Backport libfetchers from the flakes branch

This provides a pluggable mechanism for defining new fetchers. It adds
a builtin function 'fetchTree' that generalizes existing fetchers like
'fetchGit', 'fetchMercurial' and 'fetchTarball'. 'fetchTree' takes a
set of attributes, e.g.

  fetchTree {
    type = "git";
    url = "https://example.org/repo.git";
    ref = "some-branch";
    rev = "abcdef...";
  }

The existing fetchers are just wrappers around this. Note that the
input attributes to fetchTree are the same as flake input
specifications and flake lock file entries.

All fetchers share a common cache stored in
~/.cache/nix/fetcher-cache-v1.sqlite. This replaces the ad hoc caching
mechanisms in fetchGit and download.cc (e.g. ~/.cache/nix/{tarballs,git-revs*}).

This also adds support for Git worktrees (c169ea5904).

Makefile

@@ -4,6 +4,7 @@ makefiles = \
   nix-rust/local.mk \
   src/libutil/local.mk \
   src/libstore/local.mk \
+  src/libfetchers/local.mk \
   src/libmain/local.mk \
   src/libexpr/local.mk \
   src/nix/local.mk \

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);
+        return state.store->toRealPath(
-        request.unpack = true;
+            fetchers::downloadTarball(
-        return getDownloader()->downloadCached(state.store, request).path;
+                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);

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)

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);
+            res = { true, store->toRealPath(fetchers::downloadTarball(
-            request.unpack = true;
+                        store, resolveUri(elem.second), "source", false).storePath) };
-            res = { true, getDownloader()->downloadCached(store, request).path };
         } catch (DownloadError & e) {
             printError(format("warning: Nix search path entry '%1%' cannot be downloaded, ignoring") % elem.second);
             res = { false, "" };

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"
@@ -2045,68 +2044,6 @@ 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
  *************************************************************/
@@ -2289,10 +2226,6 @@ 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);

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);
 

src/libexpr/primops/fetchGit.cc

@@ -1,202 +1,17 @@
 #include "primops.hh"
 #include "eval-inline.hh"
-#include "download.hh"
 #include "store-api.hh"
-#include "pathlocks.hh"
 #include "hash.hh"
-#include "tarfile.hh"
+#include "fetchers.hh"
-
+#include "url.hh"
-#include <sys/time.h>
-
-#include <regex>
-
-#include <nlohmann/json.hpp>
-
-using namespace std::string_literals;
 
 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";
     PathSet context;
 
@@ -213,7 +28,7 @@ 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
@@ -230,17 +45,35 @@ 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");
+
+    auto parsedUrl = parseURL(
+        url.find("://") != std::string::npos
+        ? "git+" + url
+        : "git+file://" + url);
+    if (ref) parsedUrl.query.insert_or_assign("ref", *ref);
+    if (rev) parsedUrl.query.insert_or_assign("rev", rev->gitRev());
+    // 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), gitInfo.storePath, PathSet({gitInfo.storePath}));
+    auto storePath = state.store->printStorePath(tree.storePath);
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), gitInfo.rev);
+    mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), gitInfo.shortRev);
+    // Backward compatibility: set 'rev' to
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")), gitInfo.revCount);
+    // 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));
     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);

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 "fetchers.hh"
-
+#include "url.hh"
-#include <sys/time.h>
 
 #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")
+            else if (n == "rev") {
-                rev = state.forceStringNoCtx(*attr.value, *attr.pos);
+                // 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}));
+    auto storePath = state.store->printStorePath(tree.storePath);
-    mkString(*state.allocAttr(v, state.symbols.create("branch")), hgInfo.branch);
+    mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), hgInfo.rev);
+    if (input2->getRef())
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), std::string(hgInfo.rev, 0, 12));
+        mkString(*state.allocAttr(v, state.symbols.create("branch")), *input2->getRef());
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")), hgInfo.revCount);
+    // 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);

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);
+
+}

src/libfetchers/attrs.cc

@@ -0,0 +1,92 @@
+#include "attrs.hh"
+#include "fetchers.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+Attrs jsonToAttrs(const nlohmann::json & json)
+{
+    Attrs attrs;
+
+    for (auto & i : json.items()) {
+        if (i.value().is_number())
+            attrs.emplace(i.key(), i.value().get<int64_t>());
+        else if (i.value().is_string())
+            attrs.emplace(i.key(), i.value().get<std::string>());
+        else if (i.value().is_boolean())
+            attrs.emplace(i.key(), i.value().get<bool>());
+        else
+            throw Error("unsupported input attribute type in lock file");
+    }
+
+    return attrs;
+}
+
+nlohmann::json attrsToJson(const Attrs & attrs)
+{
+    nlohmann::json json;
+    for (auto & attr : attrs) {
+        if (auto v = std::get_if<int64_t>(&attr.second)) {
+            json[attr.first] = *v;
+        } else if (auto v = std::get_if<std::string>(&attr.second)) {
+            json[attr.first] = *v;
+        } else if (auto v = std::get_if<Explicit<bool>>(&attr.second)) {
+            json[attr.first] = v->t;
+        } else abort();
+    }
+    return json;
+}
+
+std::optional<std::string> maybeGetStrAttr(const Attrs & attrs, const std::string & name)
+{
+    auto i = attrs.find(name);
+    if (i == attrs.end()) return {};
+    if (auto v = std::get_if<std::string>(&i->second))
+        return *v;
+    throw Error("input attribute '%s' is not a string %s", name, attrsToJson(attrs).dump());
+}
+
+std::string getStrAttr(const Attrs & attrs, const std::string & name)
+{
+    auto s = maybeGetStrAttr(attrs, name);
+    if (!s)
+        throw Error("input attribute '%s' is missing", name);
+    return *s;
+}
+
+std::optional<int64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name)
+{
+    auto i = attrs.find(name);
+    if (i == attrs.end()) return {};
+    if (auto v = std::get_if<int64_t>(&i->second))
+        return *v;
+    throw Error("input attribute '%s' is not an integer", name);
+}
+
+int64_t getIntAttr(const Attrs & attrs, const std::string & name)
+{
+    auto s = maybeGetIntAttr(attrs, name);
+    if (!s)
+        throw Error("input attribute '%s' is missing", name);
+    return *s;
+}
+
+std::optional<bool> maybeGetBoolAttr(const Attrs & attrs, const std::string & name)
+{
+    auto i = attrs.find(name);
+    if (i == attrs.end()) return {};
+    if (auto v = std::get_if<int64_t>(&i->second))
+        return *v;
+    throw Error("input attribute '%s' is not a Boolean", name);
+}
+
+bool getBoolAttr(const Attrs & attrs, const std::string & name)
+{
+    auto s = maybeGetBoolAttr(attrs, name);
+    if (!s)
+        throw Error("input attribute '%s' is missing", name);
+    return *s;
+}
+
+}

src/libfetchers/attrs.hh

@@ -0,0 +1,37 @@
+#pragma once
+
+#include "types.hh"
+
+#include <variant>
+
+#include <nlohmann/json_fwd.hpp>
+
+namespace nix::fetchers {
+
+/* Wrap bools to prevent string literals (i.e. 'char *') from being
+   cast to a bool in Attr. */
+template<typename T>
+struct Explicit {
+    T t;
+};
+
+typedef std::variant<std::string, int64_t, Explicit<bool>> Attr;
+typedef std::map<std::string, Attr> Attrs;
+
+Attrs jsonToAttrs(const nlohmann::json & json);
+
+nlohmann::json attrsToJson(const Attrs & attrs);
+
+std::optional<std::string> maybeGetStrAttr(const Attrs & attrs, const std::string & name);
+
+std::string getStrAttr(const Attrs & attrs, const std::string & name);
+
+std::optional<int64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name);
+
+int64_t getIntAttr(const Attrs & attrs, const std::string & name);
+
+std::optional<bool> maybeGetBoolAttr(const Attrs & attrs, const std::string & name);
+
+bool getBoolAttr(const Attrs & attrs, const std::string & name);
+
+}

src/libfetchers/cache.cc

@@ -0,0 +1,121 @@
+#include "cache.hh"
+#include "sqlite.hh"
+#include "sync.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+static const char * schema = R"sql(
+
+create table if not exists Cache (
+    input     text not null,
+    info      text not null,
+    path      text not null,
+    immutable integer not null,
+    timestamp integer not null,
+    primary key (input)
+);
+)sql";
+
+struct CacheImpl : Cache
+{
+    struct State
+    {
+        SQLite db;
+        SQLiteStmt add, lookup;
+    };
+
+    Sync<State> _state;
+
+    CacheImpl()
+    {
+        auto state(_state.lock());
+
+        auto dbPath = getCacheDir() + "/nix/fetcher-cache-v1.sqlite";
+        createDirs(dirOf(dbPath));
+
+        state->db = SQLite(dbPath);
+        state->db.isCache();
+        state->db.exec(schema);
+
+        state->add.create(state->db,
+            "insert or replace into Cache(input, info, path, immutable, timestamp) values (?, ?, ?, ?, ?)");
+
+        state->lookup.create(state->db,
+            "select info, path, immutable, timestamp from Cache where input = ?");
+    }
+
+    void add(
+        ref<Store> store,
+        const Attrs & inAttrs,
+        const Attrs & infoAttrs,
+        const StorePath & storePath,
+        bool immutable) override
+    {
+        _state.lock()->add.use()
+            (attrsToJson(inAttrs).dump())
+            (attrsToJson(infoAttrs).dump())
+            (store->printStorePath(storePath))
+            (immutable)
+            (time(0)).exec();
+    }
+
+    std::optional<std::pair<Attrs, StorePath>> lookup(
+        ref<Store> store,
+        const Attrs & inAttrs) override
+    {
+        if (auto res = lookupExpired(store, inAttrs)) {
+            if (!res->expired)
+                return std::make_pair(std::move(res->infoAttrs), std::move(res->storePath));
+            debug("ignoring expired cache entry '%s'",
+                attrsToJson(inAttrs).dump());
+        }
+        return {};
+    }
+
+    std::optional<Result> lookupExpired(
+        ref<Store> store,
+        const Attrs & inAttrs) override
+    {
+        auto state(_state.lock());
+
+        auto inAttrsJson = attrsToJson(inAttrs).dump();
+
+        auto stmt(state->lookup.use()(inAttrsJson));
+        if (!stmt.next()) {
+            debug("did not find cache entry for '%s'", inAttrsJson);
+            return {};
+        }
+
+        auto infoJson = stmt.getStr(0);
+        auto storePath = store->parseStorePath(stmt.getStr(1));
+        auto immutable = stmt.getInt(2) != 0;
+        auto timestamp = stmt.getInt(3);
+
+        store->addTempRoot(storePath);
+        if (!store->isValidPath(storePath)) {
+            // FIXME: we could try to substitute 'storePath'.
+            debug("ignoring disappeared cache entry '%s'", inAttrsJson);
+            return {};
+        }
+
+        debug("using cache entry '%s' -> '%s', '%s'",
+            inAttrsJson, infoJson, store->printStorePath(storePath));
+
+        return Result {
+            .expired = !immutable && (settings.tarballTtl.get() == 0 || timestamp + settings.tarballTtl < time(0)),
+            .infoAttrs = jsonToAttrs(nlohmann::json::parse(infoJson)),
+            .storePath = std::move(storePath)
+        };
+    }
+};
+
+ref<Cache> getCache()
+{
+    static auto cache = std::make_shared<CacheImpl>();
+    return ref<Cache>(cache);
+}
+
+}

