/*
* mf.c -- mail filter subroutines
*
- * $Id$
- *
* This code is Copyright (c) 2002, by the authors of nmh. See the
* COPYRIGHT file in the root directory of the nmh distribution for
* complete copyright information.
*/
#include <h/mf.h>
-#include <ctype.h>
-#include <stdio.h>
#include <h/utils.h>
/*
* static prototypes
*/
-static char *getcpy (char *);
-static void compress (char *, unsigned char *);
-static int isat (char *);
+static int isat (const char *);
static int parse_address (void);
static int phrase (char *);
static int route_addr (char *);
static int my_lex (char *);
-static char *
-getcpy (char *s)
-{
- register char *p;
-
- if (!s) {
-/* causes compiles to blow up because the symbol _cleanup is undefined
- where did this ever come from? */
- /* _cleanup(); */
- abort();
- for(;;)
- pause();
- }
- p = mh_xmalloc ((size_t) (strlen (s) + 2));
- strcpy (p, s);
- return p;
-}
-
-
int
-isfrom(char *string)
+isfrom(const char *string)
{
- return (strncmp (string, "From ", 5) == 0
- || strncmp (string, ">From ", 6) == 0);
+ return (has_prefix(string, "From ")
+ || has_prefix(string, ">From "));
}
int
-lequal (unsigned char *a, unsigned char *b)
+lequal (const char *a, const char *b)
{
- for (; *a; a++, b++)
+ char c1, c2;
+
+ for (; *a; a++, b++) {
if (*b == 0)
return FALSE;
- else {
- char c1 = islower (*a) ? toupper (*a) : *a;
- char c2 = islower (*b) ? toupper (*b) : *b;
- if (c1 != c2)
- return FALSE;
- }
-
- return (*b == 0);
-}
-
-
-/*
- * seekadrx() is tricky. We want to cover both UUCP-style and ARPA-style
- * addresses, so for each list of addresses we see if we can find some
- * character to give us a hint.
- */
-
-
-#define CHKADR 0 /* undertermined address style */
-#define UNIXDR 1 /* UNIX-style address */
-#define ARPADR 2 /* ARPAnet-style address */
-
-
-static char *punctuators = ";<>.()[]";
-static char *vp = NULL;
-static char *tp = NULL;
-
-static struct adrx adrxs1;
-
-
-struct adrx *
-seekadrx (char *addrs)
-{
- static int state = CHKADR;
- register char *cp;
- register struct adrx *adrxp;
-
- if (state == CHKADR)
- for (state = UNIXDR, cp = addrs; *cp; cp++)
- if (strchr(punctuators, *cp)) {
- state = ARPADR;
- break;
- }
-
- switch (state) {
- case UNIXDR:
- adrxp = uucpadrx (addrs);
- break;
-
- case ARPADR:
- default:
- adrxp = getadrx (addrs);
- break;
+ c1 = toupper((unsigned char)*a);
+ c2 = toupper((unsigned char)*b);
+ if (c1 != c2)
+ return FALSE;
}
- if (adrxp == NULL)
- state = CHKADR;
-
- return adrxp;
-}
-
-
-/*
- * uucpadrx() implements a partial UUCP-style address parser. It's based
- * on the UUCP notion that addresses are separated by spaces or commas.
- */
-
-
-struct adrx *
-uucpadrx (char *addrs)
-{
- register unsigned char *cp, *wp, *xp, *yp;
- register char *zp;
- register struct adrx *adrxp = &adrxs1;
-
- if (vp == NULL) {
- vp = tp = getcpy (addrs);
- compress (addrs, vp);
- }
- else
- if (tp == NULL) {
- free (vp);
- vp = NULL;
- return NULL;
- }
-
- for (cp = tp; isspace (*cp); cp++)
- continue;
- if (*cp == 0) {
- free (vp);
- vp = tp = NULL;
- return NULL;
- }
-
- if ((wp = strchr(cp, ',')) == NULL) {
- if ((wp = strchr(cp, ' ')) != NULL) {
- xp = wp;
- while (isspace (*xp))
- xp++;
- if (*xp != 0 && isat (--xp)) {
- yp = xp + 4;
- while (isspace (*yp))
- yp++;
- if (*yp != 0) {
- if ((zp = strchr(yp, ' ')) != NULL)
- *zp = 0, tp = ++zp;
- else
- tp = NULL;
- }
- else
- *wp = 0, tp = ++wp;
- }
- else
- *wp = 0, tp = ++wp;
- }
- else
- tp = NULL;
- }
- else
- *wp = 0, tp = ++wp;
-
- if (adrxp->text)
- free (adrxp->text);
- adrxp->text = getcpy (cp);
- adrxp->mbox = cp;
- adrxp->host = adrxp->path = NULL;
- if ((wp = strrchr(cp, '@')) != NULL) {
- *wp++ = 0;
- adrxp->host = *wp ? wp : NULL;
- }
- else
- for (wp = cp + strlen (cp) - 4; wp >= cp; wp--)
- if (isat (wp)) {
- *wp++ = 0;
- adrxp->host = wp + 3;
- }
-
- adrxp->pers = adrxp->grp = adrxp->note = adrxp->err = NULL;
- adrxp->ingrp = 0;
-
- return adrxp;
-}
-
-
-static void
-compress (char *fp, unsigned char *tp)
-{
- register char c;
- register unsigned char *cp;
-
- for (c = ' ', cp = tp; (*tp = *fp++) != 0;)
- if (isspace (*tp)) {
- if (c != ' ')
- *tp++ = c = ' ';
- }
- else
- c = *tp++;
-
- if (c == ' ' && cp < tp)
- *--tp = 0;
+ return (*b == 0);
}
static int
-isat (char *p)
+isat (const char *p)
{
- return (strncmp (p, " AT ", 4)
- && strncmp (p, " At ", 4)
- && strncmp (p, " aT ", 4)
- && strncmp (p, " at ", 4) ? FALSE : TRUE);
+ return has_prefix(p, " AT ") || has_prefix(p, " At ") ||
+ has_prefix(p, " aT ") || has_prefix(p, " at ");
}
* getadrx() implements a partial 822-style address parser. The parser
* is neither complete nor correct. It does however recognize nearly all
* of the 822 address syntax. In addition it handles the majority of the
- * 733 syntax as well. Most problems arise from trying to accomodate both.
