-/* $Id: mdoc.c,v 1.69 2009/03/21 09:42:07 kristaps Exp $ */
+/* $Id: mdoc.c,v 1.162 2010/08/08 14:51:32 schwarze Exp $ */
/*
- * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@openbsd.org>
+ * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
+ * Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
*
* Permission to use, copy, modify, and distribute this software for any
- * purpose with or without fee is hereby granted, provided that the
- * above copyright notice and this permission notice appear in all
- * copies.
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
*
- * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL
- * WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED
- * WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE
- * AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
- * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
- * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
- * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
- * PERFORMANCE OF THIS SOFTWARE.
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * 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 <ctype.h>
-#include <err.h>
#include <stdarg.h>
-#include <stdlib.h>
#include <stdio.h>
+#include <stdlib.h>
#include <string.h>
+#include <time.h>
-#include "private.h"
-
-/*
- * Main caller in the libmdoc library. This begins the parsing routine,
- * handles allocation of data, and so forth. Most of the "work" is done
- * in macro.c, validate.c and action.c.
- */
-
-static struct mdoc_node *mdoc_node_alloc(const struct mdoc *);
-static int mdoc_node_append(struct mdoc *,
- struct mdoc_node *);
-
-static int parsetext(struct mdoc *, int, char *);
-static int parsemacro(struct mdoc *, int, char *);
-static int macrowarn(struct mdoc *, int, const char *);
-
+#include "mandoc.h"
+#include "libmdoc.h"
+#include "libmandoc.h"
const char *const __mdoc_macronames[MDOC_MAX] = {
- "\\\"", "Dd", "Dt", "Os",
+ "Ap", "Dd", "Dt", "Os",
"Sh", "Ss", "Pp", "D1",
"Dl", "Bd", "Ed", "Bl",
"El", "It", "Ad", "An",
"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",
"Tn", "Ux", "Xc", "Xo",
"Fo", "Fc", "Oo", "Oc",
"Bk", "Ek", "Bt", "Hf",
- "Fr", "Ud", "Lb", "Ap",
- "Lp", "Lk", "Mt", "Brq",
+ "Fr", "Ud", "Lb", "Lp",
+ "Lk", "Mt", "Brq", "Bro",
+ /* LINTED */
+ "Brc", "%C", "Es", "En",
/* LINTED */
- "Bro", "Brc", "\%C", "Es",
+ "Dx", "%Q", "br", "sp",
/* LINTED */
- "En", "Dx", "\%Q"
+ "%U", "Ta"
};
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_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,
+ enum mdoct, enum mdoc_type);
+static int node_append(struct mdoc *,
+ struct mdoc_node *);
+static int mdoc_ptext(struct mdoc *, int, char *, int);
+static int mdoc_pmacro(struct mdoc *, int, char *, int);
+static int macrowarn(struct mdoc *, int,
+ const char *, int);
+
-/*
- * Get the first (root) node of the parse tree.
- */
const struct mdoc_node *
-mdoc_node(const struct mdoc *mdoc)
+mdoc_node(const struct mdoc *m)
{
- if (MDOC_HALT & mdoc->flags)
- return(NULL);
- if (mdoc->first)
- assert(MDOC_ROOT == mdoc->first->type);
- return(mdoc->first);
+ return(MDOC_HALT & m->flags ? NULL : m->first);
}
const struct mdoc_meta *
-mdoc_meta(const struct mdoc *mdoc)
+mdoc_meta(const struct mdoc *m)
{
- if (MDOC_HALT & mdoc->flags)
- return(NULL);
- return(&mdoc->meta);
+ return(MDOC_HALT & m->flags ? NULL : &m->meta);
}
/*
- * Free up all resources contributed by a parse: the node tree, meta-data and
- * so on. Then reallocate the root node for another parse.
+ * Frees volatile resources (parse tree, meta-data, fields).
