From 995742219107dc9703db030835ea811d4554c497 Mon Sep 17 00:00:00 2001
From: Jens Steube <jens.steube@gmail.com>
Date: Wed, 19 Feb 2020 10:35:44 +0100
Subject: [PATCH] Add tokenizer option TOKEN_ATTR_SEPARATOR_FARTHEST as an
 option to deal with hash formats where separator character could be part of
 the salt data itself and at the same time does not allow escape character
 logic to be applied. This can only work if it is guaranteed by the design of
 the hash format that the separator character does not occur after the
 position of the separator character.

---
 include/shared.h           |  3 +++
 include/types.h            | 21 +++++++++++----------
 src/modules/module_02100.c |  3 ++-
 src/shared.c               | 31 ++++++++++++++++++++++++++++++-
 4 files changed, 46 insertions(+), 12 deletions(-)

diff --git a/include/shared.h b/include/shared.h
index 203a85244..96604c730 100644
--- a/include/shared.h
+++ b/include/shared.h
@@ -79,6 +79,9 @@ u64 round_up_multiple_64 (const u64 v, const u64 m);
 
 void hc_strncat (u8 *dst, const u8 *src, const size_t n);
 
+const u8 *hc_strchr_next (const u8 *input_buf, const int input_len, const u8 separator);
+const u8 *hc_strchr_last (const u8 *input_buf, const int input_len, const u8 separator);
+
 int count_char (const u8 *buf, const int len, const u8 c);
 float get_entropy (const u8 *buf, const int len);
 
diff --git a/include/types.h b/include/types.h
index dfccc083a..9f329ad0d 100644
--- a/include/types.h
+++ b/include/types.h
@@ -770,16 +770,17 @@ typedef enum user_options_map
 
 typedef enum token_attr
 {
-  TOKEN_ATTR_FIXED_LENGTH       = 1 << 0,
-  TOKEN_ATTR_OPTIONAL_ROUNDS    = 1 << 1,
-  TOKEN_ATTR_VERIFY_SIGNATURE   = 1 << 2,
-  TOKEN_ATTR_VERIFY_LENGTH      = 1 << 3,
-  TOKEN_ATTR_VERIFY_DIGIT       = 1 << 4,
-  TOKEN_ATTR_VERIFY_FLOAT       = 1 << 5,
-  TOKEN_ATTR_VERIFY_HEX         = 1 << 6,
-  TOKEN_ATTR_VERIFY_BASE64A     = 1 << 7,
-  TOKEN_ATTR_VERIFY_BASE64B     = 1 << 8,
-  TOKEN_ATTR_VERIFY_BASE64C     = 1 << 9
+  TOKEN_ATTR_FIXED_LENGTH       = 1 <<  0,
+  TOKEN_ATTR_SEPARATOR_FARTHEST = 1 <<  1,
+  TOKEN_ATTR_OPTIONAL_ROUNDS    = 1 <<  2,
+  TOKEN_ATTR_VERIFY_SIGNATURE   = 1 <<  3,
+  TOKEN_ATTR_VERIFY_LENGTH      = 1 <<  4,
+  TOKEN_ATTR_VERIFY_DIGIT       = 1 <<  5,
+  TOKEN_ATTR_VERIFY_FLOAT       = 1 <<  6,
+  TOKEN_ATTR_VERIFY_HEX         = 1 <<  7,
+  TOKEN_ATTR_VERIFY_BASE64A     = 1 <<  8,
+  TOKEN_ATTR_VERIFY_BASE64B     = 1 <<  9,
+  TOKEN_ATTR_VERIFY_BASE64C     = 1 << 10,
 
 } token_attr_t;
 
diff --git a/src/modules/module_02100.c b/src/modules/module_02100.c
index a7ef91744..f2735b482 100644
--- a/src/modules/module_02100.c
+++ b/src/modules/module_02100.c
@@ -95,7 +95,8 @@ int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSE
   token.len_min[2] = SALT_MIN;
   token.len_max[2] = SALT_MAX;
   token.sep[2]     = '#';
-  token.attr[2]    = TOKEN_ATTR_VERIFY_LENGTH;
+  token.attr[2]    = TOKEN_ATTR_VERIFY_LENGTH
+                   | TOKEN_ATTR_SEPARATOR_FARTHEST;
 
   token.len_min[3] = 32;
   token.len_max[3] = 32;
diff --git a/src/shared.c b/src/shared.c
index 0cb1323d3..05387f5e6 100644
--- a/src/shared.c
+++ b/src/shared.c
@@ -1028,6 +1028,26 @@ static int rounds_count_length (const char *input_buf, const int input_len)
   return -1;
 }
 
+const u8 *hc_strchr_next (const u8 *input_buf, const int input_len, const u8 separator)
+{
+  for (int i = 0; i < input_len; i++)
+  {
+    if (input_buf[i] == separator) return &input_buf[i];
+  }
+
+  return NULL;
+}
+
+const u8 *hc_strchr_last (const u8 *input_buf, const int input_len, const u8 separator)
+{
+  for (int i = input_len - 1; i >= 0; i--)
+  {
+    if (input_buf[i] == separator) return &input_buf[i];
+  }
+
+  return NULL;
+}
+
 int input_tokenizer (const u8 *input_buf, const int input_len, token_t *token)
 {
   int len_left = input_len;
@@ -1066,7 +1086,16 @@ int input_tokenizer (const u8 *input_buf, const int input_len, token_t *token)
         }
       }
 
-      const u8 *next_pos = (const u8 *) strchr ((const char *) token->buf[token_idx], token->sep[token_idx]);
+      const u8 *next_pos = NULL;
+
+      if (token->attr[token_idx] & TOKEN_ATTR_SEPARATOR_FARTHEST)
+      {
+        next_pos = hc_strchr_last (token->buf[token_idx], len_left, token->sep[token_idx]);
+      }
+      else
+      {
+        next_pos = hc_strchr_next (token->buf[token_idx], len_left, token->sep[token_idx]);
+      }
 
       if (next_pos == NULL) return (PARSER_SEPARATOR_UNMATCHED);