X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/49976f25a761e67a05cc493cc7ab88863a587e30..a3724ed3:/uip/mhparse.c?ds=inline diff --git a/uip/mhparse.c b/uip/mhparse.c index 5db4efed..bab4e5b6 100644 --- a/uip/mhparse.c +++ b/uip/mhparse.c @@ -1,6 +1,4 @@ - -/* - * mhparse.c -- routines to parse the contents of MIME messages +/* mhparse.c -- routines to parse the contents of MIME messages * * This code is Copyright (c) 2002, by the authors of nmh. See the * COPYRIGHT file in the root directory of the nmh distribution for @@ -15,6 +13,9 @@ #include #include #include +#include +#include "../sbr/m_mktemp.h" +#include "mhfree.h" #ifdef HAVE_ICONV # include #endif /* HAVE_ICONV */ @@ -22,10 +23,6 @@ extern int debugsw; -/* cache policies */ -extern int rcachesw; /* mhcachesbr.c */ -extern int wcachesw; /* mhcachesbr.c */ - int checksw = 0; /* check Content-MD5 field */ /* @@ -57,8 +54,8 @@ int npreferred; */ struct k2v SubText[] = { { "plain", TEXT_PLAIN }, - { "richtext", TEXT_RICHTEXT }, /* defined in RFC-1341 */ - { "enriched", TEXT_ENRICHED }, /* defined in RFC-1896 */ + { "richtext", TEXT_RICHTEXT }, /* defined in RFC 1341 */ + { "enriched", TEXT_ENRICHED }, /* defined in RFC 1896 */ { NULL, TEXT_UNKNOWN } /* this one must be last! */ }; @@ -109,17 +106,11 @@ static struct k2v EncodingType[] = { }; -/* mhcachesbr.c */ -int find_cache (CT, int, int *, char *, char *, int); - /* mhmisc.c */ int part_ok (CT); int type_ok (CT, int); void content_error (char *, CT, char *, ...); -/* mhfree.c */ -void free_encoding (CT, int); - /* * static prototypes */ @@ -209,6 +200,7 @@ parse_mime (char *file) FILE *fp; CT ct; size_t n; + struct stat statbuf; bogus_mp_content = 0; @@ -244,6 +236,13 @@ parse_mime (char *file) return NULL; } fseek (fp, 0L, SEEK_SET); + } else if (stat (file, &statbuf) == NOTOK) { + advise (file, "unable to stat"); + return NULL; + } else if (S_ISDIR(statbuf.st_mode)) { + /* Don't try to parse a directory. */ + inform("%s is a directory", file); + return NULL; } else if ((fp = fopen (file, "r")) == NULL) { advise (file, "unable to read"); return NULL; @@ -252,7 +251,7 @@ parse_mime (char *file) if (!(ct = get_content (fp, file, 1))) { if (is_stdin) (void) m_unlink (file); - advise (NULL, "unable to decode %s", file); + inform("unable to decode %s", file); return NULL; } @@ -292,7 +291,7 @@ static CT get_content (FILE *in, char *file, int toplevel) { int compnum, state; - char buf[BUFSIZ], name[NAMESZ]; + char buf[NMH_BUFSIZ], name[NAMESZ]; char *np, *vp; CT ct; HF hp; @@ -335,12 +334,25 @@ get_content (FILE *in, char *file, int toplevel) continue; case BODY: + /* There are two cases. The unusual one is when there is no + * blank line between the headers and the body. This is + * indicated by the name of the header starting with `:'. + * + * For both cases, normal first, `1' is the desired c_begin + * file position for the start of the body, and `2' is the + * file position when buf is returned. + * + * f o o : b a r \n \n b o d y \n bufsz = 6 + * 1 2 move -5 + * f o o : b a r \n b o d y \n bufsz = 4 + * 1 2 move -4 + * + * For the normal case, bufsz includes the + * header-terminating `\n', even though it is not in buf, + * but bufsz isn't affected when it's missing in the unusual + * case. */ if (name[0] == ':') { - /* Special