*/
-void
-mdoc_reset(struct mdoc *mdoc)
+static void
+mdoc_free1(struct mdoc *mdoc)
{
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);
+}
- bzero(&mdoc->meta, sizeof(struct mdoc_meta));
- mdoc->flags = 0;
- mdoc->lastnamed = mdoc->lastsec = 0;
- mdoc->first = mdoc->last =
- xcalloc(1, sizeof(struct mdoc_node));
+/*
+ * Allocate all volatile resources (parse tree, meta-data, fields).
+ */
+static void
+mdoc_alloc1(struct mdoc *mdoc)
+{
+
+ memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
+ mdoc->flags = 0;
+ mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
+ mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
+ mdoc->first = mdoc->last;
mdoc->last->type = MDOC_ROOT;
mdoc->next = MDOC_NEXT_CHILD;
}
/*
- * Completely free up all resources.
+ * Free up volatile resources (see mdoc_free1()) then re-initialises the
+ * data with mdoc_alloc1(). After invocation, parse data has been reset
+ * and the parser is ready for re-invocation on a new tree; however,
+ * cross-parse non-volatile data is kept intact.
*/
void
-mdoc_free(struct mdoc *mdoc)
+mdoc_reset(struct mdoc *mdoc)
{
- if (mdoc->first)
- mdoc_node_freelist(mdoc->first);
- if (mdoc->meta.title)
- free(mdoc->meta.title);
- if (mdoc->meta.os)
- free(mdoc->meta.os);
- if (mdoc->meta.name)
- free(mdoc->meta.name);
- if (mdoc->meta.arch)
- free(mdoc->meta.arch);
- if (mdoc->meta.vol)
- free(mdoc->meta.vol);
+ mdoc_free1(mdoc);
+ mdoc_alloc1(mdoc);
+}
+
- if (mdoc->htab)
- mdoc_tokhash_free(mdoc->htab);
+/*
+ * Completely free up all volatile and non-volatile parse resources.
+ * After invocation, the pointer is no longer usable.
+ */
+void
+mdoc_free(struct mdoc *mdoc)
+{
+ mdoc_free1(mdoc);
free(mdoc);
}
+/*
+ * Allocate volatile and non-volatile parse resources.
+ */
struct mdoc *
-mdoc_alloc(void *data, int pflags, const struct mdoc_cb *cb)
+mdoc_alloc(struct regset *regs, void *data,
+ int pflags, mandocmsg msg)
{
struct mdoc *p;
- p = xcalloc(1, sizeof(struct mdoc));
+ p = mandoc_calloc(1, sizeof(struct mdoc));
+ p->msg = msg;
p->data = data;
- if (cb)
- (void)memcpy(&p->cb, cb, sizeof(struct mdoc_cb));
-
- p->last = p->first =
- xcalloc(1, sizeof(struct mdoc_node));
- p->last->type = MDOC_ROOT;
p->pflags = pflags;
- p->next = MDOC_NEXT_CHILD;
- p->htab = mdoc_tokhash_alloc();
+ p->regs = regs;
+
+ mdoc_hash_init();
+ mdoc_alloc1(p);
return(p);
}
/*
* Climb back up the parse tree, validating open scopes. Mostly calls
- * through to macro_end in macro.c.
+ * through to macro_end() in macro.c.
*/
int
-mdoc_endparse(struct mdoc *mdoc)
+mdoc_endparse(struct mdoc *m)
{
- if (MDOC_HALT & mdoc->flags)
+ if (MDOC_HALT & m->flags)
return(0);
- if (NULL == mdoc->first)
+ else if (mdoc_macroend(m))
return(1);
-
- assert(mdoc->last);
- if ( ! macro_end(mdoc)) {
- mdoc->flags |= MDOC_HALT;
- return(0);
- }
- return(1);
+ m->flags |= MDOC_HALT;
+ return(0);
}
/*
* Main parse routine. Parses a single line -- really just hands off to
- * the macro or text parser.
+ * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
*/
int
-mdoc_parseln(struct mdoc *m, int ln, char *buf)
+mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
{
- /* If in error-mode, then we parse no more. */
-
if (MDOC_HALT & m->flags)
return(0);
- return('.' == *buf ? parsemacro(m, ln, buf) :
- parsetext(m, ln, buf));
-}
-
-
-void
-mdoc_vmsg(struct mdoc *mdoc, int ln, int pos, const char *fmt, ...)
-{
- char buf[256];
- va_list ap;
+ m->flags |= MDOC_NEWLINE;
- if (NULL == mdoc->cb.mdoc_msg)
- return;
+ /*
+ * Let the roff nS register switch SYNOPSIS mode early,
+ * such that the parser knows at all times
+ * whether this mode is on or off.
+ * Note that this mode is also switched by the Sh macro.
+ */
+ if (m->regs->regs[(int)REG_nS].set) {
+ if (m->regs->regs[(int)REG_nS].v.u)
+ m->flags |= MDOC_SYNOPSIS;
+ else
+ m->flags &= ~MDOC_SYNOPSIS;
+ }
- va_start(ap, fmt);
- (void)vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
- va_end(ap);
- (*mdoc->cb.mdoc_msg)(mdoc->data, ln, pos, buf);
+ return(('.' == buf[offs] || '\'' == buf[offs]) ?
+ mdoc_pmacro(m, ln, buf, offs) :
+ mdoc_ptext(m, ln, buf, offs));
}
int
-mdoc_verr(struct mdoc *mdoc, int ln, int pos,
- const char *fmt, ...)
+mdoc_vmsg(struct mdoc *mdoc, enum mandocerr t,
+ int ln, int pos, const char *fmt, ...)
{
char buf[256];
va_list ap;
- if (NULL == mdoc->cb.mdoc_err)
- return(0);
-
va_start(ap, fmt);
- (void)vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
+ vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
va_end(ap);
- return((*mdoc->cb.mdoc_err)(mdoc->data, ln, pos, buf));
-}
-
-
-int
-mdoc_vwarn(struct mdoc *mdoc, int ln, int pos,
- enum mdoc_warn type, const char *fmt, ...)
-{
- char buf[256];
- va_list ap;
-
- if (NULL == mdoc->cb.mdoc_warn)
- return(0);
- va_start(ap, fmt);
- (void)vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
- va_end(ap);
- return((*mdoc->cb.mdoc_warn)(mdoc->data, ln, pos, type, buf));
+ return((*mdoc->msg)(t, mdoc->data, ln, pos, buf));
}
int
-mdoc_macro(struct mdoc *m, int tok,
- int ln, int pp, int *pos, char *buf)
+mdoc_macro(MACRO_PROT_ARGS)
{
+ assert(tok < MDOC_MAX);
- /* FIXME - these should happen during validation. */
+ /* If we're in the body, deny prologue calls. */
if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
- SEC_PROLOGUE != m->lastnamed)
- return(mdoc_perr(m, ln, pp,
- "disallowed in document body"));
+ MDOC_PBODY & m->flags)
+ return(mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY));
- if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
- SEC_PROLOGUE == m->lastnamed)
- return(mdoc_perr(m, ln, pp,
- "disallowed in prologue"));
+ /* If we're in the prologue, deny "body" macros. */
- if (1 != pp && ! (MDOC_CALLABLE & mdoc_macros[tok].flags))
- return(mdoc_perr(m, ln, pp, "%s not callable",
- mdoc_macronames[tok]));
+ if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
+ ! (MDOC_PBODY & m->flags)) {
+ if ( ! mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG))
+ 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));
+ return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
}
static int
-mdoc_node_append(struct mdoc *mdoc, struct mdoc_node *p)
