X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/5dd6771b28c257af405d7248639ed0e3bcdce38b..cf57870921b26703aad420c6741c524b33736ff1:/sbr/mf.c diff --git a/sbr/mf.c b/sbr/mf.c index fd92e369..a604ce91 100644 --- a/sbr/mf.c +++ b/sbr/mf.c @@ -8,16 +8,12 @@ */ #include -#include -#include #include /* * 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 *); @@ -27,209 +23,37 @@ static int route (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; - } - - 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; + c1 = toupper((unsigned char)*a); + c2 = toupper((unsigned char)*b); + if (c1 != c2) + return FALSE; } - 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 "); } @@ -238,7 +62,7 @@ isat (char *p) * 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 * @@ -333,8 +157,8 @@ static int ingrp = 0; 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; @@ -347,29 +171,24 @@ static char adr[BUFSIZ]; 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 @@ -403,6 +222,17 @@ getadrx (char *addrs) 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) { @@ -416,10 +246,10 @@ getadrx (char *addrs) } 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; @@ -450,7 +280,7 @@ again: ; switch (my_lex (buffer)) { case LX_ATOM: case LX_QSTR: - pers = getcpy (buffer); + pers = strdup (buffer); break; case LX_SEMI: @@ -458,11 +288,10 @@ again: ; 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: @@ -478,7 +307,7 @@ again: ; return OK; /* why be choosy? */ default: - sprintf (err, "illegal address construct (%s)", buffer); + snprintf(err, sizeof err, "illegal address construct (%s)", buffer); return NOTOK; } @@ -497,13 +326,13 @@ again: ; 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); @@ -532,7 +361,7 @@ again: ; 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; } @@ -563,12 +392,13 @@ again: ; 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; } @@ -585,7 +415,7 @@ again: ; return OK; default: - sprintf (err, "missing mailbox (%s)", buffer); + snprintf(err, sizeof err, "missing mailbox (%s)", buffer); return NOTOK; } } @@ -610,7 +440,7 @@ phrase (char *buffer) static int route_addr (char *buffer) { - register char *pp = cp; + char *pp = cp; if (my_lex (buffer) == LX_AT) { if (route (buffer) == NOTOK) @@ -633,7 +463,7 @@ route_addr (char *buffer) 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; } } @@ -652,7 +482,7 @@ local_part (char *buffer) break; default: - sprintf (err, "no mailbox in local-part (%s)", buffer); + snprintf(err, sizeof err, "no mailbox in local-part (%s)", buffer); return NOTOK; } @@ -679,7 +509,7 @@ domain (char *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; } @@ -704,7 +534,7 @@ domain (char *buffer) static int route (char *buffer) { - path = getcpy ("@"); + path = strdup ("@"); for (;;) { switch (my_lex (buffer)) { @@ -714,7 +544,7 @@ route (char *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)) { @@ -730,7 +560,7 @@ route (char *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; @@ -747,7 +577,7 @@ route (char *buffer) 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; } } @@ -759,8 +589,7 @@ my_lex (char *buffer) { /* 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) @@ -772,7 +601,7 @@ my_lex (char *buffer) gotat = isat (cp); c = *cp++; - while (isspace (c)) + while (isspace ((unsigned char) c)) c = *cp++; if (c == 0) { cp = NULL; @@ -796,6 +625,7 @@ my_lex (char *buffer) continue; case '(': i++; + /* FALLTHRU */ default: ADDCHR(c); continue; @@ -804,7 +634,7 @@ my_lex (char *buffer) if (--i < 0) { *bp = 0; note = note ? add (buffer, add (" ", note)) - : getcpy (buffer); + : strdup (buffer); return my_lex (buffer); } } @@ -823,6 +653,7 @@ my_lex (char *buffer) cp = NULL; return (last_lex = LX_ERR); } + /* FALLTHRU */ default: ADDCHR(c); continue; @@ -846,6 +677,7 @@ my_lex (char *buffer) cp = NULL; return (last_lex = LX_ERR); } + /* FALLTHRU */ default: ADDCHR(c); continue; @@ -862,7 +694,7 @@ my_lex (char *buffer) 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 (;;) { @@ -871,7 +703,7 @@ my_lex (char *buffer) 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); } @@ -893,22 +725,22 @@ got_atom: ; 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; } @@ -916,7 +748,7 @@ int mfgets (FILE *in, char **bp) { int i; - register char *cp, *dp, *ep; + char *cp, *dp, *ep; static int len = 0; static char *pp = NULL; @@ -954,7 +786,8 @@ mfgets (FILE *in, char **bp) case '\t': *cp++ = '\n'; break; - } /* fall into default case */ + } + /* FALLTHRU */ default: *cp++ = i;