Merge pull request #4638 from orbekk/read_head

Resolve reference for remote git repositories (makes fetchGit work with non-'master' branch)

5 files changed, 309 insertions, 128 deletions

diff --git a/src/libfetchers/git-utils.cc b/src/libfetchers/git-utils.cc
new file mode 100644
index 000000000..060077098
--- /dev/null
+++ b/src/libfetchers/git-utils.cc
@@ -0,0 +1,25 @@
+#include "git-utils.hh"
+
+#include <regex>
+
+std::optional<std::string> parseListReferenceHeadRef(std::string_view line) {
+    const static std::regex head_ref_regex("^ref: ([^\\s]+)\\t+HEAD$");
+    std::match_results<std::string_view::const_iterator> match;
+    if (std::regex_match(line.cbegin(), line.cend(), match, head_ref_regex)) {
+        return match[1];
+    } else {
+        return std::nullopt;
+    }
+}
+
+std::optional<std::string> parseListReferenceForRev(std::string_view rev, std::string_view line) {
+    const static std::regex rev_regex("^([^\\t]+)\\t+(.*)$");
+    std::match_results<std::string_view::const_iterator> match;
+    if (!std::regex_match(line.cbegin(), line.cend(), match, rev_regex)) {
+        return std::nullopt;
+    }
+    if (rev != match[2].str()) {
+        return std::nullopt;
+    }
+    return match[1];
+}
diff --git a/src/libfetchers/git-utils.hh b/src/libfetchers/git-utils.hh
new file mode 100644
index 000000000..946a68a9e
--- /dev/null
+++ b/src/libfetchers/git-utils.hh
@@ -0,0 +1,23 @@
+#pragma once
+
+#include <string>
+#include <string_view>
+#include <optional>
+
+// Parses the HEAD ref as reported by `git ls-remote --symref`
+//
+// Returns the head branch name as reported by `git ls-remote --symref`, e.g., if
+// ls-remote returns the output below, "main" is returned based on the ref line.
+//
+//   ref: refs/heads/main       HEAD
+//
+// If the repository is in 'detached head' state (HEAD is pointing to a rev
+// instead of a branch), parseListReferenceForRev("HEAD") may be used instead.
+std::optional<std::string> parseListReferenceHeadRef(std::string_view line);
+
+// Parses a reference line from `git ls-remote --symref`, e.g.,
+// parseListReferenceForRev("refs/heads/master", line) will return 6926...
+// given the line below.
+//
+// 6926beab444c33fb57b21819b6642d032016bb1e	refs/heads/master
+std::optional<std::string> parseListReferenceForRev(std::string_view rev, std::string_view line);
diff --git a/src/libfetchers/git.cc b/src/libfetchers/git.cc
index af40990e5..9d4348cf1 100644
--- a/src/libfetchers/git.cc
+++ b/src/libfetchers/git.cc
@@ -5,15 +5,20 @@
 #include "store-api.hh"
 #include "url-parts.hh"
 #include "pathlocks.hh"
+#include "util.hh"
+#include "git-utils.hh"
 
 #include "fetch-settings.hh"
 
+#include <regex>
+#include <string.h>
 #include <sys/time.h>
 #include <sys/wait.h>
 
 using namespace std::string_literals;
 
 namespace nix::fetchers {
+namespace {
 
 // Explicit initial branch of our bare repo to suppress warnings from new version of git.
 // The value itself does not matter, since we always fetch a specific revision or branch.
@@ -21,7 +26,7 @@ namespace nix::fetchers {
 // old version of git, which will ignore unrecognized `-c` options.
 const std::string gitInitialBranch = "__nix_dummy_branch";
 
-static std::string getGitDir()
+std::string getGitDir()
 {
     auto gitDir = getEnv("GIT_DIR");
     if (!gitDir) {
@@ -30,16 +35,222 @@ static std::string getGitDir()
     return *gitDir;
 }
 