+node_append(struct mdoc *mdoc, struct mdoc_node *p)
{
assert(mdoc->last);
/* NOTREACHED */
}
+ p->parent->nchild++;
+
if ( ! mdoc_valid_pre(mdoc, p))
return(0);
if ( ! mdoc_action_pre(mdoc, p))
p->parent->tail = p;
break;
case (MDOC_BODY):
+ if (p->end)
+ break;
assert(MDOC_BLOCK == p->parent->type);
p->parent->body = p;
break;
}
mdoc->last = p;
+
+ switch (p->type) {
+ case (MDOC_TEXT):
+ if ( ! mdoc_valid_post(mdoc))
+ return(0);
+ if ( ! mdoc_action_post(mdoc))
+ return(0);
+ break;
+ default:
+ break;
+ }
+
return(1);
}
static struct mdoc_node *
-mdoc_node_alloc(const struct mdoc *mdoc)
+node_alloc(struct mdoc *m, int line, int pos,
+ enum mdoct tok, enum mdoc_type type)
{
struct mdoc_node *p;
- p = xcalloc(1, sizeof(struct mdoc_node));
- p->sec = mdoc->lastsec;
+ p = mandoc_calloc(1, sizeof(struct mdoc_node));
+ p->sec = m->lastsec;
+ p->line = line;
+ p->pos = pos;
+ p->tok = tok;
+ p->type = type;
+
+ /* Flag analysis. */
+
+ if (MDOC_SYNOPSIS & m->flags)
+ p->flags |= MDOC_SYNPRETTY;
+ else
+ p->flags &= ~MDOC_SYNPRETTY;
+ if (MDOC_NEWLINE & m->flags)
+ p->flags |= MDOC_LINE;
+ m->flags &= ~MDOC_NEWLINE;
return(p);
}
int
-mdoc_tail_alloc(struct mdoc *mdoc, int line, int pos, int tok)
+mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
{
struct mdoc_node *p;
- assert(mdoc->first);
- assert(mdoc->last);
-
- p = mdoc_node_alloc(mdoc);
-
- p->line = line;
- p->pos = pos;
- p->type = MDOC_TAIL;
- p->tok = tok;
-
- return(mdoc_node_append(mdoc, p));
+ p = node_alloc(m, line, pos, tok, MDOC_TAIL);
+ if ( ! node_append(m, p))
+ return(0);
+ m->next = MDOC_NEXT_CHILD;
+ return(1);
}
int
-mdoc_head_alloc(struct mdoc *mdoc, int line, int pos, int tok)
+mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
{
struct mdoc_node *p;
- assert(mdoc->first);
- assert(mdoc->last);
-
- p = mdoc_node_alloc(mdoc);
-
- p->line = line;
- p->pos = pos;
- p->type = MDOC_HEAD;
- p->tok = tok;
+ assert(m->first);
+ assert(m->last);
- return(mdoc_node_append(mdoc, p));
+ p = node_alloc(m, line, pos, tok, MDOC_HEAD);
+ if ( ! node_append(m, p))
+ return(0);
+ m->next = MDOC_NEXT_CHILD;
+ return(1);
}
int
-mdoc_body_alloc(struct mdoc *mdoc, int line, int pos, int tok)
+mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
{
struct mdoc_node *p;
- assert(mdoc->first);
- assert(mdoc->last);
-
- p = mdoc_node_alloc(mdoc);
-
- p->line = line;
- p->pos = pos;
- p->type = MDOC_BODY;
- p->tok = tok;
-
- return(mdoc_node_append(mdoc, p));
+ p = node_alloc(m, line, pos, tok, MDOC_BODY);
+ if ( ! node_append(m, p))
+ return(0);
+ m->next = MDOC_NEXT_CHILD;
+ return(1);
}
int
-mdoc_root_alloc(struct mdoc *mdoc)
+mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
+ struct mdoc_node *body, enum mdoc_endbody end)
{
struct mdoc_node *p;
- p = mdoc_node_alloc(mdoc);
-
- p->type = MDOC_ROOT;
-
- return(mdoc_node_append(mdoc, p));
+ p = node_alloc(m, line, pos, tok, MDOC_BODY);
+ p->pending = body;
+ p->end = end;
+ if ( ! node_append(m, p))
+ return(0);
+ m->next = MDOC_NEXT_SIBLING;
+ return(1);
}
int
-mdoc_block_alloc(struct mdoc *mdoc, int line, int pos,
- int tok, struct mdoc_arg *args)
+mdoc_block_alloc(struct mdoc *m, int line, int pos,