src/libfetchers/cache.hh

@@ -0,0 +1,34 @@
+#pragma once
+
+#include "fetchers.hh"
+
+namespace nix::fetchers {
+
+struct Cache
+{
+    virtual void add(
+        ref<Store> store,
+        const Attrs & inAttrs,
+        const Attrs & infoAttrs,
+        const StorePath & storePath,
+        bool immutable) = 0;
+
+    virtual std::optional<std::pair<Attrs, StorePath>> lookup(
+        ref<Store> store,
+        const Attrs & inAttrs) = 0;
+
+    struct Result
+    {
+        bool expired = false;
+        Attrs infoAttrs;
+        StorePath storePath;
+    };
+
+    virtual std::optional<Result> lookupExpired(
+        ref<Store> store,
+        const Attrs & inAttrs) = 0;
+};
+
+ref<Cache> getCache();
+
+}

src/libfetchers/fetchers.cc

@@ -0,0 +1,73 @@
+#include "fetchers.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+std::unique_ptr<std::vector<std::unique_ptr<InputScheme>>> inputSchemes = nullptr;
+
+void registerInputScheme(std::unique_ptr<InputScheme> && inputScheme)
+{
+    if (!inputSchemes) inputSchemes = std::make_unique<std::vector<std::unique_ptr<InputScheme>>>();
+    inputSchemes->push_back(std::move(inputScheme));
+}
+
+std::unique_ptr<Input> inputFromURL(const ParsedURL & url)
+{
+    for (auto & inputScheme : *inputSchemes) {
+        auto res = inputScheme->inputFromURL(url);
+        if (res) return res;
+    }
+    throw Error("input '%s' is unsupported", url.url);
+}
+
+std::unique_ptr<Input> inputFromURL(const std::string & url)
+{
+    return inputFromURL(parseURL(url));
+}
+
+std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs)
+{
+    for (auto & inputScheme : *inputSchemes) {
+        auto res = inputScheme->inputFromAttrs(attrs);
+        if (res) {
+            if (auto narHash = maybeGetStrAttr(attrs, "narHash"))
+                // FIXME: require SRI hash.
+                res->narHash = Hash(*narHash);
+            return res;
+        }
+    }
+    throw Error("input '%s' is unsupported", attrsToJson(attrs));
+}
+
+Attrs Input::toAttrs() const
+{
+    auto attrs = toAttrsInternal();
+    if (narHash)
+        attrs.emplace("narHash", narHash->to_string(SRI));
+    attrs.emplace("type", type());
+    return attrs;
+}
+
+std::pair<Tree, std::shared_ptr<const Input>> Input::fetchTree(ref<Store> store) const
+{
+    auto [tree, input] = fetchTreeInternal(store);
+
+    if (tree.actualPath == "")
+        tree.actualPath = store->toRealPath(tree.storePath);
+
+    if (!tree.info.narHash)
+        tree.info.narHash = store->queryPathInfo(tree.storePath)->narHash;
+
+    if (input->narHash)
+        assert(input->narHash == tree.info.narHash);
+
+    if (narHash && narHash != input->narHash)
+        throw Error("NAR hash mismatch in input '%s' (%s), expected '%s', got '%s'",
+            to_string(), tree.actualPath, narHash->to_string(SRI), input->narHash->to_string(SRI));
+
+    return {std::move(tree), input};
+}
+
+}

src/libfetchers/fetchers.hh

@@ -0,0 +1,103 @@
+#pragma once
+
+#include "types.hh"
+#include "hash.hh"
+#include "path.hh"
+#include "tree-info.hh"
+#include "attrs.hh"
+#include "url.hh"
+
+#include <memory>
+
+namespace nix { class Store; }
+
+namespace nix::fetchers {
+
+struct Input;
+
+struct Tree
+{
+    Path actualPath;
+    StorePath storePath;
+    TreeInfo info;
+};
+
+struct Input : std::enable_shared_from_this<Input>
+{
+    std::optional<Hash> narHash; // FIXME: implement
+
+    virtual std::string type() const = 0;
+
+    virtual ~Input() { }
+
+    virtual bool operator ==(const Input & other) const { return false; }
+
+    /* Check whether this is a "direct" input, that is, not
+       one that goes through a registry. */
+    virtual bool isDirect() const { return true; }
+
+    /* Check whether this is an "immutable" input, that is,
+       one that contains a commit hash or content hash. */
+    virtual bool isImmutable() const { return (bool) narHash; }
+
+    virtual bool contains(const Input & other) const { return false; }
+
+    virtual std::optional<std::string> getRef() const { return {}; }
+
+    virtual std::optional<Hash> getRev() const { return {}; }
+
+    virtual ParsedURL toURL() const = 0;
+
+    std::string to_string() const
+    {
+        return toURL().to_string();
+    }
+
+    Attrs toAttrs() const;
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTree(ref<Store> store) const;
+
+private:
+
+    virtual std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(ref<Store> store) const = 0;
+
+    virtual Attrs toAttrsInternal() const = 0;
+};
+
+struct InputScheme
+{
+    virtual ~InputScheme() { }
+
+    virtual std::unique_ptr<Input> inputFromURL(const ParsedURL & url) = 0;
+
+    virtual std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) = 0;
+};
+
+std::unique_ptr<Input> inputFromURL(const ParsedURL & url);
+
+std::unique_ptr<Input> inputFromURL(const std::string & url);
+
+std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs);
+
+void registerInputScheme(std::unique_ptr<InputScheme> && fetcher);
+
+struct DownloadFileResult
+{
+    StorePath storePath;
+    std::string etag;
+    std::string effectiveUrl;
+};
+
+DownloadFileResult downloadFile(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable);
+
+Tree downloadTarball(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable);
+
+}

