From c284e87d6b79370b9bae7be1b89886cd2465ecdb Mon Sep 17 00:00:00 2001 From: Tony Finch Date: Wed, 18 Dec 2002 20:50:44 +0000 Subject: Sync with up-stream version, including a number of bug-fixes: * The partial-evaluation of #elif sequences was broken and the spaghetti logic of its implementation was too hard to understand. I've re-done it using a straight-forward table-driven push-down automaton. * The pre-processor line parser did not allow for all of the weird places that people might put comments, which could have caused it to add syntax-errors to the output by removing a #if line containing the start- or end-marker of a comment. * The lexer didn't need to special-case the handling of string-literals or character-constants, but it did need to learn about line-continuations (backslash-newline). * The input routine was buggy and bit-rotten and trivially replacable with fgets(). I've also made the program static- and const-safe and improved the presentation-order. The formatting of the state-transition tables remains non-stylish. This commit-messsage was brought to you by code-point 45. MFC-after: one-week --- usr.bin/unifdef/unifdef.1 | 52 ++- usr.bin/unifdef/unifdef.c | 1054 ++++++++++++++++++++------------------------- 2 files changed, 506 insertions(+), 600 deletions(-) (limited to 'usr.bin/unifdef') diff --git a/usr.bin/unifdef/unifdef.1 b/usr.bin/unifdef/unifdef.1 index 23a2a1a9b1cd..4a1e2f98cec5 100644 --- a/usr.bin/unifdef/unifdef.1 +++ b/usr.bin/unifdef/unifdef.1 @@ -33,7 +33,7 @@ .\" SUCH DAMAGE. .\" .\" @(#)unifdef.1 8.2 (Berkeley) 4/1/94 -.\" $dotat: things/unifdef.1,v 1.26 2002/09/24 19:44:12 fanf2 Exp $ +.\" $dotat: things/unifdef.1,v 1.40 2002/12/13 11:33:34 fanf2 Exp $ .\" $FreeBSD$ .\" .Dd September 24, 2002 @@ -112,13 +112,9 @@ utility also understands just enough about C to know when one of the directives is inactive because it is inside a comment, -or a single or double quote. -Parsing for quotes is very simplistic: -when it finds an open quote, -it ignores everything (except escaped quotes) -until it finds a close quote, and -it will not complain if it gets -to the end of a line and finds no backslash for continuation. +or affected by a backslash-continued line. +It spots unusually-formatted preprocessor directives +and knows when the layout is too odd to handle. .Pp A script called .Nm unifdefall @@ -194,7 +190,9 @@ for creating command lines. .Pp .It Fl t -Disables parsing for C comments and quotes, which is useful +Disables parsing for C comments +and line continuations, +which is useful for plain text. .Pp .It Fl iD Ns Ar sym Ns Op = Ns Ar val @@ -209,7 +207,8 @@ or code which is under construction, then you must tell .Nm which symbols are used for that purpose so that it will not try to parse -for quotes and comments +comments +and line continuations inside those .Ic #ifdef Ns s . One specifies ignored symbols with @@ -258,12 +257,23 @@ option of .Sh DIAGNOSTICS .Bl -item .It -Inappropriate elif, else or endif. +Too many levels of nesting. +.It +Inappropriate +.Ic #elif , +.Ic #else +or +.Ic #endif . +.It +Obfuscated preprocessor control line. .It Premature .Tn EOF -with line numbers of the unterminated -.Ic #ifdef Ns s . +(with the line number of the most recent unterminated +.Ic #if Ns ). +.It +.Tn EOF +in comment. .El .Pp The @@ -273,9 +283,23 @@ utility exits 0 if the output is an exact copy of the input, .Sh BUGS Expression evaluation is very limited. .Pp -Does not work correctly if input contains null characters. +Preprocessor control lines split across more than one physical line +(because of comments or backslash-newline) +cannot be handled. +.Pp +Trigraphs are not recognized. +.Pp +There is no support for symbols with different definitions at +different points in the source file. +.Pp +The text-mode and ignore functionality doesn't correspond to modern +.Xr cpp 1 +behaviour. .Sh HISTORY The .Nm command appeared in .Bx 4.3 . +.Tn ANSI\~C +support was added in +.Fx 4.7 . diff --git a/usr.bin/unifdef/unifdef.c b/usr.bin/unifdef/unifdef.c index 553b5ef57838..c17a1962c514 100644 --- a/usr.bin/unifdef/unifdef.c +++ b/usr.bin/unifdef/unifdef.c @@ -44,7 +44,7 @@ static const char copyright[] = #ifdef __IDSTRING __IDSTRING(Berkeley, "@(#)unifdef.c 8.1 (Berkeley) 6/6/93"); __IDSTRING(NetBSD, "$NetBSD: unifdef.c,v 1.8 2000/07/03 02:51:36 matt Exp $"); -__IDSTRING(dotat, "$dotat: things/unifdef.c,v 1.75 2002/09/24 19:16:29 fanf2 Exp $"); +__IDSTRING(dotat, "$dotat: things/unifdef.c,v 1.130 2002/12/13 15:26:41 fanf2 Exp $"); #endif #ifdef __FBSDID __FBSDID("$FreeBSD$"); @@ -54,8 +54,6 @@ __FBSDID("$FreeBSD$"); /* * unifdef - remove ifdef'ed lines * - * Warning: will not work correctly if input contains nul characters. - * * Wishlist: * provide an option which will append the name of the * appropriate symbol after #else's and #endif's @@ -77,161 +75,132 @@ __FBSDID("$FreeBSD$"); /* types of input lines: */ typedef enum { LT_PLAIN, /* ordinary line */ + LT_TRUEI, /* a true #if with ignore flag */ + LT_FALSEI, /* a false #if with ignore flag */ + LT_IF, /* an unknown #if */ LT_TRUE, /* a true #if */ LT_FALSE, /* a false #if */ + LT_ELIF, /* an unknown #elif */ LT_ELTRUE, /* a true #elif */ LT_ELFALSE, /* a false #elif */ - LT_IF, /* an unknown #if */ - LT_ELIF, /* an unknown #elif */ LT_ELSE, /* #else */ LT_ENDIF, /* #endif */ - LT_EOF /* end of file */ + LT_EOF, /* end of file */ + LT_COUNT } Linetype; -typedef enum { /* 0 or 1: pass thru; 1 or 2: ignore comments */ - REJ_NO, - REJ_IGNORE, - REJ_YES -} Reject_level; +static char const * const linetype_name[] = { + "PLAIN", "TRUEI", "FALSEI", "IF", "TRUE", "FALSE", + "ELIF", "ELTRUE", "ELFALSE", "ELSE", "ENDIF", "EOF" +}; +/* state of #if processing */ typedef enum { - NO_COMMENT = false, - C_COMMENT, - CXX_COMMENT -} Comment_state; + IS_OUTSIDE, + IS_FALSE_PREFIX, /* false #if followed by false #elifs */ + IS_TRUE_PREFIX, /* first non-false #(el)if is true */ + IS_PASS_MIDDLE, /* first non-false #(el)if is unknown */ + IS_FALSE_MIDDLE, /* a false #elif after a pass state */ + IS_TRUE_MIDDLE, /* a true #elif after a pass state */ + IS_PASS_ELSE, /* an else after a pass state */ + IS_FALSE_ELSE, /* an else after a true state */ + IS_TRUE_ELSE, /* an else after only false states */ + IS_FALSE_TRAILER, /* #elifs after a true are false */ + IS_COUNT +} Ifstate; + +static char const * const ifstate_name[] = { + "OUTSIDE", "FALSE_PREFIX", "TRUE_PREFIX", + "PASS_MIDDLE", "FALSE_MIDDLE", "TRUE_MIDDLE", + "PASS_ELSE", "FALSE_ELSE", "TRUE_ELSE", + "FALSE_TRAILER" +}; +/* state of comment parser */ typedef enum { - QUOTE_NONE = false, - QUOTE_SINGLE, - QUOTE_DOUBLE -} Quote_state; - -const char *const errs[] = { -#define NO_ERR 0 - "", -#define END_ERR 1 - "", -#define ELIF_ERR 2 - "Inappropriate elif", -#define ELSE_ERR 3 - "Inappropriate else", -#define ENDIF_ERR 4 - "Inappropriate endif", -#define IEOF_ERR 5 - "Premature EOF in ifdef", -#define CEOF_ERR 6 - "Premature EOF in comment", -#define Q1EOF_ERR 7 - "Premature EOF in quoted character", -#define Q2EOF_ERR 8 - "Premature EOF in quoted string" + NO_COMMENT = false, /* outside a comment */ + C_COMMENT, /* in a comment like this one */ + CXX_COMMENT, /* between // and end of line */ + STARTING_COMMENT, /* just after slash-backslash-newline */ + FINISHING_COMMENT /* star-backslash-newline in a C comment */ +} Comment_state; + +static char const * const comment_name[] = { + "NO", "C", "CXX", "STARTING", "FINISHING" }; -/* - * These are the operators that are supported by the expression evaluator. - */ -static int op_lt(int a, int b) { return a < b; } -static int op_gt(int a, int b) { return a > b; } -static int op_le(int a, int b) { return a <= b; } -static int op_ge(int a, int b) { return a >= b; } -static int op_eq(int a, int b) { return a == b; } -static int op_ne(int a, int b) { return a != b; } -static int op_or(int a, int b) { return a || b; } -static int op_and(int a, int b) { return a && b; } +/* state of preprocessor line parser */ +typedef enum { + LS_START, /* only space and comments on this line */ + LS_HASH, /* only space, comments, and a hash */ + LS_DIRTY /* this line can't be a preprocessor line */ +} Line_state; -struct ops; +static char const * const linestate_name[] = { + "START", "HASH", "DIRTY" +}; /* - * An evaluation function takes three arguments, as follows: (1) a pointer to - * an element of the precedence table which lists the operators at the current - * level of precedence; (2) a pointer to an integer which will receive the - * value of the expression; and (3) a pointer to a char* that points to the - * expression to be evaluated and that is updated to the end of the expression - * when evaluation is complete. The function returns LT_FALSE if the value of - * the expression is zero, LT_TRUE if it is non-zero, or LT_IF if the - * expression could not be evaluated. + * Minimum translation limits from ISO/IEC 9899:1999 5.2.4.1 */ -typedef Linetype eval_fn(struct ops *, int *, const char **); - -eval_fn eval_table, eval_unary; +#define MAXDEPTH 64 /* maximum #if nesting */ +#define MAXLINE 4096 /* maximum length of line */ +#define MAXSYMS 4096 /* maximum number of symbols */ /* - * The precedence table. Expressions involving binary operators are evaluated - * in a table-driven way by eval_table. When it evaluates a subexpression it - * calls the inner function with its first argument pointing to the next - * element of the table. Innermost expressions have special non-table-driven - * handling. + * Globals. */ -struct ops { - eval_fn *inner; - struct op { - const char *str; - int (*fn)(int, int); - } op[5]; -} eval_ops[] = { - { eval_table, { { "||", op_or } } }, - { eval_table, { { "&&", op_and } } }, - { eval_table, { { "==", op_eq }, - { "!