=================================================================== RCS file: /cvs/mandoc/mdoc.c,v retrieving revision 1.97 retrieving revision 1.116 diff -u -p -r1.97 -r1.116 --- mandoc/mdoc.c 2009/07/29 08:46:06 1.97 +++ mandoc/mdoc.c 2010/01/07 10:24:43 1.116 @@ -1,4 +1,4 @@ -/* $Id: mdoc.c,v 1.97 2009/07/29 08:46:06 kristaps Exp $ */ +/* $Id: mdoc.c,v 1.116 2010/01/07 10:24:43 kristaps Exp $ */ /* * Copyright (c) 2008, 2009 Kristaps Dzonsons * @@ -14,6 +14,12 @@ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ +#ifdef HAVE_CONFIG_H +#include "config.h" +#endif + +#include + #include #include #include @@ -22,12 +28,12 @@ #include #include "libmdoc.h" +#include "libmandoc.h" const char *const __mdoc_merrnames[MERRMAX] = { "trailing whitespace", /* ETAILWS */ "unexpected quoted parameter", /* EQUOTPARM */ "unterminated quoted parameter", /* EQUOTTERM */ - "system: malloc error", /* EMALLOC */ "argument parameter suggested", /* EARGVAL */ "macro disallowed in prologue", /* EBODYPROL */ "macro disallowed in body", /* EPROLBODY */ @@ -74,8 +80,9 @@ const char *const __mdoc_merrnames[MERRMAX] = { "unclosed explicit scope", /* EOPEN */ "unterminated quoted phrase", /* EQUOTPHR */ "closure macro without prior context", /* ENOCTX */ - "invalid whitespace after control character", /* ESPACE */ - "no description found for library" /* ELIB */ + "no description found for library", /* ELIB */ + "bad child for parent context", /* EBADCHILD */ + "list arguments preceding type", /* ENOTYPE */ }; const char *const __mdoc_macronames[MDOC_MAX] = { @@ -90,11 +97,11 @@ const char *const __mdoc_macronames[MDOC_MAX] = { "Nm", "Op", "Ot", "Pa", "Rv", "St", "Va", "Vt", /* LINTED */ - "Xr", "\%A", "\%B", "\%D", + "Xr", "%A", "%B", "%D", /* LINTED */ - "\%I", "\%J", "\%N", "\%O", + "%I", "%J", "%N", "%O", /* LINTED */ - "\%P", "\%R", "\%T", "\%V", + "%P", "%R", "%T", "%V", "Ac", "Ao", "Aq", "At", "Bc", "Bf", "Bo", "Bq", "Bsx", "Bx", "Db", "Dc", @@ -111,9 +118,11 @@ const char *const __mdoc_macronames[MDOC_MAX] = { "Fr", "Ud", "Lb", "Lp", "Lk", "Mt", "Brq", "Bro", /* LINTED */ - "Brc", "\%C", "Es", "En", + "Brc", "%C", "Es", "En", /* LINTED */ - "Dx", "\%Q", "br", "sp" + "Dx", "%Q", "br", "sp", + /* LINTED */ + "%U" }; const char *const __mdoc_argnames[MDOC_ARG_MAX] = { @@ -125,14 +134,14 @@ const char *const __mdoc_argnames[MDOC_ARG_MAX] = { "ohang", "inset", "column", "width", "compact", "std", "filled", "words", "emphasis", - "symbolic", "nested" + "symbolic", "nested", "centered" }; const char * const *mdoc_macronames = __mdoc_macronames; const char * const *mdoc_argnames = __mdoc_argnames; static void mdoc_free1(struct mdoc *); -static int mdoc_alloc1(struct mdoc *); +static void mdoc_alloc1(struct mdoc *); static struct mdoc_node *node_alloc(struct mdoc *, int, int, int, enum mdoc_type); static int node_append(struct mdoc *, @@ -143,11 +152,6 @@ static int macrowarn(struct mdoc *, int, const char static int pstring(struct mdoc *, int, int, const char *, size_t); -#ifdef __linux__ -extern size_t strlcpy(char *, const char *, size_t); -#endif - - const struct mdoc_node * mdoc_node(const struct mdoc *m) { @@ -189,21 +193,17 @@ mdoc_free1(struct mdoc *mdoc) /* * Allocate all volatile resources (parse tree, meta-data, fields). */ -static int +static void mdoc_alloc1(struct mdoc *mdoc) { - bzero(&mdoc->meta, sizeof(struct mdoc_meta)); + memset(&mdoc->meta, 0, sizeof(struct