src/libfetchers/git.cc

@@ -0,0 +1,401 @@
+#include "fetchers.hh"
+#include "cache.hh"
+#include "globals.hh"
+#include "tarfile.hh"
+#include "store-api.hh"
+
+#include <sys/time.h>
+
+using namespace std::string_literals;
+
+namespace nix::fetchers {
+
+static std::string readHead(const Path & path)
+{
+    return chomp(runProgram("git", true, { "-C", path, "rev-parse", "--abbrev-ref", "HEAD" }));
+}
+
+struct GitInput : Input
+{
+    ParsedURL url;
+    std::optional<std::string> ref;
+    std::optional<Hash> rev;
+    bool shallow = false;
+
+    GitInput(const ParsedURL & url) : url(url)
+    { }
+
+    std::string type() const override { return "git"; }
+
+    bool operator ==(const Input & other) const override
+    {
+        auto other2 = dynamic_cast<const GitInput *>(&other);
+        return
+            other2
+            && url == other2->url
+            && rev == other2->rev
+            && ref == other2->ref;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) rev;
+    }
+
+    std::optional<std::string> getRef() const override { return ref; }
+
+    std::optional<Hash> getRev() const override { return rev; }
+
+    ParsedURL toURL() const override
+    {
+        ParsedURL url2(url);
+        if (url2.scheme != "git") url2.scheme = "git+" + url2.scheme;
+        if (rev) url2.query.insert_or_assign("rev", rev->gitRev());
+        if (ref) url2.query.insert_or_assign("ref", *ref);
+        if (shallow) url2.query.insert_or_assign("shallow", "1");
+        return url2;
+    }
+
+    Attrs toAttrsInternal() const override
+    {
+        Attrs attrs;
+        attrs.emplace("url", url.to_string());
+        if (ref)
+            attrs.emplace("ref", *ref);
+        if (rev)
+            attrs.emplace("rev", rev->gitRev());
+        if (shallow)
+            attrs.emplace("shallow", true);
+        return attrs;
+    }
+
+    std::pair<bool, std::string> getActualUrl() const
+    {
+        // Don't clone file:// URIs (but otherwise treat them the
+        // same as remote URIs, i.e. don't use the working tree or
+        // HEAD).
+        static bool forceHttp = getEnv("_NIX_FORCE_HTTP") == "1"; // for testing
+        bool isLocal = url.scheme == "file" && !forceHttp;
+        return {isLocal, isLocal ? url.path : url.base};
+    }
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    {
+        auto name = "source";
+
+        auto input = std::make_shared<GitInput>(*this);
+
+        assert(!rev || rev->type == htSHA1);
+
+        auto cacheType = shallow ? "git-shallow" : "git";
+
+        auto getImmutableAttrs = [&]()
+        {
+            return Attrs({
+                {"type", cacheType},
+                {"name", name},
+                {"rev", input->rev->gitRev()},
+            });
+        };
+
+        auto makeResult = [&](const Attrs & infoAttrs, StorePath && storePath)
+            -> std::pair<Tree, std::shared_ptr<const Input>>
+        {
+            assert(input->rev);
+            assert(!rev || rev == input->rev);
+            return {
+                Tree {
+                    .actualPath = store->toRealPath(storePath),
+                    .storePath = std::move(storePath),
+                    .info = TreeInfo {
+                        .revCount = shallow ? std::nullopt : std::optional(getIntAttr(infoAttrs, "revCount")),
+                        .lastModified = getIntAttr(infoAttrs, "lastModified"),
+                    },
+                },
+                input
+            };
+        };
+
+        if (rev) {
+            if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+                return makeResult(res->first, std::move(res->second));
+        }
+
+        auto [isLocal, actualUrl_] = getActualUrl();
+        auto actualUrl = actualUrl_; // work around clang bug
+
+        // If this is a local directory and no ref or revision is
+        // given, then allow the use of an unclean working tree.
+        if (!input->ref && !input->rev && isLocal) {
+            bool clean = false;
+
+            /* Check whether this repo has any commits. There are
+               probably better ways to do this. */
+            auto gitDir = actualUrl + "/.git";
+            auto commonGitDir = chomp(runProgram(
+                "git",
+                true,
+                { "-C", actualUrl, "rev-parse", "--git-common-dir" }
+            ));
+            if (commonGitDir != ".git")
+                    gitDir = commonGitDir;
+
+            bool haveCommits = !readDirectory(gitDir + "/refs/heads").empty();
+
+            try {
+                if (haveCommits) {
+                    runProgram("git", true, { "-C", actualUrl, "diff-index", "--quiet", "HEAD", "--" });
+                    clean = true;
+                }
+            } catch (ExecError & e) {
+                if (!WIFEXITED(e.status) || WEXITSTATUS(e.status) != 1) throw;
+            }
+
+            if (!clean) {
+
+                /* This is an unclean working tree. So copy all tracked files. */
+
+                if (!settings.allowDirty)
+                    throw Error("Git tree '%s' is dirty", actualUrl);
+
+                if (settings.warnDirty)
+                    warn("Git tree '%s' is dirty", actualUrl);
+
+                auto files = tokenizeString<std::set<std::string>>(
+                    runProgram("git", true, { "-C", actualUrl, "ls-files", "-z" }), "\0"s);
+
+                PathFilter filter = [&](const Path & p) -> bool {
+                    assert(hasPrefix(p, actualUrl));
+                    std::string file(p, actualUrl.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);
+                };
+
+                auto storePath = store->addToStore("source", actualUrl, true, htSHA256, filter);
+
+                auto tree = Tree {
+                    .actualPath = store->printStorePath(storePath),
+                    .storePath = std::move(storePath),
+                    .info = TreeInfo {
+                        // FIXME: maybe we should use the timestamp of the last
+                        // modified dirty file?
+                        .lastModified = haveCommits ? std::stoull(runProgram("git", true, { "-C", actualUrl, "log", "-1", "--format=%ct", "HEAD" })) : 0,
+                    }
+                };
+
+                return {std::move(tree), input};
+            }
+        }
+
+        if (!input->ref) input->ref = isLocal ? readHead(actualUrl) : "master";
+
+        Attrs mutableAttrs({
+            {"type", cacheType},
+            {"name", name},
+            {"url", actualUrl},
+            {"ref", *input->ref},
+        });
+
+        Path repoDir;
+
+        if (isLocal) {
+
+            if (!input->rev)
+                input->rev = Hash(chomp(runProgram("git", true, { "-C", actualUrl, "rev-parse", *input->ref })), htSHA1);
+
+            repoDir = actualUrl;
+
+        } else {
+
+            if (auto res = getCache()->lookup(store, mutableAttrs)) {
+                auto rev2 = Hash(getStrAttr(res->first, "rev"), htSHA1);
+                if (!rev || rev == rev2) {
+                    input->rev = rev2;
+                    return makeResult(res->first, std::move(res->second));
+                }
+            }
+
+            Path cacheDir = getCacheDir() + "/nix/gitv3/" + hashString(htSHA256, actualUrl).to_string(Base32, false);
+            repoDir = cacheDir;
+
+            if (!pathExists(cacheDir)) {
+                createDirs(dirOf(cacheDir));
+                runProgram("git", true, { "init", "--bare", repoDir });
+            }
+
+            Path localRefFile =
+                input->ref->compare(0, 5, "refs/") == 0
+                ? cacheDir + "/" + *input->ref
+                : cacheDir + "/refs/heads/" + *input->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 (input->rev) {
+                try {
+                    runProgram("git", true, { "-C", repoDir, "cat-file", "-e", input->rev->gitRev() });
+                    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'", actualUrl));
+
+                // FIXME: git stderr messes up our progress indicator, so
+                // we're using --quiet for now. Should process its stderr.
+                try {
+                    runProgram("git", true, { "-C", repoDir, "fetch", "--quiet", "--force", "--", actualUrl, fmt("%s:%s", *input->ref, *input->ref) });
+                } catch (Error & e) {
+                    if (!pathExists(localRefFile)) throw;
+                    warn("could not update local clone of Git repository '%s'; continuing with the most recent version", actualUrl);
+                }
+
+                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);
+            }
+
+            if (!input->rev)
+                input->rev = Hash(chomp(readFile(localRefFile)), htSHA1);
+        }
+
+        bool isShallow = chomp(runProgram("git", true, { "-C", repoDir, "rev-parse", "--is-shallow-repository" })) == "true";
+
+        if (isShallow && !shallow)
+            throw Error("'%s' is a shallow Git repository, but a non-shallow repository is needed", actualUrl);
+
+        // FIXME: check whether rev is an ancestor of ref.
+
+        printTalkative("using revision %s of repo '%s'", input->rev->gitRev(), actualUrl);
+
+        /* Now that we know the ref, check again whether we have it in
+           the store. */
+        if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+            return makeResult(res->first, std::move(res->second));
+
+        // FIXME: should pipe this, or find some better way to extract a
+        // revision.
+        auto source = sinkToSource([&](Sink & sink) {
+            RunOptions gitOptions("git", { "-C", repoDir, "archive", input->rev->gitRev() });
+            gitOptions.standardOut = &sink;
+            runProgram2(gitOptions);
+        });
+
+        Path tmpDir = createTempDir();
+        AutoDelete delTmpDir(tmpDir, true);
+
+        unpackTarfile(*source, tmpDir);
+
+        auto storePath = store->addToStore(name, tmpDir);
+
+        auto lastModified = std::stoull(runProgram("git", true, { "-C", repoDir, "log", "-1", "--format=%ct", input->rev->gitRev() }));
+
+        Attrs infoAttrs({
+            {"rev", input->rev->gitRev()},
+            {"lastModified", lastModified},
+        });
+
+        if (!shallow)
+            infoAttrs.insert_or_assign("revCount",
+                std::stoull(runProgram("git", true, { "-C", repoDir, "rev-list", "--count", input->rev->gitRev() })));
+
+        if (!this->rev)
+            getCache()->add(
+                store,
+                mutableAttrs,
+                infoAttrs,
+                storePath,
+                false);
+
+        getCache()->add(
+            store,
+            getImmutableAttrs(),
+            infoAttrs,
+            storePath,
+            true);
+
+        return makeResult(infoAttrs, std::move(storePath));
+    }
+};
+
+struct GitInputScheme : InputScheme
+{
+    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
+    {
+        if (url.scheme != "git" &&
+            url.scheme != "git+http" &&
+            url.scheme != "git+https" &&
+            url.scheme != "git+ssh" &&
+            url.scheme != "git+file") return nullptr;
+
+        auto url2(url);
+        if (hasPrefix(url2.scheme, "git+")) url2.scheme = std::string(url2.scheme, 4);
+        url2.query.clear();
+
+        Attrs attrs;
+        attrs.emplace("type", "git");
+
+        for (auto &[name, value] : url.query) {
+            if (name == "rev" || name == "ref")
+                attrs.emplace(name, value);
+            else
+                url2.query.emplace(name, value);
+        }
+
+        attrs.emplace("url", url2.to_string());
+
+        return inputFromAttrs(attrs);
+    }
+
+    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "git") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "url" && name != "ref" && name != "rev" && name != "shallow")
+                throw Error("unsupported Git input attribute '%s'", name);
+
+        auto input = std::make_unique<GitInput>(parseURL(getStrAttr(attrs, "url")));
+        if (auto ref = maybeGetStrAttr(attrs, "ref")) {
+            if (!std::regex_match(*ref, refRegex))
+                throw BadURL("invalid Git branch/tag name '%s'", *ref);
+            input->ref = *ref;
+        }
+        if (auto rev = maybeGetStrAttr(attrs, "rev"))
+            input->rev = Hash(*rev, htSHA1);
+
+        input->shallow = maybeGetBoolAttr(attrs, "shallow").value_or(false);
+
+        return input;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<GitInputScheme>()); });
+
+}

