=================================================================== RCS file: /cvs/mandoc/read.c,v retrieving revision 1.14 retrieving revision 1.29 diff -u -p -r1.14 -r1.29 --- mandoc/read.c 2011/04/30 10:18:24 1.14 +++ mandoc/read.c 2012/05/27 17:48:57 1.29 @@ -1,4 +1,4 @@ -/* $Id: read.c,v 1.14 2011/04/30 10:18:24 kristaps Exp $ */ +/* $Id: read.c,v 1.29 2012/05/27 17:48:57 schwarze Exp $ */ /* * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons * Copyright (c) 2010, 2011 Ingo Schwarze @@ -19,13 +19,16 @@ #include "config.h" #endif -#include -#include +#ifdef HAVE_MMAP +# include +# include +#endif #include #include #include #include +#include #include #include #include @@ -35,6 +38,7 @@ #include "libmandoc.h" #include "mdoc.h" #include "man.h" +#include "main.h" #ifndef MAP_FILE #define MAP_FILE 0 @@ -57,18 +61,18 @@ struct mparse { struct man *man; /* man parser */ struct mdoc *mdoc; /* mdoc parser */ struct roff *roff; /* roff parser (!NULL) */ - struct regset regs; /* roff registers */ int reparse_count; /* finite interp. stack */ mandocmsg mmsg; /* warning/error message handler */ void *arg; /* argument to mmsg */ const char *file; + struct buf *secondary; + char *defos; /* default operating system */ }; static void resize_buf(struct buf *, size_t); static void mparse_buf_r(struct mparse *, struct buf, int); static void mparse_readfd_r(struct mparse *, int, const char *, int); static void pset(const char *, int, struct mparse *); -static void pdesc(struct mparse *, const char *, int); static int read_whole_file(const char *, int, struct buf *, int *); static void mparse_end(struct mparse *); @@ -144,9 +148,19 @@ static const char * const mandocerrs[MANDOCERR_MAX] = "bad comment style", "bad escape sequence", "unterminated quoted string", + + /* related to equations */ + "unexpected literal in equation", "generic error", + /* related to equations */ + "unexpected equation scope closure", + "equation scope open on exit", + "overlapping equation scopes", + "unexpected end of equation", + "equation syntax error", + /* related to tables */ "bad table syntax", "bad table option", @@ -180,7 +194,6 @@ static const char * const mandocerrs[MANDOCERR_MAX] = "not a manual", "column syntax is inconsistent", "NOT IMPLEMENTED: .Bd -file", - "line scope broken, syntax violated", "argument count wrong, violates syntax", "child violates parent syntax", "argument count wrong, violates syntax", @@ -235,13 +248,14 @@ pset(const char *buf, int pos, struct mparse *curp) switch (curp->inttype) { case (MPARSE_MDOC): if (NULL == curp->pmdoc) - curp->pmdoc = mdoc_alloc(&curp->regs, curp); + curp->pmdoc = mdoc_alloc(curp->roff, curp, + curp->defos); assert(curp->pmdoc); curp->mdoc = curp->pmdoc; return; case (MPARSE_MAN): if (NULL == curp->pman) - curp->pman = man_alloc(&curp->regs, curp); + curp->pman = man_alloc(curp->roff, curp); assert(curp->pman); curp->man = curp->pman; return; @@ -251,14 +265,15 @@ pset(const char *buf, int pos, struct mparse *curp) if (pos >= 3 && 0 == memcmp(buf, ".Dd", 3)) { if (NULL == curp->pmdoc) - curp->pmdoc = mdoc_alloc(&curp->regs, curp); + curp->pmdoc = mdoc_alloc(curp->roff, curp, + curp->defos); assert(curp->pmdoc); curp->mdoc = curp->pmdoc; return; } if (NULL == curp->pman) - curp->pman = man_alloc(&curp->regs, curp); + curp->pman = man_alloc(curp->roff, curp); assert(curp->pman); curp->man = curp->pman; } @@ -314,9 +329,9 @@ mparse_buf_r(struct mparse *curp, struct buf blk, int * Warn about bogus characters. If you're using * non-ASCII encoding, you're screwing your * readers. Since I'd rather this not happen, - * I'll be helpful and drop these characters so - * we don't display gibberish. Note to manual - * writers: use special characters. + * I'll be helpful and replace these characters + * with "?", so we don't display gibberish. + * Note to manual writers: use special characters. */ c = (unsigned char) blk.buf[i]; @@ -324,8 +339,11 @@ mparse_buf_r(struct mparse *curp, struct buf blk, int if ( ! (isascii(c) && (isgraph(c) || isblank(c)))) { mandoc_msg(MANDOCERR_BADCHAR, curp, - curp->line, pos, "ignoring byte"); + curp->line, pos, NULL); i++; + if (pos >= (int)ln.sz) + resize_buf(&ln, 256); + ln.buf[pos++] = '?'; continue; } @@ -400,6 +418,27 @@ mparse_buf_r(struct mparse *curp, struct buf blk, int of = 0; + /* + * Maintain a lookaside buffer of all parsed lines. We + * only do this if mparse_keep() has been invoked (the + * buffer may be accessed with mparse_getkeep()). + */ + + if (curp->secondary) { + curp->secondary->buf = + mandoc_realloc + (curp->secondary->buf, + curp->secondary->sz + pos + 2); + memcpy(curp->secondary->buf + + curp->secondary->sz, + ln.buf, pos); + curp->secondary->sz += pos; + curp->secondary->buf + [curp->secondary->sz] = '\n'; + curp->secondary->sz++; + curp->secondary->buf + [curp->secondary->sz] = '\0'; + } rerun: rr = roff_parseln (curp->roff, curp->line, @@ -426,6 +465,13 @@ rerun: assert(MANDOCLEVEL_FATAL <= curp->file_status); break; case (ROFF_SO): + /* + * We remove `so' clauses from our lookaside + * buffer because we're going to descend into + * the file recursively. + */ + if (curp->secondary) + curp->secondary->sz -= pos + 1; mparse_readfd_r(curp, -1, ln.buf + of, 1); if (MANDOCLEVEL_FATAL <= curp->file_status) break; @@ -505,43 +551,14 @@ rerun: free(ln.buf); } -static void -pdesc(struct mparse *curp, const char *file, int fd) -{ - struct buf blk; - int with_mmap; - - /* - * Run for each opened file; may be called more than once for - * each full parse sequence if the opened file is nested (i.e., - * from `so'). Simply sucks in the whole file and moves into - * the parse phase for the file. - */ - - if ( ! read_whole_file(file, fd, &blk, &with_mmap)) { - curp->file_status = MANDOCLEVEL_SYSERR; - return; - } - - /* Line number is per-file. */ - - curp->line = 1; - - mparse_buf_r(curp, blk, 1); - - if (with_mmap) - munmap(blk.buf, blk.sz); - else - free(blk.buf); -} - static int read_whole_file(const char *file, int fd, struct buf *fb, int *with_mmap) { - struct stat st; size_t off; ssize_t ssz; +#ifdef HAVE_MMAP + struct stat st; if (-1 == fstat(fd, &st)) { perror(file); return(0); @@ -566,6 +583,7 @@ read_whole_file(const char *file, int fd, struct buf * if (fb->buf != MAP_FAILED) return(1); } +#endif /* * If this isn't a regular file (like, say, stdin), then we must @@ -628,29 +646,72 @@ mparse_end(struct mparse *curp) } static void -mparse_readfd_r(struct mparse *curp, int fd, const char *file, int re) +mparse_parse_buffer(struct mparse *curp, struct buf blk, const char *file, + int re) { const char *svfile; + /* Line number is per-file. */ + svfile = curp->file; + curp->file = file; + curp->line = 1; + + mparse_buf_r(curp, blk, 1); + + if (0 == re && MANDOCLEVEL_FATAL > curp->file_status) + mparse_end(curp); + + curp->file = svfile; +} + +enum mandoclevel +mparse_readmem(struct mparse *curp, const void *buf, size_t len, + const char *file) +{ + struct buf blk; + + blk.buf = UNCONST(buf); + blk.sz = len; + + mparse_parse_buffer(curp, blk, file, 0); + return(curp->file_status); +} + +static void +mparse_readfd_r(struct mparse *curp, int fd, const char *file, int re) +{ + struct buf blk; + int with_mmap; + if (-1 == fd) if (-1 == (fd = open(file, O_RDONLY, 0))) { perror(file); curp->file_status = MANDOCLEVEL_SYSERR; return; } + /* + * Run for each opened file; may be called more than once for + * each full parse sequence if the opened file is nested (i.e., + * from `so'). Simply sucks in the whole file and moves into + * the parse phase for the file. + */ - svfile = curp->file; - curp->file = file; + if ( ! read_whole_file(file, fd, &blk, &with_mmap)) { + curp->file_status = MANDOCLEVEL_SYSERR; + return; + } - pdesc(curp, file, fd); + mparse_parse_buffer(curp, blk, file, re); - if (0 == re && MANDOCLEVEL_FATAL > curp->file_status) - mparse_end(curp); +#ifdef HAVE_MMAP + if (with_mmap) + munmap(blk.buf, blk.sz); + else +#endif + free(blk.buf); if (STDIN_FILENO != fd && -1 == close(fd)) perror(file); - - curp->file = svfile; } enum mandoclevel @@ -662,7 +723,8 @@ mparse_readfd(struct mparse *curp, int fd, const char } struct mparse * -mparse_alloc(enum mparset inttype, enum mandoclevel wlevel, mandocmsg mmsg, void *arg) +mparse_alloc(enum mparset inttype, enum mandoclevel wlevel, + mandocmsg mmsg, void *arg, char *defos) { struct mparse *curp; @@ -674,8 +736,9 @@ mparse_alloc(enum mparset inttype, enum mandoclevel wl curp->mmsg = mmsg; curp->arg = arg; curp->inttype = inttype; + curp->defos = defos; - curp->roff = roff_alloc(&curp->regs, curp); + curp->roff = roff_alloc(curp); return(curp); } @@ -683,14 +746,14 @@ void mparse_reset(struct mparse *curp) { - memset(&curp->regs, 0, sizeof(struct regset)); - roff_reset(curp->roff); if (curp->mdoc) mdoc_reset(curp->mdoc); if (curp->man) man_reset(curp->man); + if (curp->secondary) + curp->secondary->sz = 0; curp->file_status = MANDOCLEVEL_OK; curp->mdoc = NULL; @@ -707,7 +770,10 @@ mparse_free(struct mparse *curp) man_free(curp->pman); if (curp->roff) roff_free(curp->roff); + if (curp->secondary) + free(curp->secondary->buf); + free(curp->secondary); free(curp); } @@ -766,4 +832,20 @@ const char * mparse_strlevel(enum mandoclevel lvl) { return(mandoclevels[lvl]); +} + +void +mparse_keep(struct mparse *p) +{ + + assert(NULL == p->secondary); + p->secondary = mandoc_calloc(1, sizeof(struct buf)); +} + +const char * +mparse_getkeep(const struct mparse *p) +{ + + assert(p->secondary); + return(p->secondary->sz ? p->secondary->buf : NULL); }