X-Git-Url: https://code.citadel.org/?a=blobdiff_plain;f=libcitadel%2Flib%2Fmime_parser.c;h=1fc037d43ffd77462434eb93f645001f4dbb251a;hb=8659bf61f03724755374145e6211be6bbfe74fda;hp=66f8c0c782d70f4e14871303b9060565f501f8ca;hpb=861c24fd883fcb626ade4c729aba590af48d8d14;p=citadel.git diff --git a/libcitadel/lib/mime_parser.c b/libcitadel/lib/mime_parser.c index 66f8c0c78..1fc037d43 100644 --- a/libcitadel/lib/mime_parser.c +++ b/libcitadel/lib/mime_parser.c @@ -1,11 +1,21 @@ /* - * $Id$ - * * This is the MIME parser for Citadel. * - * Copyright (c) 1998-2007 by the citadel.org development team. - * This code is distributed under the GNU General Public License v3. + * Copyright (c) 1998-2010 by the citadel.org development team. + * + * This program is open source software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 3 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ #include @@ -24,22 +34,72 @@ #include "libcitadel.h" #include "libcitadellocal.h" -void extract_key(char *target, char *source, char *key) +const unsigned char FromHexTable [256] = { + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, // 0 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, // 10 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, // 20 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, // 30 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0x00, 0x01, // 40 + 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0xFF, 0xFF, // 50 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, // 60 + 0x0F, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, // 70 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, // 80 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0x0A, 0x0B, 0x0C, // 90 + 0x0D, 0x0E, 0x0F, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //100 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //110 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //120 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //130 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //140 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //150 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //160 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //170 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //180 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //190 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //200 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //210 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //220 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //230 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, //240 + 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF //250 +}; + + +long extract_key(char *target, char *source, long sourcelen, char *key, long keylen, char KeyEnd) { - char *ptr; - char looking_for[256]; + char *sptr, *ptr = NULL; int double_quotes = 0; + long RealKeyLen = keylen; - snprintf(looking_for, sizeof looking_for, "%s=", key); + sptr = source; - ptr = bmstrcasestr(source, looking_for); + while (sptr != NULL) + { + ptr = bmstrcasestr_len(sptr, sourcelen - (sptr - source), + key, keylen); + if(ptr != NULL) + { + while (isspace(*(ptr + RealKeyLen))) + RealKeyLen ++; + if (*(ptr + RealKeyLen) == KeyEnd) + { + sptr = NULL; + RealKeyLen ++; + } + else + { + sptr = ptr + RealKeyLen + 1; + } + } + else + sptr = ptr; + } if (ptr == NULL) { - strcpy(target, ""); - return; + *target = '\0'; + return 0; } - strcpy(target, (ptr + strlen(looking_for))); + strcpy(target, (ptr + RealKeyLen)); - for (ptr=target; (*ptr != 0); ++ptr) { + for (ptr=target; (*ptr != 0); ptr++) { /* A semicolon means we've hit the end of the key, unless we're inside double quotes */ if ( (double_quotes != 1) && (*ptr == ';')) { @@ -57,6 +117,8 @@ void extract_key(char *target, char *source, char *key) } } } + *ptr = '\0'; + return ptr - target; } @@ -71,35 +133,25 @@ char *fixed_partnum(char *supplied_partnum) { } -unsigned int decode_hex(char *Source) +static