src/libfetchers/github.cc

@@ -0,0 +1,195 @@
+#include "download.hh"
+#include "cache.hh"
+#include "fetchers.hh"
+#include "globals.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+std::regex ownerRegex("[a-zA-Z][a-zA-Z0-9_-]*", std::regex::ECMAScript);
+std::regex repoRegex("[a-zA-Z][a-zA-Z0-9_-]*", std::regex::ECMAScript);
+
+struct GitHubInput : Input
+{
+    std::string owner;
+    std::string repo;
+    std::optional<std::string> ref;
+    std::optional<Hash> rev;
+
+    std::string type() const override { return "github"; }
+
+    bool operator ==(const Input & other) const override
+    {
+        auto other2 = dynamic_cast<const GitHubInput *>(&other);
+        return
+            other2
+            && owner == other2->owner
+            && repo == other2->repo
+            && rev == other2->rev
+            && ref == other2->ref;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) rev;
+    }
+
+    std::optional<std::string> getRef() const override { return ref; }
+
+    std::optional<Hash> getRev() const override { return rev; }
+
+    ParsedURL toURL() const override
+    {
+        auto path = owner + "/" + repo;
+        assert(!(ref && rev));
+        if (ref) path += "/" + *ref;
+        if (rev) path += "/" + rev->to_string(Base16, false);
+        return ParsedURL {
+            .scheme = "github",
+            .path = path,
+        };
+    }
+
+    Attrs toAttrsInternal() const override
+    {
+        Attrs attrs;
+        attrs.emplace("owner", owner);
+        attrs.emplace("repo", repo);
+        if (ref)
+            attrs.emplace("ref", *ref);
+        if (rev)
+            attrs.emplace("rev", rev->gitRev());
+        return attrs;
+    }
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    {
+        auto rev = this->rev;
+        auto ref = this->ref.value_or("master");
+
+        if (!rev) {
+            auto url = fmt("https://api.github.com/repos/%s/%s/commits/%s",
+                owner, repo, ref);
+            auto json = nlohmann::json::parse(
+                readFile(
+                    store->toRealPath(
+                        downloadFile(store, url, "source", false).storePath)));
+            rev = Hash(json["sha"], htSHA1);
+            debug("HEAD revision for '%s' is %s", url, rev->gitRev());
+        }
+
+        auto input = std::make_shared<GitHubInput>(*this);
+        input->ref = {};
+        input->rev = *rev;
+
+        Attrs immutableAttrs({
+            {"type", "git-tarball"},
+            {"rev", rev->gitRev()},
+        });
+
+        if (auto res = getCache()->lookup(store, immutableAttrs)) {
+            return {
+                Tree{
+                    .actualPath = store->toRealPath(res->second),
+                    .storePath = std::move(res->second),
+                    .info = TreeInfo {
+                        .lastModified = getIntAttr(res->first, "lastModified"),
+                    },
+                },
+                input
+            };
+        }
+
+        // FIXME: use regular /archive URLs instead? api.github.com
+        // might have stricter rate limits.
+
+        auto url = fmt("https://api.github.com/repos/%s/%s/tarball/%s",
+            owner, repo, rev->to_string(Base16, false));
+
+        std::string accessToken = settings.githubAccessToken.get();
+        if (accessToken != "")
+            url += "?access_token=" + accessToken;
+
+        auto tree = downloadTarball(store, url, "source", true);
+
+        getCache()->add(
+            store,
+            immutableAttrs,
+            {
+                {"rev", rev->gitRev()},
+                {"lastModified", *tree.info.lastModified}
+            },
+            tree.storePath,
+            true);
+
+        return {std::move(tree), input};
+    }
+};
+
+struct GitHubInputScheme : InputScheme
+{
+    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
+    {
+        if (url.scheme != "github") return nullptr;
+
+        auto path = tokenizeString<std::vector<std::string>>(url.path, "/");
+        auto input = std::make_unique<GitHubInput>();
+
+        if (path.size() == 2) {
+        } else if (path.size() == 3) {
+            if (std::regex_match(path[2], revRegex))
+                input->rev = Hash(path[2], htSHA1);
+            else if (std::regex_match(path[2], refRegex))
+                input->ref = path[2];
+            else
+                throw BadURL("in GitHub URL '%s', '%s' is not a commit hash or branch/tag name", url.url, path[2]);
+        } else
+            throw BadURL("GitHub URL '%s' is invalid", url.url);
+
+        for (auto &[name, value] : url.query) {
+            if (name == "rev") {
+                if (input->rev)
+                    throw BadURL("GitHub URL '%s' contains multiple commit hashes", url.url);
+                input->rev = Hash(value, htSHA1);
+            }
+            else if (name == "ref") {
+                if (!std::regex_match(value, refRegex))
+                    throw BadURL("GitHub URL '%s' contains an invalid branch/tag name", url.url);
+                if (input->ref)
+                    throw BadURL("GitHub URL '%s' contains multiple branch/tag names", url.url);
+                input->ref = value;
+            }
+        }
+
+        if (input->ref && input->rev)
+            throw BadURL("GitHub URL '%s' contains both a commit hash and a branch/tag name", url.url);
+
+        input->owner = path[0];
+        input->repo = path[1];
+
+        return input;
+    }
+
+    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "github") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "owner" && name != "repo" && name != "ref" && name != "rev")
+                throw Error("unsupported GitHub input attribute '%s'", name);
+
+        auto input = std::make_unique<GitHubInput>();
+        input->owner = getStrAttr(attrs, "owner");
+        input->repo = getStrAttr(attrs, "repo");
+        input->ref = maybeGetStrAttr(attrs, "ref");
+        if (auto rev = maybeGetStrAttr(attrs, "rev"))
+            input->rev = Hash(*rev, htSHA1);
+        return input;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<GitHubInputScheme>()); });
+
+}

src/libfetchers/local.mk

@@ -0,0 +1,11 @@
+libraries += libfetchers
+
+libfetchers_NAME = libnixfetchers
+
+libfetchers_DIR := $(d)
+
+libfetchers_SOURCES := $(wildcard $(d)/*.cc)
+
+libfetchers_CXXFLAGS += -I src/libutil -I src/libstore
+
+libfetchers_LIBS = libutil libstore libnixrust

src/libfetchers/mercurial.cc

@@ -0,0 +1,303 @@
+#include "fetchers.hh"
+#include "cache.hh"
+#include "globals.hh"
+#include "tarfile.hh"
+#include "store-api.hh"
+
+#include <sys/time.h>
+
+using namespace std::string_literals;
+
+namespace nix::fetchers {
+
+struct MercurialInput : Input
+{
+    ParsedURL url;
+    std::optional<std::string> ref;
+    std::optional<Hash> rev;
+
+    MercurialInput(const ParsedURL & url) : url(url)
+    { }
+
+    std::string type() const override { return "hg"; }
+
+    bool operator ==(const Input & other) const override
+    {
+        auto other2 = dynamic_cast<const MercurialInput *>(&other);
+        return
+            other2
+            && url == other2->url
+            && rev == other2->rev
+            && ref == other2->ref;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) rev;
+    }
+
+    std::optional<std::string> getRef() const override { return ref; }
+
+    std::optional<Hash> getRev() const override { return rev; }
+
+    ParsedURL toURL() const override
+    {
+        ParsedURL url2(url);
+        url2.scheme = "hg+" + url2.scheme;
+        if (rev) url2.query.insert_or_assign("rev", rev->gitRev());
+        if (ref) url2.query.insert_or_assign("ref", *ref);
+        return url;
+    }
+
+    Attrs toAttrsInternal() const override
+    {
+        Attrs attrs;
+        attrs.emplace("url", url.to_string());
+        if (ref)
+            attrs.emplace("ref", *ref);
+        if (rev)
+            attrs.emplace("rev", rev->gitRev());
+        return attrs;
+    }
+
+    std::pair<bool, std::string> getActualUrl() const
+    {
+        bool isLocal = url.scheme == "file";
+        return {isLocal, isLocal ? url.path : url.base};
+    }
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    {
+        auto name = "source";
+
+        auto input = std::make_shared<MercurialInput>(*this);
+
+        auto [isLocal, actualUrl_] = getActualUrl();
+        auto actualUrl = actualUrl_; // work around clang bug
+
+        // FIXME: return lastModified.
+
+        // FIXME: don't clone local repositories.
+
+        if (!input->ref && !input->rev && isLocal && pathExists(actualUrl + "/.hg")) {
+
+            bool clean = runProgram("hg", true, { "status", "-R", actualUrl, "--modified", "--added", "--removed" }) == "";
+
+            if (!clean) {
+
+                /* This is an unclean working tree. So copy all tracked
+                   files. */
+
+                if (!settings.allowDirty)
+                    throw Error("Mercurial tree '%s' is unclean", actualUrl);
+
+                if (settings.warnDirty)
+                    warn("Mercurial tree '%s' is unclean", actualUrl);
+
+                input->ref = chomp(runProgram("hg", true, { "branch", "-R", actualUrl }));
+
+                auto files = tokenizeString<std::set<std::string>>(
+                    runProgram("hg", true, { "status", "-R", actualUrl, "--clean", "--modified", "--added", "--no-status", "--print0" }), "\0"s);
+
+                PathFilter filter = [&](const Path & p) -> bool {
+                    assert(hasPrefix(p, actualUrl));
+                    std::string file(p, actualUrl.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);
+                };
+
+                auto storePath = store->addToStore("source", actualUrl, true, htSHA256, filter);
+
+                return {Tree {
+                    .actualPath = store->printStorePath(storePath),
+                    .storePath = std::move(storePath),
+                }, input};
+            }
+        }
+
+        if (!input->ref) input->ref = "default";
+
+        auto getImmutableAttrs = [&]()
+        {
+            return Attrs({
+                {"type", "hg"},
+                {"name", name},
+                {"rev", input->rev->gitRev()},
+            });
+        };
+
+        auto makeResult = [&](const Attrs & infoAttrs, StorePath && storePath)
+            -> std::pair<Tree, std::shared_ptr<const Input>>
+        {
+            assert(input->rev);
+            assert(!rev || rev == input->rev);
+            return {
+                Tree{
+                    .actualPath = store->toRealPath(storePath),
+                    .storePath = std::move(storePath),
+                    .info = TreeInfo {
+                        .revCount = getIntAttr(infoAttrs, "revCount"),
+                    },
+                },
+                input
+            };
+        };
+
+        if (input->rev) {
+            if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+                return makeResult(res->first, std::move(res->second));
+        }
+
+        assert(input->rev || input->ref);
+        auto revOrRef = input->rev ? input->rev->gitRev() : *input->ref;
+
+        Attrs mutableAttrs({
+            {"type", "hg"},
+            {"name", name},
+            {"url", actualUrl},
+            {"ref", *input->ref},
+        });
+
+        if (auto res = getCache()->lookup(store, mutableAttrs)) {
+            auto rev2 = Hash(getStrAttr(res->first, "rev"), htSHA1);
+            if (!rev || rev == rev2) {
+                input->rev = rev2;
+                return makeResult(res->first, std::move(res->second));
+            }
+        }
+
+        Path cacheDir = fmt("%s/nix/hg/%s", getCacheDir(), hashString(htSHA256, actualUrl).to_string(Base32, false));
+
+        /* If this is a commit hash that we already have, we don't
+           have to pull again. */
+        if (!(input->rev
+                && pathExists(cacheDir)
+                && runProgram(
+                    RunOptions("hg", { "log", "-R", cacheDir, "-r", input->rev->gitRev(), "--template", "1" })
+                    .killStderr(true)).second == "1"))
+        {
+            Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Mercurial repository '%s'", actualUrl));
+
+            if (pathExists(cacheDir)) {
+                try {
+                    runProgram("hg", true, { "pull", "-R", cacheDir, "--", actualUrl });
+                }
+                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, "--", actualUrl });
+                    } else {
+                        throw ExecError(e.status, fmt("'hg pull' %s", statusToString(e.status)));
+                    }
+                }
+            } else {
+                createDirs(dirOf(cacheDir));
+                runProgram("hg", true, { "clone", "--noupdate", "--", actualUrl, cacheDir });
+            }
+        }
+
+        auto tokens = tokenizeString<std::vector<std::string>>(
+            runProgram("hg", true, { "log", "-R", cacheDir, "-r", revOrRef, "--template", "{node} {rev} {branch}" }));
+        assert(tokens.size() == 3);
+
+        input->rev = Hash(tokens[0], htSHA1);
+        auto revCount = std::stoull(tokens[1]);
+        input->ref = tokens[2];
+
+        if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+            return makeResult(res->first, std::move(res->second));
+
+        Path tmpDir = createTempDir();
+        AutoDelete delTmpDir(tmpDir, true);
+
+        runProgram("hg", true, { "archive", "-R", cacheDir, "-r", input->rev->gitRev(), tmpDir });
+
+        deletePath(tmpDir + "/.hg_archival.txt");
+
+        auto storePath = store->addToStore(name, tmpDir);
+
+        Attrs infoAttrs({
+            {"rev", input->rev->gitRev()},
+            {"revCount", (int64_t) revCount},
+        });
+
+        if (!this->rev)
+            getCache()->add(
+                store,
+                mutableAttrs,
+                infoAttrs,
+                storePath,
+                false);
+
+        getCache()->add(
+            store,
+            getImmutableAttrs(),
+            infoAttrs,
+            storePath,
+            true);
+
+        return makeResult(infoAttrs, std::move(storePath));
+    }
+};
+
+struct MercurialInputScheme : InputScheme
+{
+    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
+    {
+        if (url.scheme != "hg+http" &&
+            url.scheme != "hg+https" &&
+            url.scheme != "hg+ssh" &&
+            url.scheme != "hg+file") return nullptr;
+
+        auto url2(url);
+        url2.scheme = std::string(url2.scheme, 3);
+        url2.query.clear();
+
+        Attrs attrs;
+        attrs.emplace("type", "hg");
+
+        for (auto &[name, value] : url.query) {
+            if (name == "rev" || name == "ref")
+                attrs.emplace(name, value);
+            else
+                url2.query.emplace(name, value);
+        }
+
+        attrs.emplace("url", url2.to_string());
+
+        return inputFromAttrs(attrs);
+    }
+
+    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "hg") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "url" && name != "ref" && name != "rev")
+                throw Error("unsupported Mercurial input attribute '%s'", name);
+
+        auto input = std::make_unique<MercurialInput>(parseURL(getStrAttr(attrs, "url")));
+        if (auto ref = maybeGetStrAttr(attrs, "ref")) {
+            if (!std::regex_match(*ref, refRegex))
+                throw BadURL("invalid Mercurial branch/tag name '%s'", *ref);
+            input->ref = *ref;
+        }
+        if (auto rev = maybeGetStrAttr(attrs, "rev"))
+            input->rev = Hash(*rev, htSHA1);
+        return input;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<MercurialInputScheme>()); });
+
+}