+ enum mdoct tok, struct mdoc_arg *args)
{
struct mdoc_node *p;
- p = mdoc_node_alloc(mdoc);
-
- p->pos = pos;
- p->line = line;
- p->type = MDOC_BLOCK;
- p->tok = tok;
+ p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
p->args = args;
-
- if (args)
+ if (p->args)
(args->refcnt)++;
-
- return(mdoc_node_append(mdoc, p));
+ if ( ! node_append(m, p))
+ return(0);
+ m->next = MDOC_NEXT_CHILD;
+ return(1);
}
int
-mdoc_elem_alloc(struct mdoc *mdoc, int line, int pos,
- int tok, struct mdoc_arg *args)
+mdoc_elem_alloc(struct mdoc *m, int line, int pos,
+ enum mdoct tok, struct mdoc_arg *args)
{
struct mdoc_node *p;
- p = mdoc_node_alloc(mdoc);
-
- p->line = line;
- p->pos = pos;
- p->type = MDOC_ELEM;
- p->tok = tok;
+ p = node_alloc(m, line, pos, tok, MDOC_ELEM);
p->args = args;
-
- if (args)
+ if (p->args)
(args->refcnt)++;
-
- return(mdoc_node_append(mdoc, p));
+ if ( ! node_append(m, p))
+ return(0);
+ m->next = MDOC_NEXT_CHILD;
+ return(1);
}
int
-mdoc_word_alloc(struct mdoc *mdoc,
- int line, int pos, const char *word)
+mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
{
- struct mdoc_node *p;
+ struct mdoc_node *n;
+ size_t sv, len;
- p = mdoc_node_alloc(mdoc);
+ len = strlen(p);
- p->line = line;
- p->pos = pos;
- p->type = MDOC_TEXT;
- p->string = xstrdup(word);
+ n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
+ n->string = mandoc_malloc(len + 1);
+ sv = strlcpy(n->string, p, len + 1);
+
+ /* Prohibit truncation. */
+ assert(sv < len + 1);
- return(mdoc_node_append(mdoc, p));
+ if ( ! node_append(m, n))
+ return(0);
+
+ m->next = MDOC_NEXT_SIBLING;
+ return(1);
}
-void
+static void
mdoc_node_free(struct mdoc_node *p)
{
+ /*
+ * XXX: if these end up being problematic in terms of memory
+ * management and dereferencing freed blocks, then make them
+ * into reference-counted double-pointers.
+ */
+
+ if (MDOC_Bd == p->tok && MDOC_BLOCK == p->type)
+ if (p->data.Bd)
+ free(p->data.Bd);
+ if (MDOC_Bl == p->tok && MDOC_BLOCK == p->type)
+ if (p->data.Bl)
+ free(p->data.Bl);
+ if (MDOC_Bf == p->tok && MDOC_HEAD == p->type)
+ if (p->data.Bf)
+ free(p->data.Bf);
+
if (p->string)
free(p->string);
if (p->args)
}
+static void
+mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
+{
+
+ /* 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_freelist(struct mdoc_node *p)
+mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
{
- if (p->child)
- mdoc_node_freelist(p->child);
- if (p->next)
- mdoc_node_freelist(p->next);
+ 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 offs)
{
+ char *c, *ws, *end;
+ struct mdoc_node *n;
+
+ /* Ignore bogus comments. */
+
+ if ('\\' == buf[offs] &&
+ '.' == buf[offs + 1] &&
+ '"' == buf[offs + 2])
+ return(mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT));
+
+ /* No text before an initial macro. */
+
+ if (SEC_NONE == m->lastnamed)
+ return(mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT));
+
+ assert(m->last);
+ n = m->last;
+
+ /*
+ * Divert directly to list processing if we're encountering a
+ * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
+ * (a MDOC_BODY means it's already open, in which case we should
+ * process within its context in the normal way).
