9 files changed, 246 insertions, 442 deletions

diff --git a/src/libexpr/attr-path.cc b/src/libexpr/attr-path.cc
index 4545bfd72..76d101b98 100644
--- a/src/libexpr/attr-path.cc
+++ b/src/libexpr/attr-path.cc
@@ -37,9 +37,6 @@ std::pair<Value *, Pos> findAlongAttrPath(EvalState & state, const string & attr
 {
     Strings tokens = parseAttrPath(attrPath);
 
-    Error attrError =
-        Error(format("attribute selection path '%1%' does not match expression") % attrPath);
-
     Value * v = &vIn;
     Pos pos = noPos;
 
diff --git a/src/libexpr/common-eval-args.cc b/src/libexpr/common-eval-args.cc
index 13950ab8d..82bfeac36 100644
--- a/src/libexpr/common-eval-args.cc
+++ b/src/libexpr/common-eval-args.cc
@@ -3,6 +3,8 @@
 #include "download.hh"
 #include "util.hh"
 #include "eval.hh"
+#include "fetchers.hh"
+#include "store-api.hh"
 
 namespace nix {
 
@@ -46,9 +48,9 @@ Bindings * MixEvalArgs::getAutoArgs(EvalState & state)
 Path lookupFileArg(EvalState & state, string s)
 {
     if (isUri(s)) {
-        CachedDownloadRequest request(s);
-        request.unpack = true;
-        return getDownloader()->downloadCached(state.store, request).path;
+        return state.store->toRealPath(
+            fetchers::downloadTarball(
+                state.store, resolveUri(s), "source", false).storePath);
     } else if (s.size() > 2 && s.at(0) == '<' && s.at(s.size() - 1) == '>') {
         Path p = s.substr(1, s.size() - 2);
         return state.findFile(p);
diff --git a/src/libexpr/local.mk b/src/libexpr/local.mk
index a4ccab376..917e8a1c7 100644
--- a/src/libexpr/local.mk
+++ b/src/libexpr/local.mk
@@ -6,9 +6,9 @@ libexpr_DIR := $(d)
 
 libexpr_SOURCES := $(wildcard $(d)/*.cc) $(wildcard $(d)/primops/*.cc) $(d)/lexer-tab.cc $(d)/parser-tab.cc
 
-libexpr_CXXFLAGS += -I src/libutil -I src/libstore -I src/libmain -I src/libexpr
+libexpr_CXXFLAGS += -I src/libutil -I src/libstore -I src/libfetchers -I src/libmain -I src/libexpr
 
-libexpr_LIBS = libutil libstore libnixrust
+libexpr_LIBS = libutil libstore libfetchers libnixrust
 
 libexpr_LDFLAGS =
 ifneq ($(OS), FreeBSD)
diff --git a/src/libexpr/parser.y b/src/libexpr/parser.y
index 08cec96dc..6f25f5cf0 100644
--- a/src/libexpr/parser.y
+++ b/src/libexpr/parser.y
@@ -545,6 +545,7 @@ formal
 
 #include "eval.hh"
 #include "download.hh"
+#include "fetchers.hh"
 #include "store-api.hh"
 
 
@@ -687,9 +688,8 @@ std::pair<bool, std::string> EvalState::resolveSearchPathElem(const SearchPathEl
 
     if (isUri(elem.second)) {
         try {
-            CachedDownloadRequest request(elem.second);
-            request.unpack = true;
-            res = { true, getDownloader()->downloadCached(store, request).path };
+            res = { true, store->toRealPath(fetchers::downloadTarball(
+                        store, resolveUri(elem.second), "source", false).storePath) };
         } catch (DownloadError & e) {
             printError(format("warning: Nix search path entry '%1%' cannot be downloaded, ignoring") % elem.second);
             res = { false, "" };
diff --git a/src/libexpr/primops.cc b/src/libexpr/primops.cc
index 81c378dff..d86ceb0e3 100644
--- a/src/libexpr/primops.cc
+++ b/src/libexpr/primops.cc
@@ -1,6 +1,5 @@
 #include "archive.hh"
 #include "derivations.hh"
-#include "download.hh"
 #include "eval-inline.hh"
 #include "eval.hh"
 #include "globals.hh"
@@ -2049,68 +2048,6 @@ static void prim_splitVersion(EvalState & state, const Pos & pos, Value * * args
 