src/libfetchers/tarball.cc

@@ -0,0 +1,277 @@
+#include "fetchers.hh"
+#include "cache.hh"
+#include "download.hh"
+#include "globals.hh"
+#include "store-api.hh"
+#include "archive.hh"
+#include "tarfile.hh"
+
+namespace nix::fetchers {
+
+DownloadFileResult downloadFile(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable)
+{
+    // FIXME: check store
+
+    Attrs inAttrs({
+        {"type", "file"},
+        {"url", url},
+        {"name", name},
+    });
+
+    auto cached = getCache()->lookupExpired(store, inAttrs);
+
+    auto useCached = [&]() -> DownloadFileResult
+    {
+        return {
+            .storePath = std::move(cached->storePath),
+            .etag = getStrAttr(cached->infoAttrs, "etag"),
+            .effectiveUrl = getStrAttr(cached->infoAttrs, "url")
+        };
+    };
+
+    if (cached && !cached->expired)
+        return useCached();
+
+    DownloadRequest request(url);
+    if (cached)
+        request.expectedETag = getStrAttr(cached->infoAttrs, "etag");
+    DownloadResult res;
+    try {
+        res = getDownloader()->download(request);
+    } catch (DownloadError & e) {
+        if (cached) {
+            warn("%s; using cached version", e.msg());
+            return useCached();
+        } else
+            throw;
+    }
+
+    // FIXME: write to temporary file.
+
+    Attrs infoAttrs({
+        {"etag", res.etag},
+        {"url", res.effectiveUri},
+    });
+
+    std::optional<StorePath> storePath;
+
+    if (res.cached) {
+        assert(cached);
+        assert(request.expectedETag == res.etag);
+        storePath = std::move(cached->storePath);
+    } else {
+        StringSink sink;
+        dumpString(*res.data, sink);
+        auto hash = hashString(htSHA256, *res.data);
+        ValidPathInfo info(store->makeFixedOutputPath(false, hash, name));
+        info.narHash = hashString(htSHA256, *sink.s);
+        info.narSize = sink.s->size();
+        info.ca = makeFixedOutputCA(false, hash);
+        store->addToStore(info, sink.s, NoRepair, NoCheckSigs);
+        storePath = std::move(info.path);
+    }
+
+    getCache()->add(
+        store,
+        inAttrs,
+        infoAttrs,
+        *storePath,
+        immutable);
+
+    if (url != res.effectiveUri)
+        getCache()->add(
+            store,
+            {
+                {"type", "file"},
+                {"url", res.effectiveUri},
+                {"name", name},
+            },
+            infoAttrs,
+            *storePath,
+            immutable);
+
+    return {
+        .storePath = std::move(*storePath),
+        .etag = res.etag,
+        .effectiveUrl = res.effectiveUri,
+    };
+}
+
+Tree downloadTarball(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable)
+{
+    Attrs inAttrs({
+        {"type", "tarball"},
+        {"url", url},
+        {"name", name},
+    });
+
+    auto cached = getCache()->lookupExpired(store, inAttrs);
+
+    if (cached && !cached->expired)
+        return Tree {
+            .actualPath = store->toRealPath(cached->storePath),
+            .storePath = std::move(cached->storePath),
+            .info = TreeInfo {
+                .lastModified = getIntAttr(cached->infoAttrs, "lastModified"),
+            },
+        };
+
+    auto res = downloadFile(store, url, name, immutable);
+
+    std::optional<StorePath> unpackedStorePath;
+    time_t lastModified;
+
+    if (cached && res.etag != "" && getStrAttr(cached->infoAttrs, "etag") == res.etag) {
+        unpackedStorePath = std::move(cached->storePath);
+        lastModified = getIntAttr(cached->infoAttrs, "lastModified");
+    } else {
+        Path tmpDir = createTempDir();
+        AutoDelete autoDelete(tmpDir, true);
+        unpackTarfile(store->toRealPath(res.storePath), tmpDir);
+        auto members = readDirectory(tmpDir);
+        if (members.size() != 1)
+            throw nix::Error("tarball '%s' contains an unexpected number of top-level files", url);
+        auto topDir = tmpDir + "/" + members.begin()->name;
+        lastModified = lstat(topDir).st_mtime;
+        unpackedStorePath = store->addToStore(name, topDir, true, htSHA256, defaultPathFilter, NoRepair);
+    }
+
+    Attrs infoAttrs({
+        {"lastModified", lastModified},
+        {"etag", res.etag},
+    });
+
+    getCache()->add(
+        store,
+        inAttrs,
+        infoAttrs,
+        *unpackedStorePath,
+        immutable);
+
+    return Tree {
+        .actualPath = store->toRealPath(*unpackedStorePath),
+        .storePath = std::move(*unpackedStorePath),
+        .info = TreeInfo {
+            .lastModified = lastModified,
+        },
+    };
+}
+
+struct TarballInput : Input
+{
+    ParsedURL url;
+    std::optional<Hash> hash;
+
+    TarballInput(const ParsedURL & url) : url(url)
+    { }
+
+    std::string type() const override { return "tarball"; }
+
+    bool operator ==(const Input & other) const override
+    {
+        auto other2 = dynamic_cast<const TarballInput *>(&other);
+        return
+            other2
+            && to_string() == other2->to_string()
+            && hash == other2->hash;
+    }
+
+    bool isImmutable() const override
+    {
+        return hash || narHash;
+    }
+
+    ParsedURL toURL() const override
+    {
+        auto url2(url);
+        // NAR hashes are preferred over file hashes since tar/zip files
+        // don't have a canonical representation.
+        if (narHash)
+            url2.query.insert_or_assign("narHash", narHash->to_string(SRI));
+        else if (hash)
+            url2.query.insert_or_assign("hash", hash->to_string(SRI));
+        return url2;
+    }
+
+    Attrs toAttrsInternal() const override
+    {
+        Attrs attrs;
+        attrs.emplace("url", url.to_string());
+        if (narHash)
+            attrs.emplace("narHash", narHash->to_string(SRI));
+        else if (hash)
+            attrs.emplace("hash", hash->to_string(SRI));
+        return attrs;
+    }
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    {
+        auto tree = downloadTarball(store, url.to_string(), "source", false);
+
+        auto input = std::make_shared<TarballInput>(*this);
+        input->narHash = store->queryPathInfo(tree.storePath)->narHash;
+
+        return {std::move(tree), input};
+    }
+};
+
+struct TarballInputScheme : InputScheme
+{
+    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
+    {
+        if (url.scheme != "file" && url.scheme != "http" && url.scheme != "https") return nullptr;
+
+        if (!hasSuffix(url.path, ".zip")
+            && !hasSuffix(url.path, ".tar")
+            && !hasSuffix(url.path, ".tar.gz")
+            && !hasSuffix(url.path, ".tar.xz")
+            && !hasSuffix(url.path, ".tar.bz2"))
+            return nullptr;
+
+        auto input = std::make_unique<TarballInput>(url);
+
+        auto hash = input->url.query.find("hash");
+        if (hash != input->url.query.end()) {
+            // FIXME: require SRI hash.
+            input->hash = Hash(hash->second);
+            input->url.query.erase(hash);
+        }
+
+        auto narHash = input->url.query.find("narHash");
+        if (narHash != input->url.query.end()) {
+            // FIXME: require SRI hash.
+            input->narHash = Hash(narHash->second);
+            input->url.query.erase(narHash);
+        }
+
+        return input;
+    }
+
+    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "tarball") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "url" && name != "hash" && name != "narHash")
+                throw Error("unsupported tarball input attribute '%s'", name);
+
+        auto input = std::make_unique<TarballInput>(parseURL(getStrAttr(attrs, "url")));
+        if (auto hash = maybeGetStrAttr(attrs, "hash"))
+            // FIXME: require SRI hash.
+            input->hash = Hash(*hash);
+
+        return input;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<TarballInputScheme>()); });
+
+}