+ * 733 syntax as well. Most problems arise from trying to accommodate both.
*
* In terms of 822, the route-specification in
*
static int last_lex = LX_END;
static char *dp = NULL;
-static unsigned char *cp = NULL;
-static unsigned char *ap = NULL;
+static char *cp = NULL;
+static char *ap = NULL;
static char *pers = NULL;
static char *mbox = NULL;
static char *host = NULL;
static struct adrx adrxs2;
+/* eai = Email Address Internationalization */
struct adrx *
-getadrx (char *addrs)
+getadrx (const char *addrs, int eai)
{
- register char *bp;
- register struct adrx *adrxp = &adrxs2;
-
- if (pers)
- free (pers);
- if (mbox)
- free (mbox);
- if (host)
- free (host);
- if (path)
- free (path);
- if (grp)
- free (grp);
- if (note)
- free (note);
+ char *bp;
+ struct adrx *adrxp = &adrxs2;
+
+ mh_xfree(pers);
+ mh_xfree(mbox);
+ mh_xfree(host);
+ mh_xfree(path);
+ mh_xfree(grp);
+ mh_xfree(note);
pers = mbox = host = path = grp = note = NULL;
err[0] = 0;
if (dp == NULL) {
- dp = cp = getcpy (addrs ? addrs : "");
+ dp = cp = strdup (addrs ? addrs : "");
glevel = 0;
}
else
break;
}
+ if (! eai) {
+ /*
+ * Reject the address if key fields contain 8bit characters
+ */
+
+ if (contains8bit(mbox, NULL) || contains8bit(host, NULL) ||
+ contains8bit(path, NULL) || contains8bit(grp, NULL)) {
+ strcpy(err, "Address contains 8-bit characters");
+ }
+ }
+
if (err[0])
for (;;) {
switch (last_lex) {
}
break;
}
- while (isspace (*ap))
+ while (isspace ((unsigned char) *ap))
ap++;
if (cp)
- sprintf (adr, "%.*s", (int)(cp - ap), ap);
+ snprintf(adr, sizeof adr, "%.*s", (int)(cp - ap), ap);
else
strcpy (adr, ap);
bp = adr + strlen (adr) - 1;
switch (my_lex (buffer)) {
case LX_ATOM:
case LX_QSTR:
- pers = getcpy (buffer);
+ pers = strdup (buffer);
break;
case LX_SEMI:
strcpy (err, "extraneous semi-colon");
return NOTOK;
}
+ /* FALLTHRU */
case LX_COMA:
- if (note) {
- free (note);
- note = NULL;
- }
+ mh_xfree(note);
+ note = NULL;
goto again;
case LX_END:
return OK; /* why be choosy? */
default:
- sprintf (err, "illegal address construct (%s)", buffer);
+ snprintf(err, sizeof err, "illegal address construct (%s)", buffer);
return NOTOK;
}
return NOTOK;
if (last_lex == LX_RBRK)
return OK;
- sprintf (err, "missing right-bracket (%s)", buffer);
+ snprintf(err, sizeof err, "missing right-bracket (%s)", buffer);
return NOTOK;
case LX_COLN:
get_group: ;
if (glevel++ > 0) {
- sprintf (err, "nested groups not allowed (%s)", pers);
+ snprintf(err, sizeof err, "nested groups not allowed (%s)", pers);
return NOTOK;
}
grp = add (": ", pers);
goto more_phrase;
default:
- sprintf (err, "no mailbox in address, only a phrase (%s%s)",
+ snprintf(err, sizeof err, "no mailbox in address, only a phrase (%s%s)",
pers, buffer);
return NOTOK;
}
strcpy (err, "extraneous semi-colon");
return NOTOK;
}
+ /* FALLTHRU */
case LX_COMA:
case LX_END:
return OK;
default:
- sprintf (err, "junk after local@domain (%s)", buffer);
+ snprintf(err, sizeof err, "junk after local@domain (%s)", buffer);
return NOTOK;
}
return OK;
default:
- sprintf (err, "missing mailbox (%s)", buffer);
+ snprintf(err, sizeof err, "missing mailbox (%s)", buffer);
return NOTOK;
}
}
static int
route_addr (char *buffer)
{