case: no blank line between header and body. The - file position indicator is on the newline at the end of the - line, but it needs to be one prior to the beginning of the - line. So subtract the length of the line, bufsz, plus 1. */ - ct->c_begin = ftell (in) - (bufsz + 1); + ct->c_begin = ftell(in) - bufsz; } else { ct->c_begin = ftell (in) - (bufsz - 1); } @@ -405,14 +417,14 @@ get_content (FILE *in, char *file, int toplevel) ucmp = !strcasecmp (cp, VRSN_VALUE); *dp = c; if (!ucmp) { - admonish (NULL, "message %s has unknown value for %s: field (%s)", + inform("message %s has unknown value for %s: field (%s), continuing...", ct->c_file, VRSN_FIELD, cp); } if (!ct->c_vrsn) { ct->c_vrsn = vrsn; } else { if (! suppress_multiple_mime_version_warning) - advise (NULL, "message %s has multiple %s: fields", + inform("message %s has multiple %s: fields", ct->c_file, VRSN_FIELD); free(vrsn); } @@ -424,7 +436,7 @@ get_content (FILE *in, char *file, int toplevel) /* Check if we've already seen a Content-Type header */ if (ct->c_ctline) { - advise (NULL, "message %s has multiple %s: fields", + inform("message %s has multiple %s: fields", ct->c_file, TYPE_FIELD); goto next_header; } @@ -455,7 +467,7 @@ get_content (FILE *in, char *file, int toplevel) * Content-Transfer-Encoding field */ if (ct->c_celine) { - advise (NULL, "message %s has multiple %s: fields", + inform("message %s has multiple %s: fields", ct->c_file, ENCODING_FIELD); goto next_header; } @@ -494,7 +506,7 @@ get_content (FILE *in, char *file, int toplevel) goto next_header; if (ct->c_digested) { - advise (NULL, "message %s has multiple %s: fields", + inform("message %s has multiple %s: fields", ct->c_file, MD5_FIELD); goto next_header; } @@ -653,12 +665,14 @@ get_ctinfo (char *cp, CT ct, int magic) for (dp = cp; istoken (*dp); dp++) continue; - c = *dp, *dp = '\0'; + c = *dp; + *dp = '\0'; ci->ci_type = mh_xstrdup(cp); /* store content type */ - *dp = c, cp = dp; + *dp = c; + cp = dp; if (!*ci->ci_type) { - advise (NULL, "invalid %s: field in message %s (empty type)", + inform("invalid %s: field in message %s (empty type)", TYPE_FIELD, ct->c_file); return NOTOK; } @@ -687,14 +701,15 @@ get_ctinfo (char *cp, CT ct, int magic) for (dp = cp; istoken (*dp); dp++) continue; - c = *dp, *dp = '\0'; + c = *dp; + *dp = '\0'; ci->ci_subtype = mh_xstrdup(cp); /* store the content subtype */ - *dp = c, cp = dp; + *dp = c; + cp = dp; if (!*ci->ci_subtype) { - advise (NULL, - "invalid %s: field in message %s (empty subtype for \"%s\")", - TYPE_FIELD, ct->c_file, ci->ci_type); + inform("invalid %s: field in message %s (empty subtype for \"%s\")", + TYPE_FIELD, ct->c_file, ci->ci_type); return NOTOK; } to_lower(ci->ci_subtype); @@ -720,7 +735,7 @@ magic_skip: mh_xfree(ct->c_id); ct->c_id = NULL; if (!(dp = strchr(ct->c_id = ++cp, '>'))) { - advise (NULL, "invalid ID in message %s", ct->c_file); + inform("invalid ID in message %s", ct->c_file); return NOTOK; } c = *dp; @@ -745,7 +760,7 @@ magic_skip: if (*dp == ']') break; if (dp < cp) { - advise (NULL, "invalid description in message %s", ct->c_file); + inform("invalid description in message %s", ct->c_file); ct->c_descr = NULL; return NOTOK; } @@ -772,7 +787,7 @@ magic_skip: if (*dp == '}') break; if (dp < cp) { - advise (NULL, "invalid disposition in message %s", ct->c_file); + inform("invalid disposition in message %s", ct->c_file); ct->c_dispo = NULL; return NOTOK; } @@ -806,7 +821,7 @@ magic_skip: cp++; if (dp == cp) { - advise (NULL, "invalid null transfer encoding specification"); + inform("invalid null transfer encoding specification"); return NOTOK; } @@ -823,7 +838,7 @@ magic_skip: } if (ct->c_reqencoding == CE_UNKNOWN) { - advise (NULL, "invalid CTE specification: \"%s\"", dp); + inform("invalid CTE specification: \"%s\"", dp); return NOTOK; } @@ -849,9 +864,8 @@ magic_skip: } } else - advise (NULL, - "extraneous information in message %s's %s: field\n%*s(%s)", - ct->c_file, TYPE_FIELD, strlen(invo_name) + 2, "", cp); + inform("extraneous information in message %s's %s: field\n%*s(%s)", + ct->c_file, TYPE_FIELD, strlen(invo_name) + 2, "", cp); } return OK; @@ -902,9 +916,11 @@ get_dispo (char *cp, CT ct, int buildflag) for (dp = cp; istoken (*dp); dp++) continue; - c = *dp, *dp = '\0'; + c = *dp; + *dp = '\0'; ct->c_dispo_type = mh_xstrdup(cp); /* store disposition type */ - *dp = c, cp = dp; + *dp = c; + cp = dp; if (*cp == '(' && get_comment (ct->c_file, DISPO_FIELD, &cp, NULL) == NOTOK) return NOTOK; @@ -917,9 +933,8 @@ get_dispo (char *cp, CT ct, int buildflag) return NOTOK; } } else if (*cp) { - advise (NULL, - "extraneous information in message %s's %s: field\n%*s(%s)", - ct->c_file, DISPO_FIELD, strlen(invo_name) + 2, "", cp); + inform("extraneous information in message %s's %s: field\n%*s(%s)", + ct->c_file, DISPO_FIELD, strlen(invo_name) + 2, "", cp); } if (buildflag) @@ -947,7 +962,7 @@ get_comment (const char *filename, const char *fieldname, char **ap, switch (c = *cp++) { case '\0': invalid: - advise (NULL, "invalid comment in message %s's %s: field", + inform("invalid comment in message %s's %s: field", filename, fieldname); return NOTOK; @@ -1086,7 +1101,7 @@ InitMultiPart (CT ct) /* * The encoding for multipart messages must be either - * 7bit, 8bit, or binary (per RFC2045). + * 7bit, 8bit, or binary (per RFC 2045). */ if (! skip_mp_cte_check && ct->c_encoding != CE_7BIT && ct->c_encoding != CE_8BIT && ct->c_encoding != CE_BINARY) { @@ -1098,13 +1113,12 @@ InitMultiPart (CT ct) while (bp >= cte && isspace ((unsigned char) *bp)) *bp-- = '\0'; for (bp = cte; *bp && isblank ((unsigned char) *bp); ++bp) continue; - admonish (NULL, - "\"%s/%s\" type in message %s must be encoded in\n" - "7bit, 8bit, or binary, per RFC 2045 (6.4). " - "mhfixmsg -fixcte can fix it, or\n" - "manually edit the file and change the \"%s\"\n" - "Content-Transfer-Encoding to one of those. For now", - ci->ci_type, ci->ci_subtype, ct->c_file, bp); + inform("\"%s/%s\" type in message %s must be encoded in\n" + "7bit, 8bit, or binary, per RFC 2045 (6.4). " + "mhfixmsg -fixcte can fix it, or\n" + "manually edit the file and change the \"%s\"\n" + "Content-Transfer-Encoding to one of those. For now, continuing...", + ci->ci_type, ci->ci_subtype, ct->c_file, bp); free (cte); return NOTOK; @@ -1127,9 +1141,8 @@ InitMultiPart (CT ct) /* complain if boundary parameter is missing */ if (!pm) { - advise (NULL, - "a \"boundary\" parameter is mandatory for \"%s/%s\" type in message %s's %s: field", - ci->ci_type, ci->ci_subtype, ct->c_file, TYPE_FIELD); + inform("a \"boundary\" parameter is mandatory for \"%s/%s\" type in message %s's %s: field", + ci->ci_type, ci->ci_subtype, ct->c_file, TYPE_FIELD); return NOTOK; } @@ -1141,7 +1154,7 @@ InitMultiPart (CT ct) for (cp = bp; isspace ((unsigned char) *cp); cp++) continue; if (!