Merge remote-tracking branch 'origin/topic/jsiwek/custom-b64-alphabet'

* origin/topic/jsiwek/custom-b64-alphabet:
  Add decode_base64_custom BiF to allow alternate base64 alphabets.

Simplified the code a little bit.

Closes #670.

src/Base64.cc

@@ -1,14 +1,27 @@
 #include "config.h"
 #include "Base64.h"
 
-static int base64_table[256];
+int Base64Decoder::default_base64_table[256];
+const string Base64Decoder::default_alphabet = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
 
-static void init_base64_table()
+int* Base64Decoder::InitBase64Table(const string& alphabet)
 	{
-	static int table_initialized = 0;
+	assert(alphabet.size() == 64);
 
-	if ( ++table_initialized > 1 )
-		return;
+	static bool default_table_initialized = false;
+
+	if ( alphabet == default_alphabet && default_table_initialized )
+		return default_base64_table;
+
+	int* base64_table = 0;
+
+	if ( alphabet == default_alphabet )
+		{
+		base64_table = default_base64_table;
+		default_table_initialized = true;
+		}
+	else
+		base64_table = new int[256];
 
 	int i;
 	for ( i = 0; i < 256; ++i )
@@ -16,28 +29,36 @@ static void init_base64_table()
 
 	for ( i = 0; i < 26; ++i )
 		{
-		base64_table['A' + i] = i;
-		base64_table['a' + i] = i + 26;
+		base64_table[int(alphabet[0 + i])] = i;
+		base64_table[int(alphabet[26 + i])] = i + 26;
 		}
 
 	for ( i = 0; i < 10; ++i )
-		base64_table['0' + i] = i + 52;
+		base64_table[int(alphabet[52 + i])] = i + 52;
 
 	// Casts to avoid compiler warnings.
-	base64_table[int('+')] = 62;
-	base64_table[int('/')] = 63;
+	base64_table[int(alphabet[62])] = 62;
+	base64_table[int(alphabet[63])] = 63;
 	base64_table[int('=')] = 0;
+
+	return base64_table;
 	}
 
-Base64Decoder::Base64Decoder(Analyzer* arg_analyzer)
+Base64Decoder::Base64Decoder(Analyzer* arg_analyzer, const string& alphabet)
 	{
-	init_base64_table();
+	base64_table = InitBase64Table(alphabet.size() ? alphabet : default_alphabet);
 	base64_group_next = 0;
 	base64_padding = base64_after_padding = 0;
 	errored = 0;
 	analyzer = arg_analyzer;
 	}
 
+Base64Decoder::~Base64Decoder()
+	{
+	if ( base64_table != default_base64_table )
+		delete base64_table;
+	}
+
 int Base64Decoder::Decode(int len, const char* data, int* pblen, char** pbuf)
 	{
 	int blen;
@@ -142,13 +163,21 @@ int Base64Decoder::Done(int* pblen, char** pbuf)
 	return 0;
 	}
 
-BroString* decode_base64(const BroString* s)
+
+BroString* decode_base64(const BroString* s, const BroString* a)
 	{
+	if ( a && a->Len() != 64 )
+		{
+		reporter->Error("base64 decoding alphabet is not 64 characters: %s",
+		                a->CheckString());
+		return 0;
+		}
+
 	int buf_len = int((s->Len() + 3) / 4) * 3 + 1;
 	int rlen2, rlen = buf_len;
 	char* rbuf2, *rbuf = new char[rlen];
 
-	Base64Decoder dec(0);
+	Base64Decoder dec(0, a ? a->CheckString() : "");
 	if ( dec.Decode(s->Len(), (const char*) s->Bytes(), &rlen, &rbuf) == -1 )
 		goto err;
 

src/Base64.h

@@ -13,11 +13,11 @@
 
 class Base64Decoder {
 public:
-	// <analyzer> is used for error reporting, and it should be zero
-	// when the decoder is called by the built-in function
-	// decode_base64().
-	Base64Decoder(Analyzer* analyzer);
-	~Base64Decoder()	{ }
+	// <analyzer> is used for error reporting, and it should be zero when
+	// the decoder is called by the built-in function decode_base64().
+	// Empty alphabet indicates the default base64 alphabet.
+	Base64Decoder(Analyzer* analyzer, const string& alphabet = "");
+	~Base64Decoder();
 
 	// A note on Decode():
 	//
@@ -57,8 +57,13 @@ protected:
 	int base64_after_padding;
 	int errored;	// if true, we encountered an error - skip further processing
 	Analyzer* analyzer;
+	int* base64_table;
+
+	static int* InitBase64Table(const string& alphabet);
+	static int default_base64_table[256];
+	static const string default_alphabet;
 };
 
-BroString* decode_base64(const BroString* s);
+BroString* decode_base64(const BroString* s, const BroString* a = 0);
 
 #endif /* base64_h */

src/bro.bif

@@ -1860,6 +1860,18 @@ function decode_base64%(s: string%): string
 		}
 	%}
 
+function decode_base64_custom%(s: string, a: string%): string
+	%{
+	BroString* t = decode_base64(s->AsString(), a->AsString());
+	if ( t )
+		return new StringVal(t);
+	else
+		{
+		reporter->Error("error in decoding string %s", s->CheckString());
+		return new StringVal("");
+		}
+	%}
+
 %%{
 #include "DCE_RPC.h"
 

testing/btest/Baseline/bifs.decode_base64/out

@@ -0,0 +1,6 @@
+bro
+bro
+bro
+bro
+bro
+bro

testing/btest/bifs/decode_base64.bro

@@ -0,0 +1,14 @@
+# @TEST-EXEC: bro -b %INPUT >out
+# @TEST-EXEC: btest-diff out
+
+global default_alphabet: string = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
+
+global my_alphabet: string = "!#$%&/(),-.:;<>@[]^ `_{|}~abcdefghijklmnopqrstuvwxyz0123456789+?";
+
+print decode_base64("YnJv");
+print decode_base64_custom("YnJv", default_alphabet);
+print decode_base64_custom("}n-v", my_alphabet);
+
+print decode_base64("YnJv");
+print decode_base64_custom("YnJv", default_alphabet);
+print decode_base64_custom("}n-v", my_alphabet);