-/* $Id: mdoc.c,v 1.118 2010/03/31 07:42:04 kristaps Exp $ */
+/* $Id: mdoc.c,v 1.138 2010/05/25 12:37:20 kristaps Exp $ */
/*
* Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
*
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
+#include <time.h>
+#include "mandoc.h"
#include "libmdoc.h"
#include "libmandoc.h"
-const char *const __mdoc_merrnames[MERRMAX] = {
- "trailing whitespace", /* ETAILWS */
- "unexpected quoted parameter", /* EQUOTPARM */
- "unterminated quoted parameter", /* EQUOTTERM */
- "argument parameter suggested", /* EARGVAL */
- "macro disallowed in prologue", /* EBODYPROL */
- "macro disallowed in body", /* EPROLBODY */
- "text disallowed in prologue", /* ETEXTPROL */
- "blank line disallowed", /* ENOBLANK */
- "text parameter too long", /* ETOOLONG */
- "invalid escape sequence", /* EESCAPE */
- "invalid character", /* EPRINT */
- "document has no body", /* ENODAT */
- "document has no prologue", /* ENOPROLOGUE */
- "expected line arguments", /* ELINE */
- "invalid AT&T argument", /* EATT */
- "default name not yet set", /* ENAME */
- "missing list type", /* ELISTTYPE */
- "missing display type", /* EDISPTYPE */
- "too many display types", /* EMULTIDISP */
- "too many list types", /* EMULTILIST */
- "NAME section must be first", /* ESECNAME */
- "badly-formed NAME section", /* ENAMESECINC */
- "argument repeated", /* EARGREP */
- "expected boolean parameter", /* EBOOL */
- "inconsistent column syntax", /* ECOLMIS */
- "nested display invalid", /* ENESTDISP */
- "width argument missing", /* EMISSWIDTH */
- "invalid section for this manual section", /* EWRONGMSEC */
- "section out of conventional order", /* ESECOOO */
- "section repeated", /* ESECREP */
- "invalid standard argument", /* EBADSTAND */
- "multi-line arguments discouraged", /* ENOMULTILINE */
- "multi-line arguments suggested", /* EMULTILINE */
- "line arguments discouraged", /* ENOLINE */
- "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 */
- "superfluous width argument", /* ENOWIDTH */
- "system: utsname error", /* EUTSNAME */
- "obsolete macro", /* EOBS */
- "end-of-line scope violation", /* EIMPBRK */
- "empty macro ignored", /* EIGNE */
- "unclosed explicit scope", /* EOPEN */
- "unterminated quoted phrase", /* EQUOTPHR */
- "closure macro without prior context", /* ENOCTX */
- "no description found for library", /* ELIB */
- "bad child for parent context", /* EBADCHILD */
- "list arguments preceding type", /* ENOTYPE */
-};
-
const char *const __mdoc_macronames[MDOC_MAX] = {
"Ap", "Dd", "Dt", "Os",
"Sh", "Ss", "Pp", "D1",
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 parsetext(struct mdoc *, int, char *);
-static int parsemacro(struct mdoc *, int, char *);
-static int macrowarn(struct mdoc *, int, const char *);
-static int pstring(struct mdoc *, int, int,
- const char *, size_t);
+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);
+
const struct mdoc_node *
mdoc_node(const struct mdoc *m)
{
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);
}
* Allocate volatile and non-volatile parse resources.
*/
struct mdoc *
-mdoc_alloc(void *data, int pflags, const struct mdoc_cb *cb)
+mdoc_alloc(void *data, int pflags, mandocmsg msg)
{
struct mdoc *p;
p = mandoc_calloc(1, sizeof(struct mdoc));
- if (cb)
- memcpy(&p->cb, cb, sizeof(struct mdoc_cb));
-
+ p->msg = msg;
p->data = data;
p->pflags = pflags;
/*
* 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)
+mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
{
if (MDOC_HALT & m->flags)
return(0);
- return('.' == *buf ? parsemacro(m, ln, buf) :
- parsetext(m, ln, buf));
-}
-
-
-int
-mdoc_verr(struct mdoc *mdoc, 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);
- va_end(ap);
-
- return((*mdoc->cb.mdoc_err)(mdoc->data, ln, pos, buf));
+ m->flags |= MDOC_NEWLINE;
+ return(('.' == buf[offs] || '\'' == buf[offs]) ?
+ mdoc_pmacro(m, ln, buf, offs) :
+ mdoc_ptext(m, ln, buf, offs));
}
int
-mdoc_vwarn(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_warn)
- 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_warn)(mdoc->data, ln, pos, buf));
-}
-
-
-int
-mdoc_err(struct mdoc *m, int line, int pos, int iserr, enum merr type)
-{
- const char *p;
-
- p = __mdoc_merrnames[(int)type];
- assert(p);
-
- if (iserr)
- return(mdoc_verr(m, line, pos, p));
-
- return(mdoc_vwarn(m, line, pos, p));
+ return((*mdoc->msg)(t, mdoc->data, ln, pos, buf));
}
mdoc_macro(struct mdoc *m, enum mdoct tok,
int ln, int pp, int *pos, char *buf)
{
-
assert(tok < MDOC_MAX);
- /*
- * If we're in the prologue, deny "body" macros. Similarly, if
- * we're in the body, deny prologue calls.