- register char *pp = cp;
+ char *pp = cp;
if (my_lex (buffer) == LX_AT) {
if (route (buffer) == NOTOK)
return OK;
default:
- sprintf (err, "no at-sign after local-part (%s)", buffer);
+ snprintf(err, sizeof err, "no at-sign after local-part (%s)", buffer);
return NOTOK;
}
}
break;
default:
- sprintf (err, "no mailbox in local-part (%s)", buffer);
+ snprintf(err, sizeof err, "no mailbox in local-part (%s)", buffer);
return NOTOK;
}
break;
default:
- sprintf (err, "no sub-domain in domain-part of address (%s)", buffer);
+ snprintf(err, sizeof err, "no sub-domain in domain-part of address (%s)", buffer);
return NOTOK;
}
static int
route (char *buffer)
{
- path = getcpy ("@");
+ path = strdup ("@");
for (;;) {
switch (my_lex (buffer)) {
break;
default:
- sprintf (err, "no sub-domain in domain-part of address (%s)", buffer);
+ snprintf(err, sizeof err, "no sub-domain in domain-part of address (%s)", buffer);
return NOTOK;
}
switch (my_lex (buffer)) {
break;
default:
- sprintf (err, "no at-sign found for next domain in route (%s)",
+ snprintf(err, sizeof err, "no at-sign found for next domain in route (%s)",
buffer);
}
break;
return OK;
default:
- sprintf (err, "no colon found to terminate route (%s)", buffer);
+ snprintf(err, sizeof err, "no colon found to terminate route (%s)", buffer);
return NOTOK;
}
}
{
/* buffer should be at least BUFSIZ bytes long */
int i, gotat = 0;
- register unsigned char c;
- register char *bp;
+ char c, *bp;
/* Add C to the buffer bp. After use of this macro *bp is guaranteed to be within the buffer. */
#define ADDCHR(C) do { *bp++ = (C); if ((bp - buffer) == (BUFSIZ-1)) goto my_lex_buffull; } while (0)
gotat = isat (cp);
c = *cp++;
- while (isspace (c))
+ while (isspace ((unsigned char) c))
c = *cp++;
if (c == 0) {
cp = NULL;
continue;
case '(':
i++;
+ /* FALLTHRU */
default:
ADDCHR(c);
continue;
if (--i < 0) {
*bp = 0;
note = note ? add (buffer, add (" ", note))
- : getcpy (buffer);
+ : strdup (buffer);
return my_lex (buffer);
}
}
cp = NULL;
return (last_lex = LX_ERR);
}
+ /* FALLTHRU */
default:
ADDCHR(c);
continue;
cp = NULL;
return (last_lex = LX_ERR);
}
+ /* FALLTHRU */
default:
ADDCHR(c);
continue;
if (c == special[i].lx_chr)
return (last_lex = special[i].lx_val);
- if (iscntrl (c))
+ if (iscntrl ((unsigned char) c))
return (last_lex = LX_ERR);
for (;;) {
for (i = 0; special[i].lx_chr != 0; i++)
if (c == special[i].lx_chr)
goto got_atom;
- if (iscntrl (c) || isspace (c))
+ if (iscntrl ((unsigned char) c) || isspace ((unsigned char) c))
break;
ADDCHR(c);
}
char *
-legal_person (char *p)
+legal_person (const char *p)
{
int i;
- register char *cp;
+ const char *cp;
static char buffer[BUFSIZ];
if (*p == '"')
- return p;
+ return (char *) p;
for (cp = p; *cp; cp++)
for (i = 0; special[i].lx_chr; i++)
if (*cp == special[i].lx_chr) {
- sprintf (buffer, "\"%s\"", p);
+ snprintf(buffer, sizeof buffer, "\"%s\"", p);
return buffer;
}
- return p;
+ return (char *) p;
}
mfgets (FILE *in, char **bp)
{
int i;
- register char *cp, *dp, *ep;
+ char *cp, *dp, *ep;
static int len = 0;
static char *pp = NULL;
case '\t':
*cp++ = '\n';
break;
- } /* fall into default case */
+ }
+ /* FALLTHRU */
default:
*cp++ = i;
projects / nmh / blobdiff