 
 /*************************************************************
- * Networking
- *************************************************************/
-
-
-void fetch(EvalState & state, const Pos & pos, Value * * args, Value & v,
-    const string & who, bool unpack, const std::string & defaultName)
-{
-    CachedDownloadRequest request("");
-    request.unpack = unpack;
-    request.name = defaultName;
-
-    state.forceValue(*args[0]);
-
-    if (args[0]->type == tAttrs) {
-
-        state.forceAttrs(*args[0], pos);
-
-        for (auto & attr : *args[0]->attrs) {
-            string n(attr.name);
-            if (n == "url")
-                request.uri = state.forceStringNoCtx(*attr.value, *attr.pos);
-            else if (n == "sha256")
-                request.expectedHash = Hash(state.forceStringNoCtx(*attr.value, *attr.pos), htSHA256);
-            else if (n == "name")
-                request.name = state.forceStringNoCtx(*attr.value, *attr.pos);
-            else
-                throw EvalError(format("unsupported argument '%1%' to '%2%', at %3%") % attr.name % who % attr.pos);
-        }
-
-        if (request.uri.empty())
-            throw EvalError(format("'url' argument required, at %1%") % pos);
-
-    } else
-        request.uri = state.forceStringNoCtx(*args[0], pos);
-
-    state.checkURI(request.uri);
-
-    if (evalSettings.pureEval && !request.expectedHash)
-        throw Error("in pure evaluation mode, '%s' requires a 'sha256' argument", who);
-
-    auto res = getDownloader()->downloadCached(state.store, request);
-
-    if (state.allowedPaths)
-        state.allowedPaths->insert(res.path);
-
-    mkString(v, res.storePath, PathSet({res.storePath}));
-}
-
-
-static void prim_fetchurl(EvalState & state, const Pos & pos, Value * * args, Value & v)
-{
-    fetch(state, pos, args, v, "fetchurl", false, "");
-}
-
-
-static void prim_fetchTarball(EvalState & state, const Pos & pos, Value * * args, Value & v)
-{
-    fetch(state, pos, args, v, "fetchTarball", true, "source");
-}
-
-
-/*************************************************************
  * Primop registration
  *************************************************************/
 
@@ -2292,10 +2229,6 @@ void EvalState::createBaseEnv()
     addPrimOp("derivationStrict", 1, prim_derivationStrict);
     addPrimOp("placeholder", 1, prim_placeholder);
 
-    // Networking
-    addPrimOp("__fetchurl", 1, prim_fetchurl);
-    addPrimOp("fetchTarball", 1, prim_fetchTarball);
-
     /* Add a wrapper around the derivation primop that computes the
        `drvPath' and `outPath' attributes lazily. */
     string path = canonPath(settings.nixDataDir + "/nix/corepkgs/derivation.nix", true);
diff --git a/src/libexpr/primops.hh b/src/libexpr/primops.hh
index c790b30f6..05d0792ef 100644
--- a/src/libexpr/primops.hh
+++ b/src/libexpr/primops.hh
@@ -20,6 +20,7 @@ struct RegisterPrimOp
    them. */
 /* Load a ValueInitializer from a DSO and return whatever it initializes */
 void prim_importNative(EvalState & state, const Pos & pos, Value * * args, Value & v);
+
 /* Execute a program and parse its output */
 void prim_exec(EvalState & state, const Pos & pos, Value * * args, Value & v);
 
diff --git a/src/libexpr/primops/fetchGit.cc b/src/libexpr/primops/fetchGit.cc
index 4aee1073e..1a8798fcc 100644
--- a/src/libexpr/primops/fetchGit.cc
+++ b/src/libexpr/primops/fetchGit.cc
@@ -1,203 +1,19 @@
 #include "primops.hh"
 #include "eval-inline.hh"
-#include "download.hh"
 #include "store-api.hh"
-#include "pathlocks.hh"
 #include "hash.hh"
-#include "tarfile.hh"
-
-#include <sys/time.h>
-
-#include <regex>
-
-#include <nlohmann/json.hpp>
-
-using namespace std::string_literals;
+#include "fetchers.hh"
+#include "url.hh"
 
