Implement S3BinaryCacheStore::queryAllValidPaths()

This allows commands like "nix verify --all" or "nix path-info --all"
to work on S3 caches.

Unfortunately, this requires some ugly hackery: when querying the
contents of the bucket, we don't want to have to read every .narinfo
file. But the S3 bucket keys only include the hash part of each store
path, not the name part. So as a special exception
queryAllValidPaths() can now return store paths *without* the name
part, and queryPathInfo() accepts such store paths (returning a
ValidPathInfo object containing the full name).

9 files changed, 113 insertions, 54 deletions

diff --git a/src/libstore/binary-cache-store.cc b/src/libstore/binary-cache-store.cc
index dacdeb3f52fd..4445bfd27a12 100644
--- a/src/libstore/binary-cache-store.cc
+++ b/src/libstore/binary-cache-store.cc
@@ -94,13 +94,15 @@ void BinaryCacheStore::addToCache(const ValidPathInfo & info,
 
     upsertFile(narInfoFile, narInfo->to_string());
 
+    auto hashPart = storePathToHash(narInfo->path);
+
     {
         auto state_(state.lock());
-        state_->pathInfoCache.upsert(narInfo->path, std::shared_ptr<NarInfo>(narInfo));
+        state_->pathInfoCache.upsert(hashPart, std::shared_ptr<NarInfo>(narInfo));
     }
 
     if (diskCache)
-        diskCache->upsertNarInfo(getUri(), std::shared_ptr<NarInfo>(narInfo));
+        diskCache->upsertNarInfo(getUri(), hashPart, std::shared_ptr<NarInfo>(narInfo));
 
     stats.narInfoWrite++;
 }
@@ -197,8 +199,6 @@ std::shared_ptr<ValidPathInfo> BinaryCacheStore::queryPathInfoUncached(const Pat
     if (!data) return 0;
 
     auto narInfo = make_ref<NarInfo>(*data, narInfoFile);
-    if (narInfo->path != storePath)
-        throw Error(format("NAR info file for store path ‘%1%’ does not match ‘%2%’") % narInfo->path % storePath);
 
     stats.narInfoRead++;
 
diff --git a/src/libstore/local-store.cc b/src/libstore/local-store.cc
index 9bc164e19f93..baf3a528b705 100644
--- a/src/libstore/local-store.cc
+++ b/src/libstore/local-store.cc
@@ -579,7 +579,7 @@ uint64_t LocalStore::addValidPath(State & state,
 
     {
         auto state_(Store::state.lock());
-        state_->pathInfoCache.upsert(info.path, std::make_shared<ValidPathInfo>(info));
+        state_->pathInfoCache.upsert(storePathToHash(info.path), std::make_shared<ValidPathInfo>(info));
     }
 
     return id;
@@ -1067,7 +1067,7 @@ void LocalStore::invalidatePath(State & state, const Path & path)
 
     {
         auto state_(Store::state.lock());
-        state_->pathInfoCache.erase(path);
+        state_->pathInfoCache.erase(storePathToHash(path));
     }
 }
 
diff --git a/src/libstore/nar-info-disk-cache.cc b/src/libstore/nar-info-disk-cache.cc
index 30ef7b36c9dd..d8b0815bf757 100644
--- a/src/libstore/nar-info-disk-cache.cc
+++ b/src/libstore/nar-info-disk-cache.cc
@@ -20,7 +20,8 @@ create table if not exists BinaryCaches (
 
 create table if not exists NARs (
     cache            integer not null,
-    storePath        text not null,
+    hashPart         text not null,
+    namePart         text not null,
     url              text,
     compression      text,
     fileHash         text,
@@ -31,7 +32,7 @@ create table if not exists NARs (
     deriver          text,
     sigs             text,
     timestamp        integer not null,
-    primary key (cache, storePath),
+    primary key (cache, hashPart),
     foreign key (cache) references BinaryCaches(id) on delete cascade
 );
 
@@ -66,7 +67,7 @@ public:
     {
         auto state(_state.lock());
 