-static std::string readHead(const Path & path)
+bool isCacheFileWithinTtl(const time_t now, const struct stat& st) {
+    return st.st_mtime + settings.tarballTtl > now;
+}
+
+bool touchCacheFile(const Path& path, const time_t& touch_time)
+{
+  struct timeval times[2];
+  times[0].tv_sec = touch_time;
+  times[0].tv_usec = 0;
+  times[1].tv_sec = touch_time;
+  times[1].tv_usec = 0;
+
+  return lutimes(path.c_str(), times) == 0;
+}
+
+Path getCachePath(std::string key)
+{
+    return getCacheDir() + "/nix/gitv3/" +
+        hashString(htSHA256, key).to_string(Base32, false);
+}
+
+// Returns the name of the HEAD branch.
+//
+// Returns the head branch name as reported by git ls-remote --symref, e.g., if
+// ls-remote returns the output below, "main" is returned based on the ref line.
+//
+//   ref: refs/heads/main       HEAD
+//   ...
+std::optional<std::string> readHead(const Path & path)
+{
+    auto [exit_code, output] = runProgram(RunOptions {
+        .program = "git",
+        .args = {"ls-remote", "--symref", path},
+    });
+    if (exit_code != 0) {
+        return std::nullopt;
+    }
+
+    std::string_view line = output;
+    line = line.substr(0, line.find("\n"));
+    if (const auto ref = parseListReferenceHeadRef(line); ref) {
+        debug("resolved HEAD ref '%s' for repo '%s'", *ref, path);
+        return *ref;
+    }
+    if (const auto rev = parseListReferenceForRev("HEAD", line); rev) {
+        debug("resolved HEAD rev '%s' for repo '%s'", *rev, path);
+        return *rev;
+    }
+    return std::nullopt;
+}
+
+// Persist the HEAD ref from the remote repo in the local cached repo.
+bool storeCachedHead(const std::string& actualUrl, const std::string& headRef)
 {
-    return chomp(runProgram("git", true, { "-C", path, "--git-dir", ".git", "rev-parse", "--abbrev-ref", "HEAD" }));
+    Path cacheDir = getCachePath(actualUrl);
+    try {
+        runProgram("git", true, { "-C", cacheDir, "symbolic-ref", "--", "HEAD", headRef });
+    } catch (ExecError &e) {
+        if (!WIFEXITED(e.status)) throw;
+        return false;
+    }
+    /* No need to touch refs/HEAD, because `git symbolic-ref` updates the mtime. */
+    return true;
 }
 
-static bool isNotDotGitDirectory(const Path & path)
+std::optional<std::string> readHeadCached(const std::string& actualUrl)
+{
+    // Create a cache path to store the branch of the HEAD ref. Append something
+    // in front of the URL to prevent collision with the repository itself.
+    Path cacheDir = getCachePath(actualUrl);
+    Path headRefFile = cacheDir + "/HEAD";
+
+    time_t now = time(0);
+    struct stat st;
+    std::optional<std::string> cachedRef;
+    if (stat(headRefFile.c_str(), &st) == 0) {
+        cachedRef = readHead(cacheDir);
+        if (cachedRef != std::nullopt &&
+            *cachedRef != gitInitialBranch &&
+            isCacheFileWithinTtl(now, st)) {
+            debug("using cached HEAD ref '%s' for repo '%s'", *cachedRef, actualUrl);
+            return cachedRef;
+        }
+    }
+
+    auto ref = readHead(actualUrl);
+    if (ref) {
+        return ref;
+    }
+
+    if (cachedRef) {
+        // If the cached git ref is expired in fetch() below, and the 'git fetch'
+        // fails, it falls back to continuing with the most recent version.
+        // This function must behave the same way, so we return the expired
+        // cached ref here.
+        warn("could not get HEAD ref for repository '%s'; using expired cached ref '%s'", actualUrl, *cachedRef);
+        return *cachedRef;
+    }
+
+    return std::nullopt;
+}
+
+bool isNotDotGitDirectory(const Path & path)
 {
     return baseNameOf(path) != ".git";
 }
 