src/libfetchers/tree-info.cc

@@ -0,0 +1,14 @@
+#include "tree-info.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+StorePath TreeInfo::computeStorePath(Store & store) const
+{
+    assert(narHash);
+    return store.makeFixedOutputPath(true, narHash, "source");
+}
+
+}

src/libfetchers/tree-info.hh

@@ -0,0 +1,29 @@
+#pragma once
+
+#include "path.hh"
+#include "hash.hh"
+
+#include <nlohmann/json_fwd.hpp>
+
+namespace nix { class Store; }
+
+namespace nix::fetchers {
+
+struct TreeInfo
+{
+    Hash narHash;
+    std::optional<uint64_t> revCount;
+    std::optional<time_t> lastModified;
+
+    bool operator ==(const TreeInfo & other) const
+    {
+        return
+            narHash == other.narHash
+            && revCount == other.revCount
+            && lastModified == other.lastModified;
+    }
+
+    StorePath computeStorePath(Store & store) const;
+};
+
+}

src/libstore/derivations.cc

@@ -378,7 +378,7 @@ Hash hashDerivationModulo(Store & store, const Derivation & drv, bool maskOutput
         if (h == drvHashes.end()) {
             assert(store.isValidPath(i.first));
             h = drvHashes.insert_or_assign(i.first.clone(), hashDerivationModulo(store,
-                readDerivation(store, store.toRealPath(store.printStorePath(i.first))), false)).first;
+                readDerivation(store, store.toRealPath(i.first)), false)).first;
         }
         inputs2.insert_or_assign(h->second.to_string(Base16, false), i.second);
     }

src/libstore/download.cc

@@ -35,10 +35,6 @@ DownloadSettings downloadSettings;
 
 static GlobalConfig::Register r1(&downloadSettings);
 
-CachedDownloadRequest::CachedDownloadRequest(const std::string & uri)
-    : uri(uri), ttl(settings.tarballTtl)
-{ }
-
 std::string resolveUri(const std::string & uri)
 {
     if (uri.compare(0, 8, "channel:") == 0)
@@ -802,139 +798,6 @@ void Downloader::download(DownloadRequest && request, Sink & sink)
     }
 }
 
-CachedDownloadResult Downloader::downloadCached(
-    ref<Store> store, const CachedDownloadRequest & request)
-{
-    auto url = resolveUri(request.uri);
-
-    auto name = request.name;
-    if (name == "") {
-        auto p = url.rfind('/');
-        if (p != string::npos) name = string(url, p + 1);
-    }
-
-    std::optional<StorePath> expectedStorePath;
-    if (request.expectedHash) {
-        expectedStorePath = store->makeFixedOutputPath(request.unpack, request.expectedHash, name);
-        if (store->isValidPath(*expectedStorePath)) {
-            CachedDownloadResult result;
-            result.storePath = store->printStorePath(*expectedStorePath);
-            result.path = store->toRealPath(result.storePath);
-            return result;
-        }
-    }
-
-    Path cacheDir = getCacheDir() + "/nix/tarballs";
-    createDirs(cacheDir);
-
-    string urlHash = hashString(htSHA256, name + std::string("\0"s) + url).to_string(Base32, false);
-
-    Path dataFile = cacheDir + "/" + urlHash + ".info";
-    Path fileLink = cacheDir + "/" + urlHash + "-file";
-
-    PathLocks lock({fileLink}, fmt("waiting for lock on '%1%'...", fileLink));
-
-    std::optional<StorePath> storePath;
-
-    string expectedETag;
-
-    bool skip = false;
-
-    CachedDownloadResult result;
-
-    if (pathExists(fileLink) && pathExists(dataFile)) {
-        storePath = store->parseStorePath(readLink(fileLink));
-        // FIXME
-        store->addTempRoot(*storePath);
-        if (store->isValidPath(*storePath)) {
-            auto ss = tokenizeString<vector<string>>(readFile(dataFile), "\n");
-            if (ss.size() >= 3 && ss[0] == url) {
-                time_t lastChecked;
-                if (string2Int(ss[2], lastChecked) && (uint64_t) lastChecked + request.ttl >= (uint64_t) time(0)) {
-                    skip = true;
-                    result.effectiveUri = request.uri;
-                    result.etag = ss[1];
-                } else if (!ss[1].empty()) {
-                    debug(format("verifying previous ETag '%1%'") % ss[1]);
-                    expectedETag = ss[1];
-                }
-            }
-        } else
-            storePath.reset();
-    }
-
-    if (!skip) {
-
-        try {
-            DownloadRequest request2(url);
-            request2.expectedETag = expectedETag;
-            auto res = download(request2);
-            result.effectiveUri = res.effectiveUri;
-            result.etag = res.etag;
-
-            if (!res.cached) {
-                StringSink sink;
-                dumpString(*res.data, sink);
-                Hash hash = hashString(request.expectedHash ? request.expectedHash.type : htSHA256, *res.data);
-                ValidPathInfo info(store->makeFixedOutputPath(false, hash, name));
-                info.narHash = hashString(htSHA256, *sink.s);
-                info.narSize = sink.s->size();
-                info.ca = makeFixedOutputCA(false, hash);
-                store->addToStore(info, sink.s, NoRepair, NoCheckSigs);
-                storePath = info.path.clone();
-            }
-
-            assert(storePath);
-            replaceSymlink(store->printStorePath(*storePath), fileLink);
-
-            writeFile(dataFile, url + "\n" + res.etag + "\n" + std::to_string(time(0)) + "\n");
-        } catch (DownloadError & e) {
-            if (!storePath) throw;
-            warn("warning: %s; using cached result", e.msg());
-            result.etag = expectedETag;
-        }
-    }
-
-    if (request.unpack) {
-        Path unpackedLink = cacheDir + "/" + ((std::string) storePath->to_string()) + "-unpacked";
-        PathLocks lock2({unpackedLink}, fmt("waiting for lock on '%1%'...", unpackedLink));
-        std::optional<StorePath> unpackedStorePath;
-        if (pathExists(unpackedLink)) {
-            unpackedStorePath = store->parseStorePath(readLink(unpackedLink));
-            // FIXME
-            store->addTempRoot(*unpackedStorePath);
-            if (!store->isValidPath(*unpackedStorePath))
-                unpackedStorePath.reset();
-        }
-        if (!unpackedStorePath) {
-            printInfo("unpacking '%s'...", url);
-            Path tmpDir = createTempDir();
-            AutoDelete autoDelete(tmpDir, true);
-            unpackTarfile(store->toRealPath(store->printStorePath(*storePath)), tmpDir);
-            auto members = readDirectory(tmpDir);
-            if (members.size() != 1)
-                throw nix::Error("tarball '%s' contains an unexpected number of top-level files", url);
-            auto topDir = tmpDir + "/" + members.begin()->name;
-            unpackedStorePath = store->addToStore(name, topDir, true, htSHA256, defaultPathFilter, NoRepair);
-        }
-        replaceSymlink(store->printStorePath(*unpackedStorePath), unpackedLink);
-        storePath = std::move(*unpackedStorePath);
-    }
-
-    if (expectedStorePath && *storePath != *expectedStorePath) {
-        unsigned int statusCode = 102;
-        Hash gotHash = request.unpack
-            ? hashPath(request.expectedHash.type, store->toRealPath(store->printStorePath(*storePath))).first
-            : hashFile(request.expectedHash.type, store->toRealPath(store->printStorePath(*storePath)));
-        throw nix::Error(statusCode, "hash mismatch in file downloaded from '%s':\n  wanted: %s\n  got:    %s",
-            url, request.expectedHash.to_string(), gotHash.to_string());
-    }
-
-    result.storePath = store->printStorePath(*storePath);
-    result.path = store->toRealPath(result.storePath);
-    return result;
-}
-
 
 bool isUri(const string & s)
 {

src/libstore/download.hh

@@ -65,28 +65,6 @@ struct DownloadResult
     uint64_t bodySize = 0;
 };
 
-struct CachedDownloadRequest
-{
-    std::string uri;
-    bool unpack = false;
-    std::string name;
-    Hash expectedHash;
-    unsigned int ttl;
-
-    CachedDownloadRequest(const std::string & uri);
-    CachedDownloadRequest() = delete;
-};
-
-struct CachedDownloadResult
-{
-    // Note: 'storePath' may be different from 'path' when using a
-    // chroot store.
-    Path storePath;
-    Path path;
-    std::optional<std::string> etag;
-    std::string effectiveUri;
-};
-
 class Store;
 
 struct Downloader
@@ -108,12 +86,6 @@ struct Downloader
        invoked on the thread of the caller. */
     void download(DownloadRequest && request, Sink & sink);
 