 namespace nix {
 
-struct GitInfo
-{
-    Path storePath;
-    std::string rev;
-    std::string shortRev;
-    uint64_t revCount = 0;
-};
-
-std::regex revRegex("^[0-9a-fA-F]{40}$");
-
-GitInfo exportGit(ref<Store> store, const std::string & uri,
-    std::optional<std::string> ref, std::string rev,
-    const std::string & name)
-{
-    if (evalSettings.pureEval && rev == "")
-        throw Error("in pure evaluation mode, 'fetchGit' requires a Git revision");
-
-    if (!ref && rev == "" && hasPrefix(uri, "/") && pathExists(uri + "/.git")) {
-
-        bool clean = true;
-
-        try {
-            runProgram("git", true, { "-C", uri, "diff-index", "--quiet", "HEAD", "--" });
-        } catch (ExecError & e) {
-            if (!WIFEXITED(e.status) || WEXITSTATUS(e.status) != 1) throw;
-            clean = false;
-        }
-
-        if (!clean) {
-
-            /* This is an unclean working tree. So copy all tracked files. */
-            GitInfo gitInfo;
-            gitInfo.rev = "0000000000000000000000000000000000000000";
-            gitInfo.shortRev = std::string(gitInfo.rev, 0, 7);
-
-            auto files = tokenizeString<std::set<std::string>>(
-                runProgram("git", true, { "-C", uri, "ls-files", "-z" }), "\0"s);
-
-            PathFilter filter = [&](const Path & p) -> bool {
-                assert(hasPrefix(p, uri));
-                std::string file(p, uri.size() + 1);
-
-                auto st = lstat(p);
-
-                if (S_ISDIR(st.st_mode)) {
-                    auto prefix = file + "/";
-                    auto i = files.lower_bound(prefix);
-                    return i != files.end() && hasPrefix(*i, prefix);
-                }
-
-                return files.count(file);
-            };
-
-            gitInfo.storePath = store->printStorePath(store->addToStore("source", uri, true, htSHA256, filter));
-
-            return gitInfo;
-        }
-
-        // clean working tree, but no ref or rev specified.  Use 'HEAD'.
-        rev = chomp(runProgram("git", true, { "-C", uri, "rev-parse", "HEAD" }));
-        ref = "HEAD"s;
-    }
-
-    if (!ref) ref = "HEAD"s;
-
-    if (rev != "" && !std::regex_match(rev, revRegex))
-        throw Error("invalid Git revision '%s'", rev);
-
-    deletePath(getCacheDir() + "/nix/git");
-
-    Path cacheDir = getCacheDir() + "/nix/gitv2/" + hashString(htSHA256, uri).to_string(Base32, false);
-
-    if (!pathExists(cacheDir)) {
-        createDirs(dirOf(cacheDir));
-        runProgram("git", true, { "init", "--bare", cacheDir });
-    }
-
-    Path localRefFile;
-    if (ref->compare(0, 5, "refs/") == 0)
-        localRefFile = cacheDir + "/" + *ref;
-    else
-        localRefFile = cacheDir + "/refs/heads/" + *ref;
-
-    bool doFetch;
-    time_t now = time(0);
-    /* If a rev was specified, we need to fetch if it's not in the
-       repo. */
-    if (rev != "") {
-        try {
-            runProgram("git", true, { "-C", cacheDir, "cat-file", "-e", rev });
-            doFetch = false;
-        } catch (ExecError & e) {
-            if (WIFEXITED(e.status)) {
-                doFetch = true;
-            } else {
-                throw;
-            }
-        }
-    } else {
-        /* If the local ref is older than ‘tarball-ttl’ seconds, do a
-           git fetch to update the local ref to the remote ref. */
-        struct stat st;
-        doFetch = stat(localRefFile.c_str(), &st) != 0 ||
-            (uint64_t) st.st_mtime + settings.tarballTtl <= (uint64_t) now;
-    }
-    if (doFetch)
-    {
-        Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Git repository '%s'", uri));
-
-        // FIXME: git stderr messes up our progress indicator, so
-        // we're using --quiet for now. Should process its stderr.
-        runProgram("git", true, { "-C", cacheDir, "fetch", "--quiet", "--force", "--", uri, fmt("%s:%s", *ref, *ref) });
-
-        struct timeval times[2];
-        times[0].tv_sec = now;
-        times[0].tv_usec = 0;
-        times[1].tv_sec = now;
-        times[1].tv_usec = 0;
-
-        utimes(localRefFile.c_str(), times);
-    }
-
-    // FIXME: check whether rev is an ancestor of ref.
-    GitInfo gitInfo;
-    gitInfo.rev = rev != "" ? rev : chomp(readFile(localRefFile));
-    gitInfo.shortRev = std::string(gitInfo.rev, 0, 7);
-
-    printTalkative("using revision %s of repo '%s'", gitInfo.rev, uri);
-
-    std::string storeLinkName = hashString(htSHA512, name + std::string("\0"s) + gitInfo.rev).to_string(Base32, false);
-    Path storeLink = cacheDir + "/" + storeLinkName + ".link";
-    PathLocks storeLinkLock({storeLink}, fmt("waiting for lock on '%1%'...", storeLink)); // FIXME: broken
-
-    try {
-        auto json = nlohmann::json::parse(readFile(storeLink));
-
-        assert(json["name"] == name && json["rev"] == gitInfo.rev);
-
-        gitInfo.storePath = json["storePath"];
-
-        if (store->isValidPath(store->parseStorePath(gitInfo.storePath))) {
-            gitInfo.revCount = json["revCount"];
-            return gitInfo;
-        }
-
-    } catch (SysError & e) {
-        if (e.errNo != ENOENT) throw;
-    }
-
-    auto source = sinkToSource([&](Sink & sink) {
-        RunOptions gitOptions("git", { "-C", cacheDir, "archive", gitInfo.rev });
-        gitOptions.standardOut = &sink;
-        runProgram2(gitOptions);
-    });
-
-    Path tmpDir = createTempDir();
-    AutoDelete delTmpDir(tmpDir, true);
-
-    unpackTarfile(*source, tmpDir);
-
-    gitInfo.storePath = store->printStorePath(store->addToStore(name, tmpDir));
-
-    gitInfo.revCount = std::stoull(runProgram("git", true, { "-C", cacheDir, "rev-list", "--count", gitInfo.rev }));
-
-    nlohmann::json json;
-    json["storePath"] = gitInfo.storePath;
-    json["uri"] = uri;
-    json["name"] = name;
-    json["rev"] = gitInfo.rev;
-    json["revCount"] = gitInfo.revCount;
-
-    writeFile(storeLink, json.dump());
-
-    return gitInfo;
-}
-
 static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
     std::string url;
     std::optional<std::string> ref;
-    std::string rev;
+    std::optional<Hash> rev;
     std::string name = "source";
+    bool fetchSubmodules = false;
     PathSet context;
 
