-/* $Id: mdoc.c,v 1.113 2009/10/30 05:58:38 kristaps Exp $ */
+/* $Id: mdoc.c,v 1.133 2010/05/15 16:24:37 kristaps Exp $ */
/*
* Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
*
* 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 <sys/types.h>
#include <assert.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
+#include <time.h>
#include "libmdoc.h"
#include "libmandoc.h"
"prologue macro out of conventional order", /* EPROLOOO */
"prologue macro repeated", /* EPROLREP */
"invalid manual section", /* EBADMSEC */
- "invalid section", /* EBADSEC */
"invalid font mode", /* EFONT */
"invalid date syntax", /* EBADDATE */
"invalid number format", /* ENUMFMT */
"no description found for library", /* ELIB */
"bad child for parent context", /* EBADCHILD */
"list arguments preceding type", /* ENOTYPE */
+ "deprecated comment style", /* EBADCOMMENT */
};
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",
"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"
+ "%U"
};
const char *const __mdoc_argnames[MDOC_ARG_MAX] = {
const char * const *mdoc_macronames = __mdoc_macronames;
const char * const *mdoc_argnames = __mdoc_argnames;
+static void mdoc_node_free(struct mdoc_node *);
+static void mdoc_node_unlink(struct mdoc *,
+ struct mdoc_node *);
static void mdoc_free1(struct mdoc *);
static void mdoc_alloc1(struct mdoc *);
static struct mdoc_node *node_alloc(struct mdoc *, int, int,
- int, enum mdoc_type);
+ enum mdoct, enum mdoc_type);
static int node_append(struct mdoc *,
struct mdoc_node *);
-static int parsetext(struct mdoc *, int, char *);
-static int parsemacro(struct mdoc *, int, char *);
+static int mdoc_ptext(struct mdoc *, int, char *);
+static int mdoc_pmacro(struct mdoc *, int, char *);
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 *
{
if (mdoc->first)
- mdoc_node_freelist(mdoc->first);
+ mdoc_node_delete(mdoc, mdoc->first);
if (mdoc->meta.title)
free(mdoc->meta.title);
if (mdoc->meta.os)
free(mdoc->meta.arch);
if (mdoc->meta.vol)
free(mdoc->meta.vol);
+ if (mdoc->meta.msec)
+ free(mdoc->meta.msec);
}
/*
* Main parse routine. Parses a single line -- really just hands off to
- * the macro (parsemacro()) or text parser (parsetext()).
+ * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
*/
int
mdoc_parseln(struct mdoc *m, int ln, char *buf)
if (MDOC_HALT & m->flags)
return(0);
- return('.' == *buf ? parsemacro(m, ln, buf) :
- parsetext(m, ln, buf));
+ m->flags |= MDOC_NEWLINE;
+ return('.' == *buf ?
+ mdoc_pmacro(m, ln, buf) :
+ mdoc_ptext(m, ln, buf));
}
int
-mdoc_macro(struct mdoc *m, int tok,
+mdoc_macro(struct mdoc *m, enum mdoct tok,
int ln, int pp, int *pos, char *buf)
{
- /*
- * If we're in the prologue, deny "body" macros. Similarly, if
- * we're in the body, deny prologue calls.