+struct WorkdirInfo
+{
+    bool clean = false;
+    bool hasHead = false;
+};
+
+// Returns whether a git workdir is clean and has commits.
+WorkdirInfo getWorkdirInfo(const Input & input, const Path & workdir)
+{
+    const bool submodules = maybeGetBoolAttr(input.attrs, "submodules").value_or(false);
+    auto gitDir = getGitDir();
+
+    auto env = getEnv();
+    // Set LC_ALL to C: because we rely on the error messages from git rev-parse to determine what went wrong
+    // that way unknown errors can lead to a failure instead of continuing through the wrong code path
+    env["LC_ALL"] = "C";
+
+    /* Check whether HEAD points to something that looks like a commit,
+       since that is the refrence we want to use later on. */
+    auto result = runProgram(RunOptions {
+        .program = "git",
+        .args = { "-C", workdir, "--git-dir", gitDir, "rev-parse", "--verify", "--no-revs", "HEAD^{commit}" },
+        .environment = env,
+        .mergeStderrToStdout = true
+    });
+    auto exitCode = WEXITSTATUS(result.first);
+    auto errorMessage = result.second;
+
+    if (errorMessage.find("fatal: not a git repository") != std::string::npos) {
+        throw Error("'%s' is not a Git repository", workdir);
+    } else if (errorMessage.find("fatal: Needed a single revision") != std::string::npos) {
+        // indicates that the repo does not have any commits
+        // we want to proceed and will consider it dirty later
+    } else if (exitCode != 0) {
+        // any other errors should lead to a failure
+        throw Error("getting the HEAD of the Git tree '%s' failed with exit code %d:\n%s", workdir, exitCode, errorMessage);
+    }
+
+    bool clean = false;
+    bool hasHead = exitCode == 0;
+
+    try {
+        if (hasHead) {
+            // Using git diff is preferrable over lower-level operations here,
+            // because its conceptually simpler and we only need the exit code anyways.
+            auto gitDiffOpts = Strings({ "-C", workdir, "diff", "HEAD", "--quiet"});
+            if (!submodules) {
+                // Changes in submodules should only make the tree dirty
+                // when those submodules will be copied as well.
+                gitDiffOpts.emplace_back("--ignore-submodules");
+            }
+            gitDiffOpts.emplace_back("--");
+            runProgram("git", true, gitDiffOpts);
+
+            clean = true;
+        }
+    } catch (ExecError & e) {
+        if (!WIFEXITED(e.status) || WEXITSTATUS(e.status) != 1) throw;
+    }
+
+    return WorkdirInfo { .clean = clean, .hasHead = hasHead };
+}
+
+std::pair<StorePath, Input> fetchFromWorkdir(ref<Store> store, Input & input, const Path & workdir, const WorkdirInfo & workdirInfo)
+{
+    const bool submodules = maybeGetBoolAttr(input.attrs, "submodules").value_or(false);
+
+    if (!fetchSettings.allowDirty)
+        throw Error("Git tree '%s' is dirty", workdir);
+
+    if (fetchSettings.warnDirty)
+        warn("Git tree '%s' is dirty", workdir);
+
+    auto gitOpts = Strings({ "-C", workdir, "ls-files", "-z" });
+    if (submodules)
+        gitOpts.emplace_back("--recurse-submodules");
+
+    auto files = tokenizeString<std::set<std::string>>(
+        runProgram("git", true, gitOpts), "\0"s);
+
+    Path actualPath(absPath(workdir));
+
+    PathFilter filter = [&](const Path & p) -> bool {
+        assert(hasPrefix(p, actualPath));
+        std::string file(p, actualPath.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(input.getName(), actualPath, FileIngestionMethod::Recursive, htSHA256, filter);
+
+    // FIXME: maybe we should use the timestamp of the last
+    // modified dirty file?
+    input.attrs.insert_or_assign(
+        "lastModified",
+        workdirInfo.hasHead ? std::stoull(runProgram("git", true, { "-C", actualPath, "log", "-1", "--format=%ct", "--no-show-signature", "HEAD" })) : 0);
+
+    return {std::move(storePath), input};
+}
+}  // end namespace
+
 struct GitInputScheme : InputScheme
 {
     std::optional<Input> inputFromURL(const ParsedURL & url) override
@@ -234,106 +445,16 @@ struct GitInputScheme : InputScheme
         auto [isLocal, actualUrl_] = getActualUrl(input);
         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 this is a local directory and no ref or revision is given,
+           allow fetching directly from a dirty workdir. */
         if (!input.getRef() && !input.getRev() && isLocal) {
-            bool clean = false;
-
-            auto env = getEnv();
-            // Set LC_ALL to C: because we rely on the error messages from git rev-parse to determine what went wrong
-            // that way unknown errors can lead to a failure instead of continuing through the wrong code path
-            env["LC_ALL"] = "C";
-
-            /* Check whether HEAD points to something that looks like a commit,
-               since that is the refrence we want to use later on. */
-            auto result = runProgram(RunOptions {
-                .program = "git",
-                .args = { "-C", actualUrl, "--git-dir", gitDir, "rev-parse", "--verify", "--no-revs", "HEAD^{commit}" },
-                .environment = env,
-                .mergeStderrToStdout = true
-            });
-            auto exitCode = WEXITSTATUS(result.first);
-            auto errorMessage = result.second;
-
-            if (errorMessage.find("fatal: not a git repository") != std::string::npos) {
-                throw Error("'%s' is not a Git repository", actualUrl);
-            } else if (errorMessage.find("fatal: Needed a single revision") != std::string::npos) {
-                // indicates that the repo does not have any commits
-                // we want to proceed and will consider it dirty later
-            } else if (exitCode != 0) {
-                // any other errors should lead to a failure
-                throw Error("getting the HEAD of the Git tree '%s' failed with exit code %d:\n%s", actualUrl, exitCode, errorMessage);
-            }
-
-            bool hasHead = exitCode == 0;
-            try {
-                if (hasHead) {
-                    // Using git diff is preferrable over lower-level operations here,
-                    // because its conceptually simpler and we only need the exit code anyways.
-                    auto gitDiffOpts = Strings({ "-C", actualUrl, "--git-dir", gitDir, "diff", "HEAD", "--quiet"});
-                    if (!submodules) {
-                        // Changes in submodules should only make the tree dirty
-                        // when those submodules will be copied as well.
-                        gitDiffOpts.emplace_back("--ignore-submodules");
-                    }
-                    gitDiffOpts.emplace_back("--");
-                    runProgram("git", true, gitDiffOpts);
-
-                    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 (!fetchSettings.allowDirty)
-                    throw Error("Git tree '%s' is dirty", actualUrl);
-
-                if (fetchSettings.warnDirty)
-                    warn("Git tree '%s' is dirty", actualUrl);
-
-                auto gitOpts = Strings({ "-C", actualUrl, "--git-dir", gitDir, "ls-files", "-z" });
-                if (submodules)
-                    gitOpts.emplace_back("--recurse-submodules");
-
-                auto files = tokenizeString<std::set<std::string>>(
-                    runProgram("git", true, gitOpts), "\0"s);
-
-                Path actualPath(absPath(actualUrl));
-
-                PathFilter filter = [&](const Path & p) -> bool {
-                    assert(hasPrefix(p, actualPath));
-                    std::string file(p, actualPath.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(input.getName(), actualPath, FileIngestionMethod::Recursive, htSHA256, filter);
-
-                // FIXME: maybe we should use the timestamp of the last
-                // modified dirty file?
-                input.attrs.insert_or_assign(
-                    "lastModified",
-                    hasHead ? std::stoull(runProgram("git", true, { "-C", actualPath, "--git-dir", gitDir, "log", "-1", "--format=%ct", "--no-show-signature", "HEAD" })) : 0);
-
-                return {std::move(storePath), input};
+            auto workdirInfo = getWorkdirInfo(input, actualUrl);
+            if (!workdirInfo.clean) {
+                return fetchFromWorkdir(store, input, actualUrl, workdirInfo);
             }
         }
 