*cp) { - advise (NULL, "invalid \"boundary\" parameter for \"%s/%s\" type in message %s's %s: field", + inform("invalid \"boundary\" parameter for \"%s/%s\" type in message %s's %s: field", ci->ci_type, ci->ci_subtype, ct->c_file, TYPE_FIELD); return NOTOK; } @@ -1211,7 +1224,7 @@ end_part: } if (! suppress_bogus_mp_content_warning) { - advise (NULL, "bogus multipart content in message %s", ct->c_file); + inform("bogus multipart content in message %s", ct->c_file); } bogus_mp_content = 1; @@ -1285,7 +1298,7 @@ last_part: * ease of choosing/displaying it later on. from a mail message on * nmh-workers, from kenh: * "Stock" MH 6.8.5 did not have a reverse_parts() function, but I - * see code in mhn that did the same thing... Acccording to the RCS + * see code in mhn that did the same thing... According to the RCS * logs, that code was around from the initial checkin of mhn.c by * John Romine in 1992, which is as far back as we have." */ @@ -1394,9 +1407,9 @@ InitMessage (CT ct) CI ci = &ct->c_ctinfo; if ((ct->c_encoding != CE_7BIT) && (ct->c_encoding != CE_8BIT)) { - admonish (NULL, - "\"%s/%s\" type in message %s should be encoded in 7bit or 8bit", - ci->ci_type, ci->ci_subtype, ct->c_file); + inform("\"%s/%s\" type in message %s should be encoded in " + "7bit or 8bit, continuing...", ci->ci_type, ci->ci_subtype, + ct->c_file); return NOTOK; } @@ -1429,10 +1442,9 @@ InitMessage (CT ct) if (sscanf (pm->pm_value, "%d", &p->pm_partno) != 1 || p->pm_partno < 1) { invalid_param: - advise (NULL, - "invalid %s parameter for \"%s/%s\" type in message %s's %s field", - pm->pm_name, ci->ci_type, ci->ci_subtype, - ct->c_file, TYPE_FIELD); + inform("invalid %s parameter for \"%s/%s\" type in message %s's %s field", + pm->pm_name, ci->ci_type, ci->ci_subtype, + ct->c_file, TYPE_FIELD); return NOTOK; } continue; @@ -1448,10 +1460,8 @@ invalid_param: if (!p->pm_partid || !p->pm_partno || (p->pm_maxno && p->pm_partno > p->pm_maxno)) { - advise (NULL, - "invalid parameters for \"%s/%s\" type in message %s's %s field", - ci->ci_type, ci->ci_subtype, - ct->c_file, TYPE_FIELD); + inform("invalid parameters for \"%s/%s\" type in message %s's %s field", + ci->ci_type, ci->ci_subtype, ct->c_file, TYPE_FIELD); return NOTOK; } } @@ -1636,9 +1646,8 @@ params_external (CT ct, int composing) } if (!e->eb_access) { - advise (NULL, - "invalid parameters for \"%s/%s\" type in message %s's %s field", - ci->ci_type, ci->ci_subtype, ct->c_file, TYPE_FIELD); + inform("invalid parameters for \"%s/%s\" type in message %s's %s field", + ci->ci_type, ci->ci_subtype, ct->c_file, TYPE_FIELD); return NOTOK; } @@ -1845,7 +1854,7 @@ openBase64 (CT ct, char **file) if (ct->c_digested) { if (memcmp(digest, ct->c_digest, - sizeof(digest) / sizeof(digest[0]))) { + sizeof digest)) { content_error (NULL, ct, "content integrity suspect (digest mismatch) -- continuing"); } else { @@ -2011,7 +2020,8 @@ openQuoted (CT ct, char **file) for (ep = (cp = bufp) + cc - 1; cp <= ep; ep--) if (!isspace ((unsigned char) *ep)) break; - *++ep = '\n', ep++; + *++ep = '\n'; + ep++; for (; cp < ep; cp++) { if (quoted > 0) { @@ -2057,11 +2067,10 @@ openQuoted (CT ct, char **file) * sequence; let's decode it (above). */ quoted = 1; continue; - } else { - /* One or both of the next 2 is out of range, making this - * an invalid escape sequence; just show the raw bytes - * (below). */ } + /* One or both of the next 2 is out of range, making this + * an invalid escape sequence; just show the raw bytes + * (below). */ } /* Just show the raw byte. */ @@ -2097,7 +2106,7 @@ openQuoted (CT ct, char **file) MD5Final (digest, &mdContext); if (memcmp((char *) digest, (char *) ct->c_digest, - sizeof(digest) / sizeof(digest[0]))) + sizeof digest)) content_error (NULL, ct, "content integrity suspect (digest mismatch) -- continuing"); else @@ -2721,7 +2730,7 @@ openMail (CT ct, char **file) default: if (pidXwait (child_id, NULL) == OK) - advise (NULL, "request sent"); + inform("request sent"); break; } @@ -2902,7 +2911,7 @@ readDigest (CT ct, char *cp) size_t len; if (decodeBase64 (cp, &digest, &len, 0, NULL) == OK) { - const size_t maxlen = sizeof ct->c_digest / sizeof ct->c_digest[0]; + const size_t maxlen = sizeof ct->c_digest; if (strlen ((char *) digest) <= maxlen) { memcpy (ct->c_digest, digest, maxlen); @@ -3282,10 +3291,8 @@ parse_header_attrs (const char *filename, const char *fieldname, if (*cp == 0) { if (! suppress_extraneous_trailing_semicolon_warning) { - advise (NULL, - "extraneous trailing ';' in message %s's %s: " - "parameter list", - filename, fieldname); + inform("extraneous trailing ';' in message %s's %s: " + "parameter list", filename, fieldname); } return DONE; } @@ -3297,10 +3304,9 @@ parse_header_attrs (const char *filename, const char *fieldname, for (up = dp; isspace ((unsigned char) *dp);) dp++; if (dp == cp || *dp != '=') { - advise (NULL, - "invalid parameter in message %s's %s: " - "field\n%*sparameter %s (error detected at offset %d)", - filename, fieldname, strlen(invo_name) + 2, "",cp, dp - cp); + inform("invalid parameter in message %s's %s: " + "field\n%*sparameter %s (error detected at offset %d)", + filename, fieldname, strlen(invo_name) + 2, "",cp, dp - cp); return NOTOK; } @@ -3324,13 +3330,14 @@ parse_header_attrs (const char *filename, const char *fieldname, if (*vp == '*' && vp < up - 1) { partial = 1; continue; - } else if (*vp == '*' && vp == up - 1) { + } + if (*vp == '*' && vp == up - 1) { encoded = 1; } else if (partial) { if (isdigit((unsigned char) *vp)) index = *vp - '0' + index * 10; else { - advise (NULL, "invalid parameter index in message %s's " + inform("invalid parameter index in message %s's " "%s: field\n%*s(parameter %s)", filename, fieldname, strlen(invo_name) + 2, "", cp); return NOTOK; @@ -3374,7 +3381,7 @@ parse_header_attrs (const char *filename, const char *fieldname, } vp++; } else { - advise(NULL, "missing charset in message %s's %s: " + inform("missing charset in message %s's %s: " "field\n%*s(parameter %s)", filename, fieldname, strlen(invo_name) + 2, "", nameptr); free(nameptr); @@ -3397,7 +3404,7 @@ parse_header_attrs (const char *filename, const char *fieldname, } vp++; } else { - advise(NULL, "missing language tag in message %s's %s: " + inform("missing language tag in message %s's %s: " "field\n%*s(parameter %s)", filename, fieldname, strlen(invo_name) + 2, "", nameptr); free(nameptr); @@ -3421,7 +3428,7 @@ parse_header_attrs (const char *filename, const char *fieldname, !isxdigit((unsigned char) *(vp + 1)) || *(vp + 2) == '\0' || !isxdigit((unsigned char) *(vp + 2))) { - advise(NULL, "invalid encoded sequence in message " + inform("invalid encoded sequence in message " "%s's %s: field\n%*s(parameter %s)", filename, fieldname, strlen(invo_name) + 2, "", nameptr); @@ -3464,11 +3471,9 @@ parse_header_attrs (const char *filename, const char *fieldname, switch (*cp++) { case '\0': bad_quote: - advise (NULL, - "invalid quoted-string in message %s's %s: " - "field\n%*s(parameter %s)", - filename, fieldname, strlen(invo_name) + 2, "", - nameptr); + inform("invalid quoted-string in message %s's %s: " + "field\n%*s(parameter %s)", filename, + fieldname, strlen(invo_name) + 2, "", nameptr); free(nameptr); mh_xfree(charset); mh_xfree(lang); @@ -3552,7 +3557,7 @@ bad_quote: } else { for (sp2 = pp->sechead; sp2 != NULL; sp2 = sp2->next) { if (sp2->index == sp->index) { - advise (NULL, "duplicate index (%d) in message " + inform("duplicate index (%d) in message " "%s's %s: field\n%*s(parameter %s)", sp->index, filename, fieldname, strlen(invo_name) + 2, "", nameptr); @@ -3567,7 +3572,7 @@ bad_quote: } if (sp2 == NULL) { - advise(NULL, "Internal error: cannot insert partial " + inform("Internal error: cannot insert partial " "param in message %s's %s: field\n%*s(parameter %s)", filename, fieldname, strlen(invo_name) + 2, "", nameptr); @@ -3610,7 +3615,7 @@ bad_quote: int pindex = 0; for (sp = pp->sechead; sp != NULL; sp = sp->next) { if (sp->index != pindex++) { - advise(NULL, "missing section %d for parameter in " + inform("missing section %d for parameter in " "message %s's %s: field\n%*s(parameter %s)", pindex - 1, filename, fieldname, strlen(invo_name) + 2, "", pp->name); @@ -3681,7 +3686,7 @@ output_params(size_t initialwidth, PM params, int *offsetout, int external) continue; if (strlen(params->pm_name) > CPERLIN) { - advise(NULL, "Parameter name \"%s\" is too long", params->pm_name); + inform("Parameter name \"%s\" is too long", params->pm_name); mh_xfree(paramout); return NULL; } @@ -3762,7 +3767,7 @@ output_params(size_t initialwidth, PM params, int *offsetout, int external) } /* - * At this point, we're either finishing a contined parameter, or + * At this point, we're either finishing a continued parameter, or * we're working on a new one. */ @@ -3985,7 +3990,7 @@ encode_param(PM pm, char *output, size_t len, size_t valuelen, output += n; outlen += n; if (output > endptr) { - advise(NULL, "Internal error: parameter buffer overflow"); + inform("Internal error: parameter buffer overflow"); return 0; } } @@ -4005,7 +4010,7 @@ encode_param(PM pm, char *output, size_t len, size_t valuelen, outlen++; } if (output > endptr) { - advise(NULL, "Internal error: parameter buffer overflow"); + inform("Internal error: parameter buffer overflow"); return 0; } } @@ -4045,13 +4050,13 @@ normal_param(PM pm, char *output, size_t len, size_t valuelen, outlen++; } if (output > endptr) { - advise(NULL, "Internal error: parameter buffer overflow"); + inform("Internal error: parameter buffer overflow"); return 0; } } if (output - 2 > endptr) { - advise(NULL, "Internal error: parameter buffer overflow"); + inform("Internal error: parameter buffer overflow"); return 0; }