-        Path dbPath = getCacheDir() + "/nix/binary-cache-v3.sqlite";
+        Path dbPath = getCacheDir() + "/nix/binary-cache-v4.sqlite";
         createDirs(dirOf(dbPath));
 
         if (sqlite3_open_v2(dbPath.c_str(), &state->db.db,
@@ -92,11 +93,11 @@ public:
             "select id, storeDir, wantMassQuery, priority from BinaryCaches where url = ?");
 
         state->insertNAR.create(state->db,
-            "insert or replace into NARs(cache, storePath, url, compression, fileHash, fileSize, narHash, "
-            "narSize, refs, deriver, sigs, timestamp) values (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)");
+            "insert or replace into NARs(cache, hashPart, namePart, url, compression, fileHash, fileSize, narHash, "
+            "narSize, refs, deriver, sigs, timestamp) values (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)");
 
         state->queryNAR.create(state->db,
-            "select * from NARs where cache = ? and storePath = ?");
+            "select * from NARs where cache = ? and hashPart = ?");
 
         state->insertNARExistence.create(state->db,
             "insert or replace into NARExistence(cache, storePath, exist, timestamp) values (?, ?, ?, ?)");
@@ -141,13 +142,13 @@ public:
     }
 
     std::pair<Outcome, std::shared_ptr<NarInfo>> lookupNarInfo(
-        const std::string & uri, const Path & storePath) override
+        const std::string & uri, const std::string & hashPart) override
     {
         auto state(_state.lock());
 
         auto queryNAR(state->queryNAR.use()
             (uriToInt(*state, uri))
-            (baseNameOf(storePath)));
+            (hashPart));
 
         if (!queryNAR.next())
             // FIXME: check NARExistence
@@ -157,26 +158,29 @@ public:
 
         // FIXME: implement TTL.
 
-        narInfo->path = storePath;
-        narInfo->url = queryNAR.getStr(2);
-        narInfo->compression = queryNAR.getStr(3);
-        if (!queryNAR.isNull(4))
-            narInfo->fileHash = parseHash(queryNAR.getStr(4));
-        narInfo->fileSize = queryNAR.getInt(5);
-        narInfo->narHash = parseHash(queryNAR.getStr(6));
-        narInfo->narSize = queryNAR.getInt(7);
-        for (auto & r : tokenizeString<Strings>(queryNAR.getStr(8), " "))
+        auto namePart = queryNAR.getStr(2);
+        narInfo->path = settings.nixStore + "/" +
+            hashPart + (namePart.empty() ? "" : "-" + namePart);
+        narInfo->url = queryNAR.getStr(3);
+        narInfo->compression = queryNAR.getStr(4);
+        if (!queryNAR.isNull(5))
+            narInfo->fileHash = parseHash(queryNAR.getStr(5));
+        narInfo->fileSize = queryNAR.getInt(6);
+        narInfo->narHash = parseHash(queryNAR.getStr(7));
+        narInfo->narSize = queryNAR.getInt(8);
+        for (auto & r : tokenizeString<Strings>(queryNAR.getStr(9), " "))
             narInfo->references.insert(settings.nixStore + "/" + r);
-        if (!queryNAR.isNull(9))
-            narInfo->deriver = settings.nixStore + "/" + queryNAR.getStr(9);
-        for (auto & sig : tokenizeString<Strings>(queryNAR.getStr(10), " "))
+        if (!queryNAR.isNull(10))
+            narInfo->deriver = settings.nixStore + "/" + queryNAR.getStr(10);
+        for (auto & sig : tokenizeString<Strings>(queryNAR.getStr(11), " "))
             narInfo->sigs.insert(sig);
 
         return {oValid, narInfo};
     }
 
     void upsertNarInfo(
-        const std::string & uri, std::shared_ptr<ValidPathInfo> info) override
+        const std::string & uri, const std::string & hashPart,
+        std::shared_ptr<ValidPathInfo> info) override
     {
         auto state(_state.lock());
 
@@ -184,9 +188,12 @@ public:
 
             auto narInfo = std::dynamic_pointer_cast<NarInfo>(info);
 
