|') np->type = NCLOBBER; else { np->type = NTO; pungetc(); } } else { /* c == '<' */ np->nfile.fd = 0; c = pgetc(); if (c == '<') { if (sizeof (struct nfile) != sizeof (struct nhere)) { np = (union node *)stalloc(sizeof (struct nhere)); np->nfile.fd = 0; } np->type = NHERE; heredoc = (struct heredoc *)stalloc(sizeof (struct heredoc)); heredoc->here = np; if ((c = pgetc()) == '-') { heredoc->striptabs = 1; } else { heredoc->striptabs = 0; pungetc(); } } else if (c == '&') np->type = NFROMFD; else if (c == '>') np->type = NFROMTO; else { np->type = NFROM; pungetc(); } } if (fd != '\0') np->nfile.fd = digit_val(fd); redirnode = np; goto parseredir_return; } /* * Parse a substitution. At this point, we have read the dollar sign * and nothing else. */ parsesub: { int subtype; int typeloc; int flags; char *p; #ifndef GDB_HACK static const char types[] = "}-+?="; #endif int bracketed_name = 0; /* used to handle ${[0-9]*} variables */ c = pgetc(); if (c != '(' && c != '{' && !is_name(c) && !is_special(c)) { USTPUTC('$', out); pungetc(); } else if (c == '(') { /* $(command) or $((arith)) */ if (pgetc() == '(') { PARSEARITH(); } else { pungetc(); PARSEBACKQNEW(); } } else { USTPUTC(CTLVAR, out); typeloc = out - stackblock(); USTPUTC(VSNORMAL, out); subtype = VSNORMAL; if (c == '{') { bracketed_name = 1; c = pgetc(); if (c == '#') { if ((c = pgetc()) == '}') c = '#'; else subtype = VSLENGTH; } else subtype = 0; } if (is_name(c)) { do { STPUTC(c, out); c = pgetc(); } while (is_in_name(c)); } else if (is_digit(c)) { if (bracketed_name) { do { STPUTC(c, out); c = pgetc(); } while (is_digit(c)); } else { STPUTC(c, out); c = pgetc(); } } else { if (! is_special(c)) badsub: synerror("Bad substitution"); USTPUTC(c, out); c = pgetc(); } STPUTC('=', out); flags = 0; if (subtype == 0) { switch (c) { case ':': flags = VSNUL; c = pgetc(); /*FALLTHROUGH*/ default: p = strchr(types, c); if (p == NULL) goto badsub; subtype = p - types + VSNORMAL; break; case '%': case '#': { int cc = c; subtype = c == '#' ? VSTRIMLEFT : VSTRIMRIGHT; c = pgetc(); if (c == cc) subtype++; else pungetc(); break; } } } else { pungetc(); } if (subtype != VSLENGTH && (dblquote || arinest)) flags |= VSQUOTE; *(stackblock() + typeloc) = subtype | flags; if (subtype != VSNORMAL) varnest++; } goto parsesub_return; } /* * Called to parse command substitutions. Newstyle is set if the command * is enclosed inside $(...); nlpp is a pointer to the head of the linked * list of commands (passed by reference), and savelen is the number of * characters on the top of the stack which must be preserved. */ parsebackq: { struct nodelist **nlpp; int savepbq; union node *n; char *volatile str; struct jmploc jmploc; struct jmploc *volatile savehandler; int savelen; int saveprompt; #if __GNUC__ /* Avoid longjmp clobbering */ (void) &saveprompt; #endif savepbq = parsebackquote; if (setjmp(jmploc.loc)) { if (str) ckfree(str); parsebackquote = 0; handler = savehandler; longjmp(handler->loc, 1); } INTOFF; str = NULL; savelen = out - stackblock(); if (savelen > 0) { str = ckmalloc(savelen); memcpy(str, stackblock(), savelen); } savehandler = handler; handler = &jmploc; INTON; if (oldstyle) { /* We must read until the closing backquote, giving special treatment to some slashes, and then push the string and reread it as input, interpreting it normally. */ char *out; int c; int savelen; char *str; STARTSTACKSTR(out); for (;;) { if (needprompt) { setprompt(2); needprompt = 0; } switch (c = pgetc()) { case '`': goto done; case '\\': if ((c = pgetc()) == '\n') { plinno++; if (doprompt) setprompt(2); else setprompt(0); /* * If eating a newline, avoid putting * the newline into the new character * stream (via the STPUTC after the * switch). */ continue; } if (c != '\\' && c != '`' && c != '$' && (!dblquote || c != '"')) STPUTC('\\', out); break; case '\n': plinno++; needprompt = doprompt; break; case PEOF: startlinno = plinno; synerror("EOF in backquote substitution"); break; default: break; } STPUTC(c, out); } done: STPUTC('\0', out); savelen = out - stackblock(); if (savelen > 0) { str = ckmalloc(savelen); memcpy(str, stackblock(), savelen); setinputstring(str, 1); } } nlpp = &bqlist; while (*nlpp) nlpp = &(*nlpp)->next; *nlpp = (struct nodelist *)stalloc(sizeof (struct nodelist)); (*nlpp)->next = NULL; parsebackquote = oldstyle; if (oldstyle) { saveprompt = doprompt; doprompt = 0; } n = list(0); if (oldstyle) doprompt = saveprompt; else { if (readtoken() != TRP) synexpect(TRP); } (*nlpp)->n = n; if (oldstyle) { /* * Start reading from old file again, ignoring any pushed back * tokens left from the backquote parsing */ popfile(); tokpushback = 0; } while (stackblocksize() <= savelen) growstackblock(); STARTSTACKSTR(out); if (str) { memcpy(out, str, savelen); STADJUST(savelen, out); INTOFF; ckfree(str); str = NULL; INTON; } parsebackquote = savepbq; handler = savehandler; if (arinest || dblquote) USTPUTC(CTLBACKQ | CTLQUOTE, out); else USTPUTC(CTLBACKQ, out); if (oldstyle) goto parsebackq_oldreturn; else goto parsebackq_newreturn; } /* * Parse an arithmetic expansion (indicate start of one and set state) */ parsearith: { if (++arinest == 1) { prevsyntax = syntax; syntax = ARISYNTAX; USTPUTC(CTLARI, out); if (dblquote) USTPUTC('"',out); else USTPUTC(' ',out); } else { /* * we collapse embedded arithmetic expansion to * parenthesis, which should be equivalent */ USTPUTC('(', out); } goto parsearith_return; } } /* end of readtoken */ #ifdef mkinit RESET { tokpushback = 0; checkkwd = 0; } #endif /* * Returns true if the text contains nothing to expand (no dollar signs * or backquotes). */ STATIC int noexpand(char *text) { char *p; char c; p = text; while ((c = *p++) != '\0') { if ( c == CTLQUOTEMARK) continue; if (c == CTLESC) p++; else if (BASESYNTAX[(int)c] == CCTL) return 0; } return 1; } /* * Return true if the argument is a legal variable name (a letter or * underscore followed by zero or more letters, underscores, and digits). */ int goodname(char *name) { char *p; p = name; if (! is_name(*p)) return 0; while (*++p) { if (! is_in_name(*p)) return 0; } return 1; } /* * Called when an unexpected token is read during the parse. The argument * is the token that is expected, or -1 if more than one type of token can * occur at this point. */ STATIC void synexpect(int token) { char msg[64]; if (token >= 0) { fmtstr(msg, 64, "%s unexpected (expecting %s)", tokname[lasttoken], tokname[token]); } else { fmtstr(msg, 64, "%s unexpected", tokname[lasttoken]); } synerror(msg); } STATIC void synerror(char *msg) { if (commandname) outfmt(&errout, "%s: %d: ", commandname, startlinno); outfmt(&errout, "Syntax error: %s\n", msg); error((char *)NULL); } STATIC void setprompt(int which) { whichprompt = which; #ifndef NO_HISTORY #ifdef EDITLINE if (!editable) #else if (!el) #endif /* EDITLINE */ #endif /* !NO_HISTORY */ out2str(getprompt(NULL)); } /* * called by editline -- any expansions to the prompt * should be added here. */ char * getprompt(void *unused __unused) { static char ps[PROMPTLEN]; char *fmt; int i, j, trim; /* * Select prompt format. */ switch (whichprompt) { case -1: fmt = pseval(); break; case 0: fmt = ""; break; case 1: fmt = ps1val(); break; case 2: fmt = ps2val(); break; default: return ""; } /* * Format prompt string. */ for (i = 0; (i < 127) && (*fmt != '\0'); i++, fmt++) if (*fmt == '\\') switch (*++fmt) { /* * Hostname. * * \h specifies just the local hostname, * \H specifies fully-qualified hostname. */ case 'h': case 'H': ps[i] == '\0'; gethostname(&ps[i], PROMPTLEN - i); /* Skip to end of hostname. */ trim = (*fmt == 'h') ? '.' : '\0'; while ((ps[i+1] != '\0') && (ps[i+1] != trim)) i++; break; /* * Working directory. * * \W specifies just the final component, * \w specifies the entire path. */ case 'W': case 'w': ps[i] == '\0'; getcwd(&ps[i], PROMPTLEN - i); if (*fmt == 'W') { /* Final path component only. */ trim = 1; for (j = i; ps[j] != '\0'; j++) if (ps[j] == '/') trim = j + 1; memmove(&ps[i], &ps[trim], j - trim + 1); } /* Skip to end of path. */ while (ps[i + 1] != '\0') i++; break; /* * Superuser status. * * '$' for normal users, '#' for root. */ case '$': ps[i] = (geteuid() != 0) ? '$' : '#'; break; /* * A literal \. */ case '\\': ps[i] = '\\'; break; /* * Emit unrecognized formats verbatim. */ default: ps[i++] = '\\'; ps[i] = *fmt; break; } else ps[i] = *fmt; ps[i] = '\0'; return (ps); } /* * $PchId: parser.c,v 1.6 2006/05/29 13:08:11 philip Exp $ */