YACC_OUTPUT="cfg_yacc.c"
LEX_OUTPUT="cfg_lex.c"
-SOURCES="upstream.c cfg_utils.c memcached.c main.c util.c worker.c fstring.c ${LEX_OUTPUT} ${YACC_OUTPUT}"
+SOURCES="upstream.c cfg_utils.c memcached.c main.c util.c worker.c fstring.c mime.c ${LEX_OUTPUT} ${YACC_OUTPUT}"
CFLAGS="$CFLAGS -W -Wall -Wpointer-arith -Wno-unused-parameter"
CFLAGS="$CFLAGS -Wno-unused-function -Wunused-variable -Wno-sign-compare"
CFLAGS="$CFLAGS -DRVERSION=\\\"${VERSION}\\\" -DHASH_COMPAT"
LDFLAGS="$LDFLAGS -L/usr/lib -L${LOCALBASE}/lib"
OPT_FLAGS="-O -pipe -fno-omit-frame-pointer"
-DEPS="cfg_file.h memcached.h util.h main.h upstream.h fstring.h ${LEX_OUTPUT} ${YACC_OUTPUT}"
+DEPS="cfg_file.h memcached.h util.h main.h upstream.h fstring.h mime.h ${LEX_OUTPUT} ${YACC_OUTPUT}"
EXEC=rspamd
USER=postfix
GROUP=postfix
*/
#define fstrfree(x) free((x)->begin); free((x))
+/*
+ * Return specified character
+ */
+#define fstridx(str, pos) *((str)->begin + (pos))
+
#endif
--- /dev/null
+#include <stdlib.h>
+#include <ctype.h>
+
+#include "mime.h"
+#include "fstring.h"
+
+/*
+ * Quoted printable and base64 decoders for mime parser
+ */
+
+static f_str_t *
+base64decode (f_str_t *src)
+{
+ int bits = 0, buf = 0, padding = 0, v;
+ size_t pos;
+ char c;
+ f_str_t *res;
+
+ res = fstralloc (src->len);
+ if (res == NULL) {
+ return NULL;
+ }
+
+ for (pos = 0; pos < src->len; pos ++) {
+ c = *(src->begin + pos);
+ if (c >= 'A' && c <= 'Z') {
+ v = c - 'A';
+ }
+ else if (c >= 'a' && c <= 'z') {
+ v = c - 'a' + 26;
+ }
+ else if (c >= '0' && c <= '9') {
+ v = c - '0' + 52;
+ }
+ else if ('+' == c) {
+ v = 62;
+ }
+ else if ('/' == c) {
+ v = 63;
+ }
+ else if ('=' == c) {
+ padding++;
+ continue;
+ }
+ else {
+ continue;
+ }
+ if (padding) {
+ padding = 0;
+ }
+ buf = buf << 6 | v;
+ bits += 6;
+ if (bits >= 8) {
+ c = 255 & (buf >> (bits - 8));
+ fstrpush (res, c);
+ }
+ }
+
+ return res;
+}
+
+static f_str_t *
+qpdecode (f_str_t *src, short header)
+{
+ f_str_t *res;
+ size_t pos;
+ char c;
+
+ res = fstralloc (src->len);
+ if (res == NULL) {
+ return NULL;
+ }
+
+ for (pos = 0; pos < src->len; pos++) {
+ c = *(src->begin + pos);
+ if (header && '_' == c) {
+ c = 0x20;
+ }
+ else if ('=' == c && pos + 3 <= src->len && isxdigit (fstridx (src, pos + 1)) && isxdigit (fstridx (src, pos + 2))) {
+ if (isdigit (fstridx (src, pos + 2))) {
+ if (isdigit (fstridx (src, pos + 1))) {
+ c = (toupper (fstridx (src, pos + 2)) - '0') | (16 * (fstridx (src, pos + 1) - '0'));
+ }
+ else {
+ c = (toupper (fstridx (src, pos + 2)) - '0') | (16 * (toupper (fstridx (src, pos + 1)) - 'A' + 10));
+ }
+ }
+ else if (isdigit (fstridx (src, pos + 1))) {
+ c = (toupper (fstridx (src, pos + 2)) - 'A' + 10) | (16 * (fstridx (src, pos + 1) - '0'));
+ }
+ else {
+ c = (toupper (fstridx (src, pos + 2)) - 'A' + 10) | (16 * (toupper (fstridx (src, pos + 1)) - 'A' + 10));