mdoc_meta)); mdoc->flags = 0; mdoc->lastnamed = mdoc->lastsec = SEC_NONE; - mdoc->last = calloc(1, sizeof(struct mdoc_node)); - if (NULL == mdoc->last) - return(0); - + mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node)); mdoc->first = mdoc->last; mdoc->last->type = MDOC_ROOT; mdoc->next = MDOC_NEXT_CHILD; - return(1); } @@ -213,12 +213,12 @@ mdoc_alloc1(struct mdoc *mdoc) * and the parser is ready for re-invocation on a new tree; however, * cross-parse non-volatile data is kept intact. */ -int +void mdoc_reset(struct mdoc *mdoc) { mdoc_free1(mdoc); - return(mdoc_alloc1(mdoc)); + mdoc_alloc1(mdoc); } @@ -231,8 +231,6 @@ mdoc_free(struct mdoc *mdoc) { mdoc_free1(mdoc); - if (mdoc->htab) - mdoc_hash_free(mdoc->htab); free(mdoc); } @@ -245,22 +243,17 @@ mdoc_alloc(void *data, int pflags, const struct mdoc_c { struct mdoc *p; - if (NULL == (p = calloc(1, sizeof(struct mdoc)))) - return(NULL); + p = mandoc_calloc(1, sizeof(struct mdoc)); + if (cb) - (void)memcpy(&p->cb, cb, sizeof(struct mdoc_cb)); + memcpy(&p->cb, cb, sizeof(struct mdoc_cb)); p->data = data; p->pflags = pflags; - if (NULL == (p->htab = mdoc_hash_alloc())) { - free(p); - return(NULL); - } else if (mdoc_alloc1(p)) - return(p); - - free(p); - return(NULL); + mdoc_hash_init(); + mdoc_alloc1(p); + return(p); } @@ -436,11 +429,7 @@ node_alloc(struct mdoc *m, int line, { struct mdoc_node *p; - if (NULL == (p = calloc(1, sizeof(struct mdoc_node)))) { - (void)mdoc_nerr(m, m->last, EMALLOC); - return(NULL); - } - + p = mandoc_calloc(1, sizeof(struct mdoc_node)); p->sec = m->lastsec; p->line = line; p->pos = pos; @@ -458,9 +447,10 @@ mdoc_tail_alloc(struct mdoc *m, int line, int pos, int struct mdoc_node *p; p = node_alloc(m, line, pos, tok, MDOC_TAIL); - if (NULL == p) + if ( ! node_append(m, p)) return(0); - return(node_append(m, p)); + m->next = MDOC_NEXT_CHILD; + return(1); } @@ -473,9 +463,10 @@ mdoc_head_alloc(struct mdoc *m, int line, int pos, int assert(m->last); p = node_alloc(m, line, pos, tok, MDOC_HEAD); - if (NULL == p) + if ( ! node_append(m, p)) return(0); - return(node_append(m, p)); + m->next = MDOC_NEXT_CHILD; + return(1); } @@ -485,9 +476,10 @@ mdoc_body_alloc(struct mdoc *m, int line, int pos, int struct mdoc_node *p; p = node_alloc(m, line, pos, tok, MDOC_BODY); - if (NULL == p) + if ( ! node_append(m, p)) return(0); - return(node_append(m, p)); + m->next = MDOC_NEXT_CHILD; + return(1); } @@ -498,12 +490,13 @@ mdoc_block_alloc(struct mdoc *m, int line, int pos, struct mdoc_node *p; p = node_alloc(m, line, pos, tok, MDOC_BLOCK); - if (NULL == p) - return(0); p->args = args; if (p->args) (args->refcnt)++; - return(node_append(m, p)); + if ( ! node_append(m, p)) + return(0); + m->next = MDOC_NEXT_CHILD; + return(1); } @@ -514,12 +507,13 @@ mdoc_elem_alloc(struct mdoc *m, int line, int pos, struct mdoc_node *p; p = node_alloc(m, line, pos, tok, MDOC_ELEM); - if (NULL == p) - return(0); p->args = args; if (p->args) (args->refcnt)++; - return(node_append(m, p)); + if ( ! node_append(m, p)) + return(0); + m->next = MDOC_NEXT_CHILD; + return(1); } @@ -530,21 +524,16 @@ pstring(struct mdoc *m, int line, int pos, const char size_t sv; n = node_alloc(m, line, pos, -1, MDOC_TEXT); - if (NULL == n) - return(mdoc_nerr(m, m->last, EMALLOC)); - - n->string = malloc(len + 1); - if (NULL == n->string) { - free(n); - return(mdoc_nerr(m, m->last, EMALLOC)); - } - + n->string = mandoc_malloc(len + 1); sv = strlcpy(n->string, p, len + 1); /* Prohibit truncation. */ assert(sv < len + 1); - return(node_append(m, n)); + if ( ! node_append(m, n)) + return(0); + m->next = MDOC_NEXT_SIBLING; + return(1); } @@ -592,6 +581,7 @@ static int parsetext(struct mdoc *m, int line, char *buf) { int i, j; + char sv; if (SEC_NONE == m->lastnamed) return(mdoc_perr(m, line, 0, ETEXTPROL)); @@ -601,18 +591,15 @@ parsetext(struct mdoc *m, int line, char *buf) * back-end, as it should be preserved as a single term. */ - if (MDOC_LITERAL & m->flags) { - if ( ! mdoc_word_alloc(m, line, 0, buf)) - return(0); - m->next = MDOC_NEXT_SIBLING; - return(1); - } + if (MDOC_LITERAL & m->flags) + return(mdoc_word_alloc(m, line, 0, buf)); /* Disallow blank/white-space lines in non-literal mode. */ for (i = 0; ' ' == buf[i]; i++) /* Skip leading whitespace. */ ; - if (0 == buf[i]) + + if ('\0' == buf[i]) return(mdoc_perr(m, line, 0, ENOBLANK)); /* @@ -628,16 +615,30 @@ parsetext(struct mdoc *m, int line, char *buf) if (i && ' ' == buf[i] && '\\' == buf[i - 1]) continue; - buf[i++] = 0; + sv = buf[i]; + buf[i++] = '\0'; + if ( ! pstring(m, line, j, &buf[j], (size_t)(i - j))) return(0); - m->next = MDOC_NEXT_SIBLING; + /* Trailing whitespace? Check at overwritten byte. */ + + if (' ' == sv && '\0' == buf[i]) + if ( ! mdoc_pwarn(m, line, i - 1, ETAILWS)) + return(0); + for ( ; ' ' == buf[i]; i++) /* Skip trailing whitespace. */ ; j = i; - if (0 == buf[i]) + + /* Trailing whitespace? */ + + if (' ' == buf[i - 1] && '\0' == buf[i]) + if ( ! mdoc_pwarn(m, line, i - 1, ETAILWS)) + return(0); + + if ('\0' == buf[i]) break; } @@ -650,7 +651,6 @@ parsetext(struct mdoc *m, int line, char *buf) - static int macrowarn(struct mdoc *m, int ln, const char *buf) { @@ -670,41 +670,50 @@ macrowarn(struct mdoc *m, int ln, const char *buf) int parsemacro(struct mdoc *m, int ln, char *buf) { - int i, c; + int i, j, c; char mac[5]; /* Empty lines are ignored. */ - if (0 == buf[1]) + if ('\0' == buf[1]) return(1); - if (' ' == buf[1]) { - i = 2; + i = 1; + + /* Accept whitespace after the initial control char. */ + + if (' ' == buf[i]) { + i++; while (buf[i] && ' ' == buf[i]) i++; - if (0 == buf[i]) + if ('\0' == buf[i]) return(1); - return(mdoc_perr(m, ln, 1, ESPACE)); } /* Copy the first word into a nil-terminated buffer. */ - for (i = 1; i < 5; i++) { - if (0 == (mac[i - 1] = buf[i])) + for (j = 0; j < 4; j++, i++) { + if ('\0' == (mac[j] = buf[i])) break; else if (' ' == buf[i]) break; + + /* Check for invalid characters. */ + + if (isgraph((u_char)buf[i])) + continue; + return(mdoc_perr(m, ln, i, EPRINT)); } - mac[i - 1] = 0; + mac[j] = 0; - if (i == 5 || i <= 2) { + if (j == 4 || j < 2) { if ( ! macrowarn(m, ln, mac)) goto err; return(1); } - if (MDOC_MAX == (c = mdoc_hash_find(m->htab, mac))) { + if (MDOC_MAX == (c = mdoc_hash_find(mac))) { if ( ! macrowarn(m, ln, mac)) goto err; return(1); @@ -715,8 +724,16 @@ parsemacro(struct mdoc *m, int ln, char *buf) while (buf[i] && ' ' == buf[i]) i++; - /* Begin recursive parse sequence. */ + /* Trailing whitespace? */ + if ('\0' == buf[i] && ' ' == buf[i - 1]) + if ( ! mdoc_pwarn(m, ln, i - 1, ETAILWS)) + goto err; + + /* + * Begin recursive parse sequence. Since we're at the start of + * the line, we don't need to do callable/parseable checks. + */ if ( ! mdoc_macro(m, c, ln, 1, &i, buf)) goto err; @@ -727,3 +744,5 @@ err: /* Error out. */ m->flags |= MDOC_HALT; return(0); } + +