=", op_ne } } }, - { eval_unary, { { "<=", op_le }, - { ">=", op_ge }, - { "<", op_lt }, - { ">", op_gt } } } -}; -FILE *input; -const char *filename; -int linenum; /* current line number */ -int stifline; /* start of current #if */ -int stqcline; /* start of current coment or quote */ -bool keepthis; /* ignore this #if's value 'cause it's const */ - -#define MAXLINE 1024 -#define KWSIZE 8 -/* tline has extra space so that it isn't overflowed when editing #elifs */ -char tline[MAXLINE+KWSIZE]; /* input buffer */ -char *keyword; /* used for editing #elif's */ - -bool complement; /* -c option in effect: do the complement */ -bool debugging; /* -d option in effect: debugging reports */ -bool killconsts; /* -k option in effect: eval constant #ifs */ -bool lnblank; /* -l option in effect: blank deleted lines */ -bool symlist; /* -s option in effect: output symbol list */ -bool text; /* -t option in effect: this is a text file */ - -int exitstat; /* program exit status */ - -#define MAXSYMS 1000 -const char *symname[MAXSYMS]; /* symbol name */ -const char *value[MAXSYMS]; /* -Dsym=value */ -bool ignore[MAXSYMS]; /* -iDsym or -iUsym */ - -int nsyms = 1; /* symbol 0 is used for tracking #ifs */ - -Reject_level reject; /* what kind of filtering we are doing */ -Comment_state incomment; /* inside C comment */ -Quote_state inquote; /* inside single or double quotes */ - -Linetype checkline(int *); -void debug(const char *, ...); -Linetype process(int); -void doif(int, Linetype, bool); -void elif2if(void); -void elif2endif(void); -void error(int, int); -void addsym(bool, bool, char *); -int findsym(const char *); -void flushline(bool); -int getline(char *, int, FILE *, bool); -Linetype ifeval(const char **); -int main(int, char **); -const char *skipcomment(const char *); -const char *skipquote(const char *, Quote_state); -const char *skipsym(const char *); -void usage(void); +static bool complement; /* -c: do the complement */ +static bool debugging; /* -d: debugging reports */ +static bool killconsts; /* -k: eval constant #ifs */ +static bool lnblank; /* -l: blank deleted lines */ +static bool symlist; /* -s: output symbol list */ +static bool text; /* -t: this is a text file */ + +static const char *symname[MAXSYMS]; /* symbol name */ +static const char *value[MAXSYMS]; /* -Dsym=value */ +static bool ignore[MAXSYMS]; /* -iDsym or -iUsym */ +static int nsyms; /* number of symbols */ + +static FILE *input; /* input file pointer */ +static const char *filename; /* input file name */ +static int linenum; /* current line number */ + +static char tline[MAXLINE+10]; /* input buffer plus space */ +static char *keyword; /* used for editing #elif's */ + +static Comment_state incomment; /* comment parser state */ +static Line_state linestate; /* #if line parser state */ +static Ifstate ifstate[MAXDEPTH]; /* #if processor state */ +static bool ignoring[MAXDEPTH]; /* ignore comments state */ +static int stifline[MAXDEPTH]; /* start of current #if */ +static int depth; /* current #if nesting */ +static bool keepthis; /* don't delete constant #if */ + +static int exitstat; /* program exit status */ + +static void addsym(bool, bool, char *); +static void debug(const char *, ...); +static void error(const char *); +static int findsym(const char *); +static void flushline(bool); +static Linetype getline(void); +static Linetype ifeval(const char **); +static void nest(void); +static void process(void); +static const char *skipcomment(const char *); +static const char *skipsym(const char *); +static void state(Ifstate); +static int strlcmp(const char *, const char *, size_t); +static void unignore(void); +static void usage(void); #define endsym(c) (!isalpha((unsigned char)c) && !isdigit((unsigned char)c) && c != '_') +/* + * The main program. + */ int main(int argc, char *argv[]) { @@ -260,24 +229,24 @@ main(int argc, char *argv[]) addsym(false, false, optarg); break; case 'I': - /* ignore for compatibility with cpp */ + /* no-op for compatibility with cpp */ break; case 'c': /* treat -D as -U and vice versa */ complement = true; break; - case 'k': /* process constant #ifs */ - killconsts = true; - break; case 'd': debugging = true; break; + case 'k': /* process constant #ifs */ + killconsts = true; + break; case 'l': /* blank deleted lines instead of omitting them */ lnblank = true; break; case 's': /* only output list of symbols that control #ifs */ symlist = true; break; - case 't': /* don't parse C comments or strings */ + case 't': /* don't parse C comments */ text = true; break; default: @@ -285,7 +254,7 @@ main(int argc, char *argv[]) } argc -= optind; argv += optind; - if (nsyms == 1 && !symlist) { + if (nsyms == 0 && !symlist) { warnx("must -D or -U at least one symbol"); usage(); } @@ -294,20 +263,20 @@ main(int argc, char *argv[]) } else if (argc == 1 && strcmp(*argv, "-") != 0) { filename = *argv; if ((input = fopen(filename, "r")) != NULL) { - (void) process(0); + process(); (void) fclose(input); } else err(2, "can't open %s", *argv); } else { filename = "[stdin]"; input = stdin; - (void) process(0); + process(); } exit(exitstat); } -void +static void usage(void) { fprintf (stderr, "usage: %s", @@ -316,305 +285,297 @@ usage(void) } /* - * This function processes #if lines and alters the pass-through - * state accordingly. All the complicated state transition suff is - * dealt with in this function, as well as checking that the - * #if/#elif/#else/#endif lines happen in the correct order. Lines - * between #if lines are handled by a recursive call to process(). + * A state transition function alters the global #if processing state + * in a particular way. The table below is indexed by the current + * processing state and the type of the current line. A NULL entry + * indicate that processing is complete. + * + * Nesting is handled by keeping a stack of states; some transition + * functions increase or decrease the depth. They also maintin the + * ignore state on a stack. In some complicated cases they have to + * alter the preprocessor directive, as follows. + * + * When we have processed a group that starts off with a known-false + * #if/#elif sequence (which has therefore been deleted) followed by a + * #elif that we don't understand and therefore must keep, we turn the + * latter into a #if to keep the nesting correct. + * + * When we find a true #elif in a group, the following block will + * always be kept and the rest of the sequence after the next #elif or + * #else will be discarded. We change the #elif to #else and the + * following directive to #endif since this has the desired behaviour. + */ +typedef void state_fn(void); + +/* report an error */ +static void Eelif (void) { error("Inappropriate #elif"); } +static void Eelse (void) { error("Inappropriate #else"); } +static void Eendif(void) { error("Inappropriate #endif"); } +static void Eeof (void) { error("Premature EOF"); } +static void Eioccc(void) { error("Obfuscated preprocessor control line"); } +/* plain line handling */ +static void print (void) { flushline(true); } +static void drop (void) { flushline(false); } +/* output lacks group's start line */ +static void Strue (void) { drop(); unignore(); state(IS_TRUE_PREFIX); } +static void Sfalse(void) { drop(); unignore(); state(IS_FALSE_PREFIX); } +static void Selse (void) { drop(); state(IS_TRUE_ELSE); } +/* print/pass this block */ +static void Pelif (void) { print(); unignore(); state(IS_PASS_MIDDLE); } +static void Pelse (void) { print(); state(IS_PASS_ELSE); } +static void Pendif(void) { print(); --depth; } +/* discard this block */ +static void Dfalse(void) { drop(); unignore(); state(IS_FALSE_TRAILER); } +static void Delif (void) { drop(); unignore(); state(IS_FALSE_MIDDLE); } +static void Delse (void) { drop(); state(IS_FALSE_ELSE); } +static void Dendif(void) { drop(); --depth; } +/* first line of group */ +static void Fdrop (void) { nest(); Dfalse(); } +static void Fpass (void) { nest(); Pelif(); } +static void Ftrue (void) { nest(); Strue(); } +static void Ffalse(void) { nest(); Sfalse(); } +/* ignore comments in this block */ +static void Idrop (void) { Fdrop(); ignore[depth] = true; } +static void Itrue (void) { Ftrue(); ignore[depth] = true; } +static void Ifalse(void) { Ffalse(); ignore[depth] = true; } +/* modify this line */ +static void +Mpass (void) { strncpy(keyword, "if ", 4); Pelif(); } +static void +Mtrue (void) { strcpy(keyword, "else\n"); print(); state(IS_TRUE_MIDDLE); } +static void +Melif (void) { strcpy(keyword, "endif\n"); print(); state(IS_FALSE_TRAILER); } +static void +Melse (void) { strcpy(keyword, "endif\n"); print(); state(IS_FALSE_ELSE); } + +static state_fn * const trans_table[IS_COUNT][LT_COUNT] = { +/* IS_OUTSIDE */ +{print,Itrue,Ifalse,Fpass,Ftrue,Ffalse,Eelif, Eelif, Eelif, Eelse,Eendif,NULL}, +/* IS_FALSE_PREFIX */ +{drop, Idrop,Idrop, Fdrop,Fdrop,Fdrop, Mpass, Strue, Sfalse,Selse,Dendif,Eeof}, +/* IS_TRUE_PREFIX */ +{print,Itrue,Ifalse,Fpass,Ftrue,Ffalse,Dfalse,Dfalse,Dfalse,Delse,Dendif,Eeof}, +/* IS_PASS_MIDDLE */ +{print,Itrue,Ifalse,Fpass,Ftrue,Ffalse,Pelif, Mtrue, Delif, Pelse,Pendif,Eeof}, +/* IS_FALSE_MIDDLE */ +{drop, Idrop,Idrop, Fdrop,Fdrop,Fdrop, Pelif, Mtrue, Delif, Pelse,Pendif,Eeof}, +/* IS_TRUE_MIDDLE */ +{print,Itrue,Ifalse,Fpass,Ftrue,Ffalse,Melif, Melif, Melif, Melse,Pendif,Eeof}, +/* IS_PASS_ELSE */ +{print,Itrue,Ifalse,Fpass,Ftrue,Ffalse,Eelif, Eelif, Eelif, Eelse,Pendif,Eeof}, +/* IS_FALSE_ELSE */ +{drop, Idrop,Idrop, Fdrop,Fdrop,Fdrop, Eelif, Eelif, Eelif, Eelse,Dendif,Eeof}, +/* IS_TRUE_ELSE */ +{print,Itrue,Ifalse,Fpass,Ftrue,Ffalse,Eelif, Eelif, Eelif, Eelse,Dendif,Eeof}, +/* IS_FALSE_TRAILER */ +{drop, Idrop,Idrop, Fdrop,Fdrop,Fdrop, Dfalse,Dfalse,Dfalse,Delse,Dendif,Eeof} +/*PLAIN TRUEI FALSEI IF TRUE FALSE