util/mime: skip over any invalid char

For certain edge case handling for spaces, spaces were handled
particularly in the remainder processing functions. Make sure that now
that as per RFC 2045, util-base64 would skip over any invalid char, the
edge cases in MIME processor also be handled the same way.

This completes the work done in e46b033.

Ticket 6135
Ticket 6207

diff --git a/src/util-base64.c b/src/util-base64.c
index 2fc560a0f06ce31e9f8593bb878321ce69d2b334..4a4a5d122c4149fcfc7bb94e78c4ac2eee46ea03 100644 (file)
--- a/src/util-base64.c
+++ b/src/util-base64.c
@@ -63,6 +63,21 @@ static inline int GetBase64Value(uint8_t c)
     return val;
 }
 
+/**
+ * \brief Checks if the given char in a byte array is Base64 alphabet
+ *
+ * \param Char that needs to be checked
+ *
+ * \return True if the char was Base64 alphabet, False otherwise
+ */
+bool IsBase64Alphabet(uint8_t encoded_byte)
+{
+    if (GetBase64Value(encoded_byte) < 0 && encoded_byte != '=') {
+        return false;
+    }
+    return true;
+}
+
 /**
  * \brief Decodes a 4-byte base64-encoded block into a 3-byte ascii-encoded block
  *

diff --git a/src/util-base64.h b/src/util-base64.h
index 66e847a1bdc2ef62505814fea4f1a0e4b4476389..53cc14c9c4e0a224fc06f415991822314845c147 100644 (file)
--- a/src/util-base64.h
+++ b/src/util-base64.h
@@ -78,6 +78,7 @@ typedef enum {
 /* Function prototypes */
 Base64Ecode DecodeBase64(uint8_t *dest, uint32_t dest_size, const uint8_t *src, uint32_t len,
         uint32_t *consumed_bytes, uint32_t *decoded_bytes, Base64Mode mode);
+bool IsBase64Alphabet(uint8_t encoded_byte);
 
 #endif
 

diff --git a/src/util-decode-mime.c b/src/util-decode-mime.c
index 141325b56ea3cb441eb222c089963d78005587bf..e13216118a0b4ca46716cfcbbc9ac74b197e98db 100644 (file)
--- a/src/util-decode-mime.c
+++ b/src/util-decode-mime.c
@@ -1184,7 +1184,7 @@ static uint32_t ProcessBase64Remainder(
 
     /* Strip spaces in remainder */
     for (uint8_t i = 0; i < state->bvr_len; i++) {
-        if (state->bvremain[i] != ' ') {
+        if (IsBase64Alphabet(state->bvremain[i])) {
             block[cnt++] = state->bvremain[i];
         }
     }
@@ -1192,7 +1192,7 @@ static uint32_t ProcessBase64Remainder(
     /* if we don't have 4 bytes see if we can fill it from `buf` */
     if (buf && len > 0 && cnt != B64_BLOCK) {
         for (uint32_t i = 0; i < len && cnt < B64_BLOCK; i++) {
-            if (buf[i] != ' ') {
+            if (IsBase64Alphabet(buf[i])) {
                 block[cnt++] = buf[i];
             }
             buf_consumed++;
@@ -1273,7 +1273,8 @@ static inline MimeDecRetCode ProcessBase64BodyLineCopyRemainder(
         return MIME_DEC_ERR_DATA;
 
     for (uint32_t i = offset; i < buf_len; i++) {
-        if (buf[i] != ' ') {
+        // Skip any characters outside of the base64 alphabet as per RFC 2045
+        if (IsBase64Alphabet(buf[i])) {
             DEBUG_VALIDATE_BUG_ON(state->bvr_len >= B64_BLOCK);
             if (state->bvr_len >= B64_BLOCK)
                 return MIME_DEC_ERR_DATA;