-        if (!input.getRef()) input.attrs.insert_or_assign("ref", isLocal ? readHead(actualUrl) : "master");
-
-        Attrs unlockedAttrs({
+        const Attrs unlockedAttrs({
             {"type", cacheType},
             {"name", name},
             {"url", actualUrl},
@@ -343,14 +464,30 @@ struct GitInputScheme : InputScheme
         Path repoDir;
 
         if (isLocal) {
+            if (!input.getRef()) {
+                auto head = readHead(actualUrl);
+                if (!head) {
+                    warn("could not read HEAD ref from repo at '%s', using 'master'", actualUrl);
+                    head = "master";
+                }
+                input.attrs.insert_or_assign("ref", *head);
+            }
 
             if (!input.getRev())
                 input.attrs.insert_or_assign("rev",
                     Hash::parseAny(chomp(runProgram("git", true, { "-C", actualUrl, "--git-dir", gitDir, "rev-parse", *input.getRef() })), htSHA1).gitRev());
 
             repoDir = actualUrl;
-
         } else {
+            const bool useHeadRef = !input.getRef();
+            if (useHeadRef) {
+                auto head = readHeadCached(actualUrl);
+                if (!head) {
+                    warn("could not read HEAD ref from repo at '%s', using 'master'", actualUrl);
+                    head = "master";
+                }
+                input.attrs.insert_or_assign("ref", *head);
+            }
 
             if (auto res = getCache()->lookup(store, unlockedAttrs)) {
                 auto rev2 = Hash::parseAny(getStrAttr(res->first, "rev"), htSHA1);
@@ -360,7 +497,7 @@ struct GitInputScheme : InputScheme
                 }
             }
 
