Add copy_symlinks keyword to Tar.tree_hash

Project.toml

@@ -16,6 +16,7 @@ Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 SimpleBufferStream = "777ac1f9-54b0-4bf8-805c-2214025038e7"
 Tar_jll = "9b64493d-8859-5bf3-93d7-7c32dd38186f"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+CodecZlib = "944b1d66-785c-5afd-91f1-9de20f533193"
 
 [targets]
-test = ["Random", "SimpleBufferStream", "Tar_jll", "Test"]
+test = ["Random", "SimpleBufferStream", "Tar_jll", "Test", "CodecZlib"]

src/Tar.jl

@@ -342,6 +342,7 @@ end
         tarball    :: Union{AbstractString, AbstractCmd, IO}
         algorithm  :: AbstractString
         skip_empty :: Bool
+        copy_symlinks :: Bool
 
 Compute a tree hash value for the file tree that the tarball contains. By
 default, this uses git's tree hashing algorithm with the SHA1 secure hash
@@ -389,21 +390,26 @@ hash, the hash value that you get will match the hash value computed by
 are hashing trees that may contain empty directories (i.e. do not come from a
 git repo), however, it is recommended that you hash them using a tool (such as
 this one) that does not ignore empty directories.
+
+If `copy_symlinks` is true, symlinks in the tarfile will be followed and the
+target hashes will be copied. This is useful for checking what the hash would
+be when using `Tar.extract` with `copy_symlinks = true`.
 """
 function tree_hash(
     predicate::Function,
     tarball::ArgRead;
     algorithm::AbstractString = "git-sha1",
     skip_empty::Bool = false,
+    copy_symlinks::Bool = false,
 )
     check_tree_hash_tarball(tarball)
     if algorithm == "git-sha1"
         return arg_read(tarball) do tar
-            git_tree_hash(predicate, tar, SHA.SHA1_CTX, skip_empty)
+            git_tree_hash(predicate, tar, SHA.SHA1_CTX, skip_empty, copy_symlinks)
         end
     elseif algorithm == "git-sha256"
         return arg_read(tarball) do tar
-            git_tree_hash(predicate, tar, SHA.SHA256_CTX, skip_empty)
+            git_tree_hash(predicate, tar, SHA.SHA256_CTX, skip_empty, copy_symlinks)
         end
     else
         error("invalid tree hashing algorithm: $algorithm")
@@ -414,12 +420,14 @@ function tree_hash(
     tarball::ArgRead;
     algorithm::AbstractString = "git-sha1",
     skip_empty::Bool = false,
+    copy_symlinks::Bool = false,
 )
     tree_hash(
         true_predicate,
         tarball,
         algorithm = algorithm,
         skip_empty = skip_empty,
+        copy_symlinks = copy_symlinks
     )
 end
 

src/extract.jl

@@ -207,12 +207,13 @@ function git_tree_hash(
     predicate::Function,
     tar::IO,
     ::Type{HashType},
-    skip_empty::Bool;
+    skip_empty::Bool,
+    copy_symlinks::Bool = false;
     buf::Vector{UInt8} = Vector{UInt8}(undef, DEFAULT_BUFFER_SIZE),
 ) where HashType <: SHA.SHA_CTX
     # build tree with leaves for files and symlinks
     tree = Dict{String,Any}()
-    read_tarball(predicate, tar; buf=buf) do hdr, parts
+    paths = read_tarball(predicate, tar; buf=buf) do hdr, parts
         isempty(parts) && return
         name = pop!(parts)
         node = tree
@@ -229,9 +230,14 @@ function git_tree_hash(
             end
             return
         elseif hdr.type == :symlink
-            mode = "120000"
-            hash = git_object_hash("blob", HashType) do io
-                write(io, hdr.link)
+            if copy_symlinks
+                mode = "120000"
+                hash = hdr.link
+            else
+                mode = "120000"
+                hash = git_object_hash("blob", HashType) do io
+                    write(io, hdr.link)
+                end
             end
         elseif hdr.type == :hardlink
             mode = iszero(hdr.mode & 0o100) ? "100644" : "100755"
@@ -249,6 +255,57 @@ function git_tree_hash(
         node[name] = (mode, hash)
     end
 