inline unsigned int _decode_hex(const char *Source) { - int ret = 0; - if (*Source < 'A') { - ret += (*Source - '0'); - } - else if (*Source > 'Z') { - ret += (*Source - 'a' + 10); - } - else { - ret += (*Source - 'A' + 10); - } + unsigned int ret = '?'; + unsigned char LO_NIBBLE; + unsigned char HI_NIBBLE; + HI_NIBBLE = FromHexTable[(unsigned char) *Source]; + LO_NIBBLE = FromHexTable[(unsigned char) *(Source+1)]; + + if ((LO_NIBBLE == 0xFF) || (LO_NIBBLE == 0xFF)) + return ret; + ret = HI_NIBBLE; ret = ret << 4; - - if (*(Source + 1) < 'A') { - ret += (*(Source + 1) - '0'); - } - else if (*(Source + 1) > 'Z') { - ret += (*(Source + 1) - 'a' + 10); - } - else { - ret += (*(Source + 1) - 'A' + 10); - } - if (ret > 255) - return '?'; + ret = ret | LO_NIBBLE; return ret; } +unsigned int decode_hex(char *Source) {return _decode_hex(Source);} + /* * Convert "quoted-printable" to binary. Returns number of bytes decoded. * according to RFC2045 section 6.7 @@ -111,20 +163,25 @@ int CtdlDecodeQuotedPrintable(char *decoded, char *encoded, int sourcelen) { while (pos < sourcelen) { - if (!strncmp(&encoded[pos], "=\r\n", 3)) - { - pos += 3; - } - else if (!strncmp(&encoded[pos], "=\n", 2)) - { - pos += 2; - } - else if (encoded[pos] == '=') + if (*(encoded + pos) == '=') { - ch = 0; - ch = decode_hex(&encoded[pos+1]); - pos += 3; - decoded[decoded_length++] = ch; + pos ++; + if (*(encoded + pos) == '\n') + { + pos ++; + } + else if (*(encoded + pos) == '\r') + { + pos ++; + if (*(encoded + pos) == '\n') + pos++; + } + else + { + ch = _decode_hex(&encoded[pos]); + pos += 2; + decoded[decoded_length++] = ch; + } } else { @@ -147,45 +204,11 @@ void mime_decode(char *partnum, char *disposition, char *id, char *name, char *filename, - void (*CallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - void (*PreMultiPartCallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - void (*PostMultiPartCallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - void *userdata, - int dont_decode -) + MimeParserCallBackType CallBack, + MimeParserCallBackType PreMultiPartCallBack, + MimeParserCallBackType PostMultiPartCallBack, + void *userdata, + int dont_decode) { char *decoded; @@ -193,18 +216,28 @@ void mime_decode(char *partnum, /* Some encodings aren't really encodings */ if (!strcasecmp(encoding, "7bit")) - strcpy(encoding, ""); + *encoding = '\0'; if (!strcasecmp(encoding, "8bit")) - strcpy(encoding, ""); + *encoding = '\0'; if (!strcasecmp(encoding, "binary")) - strcpy(encoding, ""); + *encoding = '\0'; + if (!strcasecmp(encoding, "ISO-8859-1")) + *encoding = '\0'; /* If this part is not encoded, send as-is */ if ( (strlen(encoding) == 0) || (dont_decode)) { if (CallBack != NULL) { - CallBack(name, filename, fixed_partnum(partnum), - disposition, part_start, - content_type, charset, length, encoding, id, userdata); + CallBack(name, + filename, + fixed_partnum(partnum), + disposition, + part_start, + content_type, + charset, + length, + encoding, + id, + userdata); } return; } @@ -234,145 +267,148 @@ void mime_decode(char *partnum, } if (bytes_decoded > 0) if (CallBack != NULL) { - CallBack(name, filename, fixed_partnum(partnum), - disposition, decoded, - content_type, charset, bytes_decoded, "binary", id, userdata); + char encoding_buf[SIZ]; + + strcpy(encoding_buf, "binary"); + CallBack(name, + filename, + fixed_partnum(partnum), + disposition, + decoded, + content_type, + charset, + bytes_decoded, + encoding_buf, + id, + userdata); } free(decoded); } /* - * Break out the components of a multipart message - * (This function expects to be fed HEADERS + CONTENT) - * Note: NULL can be supplied as content_end; in this case, the message is - * considered to have ended when the parser encounters a 0x00 byte. + * this is the extract of mime_decode which can be called if 'dont_decode' was set; + * to save the cpu intense process of decoding to the time when it realy wants the content. + * returns: + * - > 0 we decoded something, its on *decoded, you