+ */
+
+ if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
+ LIST_column == n->data.Bl->type) {
+ /* `Bl' is open without any children. */
+ m->flags |= MDOC_FREECOL;
+ return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
+ }
+
+ if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
+ NULL != n->parent &&
+ MDOC_Bl == n->parent->tok &&
+ LIST_column == n->parent->data.Bl->type) {
+ /* `Bl' has block-level `It' children. */
+ m->flags |= MDOC_FREECOL;
+ return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
+ }
+
+ /*
+ * Search for the beginning of unescaped trailing whitespace (ws)
+ * and for the first character not to be output (end).
+ */
+
+ /* FIXME: replace with strcspn(). */
+ ws = NULL;
+ for (c = end = buf + offs; *c; c++) {
+ switch (*c) {
+ case '-':
+ if (mandoc_hyph(buf + offs, c))
+ *c = ASCII_HYPH;
+ ws = NULL;
+ break;
+ case ' ':
+ if (NULL == ws)
+ ws = c;
+ continue;
+ case '\t':
+ /*
+ * Always warn about trailing tabs,
+ * even outside literal context,
+ * where they should be put on the next line.
+ */
+ if (NULL == ws)
+ ws = c;
+ /*
+ * Strip trailing tabs in literal context only;
+ * outside, they affect the next line.
+ */
+ if (MDOC_LITERAL & m->flags)
+ continue;
+ break;
+ case '\\':
+ /* Skip the escaped character, too, if any. */
+ if (c[1])
+ c++;
+ /* FALLTHROUGH */
+ default:
+ ws = NULL;
+ break;
+ }
+ end = c + 1;
+ }
+ *end = '\0';
+
+ if (ws)
+ if ( ! mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE))
+ return(0);
- if (SEC_PROLOGUE == m->lastnamed)
- return(mdoc_perr(m, line, 0,
- "text disallowed in prologue"));
+ if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
+ if ( ! mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN))
+ return(0);
- if (0 == buf[0] && ! (MDOC_LITERAL & m->flags))
- return(mdoc_perr(m, line, 0,
- "blank lines only in literal context"));
+ /*
+ * 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, offs, MDOC_Pp, NULL))
+ return(0);
- if ( ! mdoc_word_alloc(m, line, 0, buf))
+ m->next = MDOC_NEXT_SIBLING;
+ return(1);
+ }
+
+ if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
return(0);
- m->next = MDOC_NEXT_SIBLING;
+ if (MDOC_LITERAL & m->flags)
+ return(1);
+
+ /*
+ * 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.
+ */
+
+ assert(buf < end);
+
+ if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
+ m->last->flags |= MDOC_EOS;
+
return(1);
}
static int
-macrowarn(struct mdoc *m, int ln, const char *buf)
+macrowarn(struct mdoc *m, int ln, const char *buf, int offs)
{
- if ( ! (MDOC_IGN_MACRO & m->pflags))
- return(mdoc_perr(m, ln, 1, "unknown macro: %s%s",
- buf, strlen(buf) > 3 ? "..." : ""));
- return(mdoc_pwarn(m, ln, 1, WARN_SYNTAX,
- "unknown macro: %s%s",
- buf, strlen(buf) > 3 ? "..." : ""));
-}
+ int rc;
+ rc = mdoc_vmsg(m, MANDOCERR_MACRO, ln, offs,
+ "unknown macro: %s%s",
+ buf, strlen(buf) > 3 ? "..." : "");
+
+ /* FIXME: logic should be in driver. */
+ /* FIXME: broken, will error out and not omit a message. */
+ return(MDOC_IGN_MACRO & m->pflags ? rc : 0);
+}
/*
* Parse a macro line, that is, a line beginning with the control
* character.