+ }
+ pos += 2;
+ }
+ else if ('=' == c && pos + 2 <= src->len && ('\r' == fstridx (src, pos + 1) || '\n' == fstridx (src, pos + 1))) {
+ if ('\r' == fstridx (src, pos + 1)) {
+ if (pos + 3 <= src->len && '\n' == fstridx (src, pos + 2)) {
+ pos ++;
+ }
+ pos ++;
+ }
+ if ('\n' == fstridx (src, pos + 1)) {
+ if (pos + 3 <= src->len && '\r' == fstridx (src, pos + 2)) {
+ pos ++;
+ }
+ pos ++;
+ }
+ continue;
+ }
+ fstrpush (res, c);
+ }
+
+ return res;
+}
+
--- /dev/null
+#ifndef MIME_H
+#define MIME_H
+
+#include "fstring.h"
+#ifndef OWN_QUEUE_H
+#include <sys/queue.h>
+#else
+#include "queue.h"
+#endif
+
+/*
+ * Header types. If we reach 31, we must group the headers we need to
+ * remember at the beginning, or we should use fd_set bit sets.
+ */
+#define HDR_APPARENTLY_TO 1
+#define HDR_BCC 2
+#define HDR_CC 3
+#define HDR_CONTENT_LENGTH 4
+#define HDR_CONTENT_TRANSFER_ENCODING 5
+#define HDR_CONTENT_TYPE 6
+#define HDR_DATE 7
+#define HDR_DELIVERED_TO 8
+#define HDR_ERRORS_TO 9
+#define HDR_FROM 10
+#define HDR_MESSAGE_ID 11
+#define HDR_RECEIVED 12
+#define HDR_REPLY_TO 13
+#define HDR_RESENT_BCC 14
+#define HDR_RESENT_CC 15
+#define HDR_RESENT_DATE 16
+#define HDR_RESENT_FROM 17
+#define HDR_RESENT_MESSAGE_ID 18
+#define HDR_RESENT_REPLY_TO 19
+#define HDR_RESENT_SENDER 20
+#define HDR_RESENT_TO 21
+#define HDR_RETURN_PATH 22
+#define HDR_RETURN_RECEIPT_TO 23
+#define HDR_SENDER 24
+#define HDR_TO 25
+#define HDR_MAIL_FOLLOWUP_TO 26
+#define HDR_CONTENT_DESCRIPTION 27
+#define HDR_CONTENT_DISPOSITION 28
+#define HDR_CONTENT_ID 29
+#define HDR_MIME_VERSION 30
+#define HDR_DISP_NOTIFICATION 31
+
+#define URL_A 1
+#define URL_IMG 2
+
+/*
+ * Headers:
+ * name - header name
+ * value - decoded, translated to utf8 and normalized version
+ * type - type of header in case of known headers
+ */
+typedef struct mime_header_s {
+ f_str_t *name;
+ f_str_t *value;
+ int type;
+ LIST_ENTRY (mime_header_s) next;
+} mime_header_t;
+
+/*
+ * Body part:
+ * data - content of this part, translated to utf, decoded, normalized and deHTMLed
+ * type - content-type of this part
+ * encoding - original encoding of body part
+ */
+typedef struct mime_body_part_s {
+ f_str_t *data;
+ f_str_t *type;
+ f_str_t *encoding;
+ LIST_ENTRY (mime_body_part_s) next;
+} mime_body_part_t;
+
+/*
+ * Image and A urls:
+ * url - normalized and decoded url
+ * caption - decoded caption for this url (if any)
+ * type - image or a references
+ */
+typedef struct mime_url_s {
+ f_str_t *url;
+ f_str_t *caption;
+ int type;
+} mime_url_t;
+
+typedef struct mime_ctx_s {
+ LIST_HEAD (headersl, mime_header_s) headers;
+ LIST_HEAD (bodypartsl, mime_body_part_s) parts;
+ f_str_t *cur_content_type;
+ f_str_t *cur_encoding;
+} mime_ctx_t;
+
+#endif
projects / rspamd.git / commitdiff