+            assert(hashPart == storePathToHash(info->path));
+
             state->insertNAR.use()
                 (uriToInt(*state, uri))
-                (baseNameOf(info->path))
+                (hashPart)
+                (storePathToName(info->path))
                 (narInfo ? narInfo->url : "", narInfo != 0)
                 (narInfo ? narInfo->compression : "", narInfo != 0)
                 (narInfo && narInfo->fileHash ? narInfo->fileHash.to_string() : "", narInfo && narInfo->fileHash)
diff --git a/src/libstore/nar-info-disk-cache.hh b/src/libstore/nar-info-disk-cache.hh
index ee1aafc63a5d..f4e3fbbdcbdc 100644
--- a/src/libstore/nar-info-disk-cache.hh
+++ b/src/libstore/nar-info-disk-cache.hh
@@ -15,10 +15,11 @@ public:
     virtual bool cacheExists(const std::string & uri) = 0;
 
     virtual std::pair<Outcome, std::shared_ptr<NarInfo>> lookupNarInfo(
-        const std::string & uri, const Path & storePath) = 0;
+        const std::string & uri, const std::string & hashPart) = 0;
 
     virtual void upsertNarInfo(
-        const std::string & uri, std::shared_ptr<ValidPathInfo> narInfo) = 0;
+        const std::string & uri, const std::string & hashPart,
+        std::shared_ptr<ValidPathInfo> info) = 0;
 };
 
 /* Return a singleton cache object that can be used concurrently by
diff --git a/src/libstore/s3-binary-cache-store.cc b/src/libstore/s3-binary-cache-store.cc
index 9ac79cf416de..6fadf7be28c9 100644
--- a/src/libstore/s3-binary-cache-store.cc
+++ b/src/libstore/s3-binary-cache-store.cc
@@ -10,6 +10,7 @@
 #include <aws/s3/model/GetObjectRequest.h>
 #include <aws/s3/model/HeadObjectRequest.h>
 #include <aws/s3/model/PutObjectRequest.h>
+#include <aws/s3/model/ListObjectsRequest.h>
 
 namespace nix {
 
@@ -164,7 +165,7 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
 
     std::shared_ptr<std::string> getFile(const std::string & path)
     {
-        printMsg(lvlDebug, format("fetching ‘s3://%1%/%2%’...") % bucketName % path);
+        debug(format("fetching ‘s3://%1%/%2%’...") % bucketName % path);
 
         auto request =
             Aws::S3::Model::GetObjectRequest()
@@ -204,6 +205,38 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
         }
     }
 
+    PathSet queryAllValidPaths() override
+    {
+        PathSet paths;
+        std::string marker;
+
+        do {
+            debug(format("listing bucket ‘s3://%s’ from key ‘%s’...") % bucketName % marker);
+
+            auto res = checkAws(format("AWS error listing bucket ‘%s’") % bucketName,
+                client->ListObjects(
+                    Aws::S3::Model::ListObjectsRequest()
+                    .WithBucket(bucketName)
+                    .WithDelimiter("/")
+                    .WithMarker(marker)));
+
+            auto & contents = res.GetContents();
+
+            debug(format("got %d keys, next marker ‘%s’")
+                % contents.size() % res.GetNextMarker());
+
+            for (auto object : contents) {
+                auto & key = object.GetKey();
+                if (!hasSuffix(key, ".narinfo")) continue;
+                paths.insert(settings.nixStore + "/" + std::string(key, 0, key.size() - 8));
+            }
+
+            marker = res.GetNextMarker();
+        } while (!marker.empty());
+
+        return paths;
+    }
+
 };
 
 static RegisterStoreImplementation regStore([](const std::string & uri) -> std::shared_ptr<Store> {
diff --git a/src/libstore/store-api.cc b/src/libstore/store-api.cc
index cac137a99d7c..9870e36b5494 100644
--- a/src/libstore/store-api.cc
+++ b/src/libstore/store-api.cc
@@ -63,13 +63,16 @@ Path followLinksToStorePath(const Path & path)
 string storePathToName(const Path & path)
 {
     assertStorePath(path);
-    return string(path, settings.nixStore.size() + storePathHashLen + 2);
+    auto l = settings.nixStore.size() + 1 + storePathHashLen;
+    assert(path.size() >= l);
+    return path.size() == l ? "" : string(path, l + 1);
 }
 
 
 string storePathToHash(const Path & path)
 {
     assertStorePath(path);
+    assert(path.size() >= settings.nixStore.size() + 1 + storePathHashLen);
     return string(path, settings.nixStore.size() + 1, storePathHashLen);
 }
 