need to free it. + * - = 0 no need to decode stuff. *decoded will be NULL. + * - < 0 an error occured, either an unknown encoding, or alloc failed. no need to free. */ -void the_mime_parser(char *partnum, - char *content_start, char *content_end, - void (*CallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - void (*PreMultiPartCallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - void (*PostMultiPartCallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - void *userdata, - int dont_decode -) +int mime_decode_now (char *part_start, + size_t length, + char *encoding, + char **decoded, + size_t *bytes_decoded) { + *bytes_decoded = 0; + *decoded = NULL; + /* Some encodings aren't really encodings */ + if (!strcasecmp(encoding, "7bit")) + *encoding = '\0'; + if (!strcasecmp(encoding, "8bit")) + *encoding = '\0'; + if (!strcasecmp(encoding, "binary")) + *encoding = '\0'; - char *ptr; - char *srch = NULL; - char *part_start, *part_end = NULL; - char buf[SIZ]; - char *header; - char *boundary; - char *startary; - size_t startary_len = 0; - char *endary; - char *next_boundary; - char *content_type; - char *charset; - size_t content_length; - char *encoding; - char *disposition; - char *id; - char *name = NULL; - char *content_type_name; - char *content_disposition_name; - char *filename; - int is_multipart; - int part_seq = 0; - int i; - size_t length; - char nested_partnum[256]; - int crlf_in_use = 0; - char *evaluate_crlf_ptr = NULL; - int buflen = 0; - int headerlen = 0; - - ptr = content_start; - content_length = 0; - - boundary = malloc(SIZ); - memset(boundary, 0, SIZ); - - startary = malloc(SIZ); - memset(startary, 0, SIZ); - - endary = malloc(SIZ); - memset(endary, 0, SIZ); - - header = malloc(SIZ); - memset(header, 0, SIZ); - - content_type = malloc(SIZ); - memset(content_type, 0, SIZ); + /* If this part is not encoded, send as-is */ + if (strlen(encoding) == 0) { + return 0; + } + - charset = malloc(SIZ); - memset(charset, 0, SIZ); + /* Fail if we hit an unknown encoding. */ + if ((strcasecmp(encoding, "base64")) + && (strcasecmp(encoding, "quoted-printable"))) { + return -1; + } - encoding = malloc(SIZ); - memset(encoding, 0, SIZ); + /* + * Allocate a buffer for the decoded data. The output buffer is slightly + * larger than the input buffer; this assumes that the decoded data + * will never be significantly larger than the encoded data. This is a + * safe assumption with base64, uuencode, and quoted-printable. + */ + *decoded = malloc(length + 32768); + if (decoded == NULL) { + return -1; + } - content_type_name = malloc(SIZ); - memset(content_type_name, 0, SIZ); + if (!strcasecmp(encoding, "base64")) { + *bytes_decoded = CtdlDecodeBase64(*decoded, part_start, length); + return 1; + } + else if (!strcasecmp(encoding, "quoted-printable")) { + *bytes_decoded = CtdlDecodeQuotedPrintable(*decoded, part_start, length); + return 1; + } + return -1; +} - content_disposition_name = malloc(SIZ); - memset(content_disposition_name, 0, SIZ); +typedef enum _eIntMimeHdrs { + boundary, + startary, + endary, + content_type, + charset, + encoding, + content_type_name, + content_disposition_name, + filename, + disposition, + id, + eMax /* don't move ! */ +} eIntMimeHdrs; + +typedef struct _CBufStr { + char Key[SIZ]; + long len; +}CBufStr; + +typedef struct _interesting_mime_headers { + CBufStr b[eMax]; + long content_length; + long is_multipart; +} interesting_mime_headers; + + +static void FlushInterestingMimes(interesting_mime_headers *m) +{ + int i; + + for (i = 0; i < eMax; i++) { + m->b[i].Key[0] = '\0'; + m->b[i].len = 0; + } + m->content_length = -1; +} +static interesting_mime_headers *InitInterestingMimes(void) +{ + interesting_mime_headers *m; + m = (interesting_mime_headers*) malloc( sizeof(interesting_mime_headers)); - filename = malloc(SIZ); - memset(filename, 0, SIZ); + FlushInterestingMimes(m); - disposition = malloc(SIZ); - memset(disposition, 0, SIZ); + return m; +} - id = malloc(SIZ); - memset(id, 