- */
+
+ /* 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));
+ return(mdoc_pmsg(m, ln, pp, MANDOCERR_BADBODY));
+
+ /* 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_pmsg(m, ln, pp, 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));
}
p->pos = pos;
p->tok = tok;
p->type = type;
-
+ if (MDOC_NEWLINE & m->flags)
+ p->flags |= MDOC_LINE;
+ m->flags &= ~MDOC_NEWLINE;
return(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 offs)
{
- int i, j;
- char sv;
+ char *c, *ws, *end;
- 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.
- */
-
- if (MDOC_LITERAL & m->flags)
- return(mdoc_word_alloc(m, line, 0, buf));
+ /* Ignore bogus comments. */
- /* Disallow blank/white-space lines in non-literal mode. */
+ if ('\\' == buf[offs] &&
+ '.' == buf[offs + 1] &&
+ '"' == buf[offs + 2])
+ return(mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT));
- for (i = 0; ' ' == buf[i]; i++)
- /* Skip leading whitespace. */ ;
+ /* No text before an initial macro. */
- if ('\0' == buf[i])
- return(mdoc_perr(m, line, 0, ENOBLANK));
+ if (SEC_NONE == m->lastnamed)
+ return(mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT));
/*
- * Break apart a free-form line into tokens. Spaces are
- * stripped out of the input.
+ * Search for the beginning of unescaped trailing whitespace (ws)
+ * and for the first character not to be output (end).
*/
-
- for (j = i; buf[i]; i++) {
- if (' ' != buf[i])
- continue;
-
- /* Escaped whitespace. */
- if (i && ' ' == buf[i] && '\\' == buf[i - 1])
+ ws = NULL;
+ for (c = end = buf + offs; *c; c++) {
+ switch (*c) {
+ case '-':
+ if (mandoc_hyph(buf + offs, c))
+ *c = ASCII_HYPH;
+ 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';
- sv = buf[i];
- buf[i++] = '\0';
-
- if ( ! pstring(m, line, j, &buf[j], (size_t)(i - j)))
+ if (ws)
+ if ( ! mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE))
return(0);
- /* Trailing whitespace? Check at overwritten byte. */
+ if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
+ if ( ! mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN))
+ return(0);
- if (' ' == sv && '\0' == buf[i])
- if ( ! mdoc_pwarn(m, line, i - 1, ETAILWS))
- 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, offs, MDOC_Pp, NULL))
+ return(0);
- for ( ; ' ' == buf[i]; i++)
- /* Skip trailing whitespace. */ ;
+ m->next = MDOC_NEXT_SIBLING;
+ return(1);
+ }
- j = i;
+ if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
+ return(0);
- /* Trailing whitespace? */
+ if (MDOC_LITERAL & m->flags)
+ return(1);
- if (' ' == buf[i - 1] && '\0' == buf[i])
- if ( ! mdoc_pwarn(m, line, i - 1, ETAILWS))
- return(0);
+ /*
+ * 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.
+ */
- if ('\0' == buf[i])
- break;
- }
+ assert(buf < end);
- if (j != i && ! pstring(m, line, j, &buf[j], (size_t)(i - j)))
- return(0);
+ if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
+ m->last->flags |= MDOC_EOS;
- m->next = MDOC_NEXT_SIBLING;
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_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 ? "..." : ""));
+ int rc;
+
+ rc = mdoc_vmsg(m, MANDOCERR_MACRO, ln, offs,
+ "unknown macro: %s%s",
+ buf, strlen(buf) > 3 ? "..." : "");
+
+ /* FIXME: logic should be in driver. */
+ return(MDOC_IGN_MACRO & m->pflags ? rc : 0);
}
* character.
*/
int
-parsemacro(struct mdoc *m, int ln, char *buf)
+mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
{
- 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])
+ offs++;
+
+ if ('\0' == buf[offs])
return(1);
- i = 1;
+ i = offs;
/* Accept whitespace after the initial control char. */
return(1);
}
+ sv = i;
+
/* Copy the first word into a nil-terminated buffer. */
for (j = 0; j < 4; j++, i++) {
if (isgraph((u_char)buf[i]))
continue;
- return(mdoc_perr(m, ln, i, EPRINT));
+ if ( ! mdoc_pmsg(m, ln, i, MANDOCERR_BADCHAR))
+ return(0);
+ i--;
}
- mac[j] = 0;
+ mac[j] = '\0';
if (j == 4 || j < 2) {
- if ( ! macrowarn(m, ln, mac))
+ if ( ! macrowarn(m, ln, mac, sv))
goto err;
return(1);
}
- if (MDOC_MAX == (c = mdoc_hash_find(mac))) {
- if ( ! macrowarn(m, ln, mac))
+ if (MDOC_MAX == (tok = mdoc_hash_find(mac))) {
+ if ( ! macrowarn(m, ln, mac, sv))
goto err;
return(1);
}
while (buf[i] && ' ' == buf[i])
i++;
- /* Trailing whitespace? */
+ /*
+ * 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))
+ if ( ! mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE))
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