@@ -234,9 +237,11 @@ std::string Store::getUri()
 
 bool Store::isValidPath(const Path & storePath)
 {
+    auto hashPart = storePathToHash(storePath);
+
     {
         auto state_(state.lock());
-        auto res = state_->pathInfoCache.get(storePath);
+        auto res = state_->pathInfoCache.get(hashPart);
         if (res) {
             stats.narInfoReadAverted++;
             return *res != 0;
@@ -244,10 +249,11 @@ bool Store::isValidPath(const Path & storePath)
     }
 
     if (diskCache) {
-        auto res = diskCache->lookupNarInfo(getUri(), storePath);
+        auto res = diskCache->lookupNarInfo(getUri(), hashPart);
         if (res.first != NarInfoDiskCache::oUnknown) {
+            stats.narInfoReadAverted++;
             auto state_(state.lock());
-            state_->pathInfoCache.upsert(storePath,
+            state_->pathInfoCache.upsert(hashPart,
                 res.first == NarInfoDiskCache::oInvalid ? 0 : res.second);
             return res.first == NarInfoDiskCache::oValid;
         }
@@ -261,9 +267,11 @@ bool Store::isValidPath(const Path & storePath)
 
 ref<const ValidPathInfo> Store::queryPathInfo(const Path & storePath)
 {
+    auto hashPart = storePathToHash(storePath);
+
     {
         auto state_(state.lock());
-        auto res = state_->pathInfoCache.get(storePath);
+        auto res = state_->pathInfoCache.get(hashPart);
         if (res) {
             stats.narInfoReadAverted++;
             if (!*res)
@@ -273,12 +281,14 @@ ref<const ValidPathInfo> Store::queryPathInfo(const Path & storePath)
     }
 
     if (diskCache) {
-        auto res = diskCache->lookupNarInfo(getUri(), storePath);
+        auto res = diskCache->lookupNarInfo(getUri(), hashPart);
         if (res.first != NarInfoDiskCache::oUnknown) {
+            stats.narInfoReadAverted++;
             auto state_(state.lock());
-            state_->pathInfoCache.upsert(storePath,
+            state_->pathInfoCache.upsert(hashPart,
                 res.first == NarInfoDiskCache::oInvalid ? 0 : res.second);
-            if (res.first == NarInfoDiskCache::oInvalid)
+            if (res.first == NarInfoDiskCache::oInvalid ||
+                (res.second->path != storePath && storePathToName(storePath) != ""))
                 throw InvalidPath(format("path ‘%s’ is not valid") % storePath);
             return ref<ValidPathInfo>(res.second);
         }
@@ -287,14 +297,16 @@ ref<const ValidPathInfo> Store::queryPathInfo(const Path & storePath)
     auto info = queryPathInfoUncached(storePath);
 
     if (diskCache && info)
-        diskCache->upsertNarInfo(getUri(), info);
+        diskCache->upsertNarInfo(getUri(), hashPart, info);
 
     {
         auto state_(state.lock());
-        state_->pathInfoCache.upsert(storePath, info);
+        state_->pathInfoCache.upsert(hashPart, info);
     }
 
-    if (!info) {
+    if (!info
+        || (info->path != storePath && storePathToName(storePath) != ""))
+    {
         stats.narInfoMissing++;
         throw InvalidPath(format("path ‘%s’ is not valid") % storePath);
     }
diff --git a/src/libstore/store-api.hh b/src/libstore/store-api.hh
index eab6da91ee89..e0cc32296a9b 100644
--- a/src/libstore/store-api.hh
+++ b/src/libstore/store-api.hh
@@ -181,7 +181,7 @@ protected:
 
     struct State
     {
-        LRUCache<Path, std::shared_ptr<ValidPathInfo>> pathInfoCache{64 * 1024};
+        LRUCache<std::string, std::shared_ptr<ValidPathInfo>> pathInfoCache{64 * 1024};
     };
 