-            Path cacheDir = getCacheDir() + "/nix/gitv3/" + hashString(htSHA256, actualUrl).to_string(Base32, false);
+            Path cacheDir = getCachePath(actualUrl);
             repoDir = cacheDir;
             gitDir = ".";
 
@@ -400,7 +537,7 @@ struct GitInputScheme : InputScheme
                        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;
+                        !isCacheFileWithinTtl(now, st);
                 }
             }
 
@@ -424,13 +561,10 @@ struct GitInputScheme : InputScheme
                     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 (!touchCacheFile(localRefFile, now))
+                    warn("could not update mtime for file '%s': %s", localRefFile, strerror(errno));
+                if (useHeadRef && !storeCachedHead(actualUrl, *input.getRef()))
+                    warn("could not update cached head '%s' for '%s'", *input.getRef(), actualUrl);
             }
 
             if (!input.getRev())
diff --git a/src/libfetchers/github.cc b/src/libfetchers/github.cc
index 58b6e7c04..1bdf2759f 100644
--- a/src/libfetchers/github.cc
+++ b/src/libfetchers/github.cc
@@ -4,7 +4,7 @@
 #include "store-api.hh"
 #include "types.hh"
 #include "url-parts.hh"
-
+#include "git-utils.hh"
 #include "fetchers.hh"
 #include "fetch-settings.hh"
 
@@ -383,35 +383,29 @@ struct SourceHutInputScheme : GitArchiveInputScheme
             std::string line;
             getline(is, line);
 
-            auto ref_index = line.find("ref: ");
-            if (ref_index == std::string::npos) {
+            auto r = parseListReferenceHeadRef(line);
+            if (!r) {
                 throw BadURL("in '%d', couldn't resolve HEAD ref '%d'", input.to_string(), ref);
             }
-
-            ref_uri = line.substr(ref_index+5, line.length()-1);
-        } else
+            ref_uri = *r;
+        } else {
             ref_uri = fmt("refs/(heads|tags)/%s", ref);
+        }
 
         auto file = store->toRealPath(
             downloadFile(store, fmt("%s/info/refs", base_url), "source", false, headers).storePath);
         std::ifstream is(file);
 
         std::string line;
-        std::string id;
-        while(getline(is, line)) {
-            // Append $ to avoid partial name matches
-            std::regex pattern(fmt("%s$", ref_uri));
-
-            if (std::regex_search(line, pattern)) {
-                id = line.substr(0, line.find('\t'));
-                break;
-            }
+        std::optional<std::string> id;
+        while(!id && getline(is, line)) {
+            id = parseListReferenceForRev(ref_uri, line);
         }
 
-        if(id.empty())
+        if(!id)
             throw BadURL("in '%d', couldn't find ref '%d'", input.to_string(), ref);
 
-        auto rev = Hash::parseAny(id, htSHA1);
+        auto rev = Hash::parseAny(*id, htSHA1);
         debug("HEAD revision for '%s' is %s", fmt("%s/%s", base_url, ref), rev.gitRev());
         return rev;
     }
diff --git a/src/nix/flake.md b/src/nix/flake.md
index 7d179a6c4..c8251eb74 100644
--- a/src/nix/flake.md
+++ b/src/nix/flake.md
@@ -153,7 +153,7 @@ Currently the `type` attribute can be one of the following:
   git(+http|+https|+ssh|+git|+file|):(//<server>)?<path>(\?<params>)?
   ```
 
-  The `ref` attribute defaults to `master`.
+  The `ref` attribute defaults to resolving the `HEAD` reference.
 
   The `rev` attribute must denote a commit that exists in the branch
   or tag specified by the `ref` attribute, since Nix doesn't do a full
@@ -161,6 +161,11 @@ Currently the `type` attribute can be one of the following:
   doesn't allow fetching a `rev` without a known `ref`). The default
   is the commit currently pointed to by `ref`.
 
+  When `git+file` is used without specifying `ref` or `rev`, files are
+  fetched directly from the local `path` as long as they have been added
+  to the Git repository. If there are uncommitted changes, the reference
+  is treated as dirty and a warning is printed.
+
   For example, the following are valid Git flake references:
 
   * `git+https://example.org/my/repo`