     state.forceValue(*args[0]);
@@ -213,9 +29,11 @@ static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Va
             else if (n == "ref")
                 ref = state.forceStringNoCtx(*attr.value, *attr.pos);
             else if (n == "rev")
-                rev = state.forceStringNoCtx(*attr.value, *attr.pos);
+                rev = Hash(state.forceStringNoCtx(*attr.value, *attr.pos), htSHA1);
             else if (n == "name")
                 name = state.forceStringNoCtx(*attr.value, *attr.pos);
+            else if (n == "submodules")
+                fetchSubmodules = state.forceBool(*attr.value, *attr.pos);
             else
                 throw EvalError("unsupported argument '%s' to 'fetchGit', at %s", attr.name, *attr.pos);
         }
@@ -230,17 +48,36 @@ static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Va
     // whitelist. Ah well.
     state.checkURI(url);
 
-    auto gitInfo = exportGit(state.store, url, ref, rev, name);
+    if (evalSettings.pureEval && !rev)
+        throw Error("in pure evaluation mode, 'fetchGit' requires a Git revision");
+
+    fetchers::Attrs attrs;
+    attrs.insert_or_assign("type", "git");
+    attrs.insert_or_assign("url", url.find("://") != std::string::npos ? url : "file://" + url);
+    if (ref) attrs.insert_or_assign("ref", *ref);
+    if (rev) attrs.insert_or_assign("rev", rev->gitRev());
+    if (fetchSubmodules) attrs.insert_or_assign("submodules", true);
+    auto input = fetchers::inputFromAttrs(attrs);
+
+    // FIXME: use name?
+    auto [tree, input2] = input->fetchTree(state.store);
 
