Support SRI hashes

SRI hashes (https://www.w3.org/TR/SRI/) combine the hash algorithm and
a base-64 hash. This allows more concise and standard hash
specifications. For example, instead of

  import <nix/fetchurl.nl> {
    url = https://nixos.org/releases/nix/nix-2.1.3/nix-2.1.3.tar.xz;
    sha256 = "5d22dad058d5c800d65a115f919da22938c50dd6ba98c5e3a183172d149840a4";
  };

you can write

  import <nix/fetchurl.nl> {
    url = https://nixos.org/releases/nix/nix-2.1.3/nix-2.1.3.tar.xz;
    hash = "sha256-XSLa0FjVyADWWhFfkZ2iKTjFDda6mMXjoYMXLRSYQKQ=";
  };

In fixed-output derivations, the outputHashAlgo is no longer mandatory
if outputHash specifies the hash (either as an SRI or in the old
"<type>:<hash>" format).

'nix hash-{file,path}' now print hashes in SRI format by default. I
also reverted them to use SHA-256 by default because that's what we're
using most of the time in Nixpkgs.

Suggested by @zimbatm.

5 files changed, 45 insertions, 32 deletions

diff --git a/src/libexpr/primops.cc b/src/libexpr/primops.cc
index 7372134e2c9b..60698f7402e0 100644
--- a/src/libexpr/primops.cc
+++ b/src/libexpr/primops.cc
@@ -724,16 +724,14 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
         if (outputs.size() != 1 || *(outputs.begin()) != "out")
             throw Error(format("multiple outputs are not supported in fixed-output derivations, at %1%") % posDrvName);
 
-        HashType ht = parseHashType(outputHashAlgo);
-        if (ht == htUnknown)
-            throw EvalError(format("unknown hash algorithm '%1%', at %2%") % outputHashAlgo % posDrvName);
+        HashType ht = outputHashAlgo.empty() ? htUnknown : parseHashType(outputHashAlgo);
         Hash h(*outputHash, ht);
-        outputHash = h.to_string(Base16, false);
-        if (outputHashRecursive) outputHashAlgo = "r:" + outputHashAlgo;
 
         Path outPath = state.store->makeFixedOutputPath(outputHashRecursive, h, drvName);
         if (!jsonObject) drv.env["out"] = outPath;
-        drv.outputs["out"] = DerivationOutput(outPath, outputHashAlgo, *outputHash);
+        drv.outputs["out"] = DerivationOutput(outPath,
+            (outputHashRecursive ? "r:" : "") + printHashType(h.type),
+            h.to_string(Base16, false));
     }
 
     else {
diff --git a/src/libutil/hash.cc b/src/libutil/hash.cc
index 9d82f13a5e38..1c14ebb187cc 100644
--- a/src/libutil/hash.cc
+++ b/src/libutil/hash.cc
@@ -105,9 +105,9 @@ string printHash16or32(const Hash & hash)
 std::string Hash::to_string(Base base, bool includeType) const
 {
     std::string s;
-    if (includeType) {
+    if (base == SRI || includeType) {
         s += printHashType(type);
-        s += ':';
+        s += base == SRI ? '-' : ':';
     }
     switch (base) {
     case Base16:
@@ -117,6 +117,7 @@ std::string Hash::to_string(Base base, bool includeType) const
         s += printHash32(*this);
         break;
     case Base64:
+    case SRI:
         s += base64Encode(std::string((const char *) hash, hashSize));
         break;
     }
@@ -127,28 +128,33 @@ std::string Hash::to_string(Base base, bool includeType) const
 Hash::Hash(const std::string & s, HashType type)
     : type(type)
 {
-    auto colon = s.find(':');
-
     size_t pos = 0;
-
-    if (colon == string::npos) {
-        if (type == htUnknown)
+    bool isSRI = false;
+
+    auto sep = s.find(':');
+    if (sep == string::npos) {
+        sep = s.find('-');
+        if (sep != string::npos) {
+            isSRI = true;
+        } else if (type == htUnknown)
             throw BadHash("hash '%s' does not include a type", s);
-    } else {
-        string hts = string(s, 0, colon);
+    }
+
+    if (sep != string::npos) {
+        string hts = string(s, 0, sep);
         this->type = parseHashType(hts);
         if (this->type == htUnknown)
             throw BadHash("unknown hash type '%s'", hts);
         if (type != htUnknown && type != this->type)
             throw BadHash("hash '%s' should have type '%s'", s, printHashType(type));
-        pos = colon + 1;
+        pos = sep + 1;
     }
 
     init();
 
     size_t size = s.size() - pos;
 
-    if (size == base16Len()) {
+    if (!isSRI && size == base16Len()) {
 
         auto parseHexDigit = [&](char c) {
             if (c >= '0' && c <= '9') return c - '0';
@@ -164,7 +170,7 @@ Hash::Hash(const std::string & s, HashType type)
         }
     }
 
-    else if (size == base32Len()) {
+    else if (!isSRI && size == base32Len()) {
 
         for (unsigned int n = 0; n < size; ++n) {
             char c = s[pos + size - n - 1];
@@ -187,10 +193,10 @@ Hash::Hash(const std::string & s, HashType type)
         }
     }
 
