Compact serialization for scripts

Special serializers for script which detect common cases and encode
them much more efficiently. 3 special cases are defined:
* Pay to pubkey hash (encoded as 21 bytes)
* Pay to script hash (encoded as 21 bytes)
* Pay to pubkey starting with 0x02, 0x03 or 0x04 (encoded as 33 bytes)

Other scripts up to 121 bytes require 1 byte + script length. Above
that, scripts up to 16505 bytes require 2 bytes + script length.

5 files changed, 213 insertions, 4 deletions

diff --git a/src/key.cpp b/src/key.cpp
index 23f315203e..20114e6bb2 100644
--- a/src/key.cpp
+++ b/src/key.cpp
@@ -120,9 +120,9 @@ err:
     return ret;
 }
 
-void CKey::SetCompressedPubKey()
+void CKey::SetCompressedPubKey(bool fCompressed)
 {
-    EC_KEY_set_conv_form(pkey, POINT_CONVERSION_COMPRESSED);
+    EC_KEY_set_conv_form(pkey, fCompressed ? POINT_CONVERSION_COMPRESSED : POINT_CONVERSION_UNCOMPRESSED);
     fCompressedPubKey = true;
 }
 
diff --git a/src/key.h b/src/key.h
index c98f52ed04..2a2723bbee 100644
--- a/src/key.h
+++ b/src/key.h
@@ -113,9 +113,8 @@ protected:
     bool fSet;
     bool fCompressedPubKey;
 
-    void SetCompressedPubKey();
-
 public:
+    void SetCompressedPubKey(bool fCompressed = true);
 
     void Reset();
 
diff --git a/src/main.h b/src/main.h
index ea71baf902..1af781f45c 100644
--- a/src/main.h
+++ b/src/main.h
@@ -647,6 +647,20 @@ protected:
     const CTxOut& GetOutputFor(const CTxIn& input, const MapPrevTx& inputs) const;
 };
 
+/** wrapper for CTxOut that provides a more compact serialization */
+class CTxOutCompressor
+{
+private:
+    CTxOut &txout;
+public:
+    CTxOutCompressor(CTxOut &txoutIn) : txout(txoutIn) { }
+
+    IMPLEMENT_SERIALIZE(
+        READWRITE(VARINT(txout.nValue));
+        CScriptCompressor cscript(REF(txout.scriptPubKey));
+        READWRITE(cscript);
+    )
+};
 
 
 
diff --git a/src/script.cpp b/src/script.cpp
index 89bc5632bc..a840bb1c0d 100644
--- a/src/script.cpp
+++ b/src/script.cpp
@@ -1951,3 +1951,128 @@ void CScript::SetMultisig(int nRequired, const std::vector<CKey>& keys)
         *this << key.GetPubKey();
     *this << EncodeOP_N(keys.size()) << OP_CHECKMULTISIG;
 }
+
+bool CScriptCompressor::IsToKeyID(CKeyID &hash) const
+{
+    if (script.size() == 25 && script[0] == OP_DUP && script[1] == OP_HASH160 
+                            && script[2] == 20 && script[23] == OP_EQUALVERIFY
+                            && script[24] == OP_CHECKSIG) {
+        memcpy(&hash, &script[3], 20);
+        return true;
+    }
+    return false;
+}
+
+bool CScriptCompressor::IsToScriptID(CScriptID &hash) const
+{
+    if (script.size() == 23 && script[0] == OP_HASH160 && script[1] == 20
+                            && script[22] == OP_EQUAL) {
+        memcpy(&hash, &script[2], 20);
+        return true;
+    }
+    return false;
+}
+
+bool CScriptCompressor::IsToPubKey(std::vector<unsigned char> &pubkey) const
+{
+    if (script.size() == 35 && script[0] == 33 && script[34] == OP_CHECKSIG
+                            && (script[1] == 0x02 || script[1] == 0x03)) {
+        pubkey.resize(33);
+        memcpy(&pubkey[0], &script[1], 33);
+        return true;
+    }
+    if (script.size() == 67 && script[0] == 65 && script[66] == OP_CHECKSIG
+                            && script[1] == 0x04) {
+        pubkey.resize(65);
+        memcpy(&pubkey[0], &script[1], 65);
+        CKey key;
+        return (key.SetPubKey(CPubKey(pubkey))); // SetPubKey fails if this is not a valid public key, a case that would not be compressible
+    }
+    return false;
+}
+
+bool CScriptCompressor::Compress(std::vector<unsigned char> &out) const
+{
+    CKeyID keyID;
+    if (IsToKeyID(keyID)) {
+        out.resize(21);
+        out[0] = 0x00;
+        memcpy(&out[1], &keyID, 20);
+        return true;
+    }
+    CScriptID scriptID;
+    if (IsToScriptID(scriptID)) {
+        out.resize(21);
+        out[0] = 0x01;
+        memcpy(&out[1], &scriptID, 20);
+        return true;
+    }
+    std::vector<unsigned char> pubkey;
+    if (IsToPubKey(pubkey)) {
+        out.resize(33);
+        memcpy(&out[1], &pubkey[1], 32);
+        if (pubkey[0] == 0x02 || pubkey[0] == 0x03) {
+            out[0] = pubkey[0];
+            return true;
+        } else if (pubkey[0] == 0x04) {
+            out[0] = 0x04 | (pubkey[64] & 0x01);
+            return true;
+        }
+    }
+    return false;
+}
+
+unsigned int CScriptCompressor::GetSpecialSize(unsigned int nSize) const
+{
+    if (nSize == 0 || nSize == 1)
+        return 20;
+    if (nSize == 2 || nSize == 3 || nSize == 4 || nSize == 5)
+        return 32;
+    return 0;
+}
+
+bool CScriptCompressor::Decompress(unsigned int nSize, const std::vector<unsigned char> &in)
+{
+    switch(nSize) {
+    case 0x00:
+        script.resize(25);
+        script[0] = OP_DUP;
+        script[1] = OP_HASH160;
+        script[2] = 20;
+        memcpy(&script[3], &in[0], 20);
+        script[23] = OP_EQUALVERIFY;
+        script[24] = OP_CHECKSIG;
+        return true;
+    case 0x01:
+        script.resize(23);
+        script[0] = OP_HASH160;
+        script[1] = 20;
+        memcpy(&script[2], &in[0], 20);
+        script[22] = OP_EQUAL;
+        return true;
+    case 0x02:
+    case 0x03:
+        script.resize(35);
+        script[0] = 33;
+        script[1] = nSize;
+        memcpy(&script[2], &in[0], 32);
+        script[34] = OP_CHECKSIG;
+        return true;
+    case 0x04:
+    case 0x05:
+        std::vector<unsigned char> vch(33, 0x00);
+        vch[0] = nSize - 2;
+        memcpy(&vch[1], &in[0], 32);
+        CKey key;
+        if (!key.SetPubKey(CPubKey(vch)))
+            return false;
+        key.SetCompressedPubKey(false); // Decompress public key
+        CPubKey pubkey = key.GetPubKey();
+        script.resize(67);
+        script[0] = 65;
+        memcpy(&script[1], &pubkey.Raw()[0], 65);
+        script[66] = OP_CHECKSIG;
+        return true;
+    }
+    return false;
+}
diff --git a/src/script.h b/src/script.h
index 51e3b5eb84..e7b52d95e1 100644
--- a/src/script.h
+++ b/src/script.h
@@ -579,7 +579,78 @@ public:
     }
 };
 