*/
-int
-parsemacro(struct mdoc *m, int ln, char *buf)
+static int
+mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
{
- int i, c;
+ enum mdoct tok;
+ int i, j, sv;
char mac[5];
+ struct mdoc_node *n;
- /* Comments and empties are quickly ignored. */
+ /* Empty lines are ignored. */
- if (0 == buf[1])
+ offs++;
+
+ if ('\0' == buf[offs])
return(1);
- if (' ' == buf[1]) {
- i = 2;
- while (buf[i] && ' ' == buf[i])
+ i = offs;
+
+ /* Accept tabs/whitespace after the initial control char. */
+
+ if (' ' == buf[i] || '\t' == buf[i]) {
+ i++;
+ while (buf[i] && (' ' == buf[i] || '\t' == buf[i]))
i++;
- if (0 == buf[i])
+ if ('\0' == buf[i])
return(1);
- return(mdoc_perr(m, ln, 1, "invalid syntax"));
}
- if (buf[1] && '\\' == buf[1])
- if (buf[2] && '\"' == buf[2])
- return(1);
-
- /* Copy the first word into a nil-terminated buffer. */
+ sv = i;
- for (i = 1; i < 5; i++) {
- if (0 == (mac[i - 1] = buf[i]))
- break;
- else if (' ' == buf[i])
- break;
- }
+ /*
+ * Copy the first word into a nil-terminated buffer.
+ * Stop copying when a tab, space, or eoln is encountered.
+ */
- mac[i - 1] = 0;
+ j = 0;
+ while (j < 4 && '\0' != buf[i] && ' ' != buf[i] && '\t' != buf[i])
+ mac[j++] = buf[i++];
+ mac[j] = '\0';
- if (i == 5 || i <= 2) {
- if ( ! macrowarn(m, ln, mac))
+ if (j == 4 || j < 2) {
+ if ( ! macrowarn(m, ln, mac, sv))
goto err;
return(1);
}
- if (MDOC_MAX == (c = mdoc_tokhash_find(m->htab, mac))) {
- if ( ! macrowarn(m, ln, mac))
+ if (MDOC_MAX == (tok = mdoc_hash_find(mac))) {
+ if ( ! macrowarn(m, ln, mac, sv))
goto err;
return(1);
}
- /* The macro is sane. Jump to the next word. */
+ /* Disregard the first trailing tab, if applicable. */
+
+ if ('\t' == buf[i])
+ i++;
+
+ /* Jump to the next non-whitespace word. */
while (buf[i] && ' ' == buf[i])
i++;
- /* Begin recursive parse sequence. */
+ /*
+ * Trailing whitespace. Note that tabs are allowed to be passed
+ * into the parser as "text", so we only warn about spaces here.
+ */
- if ( ! mdoc_macro(m, c, ln, 1, &i, buf))
- goto err;
+ if ('\0' == buf[i] && ' ' == buf[i - 1])
+ if ( ! mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE))
+ goto err;
+
+ /*
+ * If an initial macro or a list invocation, divert directly
+ * into macro processing.
+ */
+
+ if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
+ if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
+ goto err;
+ return(1);
+ }
+
+ n = m->last;
+ assert(m->last);
/*
- * If we're in literal mode, then add a newline to the end of
- * macro lines. Our frontends will interpret this correctly
- * (it's documented in mdoc.3).
+ * If the first macro of a `Bl -column', open an `It' block
+ * context around the parsed macro.
*/
+ if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
+ LIST_column == n->data.Bl->type) {
+ m->flags |= MDOC_FREECOL;
+ if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
+ goto err;
+ return(1);
+ }
+
+ /*
+ * If we're following a block-level `It' within a `Bl -column'
+ * context (perhaps opened in the above block or in ptext()),
+ * then open an `It' block context around the parsed macro.
+ */
+
+ if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
+ NULL != n->parent &&
+ MDOC_Bl == n->parent->tok &&
+ LIST_column == n->parent->data.Bl->type) {
+ m->flags |= MDOC_FREECOL;
+ if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
+ goto err;
+ return(1);
+ }
+
+ /* Normal processing of a macro. */
+
+ if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
+ goto err;
+
return(1);
err: /* Error out. */
m->flags |= MDOC_HALT;
return(0);
}
+
+
git.ckatri.com / mandoc.git / blobdiff