+    if copy_symlinks
+        # resolve the internal targets of symlinks
+        for (path, what) in paths
+            what isa String || continue
+            target = link_target(paths, path, what)
+            paths[path] = something(target, :symlink)
+        end
+
+        for (path, what) in paths
+            what isa AbstractString || continue
+            paths[path] = follow_symlink_chain([path], what, paths)
+        end
+
+        # use paths to index into the tree
+        function get_tree_index(tree::Dict, path::AbstractString)
+            node = tree
+            parts = splitpath(path)
+            for part in parts
+                node = node[part]
+            end
+            return node
+        end
+        function set_tree_index!(tree::Dict, value, path::AbstractString)
+            node = tree
+            parts = splitpath(path)
+            for part in parts[1:end-1]
+                node = node[part]
+            end
+            node[parts[end]] = value
+        end
+        function prune_tree_index!(tree::Dict, path::AbstractString)
+            node = tree
+            parts = splitpath(path)
+            for part in parts[1:end-1]
+                node = node[part]
+            end
+            delete!(node, parts[end])
+        end
+
+        # copy hashes
+        for (path, what) in paths
+            if what isa AbstractString
+                what_hash = get_tree_index(tree, what)
+                set_tree_index!(tree, what_hash, path)
+            elseif what == :symlink
+                # external symlink
+                prune_tree_index!(tree, path)
+            end
+        end
+    end
+
     # prune directories that don't contain any files
     if skip_empty
         prune_empty!(node::Tuple) = true

test/runtests.jl

@@ -83,11 +83,25 @@ end
     @testset "Tar.tree_hash" begin
         arg_readers(tarball) do tar
             @arg_test tar @test Tar.tree_hash(tar) == hash
+            @arg_test tar @test Tar.tree_hash(tar; copy_symlinks=true) != hash
             @arg_test tar @test empty_tree_sha1 == Tar.tree_hash(hdr->false, tar)
+            @arg_test tar @test empty_tree_sha1 == Tar.tree_hash(hdr->false, tar; copy_symlinks=true)
             @arg_test tar @test empty_tree_sha1 ==
                 Tar.tree_hash(hdr->false, tar, algorithm="git-sha1")
+            @arg_test tar @test empty_tree_sha1 ==
+                Tar.tree_hash(hdr->false, tar, algorithm="git-sha1", copy_symlinks=true)
             @arg_test tar @test empty_tree_sha256 ==
                 Tar.tree_hash(hdr->false, tar, algorithm="git-sha256")
+            @arg_test tar @test empty_tree_sha256 ==
+                Tar.tree_hash(hdr->false, tar, algorithm="git-sha256", copy_symlinks=true)
+        end
+        NON_STDLIB_TESTS && begin
+            open(GzipDecompressorStream, "data/iso_codes.v4.11.0.any.tar.gz") do io
+                @test Tar.tree_hash(io) == "71f68a3d55d73f2e15a3969c241fae2349b1feb5"
+            end
+            open(GzipDecompressorStream, "data/iso_codes.v4.11.0.any.tar.gz") do io
+                @test Tar.tree_hash(io; copy_symlinks=true) == "409d6ac4c02dae43ff4fe576b5c5820d0386fb3f"
+            end
         end
     end
     @testset "Tar.list & check properties" begin

test/setup.jl

@@ -10,6 +10,7 @@ const NON_STDLIB_TESTS = Main == @__MODULE__
 
 if NON_STDLIB_TESTS
     using SimpleBufferStream
+    using CodecZlib
 
     using Tar_jll
     if isdefined(Tar_jll, :tar)