+/** Compact serializer for scripts.
+ *
+ *  It detects common cases and encodes them much more efficiently.
+ *  3 special cases are defined:
+ *  * Pay to pubkey hash (encoded as 21 bytes)
+ *  * Pay to script hash (encoded as 21 bytes)
+ *  * Pay to pubkey starting with 0x02, 0x03 or 0x04 (encoded as 33 bytes)
+ *
+ *  Other scripts up to 121 bytes require 1 byte + script length. Above
+ *  that, scripts up to 16505 bytes require 2 bytes + script length.
+ */
+class CScriptCompressor
+{
+private:
+    // make this static for now (there are only 6 special scripts defined)
+    // this can potentially be extended together with a new nVersion for
+    // transactions, in which case this value becomes dependent on nVersion
+    // and nHeight of the enclosing transaction.
+    static const unsigned int nSpecialScripts = 6;
+
+    CScript &script;
+protected:
+    // These check for scripts for which a special case with a shorter encoding is defined.
+    // They are implemented separately from the CScript test, as these test for exact byte
+    // sequence correspondences, and are more strict. For example, IsToPubKey also verifies
+    // whether the public key is valid (as invalid ones cannot be represented in compressed
+    // form).
+    bool IsToKeyID(CKeyID &hash) const;
+    bool IsToScriptID(CScriptID &hash) const;
+    bool IsToPubKey(std::vector<unsigned char> &pubkey) const;
+
+    bool Compress(std::vector<unsigned char> &out) const;
+    unsigned int GetSpecialSize(unsigned int nSize) const;
+    bool Decompress(unsigned int nSize, const std::vector<unsigned char> &out);
+public:
+    CScriptCompressor(CScript &scriptIn) : script(scriptIn) { }
+
+    unsigned int GetSerializeSize(int nType, int nVersion) const {
+        std::vector<unsigned char> compr;
+        if (Compress(compr))
+            return compr.size();
+        unsigned int nSize = script.size() + nSpecialScripts;
+        return script.size() + VARINT(nSize).GetSerializeSize(nType, nVersion);
+    }
 
+    template<typename Stream>
+    void Serialize(Stream &s, int nType, int nVersion) const {
+        std::vector<unsigned char> compr;
+        if (Compress(compr)) {
+            s << CFlatData(&compr[0], &compr[compr.size()]);
+            return;
+        }
+        unsigned int nSize = script.size() + nSpecialScripts;
+        s << VARINT(nSize);
+        s << CFlatData(&script[0], &script[script.size()]);
+    }
+
+    template<typename Stream>
+    void Unserialize(Stream &s, int nType, int nVersion) {
+        unsigned int nSize;
+        s >> VARINT(nSize);
+        if (nSize < nSpecialScripts) {
+            std::vector<unsigned char> vch(GetSpecialSize(nSize), 0x00);
+            s >> REF(CFlatData(&vch[0], &vch[vch.size()]));
+            Decompress(nSize, vch);
+            return;
+        }
+        nSize -= nSpecialScripts;
+        script.resize(nSize);
+        s >> REF(CFlatData(&script[0], &script[script.size()]));
+    }
+};
 
 bool IsCanonicalPubKey(const std::vector<unsigned char> &vchPubKey);
 bool IsCanonicalSignature(const std::vector<unsigned char> &vchSig);