     state.mkAttrs(v, 8);
-    mkString(*state.allocAttr(v, state.sOutPath), gitInfo.storePath, PathSet({gitInfo.storePath}));
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), gitInfo.rev);
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), gitInfo.shortRev);
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")), gitInfo.revCount);
+    auto storePath = state.store->printStorePath(tree.storePath);
+    mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
+    // Backward compatibility: set 'rev' to
+    // 0000000000000000000000000000000000000000 for a dirty tree.
+    auto rev2 = input2->getRev().value_or(Hash(htSHA1));
+    mkString(*state.allocAttr(v, state.symbols.create("rev")), rev2.gitRev());
+    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), rev2.gitShortRev());
+    // Backward compatibility: set 'revCount' to 0 for a dirty tree.
+    mkInt(*state.allocAttr(v, state.symbols.create("revCount")),
+        tree.info.revCount.value_or(0));
+    mkBool(*state.allocAttr(v, state.symbols.create("submodules")), fetchSubmodules);
     v.attrs->sort();
 
     if (state.allowedPaths)
-        state.allowedPaths->insert(state.store->toRealPath(gitInfo.storePath));
+        state.allowedPaths->insert(tree.actualPath);
 }
 
 static RegisterPrimOp r("fetchGit", 1, prim_fetchGit);
diff --git a/src/libexpr/primops/fetchMercurial.cc b/src/libexpr/primops/fetchMercurial.cc
index db274fa4f..f18351646 100644
--- a/src/libexpr/primops/fetchMercurial.cc
+++ b/src/libexpr/primops/fetchMercurial.cc
@@ -1,174 +1,18 @@
 #include "primops.hh"
 #include "eval-inline.hh"
-#include "download.hh"
 #include "store-api.hh"
-#include "pathlocks.hh"
-
-#include <sys/time.h>
+#include "fetchers.hh"
+#include "url.hh"
 
 #include <regex>
 
-#include <nlohmann/json.hpp>
-
-using namespace std::string_literals;
-
 namespace nix {
 