- */
+ assert(tok < MDOC_MAX);
+
+ /* If we're in the body, deny prologue calls. */
+
if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
MDOC_PBODY & m->flags)
return(mdoc_perr(m, ln, pp, EPROLBODY));
+
+ /* If we're in the prologue, deny "body" macros. */
+
if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
- ! (MDOC_PBODY & m->flags))
- return(mdoc_perr(m, ln, pp, EBODYPROL));
+ ! (MDOC_PBODY & m->flags)) {
+ if ( ! mdoc_pwarn(m, ln, pp, EBODYPROL))
+ return(0);
+ if (NULL == m->meta.title)
+ m->meta.title = mandoc_strdup("unknown");
+ if (NULL == m->meta.vol)
+ m->meta.vol = mandoc_strdup("local");
+ if (NULL == m->meta.os)
+ m->meta.os = mandoc_strdup("local");
+ if (0 == m->meta.date)
+ m->meta.date = time(NULL);
+ m->flags |= MDOC_PBODY;
+ }
return((*mdoc_macros[tok].fp)(m, tok, ln, pp, pos, buf));
}
static struct mdoc_node *
-node_alloc(struct mdoc *m, int line,
- int pos, int tok, enum mdoc_type type)
+node_alloc(struct mdoc *m, int line, int pos,
+ enum mdoct tok, enum mdoc_type type)
{
struct mdoc_node *p;
p->line = line;
p->pos = pos;
p->tok = tok;
- if (MDOC_TEXT != (p->type = type))
- assert(p->tok >= 0);
-
+ p->type = type;
+ if (MDOC_NEWLINE & m->flags)
+ p->flags |= MDOC_LINE;
+ m->flags &= ~MDOC_NEWLINE;
return(p);
}
int
-mdoc_tail_alloc(struct mdoc *m, int line, int pos, int tok)
+mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
{
struct mdoc_node *p;
int
-mdoc_head_alloc(struct mdoc *m, int line, int pos, int tok)
+mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
{
struct mdoc_node *p;
int
-mdoc_body_alloc(struct mdoc *m, int line, int pos, int tok)
+mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
{
struct mdoc_node *p;
int
mdoc_block_alloc(struct mdoc *m, int line, int pos,
- int tok, struct mdoc_arg *args)
+ enum mdoct tok, struct mdoc_arg *args)
{
struct mdoc_node *p;
int
mdoc_elem_alloc(struct mdoc *m, int line, int pos,
- int tok, struct mdoc_arg *args)
+ enum mdoct tok, struct mdoc_arg *args)
{
struct mdoc_node *p;
}
-static int
-pstring(struct mdoc *m, int line, int pos, const char *p, size_t len)
+int
+mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
{
struct mdoc_node *n;
- size_t sv;
+ size_t sv, len;
+
+ len = strlen(p);
- n = node_alloc(m, line, pos, -1, MDOC_TEXT);
+ n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
n->string = mandoc_malloc(len + 1);
sv = strlcpy(n->string, p, len + 1);
if ( ! node_append(m, n))
return(0);
+
m->next = MDOC_NEXT_SIBLING;
return(1);
}
-int
-mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
-{
-
- return(pstring(m, line, pos, p, strlen(p)));
-}
-
-
void
mdoc_node_free(struct mdoc_node *p)
{
- if (p->parent)
- p->parent->nchild--;
if (p->string)
free(p->string);
if (p->args)
}
-void
-mdoc_node_freelist(struct mdoc_node *p)
+static void
+mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
{
- if (p->child)
- mdoc_node_freelist(p->child);
- if (p->next)
- mdoc_node_freelist(p->next);
+ /* Adjust siblings. */
+
+ if (n->prev)
+ n->prev->next = n->next;
+ if (n->next)
+ n->next->prev = n->prev;
+
+ /* Adjust parent. */
+
+ if (n->parent) {
+ n->parent->nchild--;
+ if (n->parent->child == n)
+ n->parent->child = n->prev ? n->prev : n->next;
+ }
+
+ /* Adjust parse point, if applicable. */
+
+ if (m && m->last == n) {
+ if (n->prev) {
+ m->last = n->prev;
+ m->next = MDOC_NEXT_SIBLING;
+ } else {
+ m->last = n->parent;
+ m->next = MDOC_NEXT_CHILD;
+ }
+ }
+
+ if (m && m->first == n)
+ m->first = NULL;
+}
+
+
+void
+mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
+{
+ while (p->child) {
+ assert(p->nchild);
+ mdoc_node_delete(m, p->child);
+ }
assert(0 == p->nchild);
+
+ mdoc_node_unlink(m, p);
mdoc_node_free(p);
}
* control character.
*/
static int
-parsetext(struct mdoc *m, int line, char *buf)
+mdoc_ptext(struct mdoc *m, int line, char *buf)
{
- int i, j;
+ int i;
+
+ /* Ignore bogus comments. */
+
+ if ('\\' == buf[0] && '.' == buf[1] && '\"' == buf[2])
+ return(mdoc_pwarn(m, line, 0, EBADCOMMENT));
+
+ /* No text before an initial macro. */
if (SEC_NONE == m->lastnamed)
return(mdoc_perr(m, line, 0, ETEXTPROL));
-
- /*
- * If in literal mode, then pass the buffer directly to the
- * back-end, as it should be preserved as a single term.
- */
+ /* Literal just gets pulled in as-is. */
+
if (MDOC_LITERAL & m->flags)
return(mdoc_word_alloc(m, line, 0, buf));
- /* Disallow blank/white-space lines in non-literal mode. */
+ /* Check for a blank line, which may also consist of spaces. */
for (i = 0; ' ' == buf[i]; i++)
- /* Skip leading whitespace. */ ;
- if (0 == buf[i])
- return(mdoc_perr(m, line, 0, ENOBLANK));
+ /* Skip to first non-space. */ ;
- /*
- * Break apart a free-form line into tokens. Spaces are
- * stripped out of the input.