-    else if (size == base64Len()) {
+    else if (isSRI || size == base64Len()) {
         auto d = base64Decode(std::string(s, pos));
         if (d.size() != hashSize)
-            throw BadHash("invalid base-64 hash '%s'", s);
+            throw BadHash("invalid %s hash '%s'", isSRI ? "SRI" : "base-64", s);
         assert(hashSize);
         memcpy(hash, d.data(), hashSize);
     }
diff --git a/src/libutil/hash.hh b/src/libutil/hash.hh
index fd7a61df8e46..2dbc3b630814 100644
--- a/src/libutil/hash.hh
+++ b/src/libutil/hash.hh
@@ -20,7 +20,7 @@ const int sha512HashSize = 64;
 
 extern const string base32Chars;
 
-enum Base : int { Base64, Base32, Base16 };
+enum Base : int { Base64, Base32, Base16, SRI };
 
 
 struct Hash
@@ -38,8 +38,9 @@ struct Hash
     Hash(HashType type) : type(type) { init(); };
 
     /* Initialize the hash from a string representation, in the format
-       "[<type>:]<base16|base32|base64>". If the 'type' argument is
-       htUnknown, then the hash type must be specified in the
+       "[<type>:]<base16|base32|base64>" or "<type>-<base64>" (a
+       Subresource Integrity hash expression). If the 'type' argument
+       is htUnknown, then the hash type must be specified in the
        string. */
     Hash(const std::string & s, HashType type = htUnknown);
 
diff --git a/src/nix-store/nix-store.cc b/src/nix-store/nix-store.cc
index a9ad14762e62..5b37237eb165 100644
--- a/src/nix-store/nix-store.cc
+++ b/src/nix-store/nix-store.cc
@@ -1000,6 +1000,9 @@ static int _main(int argc, char * * argv)
         Strings opFlags, opArgs;
         Operation op = 0;
 
+        Hash h("sha512-Q2bFTOhEALkN8hOms2FKTDLy7eugP2zFZ1T8LCvX42Fp3WoNr3bjZSAHeOsHrbV1Fu9/A0EzCinRE7Af1ofPrw==");
+        printError("GOT HASH %s", h.to_string(Base64));
+
         parseCmdLine(argc, argv, [&](Strings::iterator & arg, const Strings::iterator & end) {
             Operation oldOp = op;
 
diff --git a/src/nix/hash.cc b/src/nix/hash.cc
index 64062fb97955..af4105e28904 100644
--- a/src/nix/hash.cc
+++ b/src/nix/hash.cc
@@ -9,13 +9,14 @@ struct CmdHash : Command
 {
     enum Mode { mFile, mPath };
     Mode mode;
-    Base base = Base16;
+    Base base = SRI;
     bool truncate = false;
-    HashType ht = htSHA512;
+    HashType ht = htSHA256;
     std::vector<std::string> paths;
 
     CmdHash(Mode mode) : mode(mode)
     {
+        mkFlag(0, "sri", "print hash in SRI format", &base, SRI);
         mkFlag(0, "base64", "print hash in base-64", &base, Base64);
         mkFlag(0, "base32", "print hash in base-32 (Nix-specific)", &base, Base32);
         mkFlag(0, "base16", "print hash in base-16", &base, Base16);
@@ -43,7 +44,7 @@ struct CmdHash : Command
             Hash h = mode == mFile ? hashFile(ht, path) : hashPath(ht, path).first;
             if (truncate && h.hashSize > 20) h = compressHash(h, 20);
             std::cout << format("%1%\n") %
-                h.to_string(base, false);
+                h.to_string(base, base == SRI);
         }
     }
 };
@@ -54,7 +55,7 @@ static RegisterCommand r2(make_ref<CmdHash>(CmdHash::mPath));
 struct CmdToBase : Command
 {
     Base base;
-    HashType ht = htSHA512;
+    HashType ht = htUnknown;
     std::vector<std::string> args;
 
     CmdToBase(Base base) : base(base)
@@ -70,26 +71,30 @@ struct CmdToBase : Command
         return
             base == Base16 ? "to-base16" :
             base == Base32 ? "to-base32" :
-            "to-base64";
+            base == Base64 ? "to-base64" :
+            "to-sri";
     }
 
     std::string description() override
     {
-        return fmt("convert a hash to base-%d representation",
-            base == Base16 ? 16 :
-            base == Base32 ? 32 : 64);
+        return fmt("convert a hash to %s representation",
+            base == Base16 ? "base-16" :
+            base == Base32 ? "base-32" :
+            base == Base64 ? "base-64" :
+            "SRI");
     }
 
     void run() override
     {
         for (auto s : args)
-            std::cout << fmt("%s\n", Hash(s, ht).to_string(base, false));
+            std::cout << fmt("%s\n", Hash(s, ht).to_string(base, base == SRI));
     }
 };
 
 static RegisterCommand r3(make_ref<CmdToBase>(Base16));
 static RegisterCommand r4(make_ref<CmdToBase>(Base32));
 static RegisterCommand r5(make_ref<CmdToBase>(Base64));
+static RegisterCommand r6(make_ref<CmdToBase>(SRI));
 
 /* Legacy nix-hash command. */
 static int compatNixHash(int argc, char * * argv)