-struct HgInfo
-{
-    Path storePath;
-    std::string branch;
-    std::string rev;
-    uint64_t revCount = 0;
-};
-
-std::regex commitHashRegex("^[0-9a-fA-F]{40}$");
-
-HgInfo exportMercurial(ref<Store> store, const std::string & uri,
-    std::string rev, const std::string & name)
-{
-    if (evalSettings.pureEval && rev == "")
-        throw Error("in pure evaluation mode, 'fetchMercurial' requires a Mercurial revision");
-
-    if (rev == "" && hasPrefix(uri, "/") && pathExists(uri + "/.hg")) {
-
-        bool clean = runProgram("hg", true, { "status", "-R", uri, "--modified", "--added", "--removed" }) == "";
-
-        if (!clean) {
-
-            /* This is an unclean working tree. So copy all tracked
-               files. */
-
-            printTalkative("copying unclean Mercurial working tree '%s'", uri);
-
-            HgInfo hgInfo;
-            hgInfo.rev = "0000000000000000000000000000000000000000";
-            hgInfo.branch = chomp(runProgram("hg", true, { "branch", "-R", uri }));
-
-            auto files = tokenizeString<std::set<std::string>>(
-                runProgram("hg", true, { "status", "-R", uri, "--clean", "--modified", "--added", "--no-status", "--print0" }), "\0"s);
-
-            PathFilter filter = [&](const Path & p) -> bool {
-                assert(hasPrefix(p, uri));
-                std::string file(p, uri.size() + 1);
-
-                auto st = lstat(p);
-
-                if (S_ISDIR(st.st_mode)) {
-                    auto prefix = file + "/";
-                    auto i = files.lower_bound(prefix);
-                    return i != files.end() && hasPrefix(*i, prefix);
-                }
-
-                return files.count(file);
-            };
-
-            hgInfo.storePath = store->printStorePath(store->addToStore("source", uri, true, htSHA256, filter));
-
-            return hgInfo;
-        }
-    }
-
-    if (rev == "") rev = "default";
-
-    Path cacheDir = fmt("%s/nix/hg/%s", getCacheDir(), hashString(htSHA256, uri).to_string(Base32, false));
-
-    Path stampFile = fmt("%s/.hg/%s.stamp", cacheDir, hashString(htSHA512, rev).to_string(Base32, false));
-
-    /* If we haven't pulled this repo less than ‘tarball-ttl’ seconds,
-       do so now. */
-    time_t now = time(0);
-    struct stat st;
-    if (stat(stampFile.c_str(), &st) != 0 ||
-        (uint64_t) st.st_mtime + settings.tarballTtl <= (uint64_t) now)
-    {
-        /* Except that if this is a commit hash that we already have,
-           we don't have to pull again. */
-        if (!(std::regex_match(rev, commitHashRegex)
-                && pathExists(cacheDir)
-                && runProgram(
-                    RunOptions("hg", { "log", "-R", cacheDir, "-r", rev, "--template", "1" })
-                    .killStderr(true)).second == "1"))
-        {
-            Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Mercurial repository '%s'", uri));
-
-            if (pathExists(cacheDir)) {
-                try {
-                    runProgram("hg", true, { "pull", "-R", cacheDir, "--", uri });
-                }
-                catch (ExecError & e) {
-                    string transJournal = cacheDir + "/.hg/store/journal";
-                    /* hg throws "abandoned transaction" error only if this file exists */
-                    if (pathExists(transJournal)) {
-                        runProgram("hg", true, { "recover", "-R", cacheDir });
-                        runProgram("hg", true, { "pull", "-R", cacheDir, "--", uri });
-                    } else {
-                        throw ExecError(e.status, fmt("'hg pull' %s", statusToString(e.status)));
-                    }
-                }
-            } else {
-                createDirs(dirOf(cacheDir));
-                runProgram("hg", true, { "clone", "--noupdate", "--", uri, cacheDir });
-            }
-        }
-
-        writeFile(stampFile, "");
-    }
-
-    auto tokens = tokenizeString<std::vector<std::string>>(
-        runProgram("hg", true, { "log", "-R", cacheDir, "-r", rev, "--template", "{node} {rev} {branch}" }));
-    assert(tokens.size() == 3);
-
-    HgInfo hgInfo;
-    hgInfo.rev = tokens[0];
-    hgInfo.revCount = std::stoull(tokens[1]);
-    hgInfo.branch = tokens[2];
-
-    std::string storeLinkName = hashString(htSHA512, name + std::string("\0"s) + hgInfo.rev).to_string(Base32, false);
-    Path storeLink = fmt("%s/.hg/%s.link", cacheDir, storeLinkName);
-
-    try {
-        auto json = nlohmann::json::parse(readFile(storeLink));
-
-        assert(json["name"] == name && json["rev"] == hgInfo.rev);
-
-        hgInfo.storePath = json["storePath"];
-
-        if (store->isValidPath(store->parseStorePath(hgInfo.storePath))) {
-            printTalkative("using cached Mercurial store path '%s'", hgInfo.storePath);
-            return hgInfo;
-        }
-
-    } catch (SysError & e) {
-        if (e.errNo != ENOENT) throw;
-    }
-
-    Path tmpDir = createTempDir();
-    AutoDelete delTmpDir(tmpDir, true);
-
-    runProgram("hg", true, { "archive", "-R", cacheDir, "-r", rev, tmpDir });
-
-    deletePath(tmpDir + "/.hg_archival.txt");
-
-    hgInfo.storePath = store->printStorePath(store->addToStore(name, tmpDir));
-
-    nlohmann::json json;
-    json["storePath"] = hgInfo.storePath;
-    json["uri"] = uri;
-    json["name"] = name;
-    json["branch"] = hgInfo.branch;
-    json["rev"] = hgInfo.rev;
-    json["revCount"] = hgInfo.revCount;
-
-    writeFile(storeLink, json.dump());
-
-    return hgInfo;
-}
-
 static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
     std::string url;
-    std::string rev;
+    std::optional<Hash> rev;
+    std::optional<std::string> ref;
     std::string name = "source";
     PathSet context;
 