+ if ('\0' == buf[i]) {
+ if ( ! mdoc_pwarn(m, line, 0, ENOBLANK))
+ return(0);
+
+ /*
+ * Insert a `Pp' in the case of a blank line. Technically,
+ * blank lines aren't allowed, but enough manuals assume this
+ * behaviour that we want to work around it.
+ */
+ if ( ! mdoc_elem_alloc(m, line, 0, MDOC_Pp, NULL))
+ return(0);
+
+ m->next = MDOC_NEXT_SIBLING;
+ return(1);
+ }
+
+ /*
+ * Warn if the last un-escaped character is whitespace. Then
+ * strip away the remaining spaces (tabs stay!).
*/
- for (j = i; buf[i]; i++) {
- if (' ' != buf[i])
- continue;
+ i = (int)strlen(buf);
+ assert(i);
- /* Escaped whitespace. */
- if (i && ' ' == buf[i] && '\\' == buf[i - 1])
- continue;
+ if (' ' == buf[i - 1] || '\t' == buf[i - 1]) {
+ if (i > 1 && '\\' != buf[i - 2])
+ if ( ! mdoc_pwarn(m, line, i - 1, ETAILWS))
+ return(0);
- buf[i++] = 0;
- if ( ! pstring(m, line, j, &buf[j], (size_t)(i - j)))
- return(0);
+ for (--i; i && ' ' == buf[i]; i--)
+ /* Spin back to non-space. */ ;
- for ( ; ' ' == buf[i]; i++)
- /* Skip trailing whitespace. */ ;
+ /* Jump ahead of escaped whitespace. */
+ i += '\\' == buf[i] ? 2 : 1;
- j = i;
- if (0 == buf[i])
- break;
+ buf[i] = '\0';
}
- if (j != i && ! pstring(m, line, j, &buf[j], (size_t)(i - j)))
+ /* Allocate the whole word. */
+
+ if ( ! mdoc_word_alloc(m, line, 0, buf))
return(0);
- m->next = MDOC_NEXT_SIBLING;
+ /*
+ * End-of-sentence check. If the last character is an unescaped
+ * EOS character, then flag the node as being the end of a
+ * sentence. The front-end will know how to interpret this.
+ */
+
+ /* FIXME: chain of close delims. */
+
+ assert(i);
+
+ if (mandoc_eos(buf, (size_t)i))
+ m->last->flags |= MDOC_EOS;
+
return(1);
}
-
static int
macrowarn(struct mdoc *m, int ln, const char *buf)
{
if ( ! (MDOC_IGN_MACRO & m->pflags))
- return(mdoc_verr(m, ln, 0,
- "unknown macro: %s%s",
+ return(mdoc_verr(m, ln, 0, "unknown macro: %s%s",
buf, strlen(buf) > 3 ? "..." : ""));
return(mdoc_vwarn(m, ln, 0, "unknown macro: %s%s",
buf, strlen(buf) > 3 ? "..." : ""));
* character.
*/
int
-parsemacro(struct mdoc *m, int ln, char *buf)
+mdoc_pmacro(struct mdoc *m, int ln, char *buf)
{
- int i, j, c;
- char mac[5];
+ enum mdoct tok;
+ int i, j, sv;
+ char mac[5];
/* Empty lines are ignored. */
- if (0 == buf[1])
+ if ('\0' == buf[1])
return(1);
i = 1;
i++;
while (buf[i] && ' ' == buf[i])
i++;
- if (0 == buf[i])
+ if ('\0' == buf[i])
return(1);
}
+ sv = i;
+
/* Copy the first word into a nil-terminated buffer. */
for (j = 0; j < 4; j++, i++) {
- if (0 == (mac[j] = buf[i]))
+ if ('\0' == (mac[j] = buf[i]))
break;
else if (' ' == buf[i])
break;
return(1);
}
- if (MDOC_MAX == (c = mdoc_hash_find(mac))) {
+ if (MDOC_MAX == (tok = mdoc_hash_find(mac))) {
if ( ! macrowarn(m, ln, mac))
goto err;
return(1);
while (buf[i] && ' ' == buf[i])
i++;
+ /*
+ * Trailing whitespace. Note that tabs are allowed to be passed
+ * into the parser as "text", so we only warn about spaces here.
+ */
+
+ 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))
+ if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
goto err;
return(1);
git.ckatri.com / mandoc.git / blobdiff