0, SIZ); - /* If the caller didn't supply an endpointer, generate one by measure */ - if (content_end == NULL) { - content_end = &content_start[strlen(content_start)]; - } +static long parse_MimeHeaders(interesting_mime_headers *m, + char** pcontent_start, + char *content_end) +{ + char buf[SIZ]; + char header[SIZ]; + long headerlen; + char *ptr, *pch; + int buflen = 0; + int i; /* Learn interesting things from the headers */ - strcpy(header, ""); + ptr = *pcontent_start; + *header = '\0'; headerlen = 0; do { ptr = memreadlinelen(ptr, buf, SIZ, &buflen); - if (ptr >= content_end) { - goto end_parser; - } for (i = 0; i < buflen; ++i) { if (isspace(buf[i])) { @@ -380,41 +416,58 @@ void the_mime_parser(char *partnum, } } - if (!isspace(buf[0])) { + if (!isspace(buf[0]) && (headerlen > 0)) { if (!strncasecmp(header, "Content-type:", 13)) { - strcpy(content_type, &header[13]); - striplt(content_type); - extract_key(content_type_name, content_type, "name"); - extract_key(charset, content_type, "charset"); - extract_key(boundary, header, "boundary"); + memcpy (m->b[content_type].Key, &header[13], headerlen - 12); + m->b[content_type].Key[headerlen - 12] = '\0'; + m->b[content_type].len = striplt (m->b[content_type].Key); + + m->b[content_type_name].len = extract_key(m->b[content_type_name].Key, CKEY(m->b[content_type]), HKEY("name"), '='); + m->b[charset].len = extract_key(m->b[charset].Key, CKEY(m->b[content_type]), HKEY("charset"), '='); + m->b[boundary].len = extract_key(m->b[boundary].Key, header, headerlen, HKEY("boundary"), '='); + /* Deal with weird headers */ - if (strchr(content_type, ' ')) - *(strchr(content_type, ' ')) = '\0'; - if (strchr(content_type, ';')) - *(strchr(content_type, ';')) = '\0'; + pch = strchr(m->b[content_type].Key, ' '); + if (pch != NULL) { + *pch = '\0'; + m->b[content_type].len = m->b[content_type].Key - pch; + } + pch = strchr(m->b[content_type].Key, ';'); + if (pch != NULL) { + *pch = '\0'; + m->b[content_type].len = m->b[content_type].Key - pch; + } } - if (!strncasecmp(header, "Content-Disposition:", 20)) { - strcpy(disposition, &header[20]); - striplt(disposition); - extract_key(content_disposition_name, disposition, "name"); - extract_key(filename, disposition, "filename"); + else if (!strncasecmp(header, "Content-Disposition:", 20)) { + memcpy (m->b[disposition].Key, &header[20], headerlen - 19); + m->b[disposition].Key[headerlen - 19] = '\0'; + m->b[disposition].len = striplt(m->b[disposition].Key); + + m->b[content_disposition_name].len = extract_key(m->b[content_disposition_name].Key, CKEY(m->b[disposition]), HKEY("name"), '='); + m->b[filename].len = extract_key(m->b[filename].Key, CKEY(m->b[disposition]), HKEY("filename"), '='); + pch = strchr(m->b[disposition].Key, ';'); + if (pch != NULL) *pch = '\0'; + m->b[disposition].len = striplt(m->b[disposition].Key); } - if (!strncasecmp(header, "Content-ID:", 11)) { - strcpy(id, &header[11]); - striplt(id); - stripallbut(id, '<', '>'); + else if (!strncasecmp(header, "Content-ID:", 11)) { + memcpy(m->b[id].Key, &header[11], headerlen - 11); + m->b[id].Key[headerlen - 11] = '\0'; + striplt(m->b[id].Key); + m->b[id].len = stripallbut(m->b[id].Key, '<', '>'); } - if (!strncasecmp(header, "Content-length: ", 15)) { - char clbuf[10]; - safestrncpy(clbuf, &header[15], sizeof clbuf); - striplt(clbuf); - content_length = (size_t) atol(clbuf); + else if (!strncasecmp(header, "Content-length: ", 15)) { + char *clbuf; + clbuf = &header[15]; + while (isspace(*clbuf)) + clbuf ++; + m->content_length = (size_t) atol(clbuf); } - if (!strncasecmp(header, "Content-transfer-encoding: ", 26)) { - strcpy(encoding, &header[26]); - striplt(encoding); + else if (!strncasecmp(header, "Content-transfer-encoding: ", 26)) { + memcpy(m->b[encoding].Key, &header[26], headerlen - 26); + m->b[encoding].Key[headerlen - 26] = '\0'; + m->b[encoding].len = striplt(m->b[encoding].Key); } - strcpy(header, ""); + *header = '\0'; headerlen = 0; } if ((headerlen + buflen + 