ELIF ELTRUE ELFALSE ELSE ENDIF EOF*/ +}; + +/* + * State machine utility functions */ -void -doif(int depth, Linetype lineval, bool ignoring) +static void +nest(void) { - Reject_level savereject; - bool active; - bool donetrue; - bool inelse; - int saveline; - - debug("#if line %d code %d depth %d", - linenum, lineval, depth); - saveline = stifline; - stifline = linenum; - savereject = reject; - inelse = false; - donetrue = false; - if (lineval == LT_IF || reject != REJ_NO) { - active = false; - ignoring = false; - flushline(true); - } else if (ignoring) { - active = false; - flushline(true); - if (lineval == LT_FALSE) - reject = REJ_IGNORE; - else - donetrue = true; - } else { - active = true; - flushline(false); - if (lineval == LT_FALSE) - reject = REJ_YES; - else - donetrue = true; - } - debug("active %d ignore %d", active, ignoring); - for (;;) { - switch (lineval = process(depth)) { - case LT_ELIF: - debug("#elif start %d line %d code %d depth %d", - stifline, linenum, lineval, depth); - if (inelse) - error(ELIF_ERR, depth); - donetrue = false; - reject = savereject; - if (active) { - active = false; - elif2if(); - flushline(true); - } else { - ignoring = false; - flushline(true); - } - debug("active %d ignore %d", active, ignoring); - break; - case LT_ELTRUE: - case LT_ELFALSE: - debug("#elif start %d line %d code %d depth %d", - stifline, linenum, lineval, depth); - if (inelse) - error(ELIF_ERR, depth); - if (active) { - flushline(false); - } else { - ignoring = false; - active = true; - elif2endif(); - flushline(true); - } - if (lineval == LT_ELFALSE) - reject = REJ_YES; - else { - reject = REJ_NO; - donetrue = true; - } - debug("active %d ignore %d", active, ignoring); - break; - case LT_ELSE: - debug("#else start %d line %d code %d depth %d", - stifline, linenum, lineval, depth); - if (inelse) - error(ELSE_ERR, depth); - if (active) { - flushline(false); - if (reject == REJ_YES && !donetrue) - reject = REJ_NO; - else - reject = REJ_YES; - } else { - flushline(true); - if (ignoring) { - if (reject == REJ_IGNORE) - reject = REJ_NO; - else - reject = REJ_IGNORE; - } - } - inelse = true; - debug("active %d ignore %d", active, ignoring); - break; - case LT_ENDIF: - debug("#endif start %d line %d code %d depth %d", - stifline, linenum, lineval, depth); - if (active) - flushline(false); - else - flushline(true); - reject = savereject; - stifline = saveline; - return; - default: - /* bug */ - abort(); - } + depth += 1; + if (depth >= MAXDEPTH) + error("Too many levels of nesting"); + stifline[depth] = linenum; +} +static void +state(Ifstate is) +{ + ifstate[depth] = is; +} +static void +unignore(void) +{ + ignore[depth] = ignore[depth-1]; +} + +/* + * Write a line to the output or not, according to command line options. + */ +static void +flushline(bool keep) +{ + if (symlist) + return; + if (keep ^ complement) + fputs(tline, stdout); + else { + if (lnblank) + putc('\n', stdout); + exitstat = 1; } } /* - * The main file processing routine. This function deals with passing - * through normal non-#if lines, correct nesting of #if sections, and - * checking that things terminate correctly at the end of file. The - * complicated stuff is delegated to doif(). + * The driver for the state machine. */ -Linetype -process(int depth) +static void +process(void) { Linetype lineval; - int cursym; + state_fn *trans; for (;;) { linenum++; - if (getline(tline, MAXLINE, input, false) == EOF) { - if (incomment) - error(CEOF_ERR, depth); - if (inquote == QUOTE_SINGLE) - error(Q1EOF_ERR, depth); - if (inquote == QUOTE_DOUBLE) - error(Q2EOF_ERR, depth); - if (depth != 0) - error(IEOF_ERR, depth); - return LT_EOF; - } - switch (lineval = checkline(&cursym)) { - case LT_PLAIN: - flushline(true); + lineval = getline(); + trans = trans_table[ifstate[depth]][lineval]; + if (trans == NULL) break; - case LT_IF: - case LT_TRUE: - case LT_FALSE: - doif(depth + 1, lineval, ignore[cursym]); - break; - case LT_ELIF: - case LT_ELTRUE: - case LT_ELFALSE: - case LT_ELSE: - case LT_ENDIF: - if (depth != 0) - return lineval; - if (lineval == LT_ENDIF) - error(ENDIF_ERR, depth); - if (lineval == LT_ELSE) - error(ELSE_ERR, depth); - error(ELIF_ERR, depth); - default: - /* bug */ - abort(); - } + trans(); + debug("process %s -> %s depth %d", + linetype_name[lineval], + ifstate_name[ifstate[depth]], depth); } + if (incomment) + error("EOF in comment"); } /* - * Parse a line and determine its type. + * Parse a line and determine its type. We keep the preprocessor line + * parser state between calls in a global variable. */ -Linetype -checkline(int *cursym) +static Linetype +getline(void) { const char *cp; - char *symp; + int cursym; + int kwlen; Linetype retval; - char kw[KWSIZE]; + Comment_state wascomment; + if (fgets(tline, MAXLINE, input) == NULL) + return LT_EOF; retval = LT_PLAIN; + wascomment = incomment; cp = skipcomment(tline); - if (*cp != '#' - || incomment - || inquote == QUOTE_SINGLE - || inquote == QUOTE_DOUBLE - ) - goto eol; - - cp = skipcomment(++cp); - keyword = (char *)cp; - symp = kw; - while (!endsym(*cp)) { - *symp = *cp++; - if (++symp >= &kw[KWSIZE]) - goto eol; + if (linestate == LS_START) { + if (*cp == '#') { + linestate = LS_HASH; + cp = skipcomment(cp + 1); + } else if (*cp != '\0') + linestate = LS_DIRTY; } - *symp = '\0'; - - if (strcmp(kw, "ifdef") == 0) { - retval = LT_TRUE; - goto ifdef; - } else if (strcmp(kw, "ifndef") == 0) { - retval = LT_FALSE; - ifdef: - cp = skipcomment(++cp); - if (incomment) { + if (!incomment && linestate == LS_HASH) { + keyword = tline + (cp - tline); + cp = skipsym(cp); + kwlen = cp - keyword; + if (strncmp(cp, "\\\n", 2) == 0) + Eioccc(); + if (strlcmp("ifdef", keyword, kwlen) == 0 || + strlcmp("ifndef", keyword, kwlen) == 0) { + cp = skipcomment(cp); + if ((cursym = findsym(cp)) < 0) + retval = LT_IF; + else { + retval = (keyword[2] == 'n') + ? LT_FALSE : LT_TRUE; + if (value[cursym] == NULL) + retval = (retval == LT_TRUE) + ? LT_FALSE : LT_TRUE; + if (ignore[cursym]) + retval = (retval == LT_TRUE) + ? LT_TRUEI : LT_FALSEI; + } + cp = skipsym(cp); + } else if (strlcmp("if", keyword, kwlen) == 0) + retval = ifeval(&cp); + else if (strlcmp("elif", keyword, kwlen) == 0) + retval = ifeval(&cp) - LT_IF + LT_ELIF; + else if (strlcmp("else", keyword, kwlen) == 0) + retval = LT_ELSE; + else if (strlcmp("endif", keyword, kwlen) == 0) + retval = LT_ENDIF; + else { + linestate = LS_DIRTY; retval = LT_PLAIN; - goto eol; } - if ((*cursym = findsym(cp)) == 0) - retval = LT_IF; - else if (value[*cursym] == NULL) - retval = (retval == LT_TRUE) - ? LT_FALSE : LT_TRUE; - } else if (strcmp(kw, "if") == 0) { - retval = ifeval(&cp); - cp = skipcomment(cp); - if (*cp != '\n' || keepthis) - retval = LT_IF; - *cursym = 0; - } else if (strcmp(kw, "elif") == 0) { - retval = ifeval(&cp); cp = skipcomment(cp); - if (*cp != '\n' || keepthis) - retval = LT_ELIF; - if (retval == LT_IF) - retval = LT_ELIF; - if (retval == LT_TRUE) - retval = LT_ELTRUE; - if (retval == LT_FALSE) - retval = LT_ELFALSE; - *cursym = 0; - } else if (strcmp(kw, "else") == 0) - retval = LT_ELSE; - else if (strcmp(kw, "endif") == 0) - retval = LT_ENDIF; - -eol: - if (!text && reject != REJ_IGNORE) - for (; *cp;) { - if (incomment) - cp = skipcomment(cp); - else if (inquote == QUOTE_SINGLE) - cp = skipquote(cp, QUOTE_SINGLE); - else if (inquote == QUOTE_DOUBLE) - cp = skipquote(cp, QUOTE_DOUBLE); - else if (*cp == '/' && (cp[1] == '*' || cp[1] == '/')) - cp = skipcomment(cp); - else if (*cp == '\'') - cp = skipquote(cp, QUOTE_SINGLE); - else if (*cp == '"') - cp = skipquote(cp, QUOTE_DOUBLE); - else - cp++; + if (*cp != '\0') { + linestate = LS_DIRTY; + if (retval == LT_TRUE || retval == LT_FALSE || + retval == LT_TRUEI || retval == LT_FALSEI) + retval = LT_IF; + if (retval == LT_ELTRUE || retval == LT_ELFALSE) + retval = LT_ELIF; } + if (retval != LT_PLAIN && (wascomment || incomment)) + Eioccc(); + if (linestate == LS_HASH) + abort(); /* bug */ + } + if (linestate == LS_DIRTY) { + while (*cp != '\0') + cp = skipcomment(cp + 1); + } + debug("parser %s comment %s line", + comment_name[incomment], linestate_name[linestate]); return retval; } /* - * Turn a #elif line into a #if. This function is used when we are - * processing a #if/#elif/#else/#endif sequence that starts off with a - * #if that we understand (and therefore it has been deleted) which is - * followed by a #elif that we don't understand and therefore must be - * kept. We turn it into a #if to keep the nesting correct. + * These are the operators that are supported by the expression evaluator. */ -void -elif2if(void) -{ - strncpy(keyword, "if ", 4); -} +static int op_lt(int a, int b) { return a < b; } +static int op_gt(int a, int b) { return a > b; } +static int op_le(int a, int b) { return a <= b; } +static int op_ge(int a, int b) { return a >= b; } +static int op_eq(int a, int b) { return a == b; } +static int op_ne(int a, int b) { return a != b; } +static int op_or(int a, int b) { return a || b; } +static int op_and(int a, int b) { return a && b; } /* - * Turn a #elif line into a #endif. This is used in the opposite - * situation to elif2if, i.e. a #if that we don't understand is - * followed by a #elif that we do; rather than deleting the #elif (as - * we would for a #if) we turn it into a #endif to keep the nesting - * correct. + * An evaluation function takes three arguments, as follows: (1) a pointer to + * an element of the precedence table which lists the operators at the current + * level of precedence; (2) a pointer to an integer which will receive the + * value of the expression; and (3) a pointer to a char* that points to the + * expression to be evaluated and that is updated to the end of the expression + * when evaluation is