-    /* Check if the specified file is already in ~/.cache/nix/tarballs
-       and is more recent than ‘tarball-ttl’ seconds. Otherwise,
-       use the recorded ETag to verify if the server has a more
-       recent version, and if so, download it to the Nix store. */
-    CachedDownloadResult downloadCached(ref<Store> store, const CachedDownloadRequest & request);
-
     enum Error { NotFound, Forbidden, Misc, Transient, Interrupted };
 };
 
@@ -135,4 +107,7 @@ public:
 
 bool isUri(const string & s);
 
+/* Resolve deprecated 'channel:<foo>' URLs. */
+std::string resolveUri(const std::string & uri);
+
 }

src/libstore/globals.hh

@@ -351,12 +351,21 @@ public:
     Setting<Paths> pluginFiles{this, {}, "plugin-files",
         "Plugins to dynamically load at nix initialization time."};
 
+    Setting<std::string> githubAccessToken{this, "", "github-access-token",
+        "GitHub access token to get access to GitHub data through the GitHub API for github:<..> flakes."};
+
     Setting<Strings> experimentalFeatures{this, {}, "experimental-features",
         "Experimental Nix features to enable."};
 
     bool isExperimentalFeatureEnabled(const std::string & name);
 
     void requireExperimentalFeature(const std::string & name);
+
+    Setting<bool> allowDirty{this, true, "allow-dirty",
+        "Whether to allow dirty Git/Mercurial trees."};
+
+    Setting<bool> warnDirty{this, true, "warn-dirty",
+        "Whether to warn about dirty Git/Mercurial trees."};
 };
 
 

src/libstore/store-api.cc

@@ -6,6 +6,7 @@
 #include "thread-pool.hh"
 #include "json.hh"
 #include "derivations.hh"
+#include "url.hh"
 
 #include <future>
 
@@ -40,7 +41,7 @@ Path Store::followLinksToStore(std::string_view _path) const
         path = absPath(target, dirOf(path));
     }
     if (!isInStore(path))
-        throw Error(format("path '%1%' is not in the Nix store") % path);
+        throw NotInStore("path '%1%' is not in the Nix store", path);
     return path;
 }
 
@@ -866,27 +867,7 @@ std::pair<std::string, Store::Params> splitUriAndParams(const std::string & uri_
     Store::Params params;
     auto q = uri.find('?');
     if (q != std::string::npos) {
-        for (auto s : tokenizeString<Strings>(uri.substr(q + 1), "&")) {
+        params = decodeQuery(uri.substr(q + 1));
-            auto e = s.find('=');
-            if (e != std::string::npos) {
-                auto value = s.substr(e + 1);
-                std::string decoded;
-                for (size_t i = 0; i < value.size(); ) {
-                    if (value[i] == '%') {
-                        if (i + 2 >= value.size())
-                            throw Error("invalid URI parameter '%s'", value);
-                        try {
-                            decoded += std::stoul(std::string(value, i + 1, 2), 0, 16);
-                            i += 3;
-                        } catch (...) {
-                            throw Error("invalid URI parameter '%s'", value);
-                        }
-                    } else
-                        decoded += value[i++];
-                }
-                params[s.substr(0, e)] = decoded;
-            }
-        }
         uri = uri_.substr(0, q);
     }
     return {uri, params};

src/libstore/store-api.hh

@@ -28,6 +28,7 @@ MakeError(InvalidPath, Error);
 MakeError(Unsupported, Error);
 MakeError(SubstituteGone, Error);
 MakeError(SubstituterDisabled, Error);
+MakeError(NotInStore, Error);
 
 
 struct BasicDerivation;

src/libutil/types.hh

@@ -157,4 +157,12 @@ typedef list<Path> Paths;
 typedef set<Path> PathSet;
 
 
+/* Helper class to run code at startup. */
+template<typename T>
+struct OnStartup
+{
+    OnStartup(T && t) { t(); }
+};
+
+
 }

src/libutil/url.cc

@@ -0,0 +1,137 @@
+#include "url.hh"
+#include "util.hh"
+
+namespace nix {
+
+std::regex refRegex(refRegexS, std::regex::ECMAScript);
+std::regex revRegex(revRegexS, std::regex::ECMAScript);
+std::regex flakeIdRegex(flakeIdRegexS, std::regex::ECMAScript);
+
+ParsedURL parseURL(const std::string & url)
+{
+    static std::regex uriRegex(
+        "((" + schemeRegex + "):"
+        + "(?:(?://(" + authorityRegex + ")(" + absPathRegex + "))|(/?" + pathRegex + ")))"
+        + "(?:\\?(" + queryRegex + "))?"
+        + "(?:#(" + queryRegex + "))?",
+        std::regex::ECMAScript);
+
+    std::smatch match;
+
+    if (std::regex_match(url, match, uriRegex)) {
+        auto & base = match[1];
+        std::string scheme = match[2];
+        auto authority = match[3].matched
+            ? std::optional<std::string>(match[3]) : std::nullopt;
+        std::string path = match[4].matched ? match[4] : match[5];
+        auto & query = match[6];
+        auto & fragment = match[7];
+
+        auto isFile = scheme.find("file") != std::string::npos;
+
+        if (authority && *authority != "" && isFile)
+            throw Error("file:// URL '%s' has unexpected authority '%s'",
+                url, *authority);
+
+        if (isFile && path.empty())
+            path = "/";
+
+        return ParsedURL{
+            .url = url,
+            .base = base,
+            .scheme = scheme,
+            .authority = authority,
+            .path = path,
+            .query = decodeQuery(query),
+            .fragment = percentDecode(std::string(fragment))
+        };
+    }
+
+    else
+        throw BadURL("'%s' is not a valid URL", url);
+}
+
+std::string percentDecode(std::string_view in)
+{
+    std::string decoded;
+    for (size_t i = 0; i < in.size(); ) {
+        if (in[i] == '%') {
+            if (i + 2 >= in.size())
+                throw BadURL("invalid URI parameter '%s'", in);
+            try {
+                decoded += std::stoul(std::string(in, i + 1, 2), 0, 16);
+                i += 3;
+            } catch (...) {
+                throw BadURL("invalid URI parameter '%s'", in);
+            }
+        } else
+            decoded += in[i++];
+    }
+    return decoded;
+}
+
+std::map<std::string, std::string> decodeQuery(const std::string & query)
+{
+    std::map<std::string, std::string> result;
+
+    for (auto s : tokenizeString<Strings>(query, "&")) {
+        auto e = s.find('=');
+        if (e != std::string::npos)
+            result.emplace(
+                s.substr(0, e),
+                percentDecode(std::string_view(s).substr(e + 1)));
+    }
+
+    return result;
+}
+
+std::string percentEncode(std::string_view s)
+{
+    std::string res;
+    for (auto & c : s)
+        if ((c >= 'a' && c <= 'z')
+            || (c >= 'A' && c <= 'Z')
+            || (c >= '0' && c <= '9')
+            || strchr("-._~!$&'()*+,;=:@", c))
+            res += c;
+        else
+            res += fmt("%%%02x", (unsigned int) c);
+    return res;
+}
+
+std::string encodeQuery(const std::map<std::string, std::string> & ss)
+{
+    std::string res;
+    bool first = true;
+    for (auto & [name, value] : ss) {
+        if (!first) res += '&';
+        first = false;
+        res += percentEncode(name);
+        res += '=';
+        res += percentEncode(value);
+    }
+    return res;
+}
+
+std::string ParsedURL::to_string() const
+{
+    return
+        scheme
+        + ":"
+        + (authority ? "//" + *authority : "")
+        + path
+        + (query.empty() ? "" : "?" + encodeQuery(query))
+        + (fragment.empty() ? "" : "#" + percentEncode(fragment));
+}
+
+bool ParsedURL::operator ==(const ParsedURL & other) const
+{
+    return
+        scheme == other.scheme
+        && authority == other.authority
+        && path == other.path
+        && query == other.query
+        && fragment == other.fragment;
+}
+
+}

src/libutil/url.hh

@@ -0,0 +1,62 @@
+#pragma once
+
+#include "types.hh"
+
+#include <regex>
+
+namespace nix {
+
+struct ParsedURL
+{
+    std::string url;
+    std::string base; // URL without query/fragment
+    std::string scheme;
+    std::optional<std::string> authority;
+    std::string path;
+    std::map<std::string, std::string> query;
+    std::string fragment;
+
+    std::string to_string() const;
+
+    bool operator ==(const ParsedURL & other) const;
+};
+
+MakeError(BadURL, Error);
+
+std::string percentDecode(std::string_view in);
+
+std::map<std::string, std::string> decodeQuery(const std::string & query);
+
+ParsedURL parseURL(const std::string & url);
+
+// URI stuff.
+const static std::string pctEncoded = "(?:%[0-9a-fA-F][0-9a-fA-F])";
+const static std::string schemeRegex = "(?:[a-z+]+)";
+const static std::string ipv6AddressRegex = "(?:\\[[0-9a-fA-F:]+\\])";
+const static std::string unreservedRegex = "(?:[a-zA-Z0-9-._~])";
+const static std::string subdelimsRegex = "(?:[!$&'\"()*+,;=])";
+const static std::string hostnameRegex = "(?:(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + ")*)";
+const static std::string hostRegex = "(?:" + ipv6AddressRegex + "|" + hostnameRegex + ")";
+const static std::string userRegex = "(?:(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + "|:)*)";
+const static std::string authorityRegex = "(?:" + userRegex + "@)?" + hostRegex + "(?::[0-9]+)?";
+const static std::string pcharRegex = "(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + "|[:@])";
+const static std::string queryRegex = "(?:" + pcharRegex + "|[/? \"])*";
+const static std::string segmentRegex = "(?:" + pcharRegex + "+)";
+const static std::string absPathRegex = "(?:(?:/" + segmentRegex + ")*/?)";
+const static std::string pathRegex = "(?:" + segmentRegex + "(?:/" + segmentRegex + ")*/?)";
+
+// A Git ref (i.e. branch or tag name).
+const static std::string refRegexS = "[a-zA-Z0-9][a-zA-Z0-9_.-]*"; // FIXME: check
+extern std::regex refRegex;
+
+// A Git revision (a SHA-1 commit hash).
+const static std::string revRegexS = "[0-9a-fA-F]{40}";
+extern std::regex revRegex;
+
+// A ref or revision, or a ref followed by a revision.
+const static std::string refAndOrRevRegex = "(?:(" + revRegexS + ")|(?:(" + refRegexS + ")(?:/(" + revRegexS + "))?))";
+
+const static std::string flakeIdRegexS = "[a-zA-Z][a-zA-Z0-9_-]*";
+extern std::regex flakeIdRegex;
+
+}

src/nix-channel/nix-channel.cc

@@ -3,6 +3,7 @@
 #include "download.hh"
 #include "store-api.hh"
 #include "../nix/legacy.hh"
+#include "fetchers.hh"
 