2) < SIZ) { @@ -422,45 +475,186 @@ void the_mime_parser(char *partnum, headerlen += buflen; header[headerlen] = '\0'; } + if (ptr >= content_end) { + return -1; + } } while ((!IsEmptyStr(buf)) && (*ptr != 0)); - if (strchr(disposition, ';')) - *(strchr(disposition, ';')) = '\0'; - striplt(disposition); - if (strchr(content_type, ';')) - *(strchr(content_type, ';')) = '\0'; - striplt(content_type); - - if (!IsEmptyStr(boundary)) { - is_multipart = 1; - } else { - is_multipart = 0; + m->is_multipart = m->b[boundary].len != 0; + *pcontent_start = ptr; + + return 0; +} + + +static int IsAsciiEncoding(interesting_mime_headers *m) +{ + + if ((m->b[encoding].len != 0) && + (strcasecmp(m->b[encoding].Key, "base64") == 0)) + return 1; + if ((m->b[encoding].len != 0) && + (strcmp(m->b[encoding].Key, "quoted-printable") == 0)) + return 1; + + return 0; +} + +static char *FindNextContent(char *ptr, + char *content_end, + interesting_mime_headers *SubMimeHeaders, + interesting_mime_headers *m) +{ + char *next_boundary; + char tmp; + + if (IsAsciiEncoding(SubMimeHeaders)) { + tmp = *content_end; + *content_end = '\0'; + + /** + * ok, if we have a content length of the mime part, + * try skipping the content on the search for the next + * boundary. since we don't trust the content_length + * to be all accurate, and suspect it to lose one digit + * per line with a line length of 80 chars, we need + * to start searching a little before.. + */ + + if ((SubMimeHeaders->content_length != -1) && + (SubMimeHeaders->content_length > 10)) + { + char *pptr; + long lines; + + lines = SubMimeHeaders->content_length / 80; + pptr = ptr + SubMimeHeaders->content_length - lines - 10; + if (pptr < content_end) + ptr = pptr; + } + + next_boundary = strstr(ptr, m->b[startary].Key); + *content_end = tmp; } + else { + char *srch; + /** + * ok, if we have a content length of the mime part, + * try skipping the content on the search for the next + * boundary. since we don't trust the content_length + * to be all accurate, start searching a little before.. + */ + + if ((SubMimeHeaders->content_length != -1) && + (SubMimeHeaders->content_length > 10)) + { + char *pptr; + pptr = ptr + SubMimeHeaders->content_length - 10; + if (pptr < content_end) + ptr = pptr; + } + + + srch = next_boundary = NULL; + for (srch = memchr(ptr, '-', content_end - ptr); + (srch != NULL) && (srch < content_end); + srch = memchr(srch, '-', content_end - srch)) + { + if (!memcmp(srch, + m->b[startary].Key, + m->b[startary].len)) + { + next_boundary = srch; + srch = content_end; + } + else srch ++; + + } + + } + return next_boundary; +} + +/* + * Break out the components of a multipart message + * (This function expects to be fed HEADERS + CONTENT) + * Note: NULL can be supplied as content_end; in this case, the message is + * considered to have ended when the parser encounters a 0x00 byte. + */ +static void recurseable_mime_parser(char *partnum, + char *content_start, char *content_end, + MimeParserCallBackType CallBack, + MimeParserCallBackType PreMultiPartCallBack, + MimeParserCallBackType PostMultiPartCallBack, + void *userdata, + int dont_decode, + interesting_mime_headers *m) +{ + interesting_mime_headers *SubMimeHeaders; + char *ptr; + char *part_start; + char *part_end = NULL; + char *evaluate_crlf_ptr = NULL; + char *next_boundary; + char nested_partnum[256]; + int crlf_in_use = 0; + int part_seq = 0; + CBufStr *chosen_name; + /* If this is a multipart message, then recursively process it */ + ptr = content_start; part_start = NULL; - if (is_multipart) { + if (m->is_multipart) { /* Tell the client about this message's multipartedness */ if (PreMultiPartCallBack != NULL) { - PreMultiPartCallBack("", "", partnum, "", - NULL, content_type, charset, - 0, encoding, id, userdata); + PreMultiPartCallBack("", + "", + partnum, + "", + NULL, + m->b[content_type].Key, + m->b[charset].Key, + 0, + m->b[encoding].Key, + m->b[id].Key, + userdata); } /* Figure out where the boundaries are */ - snprintf(startary, SIZ, "--%s", boundary); - snprintf(endary, SIZ, "--%s--", boundary); - startary_len = strlen(startary); + m->b[startary].len = snprintf(m->b[startary].Key, SIZ, "--%s", m->b[boundary].Key); + SubMimeHeaders = InitInterestingMimes (); + + while ((*ptr == '\r') || (*ptr == '\n')) ptr ++; + + if (strncmp(ptr, m->b[startary].Key, m->b[startary].len) == 0) + ptr += m->b[startary].len; + + while ((*ptr == '\r') || (*ptr == '\n')) ptr ++; part_start = NULL; do { - next_boundary = NULL; - for (srch=ptr; srch 2) && + (*(ptr - 2) == '\r')) + crlf_in_use = 1; + + part_start = ptr; + + next_boundary = FindNextContent(ptr, + content_end, + SubMimeHeaders, + m); + if ((next_boundary != NULL) && + (next_boundary - part_start < 3)) { + FlushInterestingMimes(SubMimeHeaders); + + continue; } if ( (part_start != NULL) && (next_boundary != NULL) ) { @@ -481,19 +675,22 @@ void the_mime_parser(char *partnum, sizeof nested_partnum, "%d", ++part_seq); } - the_mime_parser(nested_partnum, - part_start, part_end, - CallBack, - PreMultiPartCallBack, - PostMultiPartCallBack, - userdata, - dont_decode); + recurseable_mime_parser(nested_partnum, + part_start, + part_end, + CallBack, + PreMultiPartCallBack, + PostMultiPartCallBack, + userdata, + dont_decode, + SubMimeHeaders); } if (next_boundary != NULL) { /* If we pass out of scope, don't attempt to * read past the end boundary. */ - if (!strcmp(next_boundary, endary)) { + if ((*(next_boundary + m->b[startary].len) == '-') && + (*(next_boundary + m->b[startary].len + 1) == '-') ){ ptr = content_end; } else { @@ -503,7 +700,9 @@ void the_mime_parser(char *partnum, /* Determine whether newlines are LF or CRLF */ evaluate_crlf_ptr = part_start; --evaluate_crlf_ptr; - if (!memcmp(evaluate_crlf_ptr, "\r\n", 2)) { + if ((*evaluate_crlf_ptr == '\r') && + (*(evaluate_crlf_ptr + 1) == '\n')) + { crlf_in_use = 1; } else { @@ -519,24 +718,30 @@ void the_mime_parser(char *partnum, /* Invalid end of multipart. Bail out! */ ptr = content_end; } + FlushInterestingMimes(SubMimeHeaders); } while ( (ptr < content_end) && (next_boundary != NULL) ); + free(SubMimeHeaders); + if (PostMultiPartCallBack != NULL) { - PostMultiPartCallBack("", "", partnum, "", NULL, - content_type, charset, 0, encoding, id, userdata); + PostMultiPartCallBack("", + "", + partnum, + "", + NULL, + m->b[content_type].Key, + m->b[charset].Key, + 0, + m->b[encoding].Key, + m->b[id].Key, + userdata); } - goto end_parser; - } - - /* If it's not a multipart message, then do something with it */ - if (!is_multipart) { + } /* If it's not a multipart message, then do something with it */ + else { + size_t length; part_start = ptr; - length = 0; - while (ptr < content_end) { - ++ptr; - ++length; - } - part_end = content_end; + length = content_end - part_start; + ptr = part_end = content_end; /* The following code will truncate the MIME part to the size @@ -552,32 +757,48 @@ void the_mime_parser(char *partnum, * and sometimes it's tacked on to Content-disposition. Use * whichever one we have. */ - if (strlen(content_disposition_name) > strlen(content_type_name)) { - name = content_disposition_name; + if (m->b[content_disposition_name].len > m->b[content_type_name].len) { + chosen_name = &m->b[content_disposition_name]; } else { - name = content_type_name; + chosen_name = &m->b[content_type_name]; } /* Ok, we've got a non-multipart part here, so do something with it. */ mime_decode(partnum, - part_start, length, - content_type, charset, encoding, disposition, id, - name, filename, - CallBack, NULL, NULL, - userdata, dont_decode - ); + part_start, + length, + m->b[content_type].Key, + m->b[charset].Key, + m->b[encoding].Key, + m->b[disposition].Key, + m->b[id].Key, + chosen_name->Key, + m->b[filename].Key, + CallBack, + NULL, NULL, + userdata, + dont_decode + ); /* * Now if it's an encapsulated message/rfc822 then we have to recurse into it */ - if (!strcasecmp(content_type, "message/rfc822")) { + if (!strcasecmp(&m->b[content_type].Key[0], "message/rfc822")) { if (PreMultiPartCallBack != NULL) { - PreMultiPartCallBack("", "", partnum, "", - NULL, content_type, charset, - 0, encoding, id, userdata); + PreMultiPartCallBack("", + "", + partnum, + "", + NULL, + m->b[content_type].Key, + m->b[charset].Key, + 0, + m->b[encoding].Key, + m->b[id].Key, + userdata); } if (CallBack != NULL) { if (strlen(partnum) > 0) { @@ -592,17 +813,27 @@ void the_mime_parser(char *partnum, "%d", ++part_seq); } the_mime_parser(nested_partnum, - part_start, part_end, - CallBack, - PreMultiPartCallBack, - PostMultiPartCallBack, - userdata, - dont_decode - ); + part_start, + part_end, + CallBack, + PreMultiPartCallBack, + PostMultiPartCallBack, + userdata, + dont_decode + ); } if (PostMultiPartCallBack != NULL) { - PostMultiPartCallBack("", "", partnum, "", NULL, - content_type, charset, 0, encoding, id, userdata); + PostMultiPartCallBack("", + "", + partnum, + "", + NULL, + m->b[content_type].Key, + m->b[charset].Key, + 0, + m->b[encoding].Key, + m->b[id].Key, + userdata); } @@ -610,22 +841,45 @@ void the_mime_parser(char *partnum, } -end_parser: /* free the buffers! end the oppression!! */ - free(boundary); - free(startary); - free(endary); - free(header); - free(content_type); - free(charset); - free(encoding); - free(content_type_name); - free(content_disposition_name); - free(filename); - free(disposition); - free(id); } +/* + * Break out the components of a multipart message + * (This function expects to be fed HEADERS + CONTENT) + * Note: NULL can be supplied as content_end; in this case, the message is + * considered to have ended when the parser encounters a 0x00 byte. + */ +void the_mime_parser(char *partnum, + char *content_start, char *content_end, + MimeParserCallBackType CallBack, + MimeParserCallBackType PreMultiPartCallBack, + MimeParserCallBackType PostMultiPartCallBack, + void *userdata, + int dont_decode) +{ + interesting_mime_headers *m; + + /* If the caller didn't supply an endpointer, generate one by measure */ + if (content_end == NULL) { + content_end = &content_start[strlen(content_start)]; + } + + m = InitInterestingMimes(); + if (!parse_MimeHeaders(m, &content_start, content_end)) + { + + recurseable_mime_parser(partnum, + content_start, content_end, + CallBack, + PreMultiPartCallBack, + PostMultiPartCallBack, + userdata, + dont_decode, + m); + } + free(m); +} /* * Entry point for the MIME parser. @@ -634,50 +888,12 @@ end_parser: /* free the buffers! end the oppression!! */ * considered to have ended when the parser encounters a 0x00 byte. */ void mime_parser(char *content_start, - char *content_end, - - void (*CallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - - void (*PreMultiPartCallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - - void (*PostMultiPartCallBack) - (char *cbname, - char *cbfilename, - char *cbpartnum, - char *cbdisp, - void *cbcontent, - char *cbtype, - char *cbcharset, - size_t cblength, - char *cbencoding, - char *cbid, - void *cbuserdata), - - void *userdata, - int dont_decode -) + char *content_end, + MimeParserCallBackType CallBack, + MimeParserCallBackType PreMultiPartCallBack, + MimeParserCallBackType PostMultiPartCallBack, + void *userdata, + int dont_decode) { the_mime_parser("", content_start, content_end, @@ -762,6 +978,8 @@ const char* GuessMimeByFilename(const char *what, size_t len) return "text/plain"; else if ((len > 3) && !strncasecmp(&what[len - 4], ".css", 4)) return "text/css"; + else if ((len > 3) && !strncasecmp(&what[len - 4], ".htc", 4)) + return "text/x-component"; else if ((len > 3) && !strncasecmp(&what[len - 4], ".jpg", 4)) return "image/jpeg"; else if ((len > 3) && !strncasecmp(&what[len - 4], ".png", 4)) @@ -839,8 +1057,8 @@ int LoadIconDir(const char *DirName) { char *MinorPtr; char *PStart; -#ifdef _DIRENT_HAVE_D_NAMELEN - d_namelen = filedir_entry->d_namelen; +#ifdef _DIRENT_HAVE_D_NAMLEN + d_namelen = filedir_entry->d_namlen; #else d_namelen = strlen(filedir_entry->d_name); #endif