complete. The function returns LT_FALSE if the value of + * the expression is zero, LT_TRUE if it is non-zero, or LT_IF if the + * expression could not be evaluated. */ -void -elif2endif(void) -{ - strcpy(keyword, "endif\n"); -} +struct ops; + +typedef Linetype eval_fn(const struct ops *, int *, const char **); + +static eval_fn eval_table, eval_unary; + +/* + * The precedence table. Expressions involving binary operators are evaluated + * in a table-driven way by eval_table. When it evaluates a subexpression it + * calls the inner function with its first argument pointing to the next + * element of the table. Innermost expressions have special non-table-driven + * handling. + */ +static const struct ops { + eval_fn *inner; + struct op { + const char *str; + int (*fn)(int, int); + } op[5]; +} eval_ops[] = { + { eval_table, { { "||", op_or } } }, + { eval_table, { { "&&", op_and } } }, + { eval_table, { { "==", op_eq }, + { "!=", op_ne } } }, + { eval_unary, { { "<=", op_le }, + { ">=", op_ge }, + { "<", op_lt }, + { ">", op_gt } } } +}; /* * Function for evaluating the innermost parts of expressions, * viz. !expr (expr) defined(symbol) symbol number * We reset the keepthis flag when we find a non-constant subexpression. */ -Linetype -eval_unary(struct ops *ops, int *valp, const char **cpp) +static Linetype +eval_unary(const struct ops *ops, int *valp, const char **cpp) { const char *cp; char *ep; @@ -646,7 +607,7 @@ eval_unary(struct ops *ops, int *valp, const char **cpp) return LT_IF; cp = skipcomment(cp); sym = findsym(cp); - if (sym == 0 && !symlist) + if (sym < 0 && !symlist) return LT_IF; *valp = (value[sym] != NULL); cp = skipsym(cp); @@ -657,7 +618,7 @@ eval_unary(struct ops *ops, int *valp, const char **cpp) } else if (!endsym(*cp)) { debug("eval%d symbol", ops - eval_ops); sym = findsym(cp); - if (sym == 0 && !symlist) + if (sym < 0 && !symlist) return LT_IF; if (value[sym] == NULL) *valp = 0; @@ -679,11 +640,11 @@ eval_unary(struct ops *ops, int *valp, const char **cpp) /* * Table-driven evaluation of binary operators. */ -Linetype -eval_table(struct ops *ops, int *valp, const char **cpp) +static Linetype +eval_table(const struct ops *ops, int *valp, const char **cpp) { + const struct op *op; const char *cp; - struct op *op; int val; debug("eval%d", ops - eval_ops); @@ -712,104 +673,103 @@ eval_table(struct ops *ops, int *valp, const char **cpp) /* * Evaluate the expression on a #if or #elif line. If we can work out * the result we return LT_TRUE or LT_FALSE accordingly, otherwise we - * return just a generic LT_IF. If the expression is constant and - * we are not processing constant #ifs then the keepthis flag is true. + * return just a generic LT_IF. */ -Linetype +static Linetype ifeval(const char **cpp) { + int ret; int val; + debug("eval %s", *cpp); keepthis = killconsts ? false : true; - return eval_table(eval_ops, &val, cpp); + ret = eval_table(eval_ops, &val, cpp); + return keepthis ? LT_IF : ret; } /* * Skip over comments and stop at the next character position that is - * not whitespace. + * not whitespace. Between calls we keep the comment state in a global + * variable, and we also make a note when we get a proper end-of-line. + * XXX: doesn't cope with the buffer splitting inside a state transition. */ -const char * +static const char * skipcomment(const char *cp) { - if (incomment) - goto inside; - for (;; cp++) { - while (*cp == ' ' || *cp == '\t') - cp++; - if (text) - return cp; - if (cp[0] != '/') - return cp; - - if (cp[1] == '*') { - if (!incomment) { + if (text || ignoring[depth]) { + while (isspace((unsigned char)*cp)) + cp += 1; + return cp; + } + while (*cp != '\0') + if (strncmp(cp, "\\\n", 2) == 0) + cp += 2; + else switch (incomment) { + case NO_COMMENT: + if (strncmp(cp, "/\\\n", 3) == 0) { + incomment = STARTING_COMMENT; + cp += 3; + } else if (strncmp(cp, "/*", 2) == 0) { incomment = C_COMMENT; - stqcline = linenum; - } - } else if (cp[1] == '/') { - if (!incomment) { + cp += 2; + } else if (strncmp(cp, "//", 2) == 0) { incomment = CXX_COMMENT; - stqcline = linenum; + cp += 2; + } else if (strncmp(cp, "\n", 1) == 0) { + linestate = LS_START; + cp += 1; + } else if (strchr(" \t", *cp) != NULL) { + cp += 1; + } else + return cp; + continue; + case CXX_COMMENT: + if (strncmp(cp, "\n", 1) == 0) { + incomment = NO_COMMENT; + linestate = LS_START; } - } else - return cp; - - cp += 2; -inside: - if (incomment == C_COMMENT) { - for (;;) { - for (; *cp != '*'; cp++) - if (*cp == '\0') - return cp; - if (*++cp == '/') { - incomment = NO_COMMENT; - break; - } + cp += 1; + continue; + case C_COMMENT: + if (strncmp(cp, "*\\\n", 3) == 0) { + incomment = FINISHING_COMMENT; + cp += 3; + } else if (strncmp(cp, "*/", 2) == 0) { + incomment = NO_COMMENT; + cp += 2; + } else + cp += 1; + continue; + case STARTING_COMMENT: + if (*cp == '*') { + incomment = C_COMMENT; + cp += 