     Sync<State> state;
@@ -206,10 +206,15 @@ public:
     /* Query which of the given paths is valid. */
     virtual PathSet queryValidPaths(const PathSet & paths) = 0;
 
-    /* Query the set of all valid paths. */
+    /* Query the set of all valid paths. Note that for some store
+       backends, the name part of store paths may be omitted
+       (i.e. you'll get /nix/store/<hash> rather than
+       /nix/store/<hash>-<name>). Use queryPathInfo() to obtain the
+       full store path. */
     virtual PathSet queryAllValidPaths() = 0;
 
-    /* Query information about a valid path. */
+    /* Query information about a valid path. It is permitted to omit
+       the name part of the store path. */
     ref<const ValidPathInfo> queryPathInfo(const Path & path);
 
 protected:
diff --git a/src/nix/path-info.cc b/src/nix/path-info.cc
index 8497d708c3f0..d144ef082977 100644
--- a/src/nix/path-info.cc
+++ b/src/nix/path-info.cc
@@ -4,6 +4,7 @@
 #include "store-api.hh"
 
 #include <iomanip>
+#include <algorithm>
 
 using namespace nix;
 
@@ -48,14 +49,14 @@ struct CmdPathInfo : StorePathsCommand
         for (auto & storePath : storePaths)
             pathLen = std::max(pathLen, storePath.size());
 
-        for (auto & storePath : storePaths) {
-            if (!store->isValidPath(storePath))
-                throw Error(format("path ‘%s’ is not valid") % storePath);
+        for (auto storePath : storePaths) {
+            auto info = store->queryPathInfo(storePath);
+            storePath = info->path; // FIXME: screws up padding
 
-            std::cout << storePath << std::string(pathLen - storePath.size(), ' ');
+            std::cout << storePath << std::string(std::max(0, (int) pathLen - (int) storePath.size()), ' ');
 
             if (showSize) {
-                std::cout << '\t' << std::setw(11) << store->queryPathInfo(storePath)->narSize;
+                std::cout << '\t' << std::setw(11) << info->narSize;
             }
 
             if (showClosureSize) {
diff --git a/src/nix/verify.cc b/src/nix/verify.cc
index fdbc2b0fde33..3844535e77df 100644
--- a/src/nix/verify.cc
+++ b/src/nix/verify.cc
@@ -98,7 +98,7 @@ struct CmdVerify : StorePathsCommand
                 if (!noContents) {
 
                     HashSink sink(info->narHash.type);
-                    store->narFromPath(storePath, sink);
+                    store->narFromPath(info->path, sink);
 
                     auto hash = sink.finish();
 
@@ -106,7 +106,7 @@ struct CmdVerify : StorePathsCommand
                         corrupted = 1;
                         printMsg(lvlError,
                             format("path ‘%s’ was modified! expected hash ‘%s’, got ‘%s’")
-                            % storePath % printHash(info->narHash) % printHash(hash.first));
+                            % info->path % printHash(info->narHash) % printHash(hash.first));
                     }
 
                 }
@@ -138,7 +138,7 @@ struct CmdVerify : StorePathsCommand
                         for (auto & store2 : substituters) {
                             if (validSigs >= actualSigsNeeded) break;
                             try {
-                                doSigs(store2->queryPathInfo(storePath)->sigs);
+                                doSigs(store2->queryPathInfo(info->path)->sigs);
                             } catch (InvalidPath &) {
                             } catch (Error & e) {
                                 printMsg(lvlError, format(ANSI_RED "error:" ANSI_NORMAL " %s") % e.what());
@@ -151,7 +151,7 @@ struct CmdVerify : StorePathsCommand
 
                     if (!good) {
                         untrusted++;
-                        printMsg(lvlError, format("path ‘%s’ is untrusted") % storePath);
+                        printMsg(lvlError, format("path ‘%s’ is untrusted") % info->path);
                     }
 
                 }