 #include <fcntl.h>
 #include <regex>
@@ -86,12 +87,9 @@ static void update(const StringSet & channelNames)
         // We want to download the url to a file to see if it's a tarball while also checking if we
         // got redirected in the process, so that we can grab the various parts of a nix channel
         // definition from a consistent location if the redirect changes mid-download.
-        CachedDownloadRequest request(url);
+        auto result = fetchers::downloadFile(store, url, std::string(baseNameOf(url)), false);
-        request.ttl = 0;
+        auto filename = store->toRealPath(result.storePath);
-        auto dl = getDownloader();
+        url = result.effectiveUrl;
-        auto result = dl->downloadCached(store, request);
-        auto filename = result.path;
-        url = chomp(result.effectiveUri);
 
         // If the URL contains a version number, append it to the name
         // attribute (so that "nix-env -q" on the channels profile
@@ -114,11 +112,10 @@ static void update(const StringSet & channelNames)
         if (!unpacked) {
             // Download the channel tarball.
             try {
-                filename = dl->downloadCached(store, CachedDownloadRequest(url + "/nixexprs.tar.xz")).path;
+                filename = store->toRealPath(fetchers::downloadFile(store, url + "/nixexprs.tar.xz", "nixexprs.tar.xz", false).storePath);
             } catch (DownloadError & e) {
-                filename = dl->downloadCached(store, CachedDownloadRequest(url + "/nixexprs.tar.bz2")).path;
+                filename = store->toRealPath(fetchers::downloadFile(store, url + "/nixexprs.tar.bz2", "nixexprs.tar.bz2", false).storePath);
             }
-            chomp(filename);
         }
 
         // Regardless of where it came from, add the expression representing this channel to accumulated expression
@@ -185,6 +182,8 @@ static int _main(int argc, char ** argv)
             } else if (*arg == "--rollback") {
                 cmd = cRollback;
             } else {
+                if (hasPrefix(*arg, "-"))
+                    throw UsageError("unsupported argument '%s'", *arg);
                 args.push_back(std::move(*arg));
             }
             return true;

src/nix/local.mk

@@ -15,9 +15,9 @@ nix_SOURCES := \
   $(wildcard src/nix-prefetch-url/*.cc) \
   $(wildcard src/nix-store/*.cc) \
 
-nix_CXXFLAGS += -I src/libutil -I src/libstore -I src/libexpr -I src/libmain
+nix_CXXFLAGS += -I src/libutil -I src/libstore -I src/libfetchers -I src/libexpr -I src/libmain
 
-nix_LIBS = libexpr libmain libstore libutil libnixrust
+nix_LIBS = libexpr libmain libfetchers libstore libutil libnixrust
 
 nix_LDFLAGS = -pthread $(SODIUM_LIBS) $(EDITLINE_LIBS) $(BOOST_LDFLAGS) -lboost_context -lboost_thread -lboost_system
 

tests/fetchGit.sh

@@ -11,7 +11,7 @@ repo=$TEST_ROOT/git
 
 export _NIX_FORCE_HTTP=1
 
-rm -rf $repo ${repo}-tmp $TEST_HOME/.cache/nix/gitv2
+rm -rf $repo ${repo}-tmp $TEST_HOME/.cache/nix $TEST_ROOT/worktree $TEST_ROOT/shallow
 
 git init $repo
 git -C $repo config user.email "foobar@example.com"
@@ -25,8 +25,16 @@ rev1=$(git -C $repo rev-parse HEAD)
 
 echo world > $repo/hello
 git -C $repo commit -m 'Bla2' -a
+git -C $repo worktree add $TEST_ROOT/worktree
+echo hello >> $TEST_ROOT/worktree/hello
 rev2=$(git -C $repo rev-parse HEAD)
 
+# Fetch a worktree
+unset _NIX_FORCE_HTTP
+path0=$(nix eval --raw "(builtins.fetchGit file://$TEST_ROOT/worktree).outPath")
+export _NIX_FORCE_HTTP=1
+[[ $(tail -n 1 $path0/hello) = "hello" ]]
+
 # Fetch the default branch.
 path=$(nix eval --raw "(builtins.fetchGit file://$repo).outPath")
 [[ $(cat $path/hello) = world ]]
@@ -50,9 +58,6 @@ path2=$(nix eval --raw "(builtins.fetchGit file://$repo).outPath")
 [[ $(nix eval "(builtins.fetchGit file://$repo).revCount") = 2 ]]
 [[ $(nix eval --raw "(builtins.fetchGit file://$repo).rev") = $rev2 ]]
 
-# But with TTL 0, it should fail.
-(! nix eval --tarball-ttl 0 "(builtins.fetchGit file://$repo)" -vvvvv)
-
 # Fetching with a explicit hash should succeed.
 path2=$(nix eval --tarball-ttl 0 --raw "(builtins.fetchGit { url = file://$repo; rev = \"$rev2\"; }).outPath")
 [[ $path = $path2 ]]
@@ -74,6 +79,7 @@ echo bar > $repo/dir2/bar
 git -C $repo add dir1/foo
 git -C $repo rm hello
 
+unset _NIX_FORCE_HTTP
 path2=$(nix eval --raw "(builtins.fetchGit $repo).outPath")
 [ ! -e $path2/hello ]
 [ ! -e $path2/bar ]
@@ -110,9 +116,9 @@ path=$(nix eval --raw "(builtins.fetchGit file://$repo).outPath")
 git -C $repo checkout $rev2 -b dev
 echo dev > $repo/hello
 
-# File URI uses 'master' unless specified otherwise
+# File URI uses dirty tree unless specified otherwise
 path2=$(nix eval --raw "(builtins.fetchGit file://$repo).outPath")
-[[ $path = $path2 ]]
+[ $(cat $path2/hello) = dev ]
 
 # Using local path with branch other than 'master' should work when clean or dirty
 path3=$(nix eval --raw "(builtins.fetchGit $repo).outPath")
@@ -131,9 +137,9 @@ path5=$(nix eval --raw "(builtins.fetchGit { url = $repo; ref = \"dev\"; }).outP
 
 
 # Nuke the cache
-rm -rf $TEST_HOME/.cache/nix/gitv2
+rm -rf $TEST_HOME/.cache/nix
 
-# Try again, but without 'git' on PATH
+# Try again, but without 'git' on PATH. This should fail.
 NIX=$(command -v nix)
 # This should fail
 (! PATH= $NIX eval --raw "(builtins.fetchGit { url = $repo; ref = \"dev\"; }).outPath" )
@@ -141,3 +147,13 @@ NIX=$(command -v nix)
 # Try again, with 'git' available.  This should work.
 path5=$(nix eval --raw "(builtins.fetchGit { url = $repo; ref = \"dev\"; }).outPath")
 [[ $path3 = $path5 ]]
+
+# Fetching a shallow repo shouldn't work by default, because we can't
+# return a revCount.
+git clone --depth 1 file://$repo $TEST_ROOT/shallow
+(! nix eval --raw "(builtins.fetchGit { url = $TEST_ROOT/shallow; ref = \"dev\"; }).outPath")
+
+# But you can request a shallow clone, which won't return a revCount.
+path6=$(nix eval --raw "(builtins.fetchTree { type = \"git\"; url = \"file://$TEST_ROOT/shallow\"; ref = \"dev\"; shallow = true; }).outPath")
+[[ $path3 = $path6 ]]
+[[ $(nix eval "(builtins.fetchTree { type = \"git\"; url = \"file://$TEST_ROOT/shallow\"; ref = \"dev\"; shallow = true; }).revCount or 123") == 123 ]]

tests/init.sh

@@ -17,7 +17,7 @@ cat > "$NIX_CONF_DIR"/nix.conf <<EOF
 build-users-group =
 keep-derivations = false
 sandbox = false
-experimental-features = nix-command
+experimental-features = nix-command flakes
 include nix.conf.extra
 EOF
 

tests/tarball.sh

@@ -10,6 +10,8 @@ mkdir -p $tarroot
 cp dependencies.nix $tarroot/default.nix
 cp config.nix dependencies.builder*.sh $tarroot/
 
+hash=$(nix hash-path $tarroot)
+
 test_tarball() {
     local ext="$1"
     local compressor="$2"
@@ -25,6 +27,11 @@ test_tarball() {
 
     nix-build -o $TEST_ROOT/result -E "import (fetchTarball file://$tarball)"
 
+    nix-build --experimental-features flakes -o $TEST_ROOT/result -E "import (fetchTree file://$tarball)"
+    nix-build --experimental-features flakes -o $TEST_ROOT/result -E "import (fetchTree { type = \"tarball\"; url = file://$tarball; })"
+    nix-build --experimental-features flakes -o $TEST_ROOT/result -E "import (fetchTree { type = \"tarball\"; url = file://$tarball; narHash = \"$hash\"; })"
+    nix-build --experimental-features flakes -o $TEST_ROOT/result -E "import (fetchTree { type = \"tarball\"; url = file://$tarball; narHash = \"sha256-xdKv2pq/IiwLSnBBJXW8hNowI4MrdZfW+SYqDQs7Tzc=\"; })" 2>&1 | grep 'NAR hash mismatch in input'
+
     nix-instantiate --eval -E '1 + 2' -I fnord=file://no-such-tarball.tar$ext
     nix-instantiate --eval -E 'with <fnord/xyzzy>; 1 + 2' -I fnord=file://no-such-tarball$ext
     (! nix-instantiate --eval -E '<fnord/xyzzy> 1' -I fnord=file://no-such-tarball$ext)