1; + } else if (*cp == '/') { + incomment = CXX_COMMENT; + cp += 1; + } else { + incomment = NO_COMMENT; + linestate = LS_DIRTY; } + continue; + case FINISHING_COMMENT: + if (*cp == '/') { + incomment = NO_COMMENT; + cp += 1; + } else + incomment = C_COMMENT; + continue; + default: + /* bug */ + abort(); } - else if (incomment == CXX_COMMENT) { - for (; *cp != '\n'; cp++) - if (*cp == '\0') - return cp; - incomment = NO_COMMENT; - } - } -} - -/* - * Skip over a quoted string or character and stop at the next charaacter - * position that is not whitespace. - */ -const char * -skipquote(const char *cp, Quote_state type) -{ - char qchar; - - qchar = type == QUOTE_SINGLE ? '\'' : '"'; - - if (inquote == type) - goto inside; - for (;; cp++) { - if (*cp != qchar) - return cp; - cp++; - inquote = type; - stqcline = linenum; -inside: - for (;; cp++) { - if (*cp == qchar) - break; - if (*cp == '\0' || (*cp == '\\' && *++cp == '\0')) - return cp; - } - inquote = QUOTE_NONE; - } + return cp; } /* * Skip over an identifier. */ -const char * +static const char * skipsym(const char *cp) { while (!endsym(*cp)) @@ -819,53 +779,47 @@ skipsym(const char *cp) /* * Look for the symbol in the symbol table. If is is found, we return - * the symbol table index, else we return 0. + * the symbol table index, else we return -1. */ -int +static int findsym(const char *str) { const char *cp; - const char *symp; int symind; - if (symlist) { - for (cp = str; !endsym(*cp); cp++) - continue; + cp = skipsym(str); + if (cp == str) + return -1; + if (symlist) printf("%.*s\n", cp-str, str); - } - for (symind = 1; symind < nsyms; ++symind) { - for (cp = str, symp = symname[symind] - ; *cp && *symp && *cp == *symp - ; cp++, symp++ - ) - continue; - if (*symp == '\0' && endsym(*cp)) { + for (symind = 0; symind < nsyms; ++symind) { + if (strlcmp(symname[symind], str, cp-str) == 0) { debug("findsym %s %s", symname[symind], value[symind] ? value[symind] : ""); return symind; } } - return 0; + return -1; } /* * Add a symbol to the symbol table. */ -void +static void addsym(bool ignorethis, bool definethis, char *sym) { int symind; char *val; symind = findsym(sym); - if (symind == 0) { + if (symind < 0) { if (nsyms >= MAXSYMS) errx(2, "too many symbols"); symind = nsyms++; } symname[symind] = sym; ignore[symind] = ignorethis; - val = (char *)skipsym(sym); + val = sym + (skipsym(sym) - sym); if (definethis) { if (*val == '=') { value[symind] = val+1; @@ -882,94 +836,24 @@ addsym(bool ignorethis, bool definethis, char *sym) } /* - * Read a line from the input and expand tabs if requested and (if - * compiled in) treats form-feed as an end-of-line. + * Compare s with n characters of t. + * The same as strncmp() except that it checks that s[n] == '\0'. */ -int -getline(char *line, int maxline, FILE *inp, bool expandtabs) +static int +strlcmp(const char *s, const char *t, size_t n) { - int tmp; - int num; - int chr; -#ifdef FFSPECIAL - static bool havechar = false; /* have leftover char from last time */ - static char svchar; -#endif /* FFSPECIAL */ - - num = 0; -#ifdef FFSPECIAL - if (havechar) { - havechar = false; - chr = svchar; - goto ent; - } -#endif /* FFSPECIAL */ - while (num + 8 < maxline) { /* leave room for tab */ - chr = getc(inp); - if (isprint(chr)) { -#ifdef FFSPECIAL - ent: -#endif /* FFSPECIAL */ - *line++ = chr; - num++; - } else - switch (chr) { - case EOF: - return EOF; - - case '\t': - if (expandtabs) { - num += tmp = 8 - (num & 7); - do - *line++ = ' '; - while (--tmp); - break; - } - default: - *line++ = chr; - num++; - break; - - case '\n': - *line = '\n'; - num++; - goto end; - -#ifdef FFSPECIAL - case '\f': - if (++num == 1) - *line = '\f'; - else { - *line = '\n'; - havechar = true; - svchar = chr; - } - goto end; -#endif /* FFSPECIAL */ - } - } -end: - *++line = '\0'; - return num; + while (n-- && *t != '\0') + if (*s != *t) + return (unsigned char)*s - (unsigned char)*t; + else + ++s, ++t; + return (unsigned char)*s; } /* - * Write a line to the output or not, according to the current - * filtering state. + * Diagnostics. */ -void -flushline(bool keep) -{ - if (symlist) - return; - if ((keep && reject != REJ_YES) ^ complement) - fputs(tline, stdout); - else if (lnblank) - putc('\n', stdout); - return; -} - -void +static void debug(const char *msg, ...) { va_list ap; @@ -981,14 +865,12 @@ debug(const char *msg, ...) } } -void -error(int code, int depth) +static void +error(const char *msg) { - if (incomment || inquote) - errx(2, "error in %s line %d: %s (#if depth %d)", - filename, stqcline, errs[code], depth); + if (depth == 0) + errx(2, "%s: %d: %s", filename, linenum, msg); else - errx(2, "error in %s line %d: %s" - " (#if depth %d start line %d)", - filename, linenum, errs[code], depth, stifline); + errx(2, "%s: %d: %s (#if line %d depth %d)", + filename, linenum, msg, stifline[depth], depth); } -- cgit v1.2.3