@@ -182,8 +26,15 @@ static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * ar
             string n(attr.name);
             if (n == "url")
                 url = state.coerceToString(*attr.pos, *attr.value, context, false, false);
-            else if (n == "rev")
-                rev = state.forceStringNoCtx(*attr.value, *attr.pos);
+            else if (n == "rev") {
+                // Ugly: unlike fetchGit, here the "rev" attribute can
+                // be both a revision or a branch/tag name.
+                auto value = state.forceStringNoCtx(*attr.value, *attr.pos);
+                if (std::regex_match(value, revRegex))
+                    rev = Hash(value, htSHA1);
+                else
+                    ref = value;
+            }
             else if (n == "name")
                 name = state.forceStringNoCtx(*attr.value, *attr.pos);
             else
@@ -200,18 +51,36 @@ static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * ar
     // whitelist. Ah well.
     state.checkURI(url);
 
-    auto hgInfo = exportMercurial(state.store, url, rev, name);
+    if (evalSettings.pureEval && !rev)
+        throw Error("in pure evaluation mode, 'fetchMercurial' requires a Mercurial revision");
+
+    auto parsedUrl = parseURL(
+        url.find("://") != std::string::npos
+        ? "hg+" + url
+        : "hg+file://" + url);
+    if (rev) parsedUrl.query.insert_or_assign("rev", rev->gitRev());
+    if (ref) parsedUrl.query.insert_or_assign("ref", *ref);
+    // FIXME: use name
+    auto input = fetchers::inputFromURL(parsedUrl);
+
+    auto [tree, input2] = input->fetchTree(state.store);
 
     state.mkAttrs(v, 8);
-    mkString(*state.allocAttr(v, state.sOutPath), hgInfo.storePath, PathSet({hgInfo.storePath}));
-    mkString(*state.allocAttr(v, state.symbols.create("branch")), hgInfo.branch);
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), hgInfo.rev);
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), std::string(hgInfo.rev, 0, 12));
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")), hgInfo.revCount);
+    auto storePath = state.store->printStorePath(tree.storePath);
+    mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
+    if (input2->getRef())
+        mkString(*state.allocAttr(v, state.symbols.create("branch")), *input2->getRef());
+    // Backward compatibility: set 'rev' to
+    // 0000000000000000000000000000000000000000 for a dirty tree.
+    auto rev2 = input2->getRev().value_or(Hash(htSHA1));
+    mkString(*state.allocAttr(v, state.symbols.create("rev")), rev2.gitRev());
+    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), std::string(rev2.gitRev(), 0, 12));
+    if (tree.info.revCount)
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *tree.info.revCount);
     v.attrs->sort();
 
     if (state.allowedPaths)
-        state.allowedPaths->insert(state.store->toRealPath(hgInfo.storePath));
+        state.allowedPaths->insert(tree.actualPath);
 }
 
