X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/8a14191c0a0ad15bb8f35b49748c75c1e0a8c2f4..a3724ed3:/uip/mhparse.c?ds=sidebyside diff --git a/uip/mhparse.c b/uip/mhparse.c index 2be97603..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 */ @@ -245,8 +236,8 @@ parse_mime (char *file) return NULL; } fseek (fp, 0L, SEEK_SET); - } else if (lstat (file, &statbuf) == NOTOK) { - advise (file, "unable to lstat"); + } 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. */ @@ -300,7 +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; @@ -343,12 +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); } @@ -413,7 +417,7 @@ 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) { @@ -661,9 +665,11 @@ 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) { inform("invalid %s: field in message %s (empty type)", @@ -695,9 +701,11 @@ 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) { inform("invalid %s: field in message %s (empty subtype for \"%s\")", @@ -908,9 +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; @@ -1091,7 +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) { @@ -1103,13 +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; @@ -1289,7 +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." */ @@ -1398,9 +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; } @@ -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 @@ -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); @@ -3321,7 +3330,8 @@ 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)) @@ -3757,7 +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. */