 static RegisterPrimOp r("fetchMercurial", 1, prim_fetchMercurial);
diff --git a/src/libexpr/primops/fetchTree.cc b/src/libexpr/primops/fetchTree.cc
new file mode 100644
index 000000000..9586f71ed
--- /dev/null
+++ b/src/libexpr/primops/fetchTree.cc
@@ -0,0 +1,165 @@
+#include "primops.hh"
+#include "eval-inline.hh"
+#include "store-api.hh"
+#include "fetchers.hh"
+#include "download.hh"
+
+#include <ctime>
+#include <iomanip>
+
+namespace nix {
+
+void emitTreeAttrs(
+    EvalState & state,
+    const fetchers::Tree & tree,
+    std::shared_ptr<const fetchers::Input> input,
+    Value & v)
+{
+    state.mkAttrs(v, 8);
+
+    auto storePath = state.store->printStorePath(tree.storePath);
+
+    mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
+
+    assert(tree.info.narHash);
+    mkString(*state.allocAttr(v, state.symbols.create("narHash")),
+        tree.info.narHash.to_string(SRI));
+
+    if (input->getRev()) {
+        mkString(*state.allocAttr(v, state.symbols.create("rev")), input->getRev()->gitRev());
+        mkString(*state.allocAttr(v, state.symbols.create("shortRev")), input->getRev()->gitShortRev());
+    }
+
+    if (tree.info.revCount)
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *tree.info.revCount);
+
+    if (tree.info.lastModified)
+        mkString(*state.allocAttr(v, state.symbols.create("lastModified")),
+            fmt("%s", std::put_time(std::gmtime(&*tree.info.lastModified), "%Y%m%d%H%M%S")));
+
+    v.attrs->sort();
+}
+
+static void prim_fetchTree(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    settings.requireExperimentalFeature("flakes");
+
+    std::shared_ptr<const fetchers::Input> input;
+    PathSet context;
+
+    state.forceValue(*args[0]);
+
+    if (args[0]->type == tAttrs) {
+        state.forceAttrs(*args[0], pos);
+
+        fetchers::Attrs attrs;
+
+        for (auto & attr : *args[0]->attrs) {
+            state.forceValue(*attr.value);
+            if (attr.value->type == tString)
+                attrs.emplace(attr.name, attr.value->string.s);
+            else if (attr.value->type == tBool)
+                attrs.emplace(attr.name, attr.value->boolean);
+            else
+                throw TypeError("fetchTree argument '%s' is %s while a string or Boolean is expected",
+                    attr.name, showType(*attr.value));
+        }
+
+        if (!attrs.count("type"))
+            throw Error("attribute 'type' is missing in call to 'fetchTree', at %s", pos);
+
+        input = fetchers::inputFromAttrs(attrs);
+    } else
+        input = fetchers::inputFromURL(state.coerceToString(pos, *args[0], context, false, false));
+
+    if (evalSettings.pureEval && !input->isImmutable())
+        throw Error("in pure evaluation mode, 'fetchTree' requires an immutable input");
+
+    // FIXME: use fetchOrSubstituteTree
+    auto [tree, input2] = input->fetchTree(state.store);
+
+    if (state.allowedPaths)
+        state.allowedPaths->insert(tree.actualPath);
+
+    emitTreeAttrs(state, tree, input2, v);
+}
+
+static RegisterPrimOp r("fetchTree", 1, prim_fetchTree);
+
+static void fetch(EvalState & state, const Pos & pos, Value * * args, Value & v,
+    const string & who, bool unpack, std::string name)
+{
+    std::optional<std::string> url;
+    std::optional<Hash> expectedHash;
+
+    state.forceValue(*args[0]);
+
+    if (args[0]->type == tAttrs) {
+
+        state.forceAttrs(*args[0], pos);
+
+        for (auto & attr : *args[0]->attrs) {
+            string n(attr.name);
+            if (n == "url")
+                url = state.forceStringNoCtx(*attr.value, *attr.pos);
+            else if (n == "sha256")
+                expectedHash = Hash(state.forceStringNoCtx(*attr.value, *attr.pos), htSHA256);
+            else if (n == "name")
+                name = state.forceStringNoCtx(*attr.value, *attr.pos);
+            else
+                throw EvalError("unsupported argument '%s' to '%s', at %s",
+                    attr.name, who, attr.pos);
+        }
+
+        if (!url)
+            throw EvalError("'url' argument required, at %s", pos);
+
+    } else
+        url = state.forceStringNoCtx(*args[0], pos);
+
+    url = resolveUri(*url);
+
+    state.checkURI(*url);
+
+    if (name == "")
+        name = baseNameOf(*url);
+
+    if (evalSettings.pureEval && !expectedHash)
+        throw Error("in pure evaluation mode, '%s' requires a 'sha256' argument", who);
+
+    auto storePath =
+        unpack
+        ? fetchers::downloadTarball(state.store, *url, name, (bool) expectedHash).storePath
+        : fetchers::downloadFile(state.store, *url, name, (bool) expectedHash).storePath;
+
+    auto path = state.store->toRealPath(storePath);
+
+    if (expectedHash) {
+        auto hash = unpack
+            ? state.store->queryPathInfo(storePath)->narHash
+            : hashFile(htSHA256, path);
+        if (hash != *expectedHash)
+            throw Error((unsigned int) 102, "hash mismatch in file downloaded from '%s':\n  wanted: %s\n  got:    %s",
+                *url, expectedHash->to_string(), hash.to_string());
+    }
+
+    if (state.allowedPaths)
+        state.allowedPaths->insert(path);
+
+    mkString(v, path, PathSet({path}));
+}
+
+static void prim_fetchurl(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    fetch(state, pos, args, v, "fetchurl", false, "");
+}
+
+static void prim_fetchTarball(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    fetch(state, pos, args, v, "fetchTarball", true, "source");
+}
+
+static RegisterPrimOp r2("__fetchurl", 1, prim_fetchurl);
+static RegisterPrimOp r3("fetchTarball", 1, prim_fetchTarball);
+
+}