Push options from cli to the Nar class

That will ease testing. The cli becomes a pass-through.

This also bootstraps tests reusing existing test cases from before (using the nix
binary).

Refs. swh/meta#4979

swh/loader/core/nar.py

@@ -8,7 +8,6 @@ import io
 import os
 from pathlib import Path
 import stat
-from typing import Callable
 
 import click
 
@@ -84,9 +83,33 @@ class Nar:
 
     """
 
-    def __init__(self, exclude_vcs: bool, updater: Callable, debug: bool = False):
+    def __init__(
+        self,
+        hash_algo: str,
+        format_output: str = "hex",
+        exclude_vcs: bool = False,
+        debug: bool = False,
+    ):
+        self.updater = hashlib.sha256() if hash_algo == "sha256" else hashlib.sha1()
+        format_output = format_output.lower()
         self.exclude_vcs = exclude_vcs
-        self.updater_fn = updater
+
+        def identity(hsh: str) -> str:
+            return hsh
+
+        def convert_b64(hsh: str) -> str:
+            return base64.b64encode(bytes.fromhex(hsh)).decode().lower()
+
+        def convert_b32(hsh: str) -> str:
+            return base64.b32encode(bytes.fromhex(hsh)).decode().lower()
+
+        convert_fn = {
+            "hex": identity,
+            "base64": convert_b64,
+            "base32": convert_b32,
+        }
+
+        self.convert_fn = convert_fn[format_output]
 
         self.__debug = debug
         self.__indent = 0
@@ -114,14 +137,14 @@ class Nar:
             raise ValueError("not string nor file")
 
         blen = length.to_bytes(8, byteorder="little")  # 64-bit little endian
-        self.updater_fn(blen)
+        self.updater.update(blen)
 
         # first part of 'pad'
         if isinstance(thing, str):
-            self.updater_fn(byte_sequence)
+            self.updater.update(byte_sequence)
         elif isinstance(thing, io.BufferedReader):
             for chunk in iter(lambda: thing.read(CHUNK_SIZE), b""):
-                self.updater_fn(chunk)
+                self.updater.update(chunk)
 
         # second part of 'pad
         m = length % 8
@@ -130,7 +153,7 @@ class Nar:
         else:
             offset = 8 - m
         boffset = bytearray(offset)
-        self.updater_fn(boffset)
+        self.updater.update(boffset)
 
     def _filter_and_serialize(self, fso: Path) -> None:
         """On the first level of the main tree, we may have to skip some paths (e.g.
@@ -167,7 +190,7 @@ class Nar:
             if os.access(fso, os.X_OK):
                 self.str_(["executable", ""])
             self.str_("contents")
-            with fso.open("rb") as f:
+            with open(str(fso), "rb") as f:
                 self.str_(f)
 
         elif stat.S_ISLNK(mode):
@@ -205,7 +228,10 @@ class Nar:
             else []
         )
         self._serialize(fso, first_level=True)
-        return
+
+    def result(self):
+        """Compute the result in the required format."""
+        return self.convert_fn(self.updater.hexdigest())
 
 
 @swh_cli_group.command(name="nar", context_settings=CONTEXT_SETTINGS)
@@ -225,25 +251,6 @@ class Nar:
 @click.option("--debug/--no-debug", default=lambda: os.environ.get("DEBUG", False))
 def cli(exclude_vcs, directory, hash_algo, format_output, debug):
     """Compute NAR hashes on a directory."""
-    h = hashlib.sha256() if hash_algo == "sha256" else hashlib.sha1()
-    updater = h.update
-    format_output = format_output.lower()
-
-    def identity(hsh):
-        return hsh
-
-    def convert_b64(hsh: str):
-        return base64.b64encode(bytes.fromhex(hsh)).decode().lower()
-
-    def convert_b32(hsh: str):
-        return base64.b32encode(bytes.fromhex(hsh)).decode().lower()
-
-    convert_fn = {
-        "hex": identity,
-        "base64": convert_b64,
-        "base32": convert_b32,
-    }
-
-    nar = Nar(exclude_vcs, updater, debug=debug)
+    nar = Nar(hash_algo, format_output, exclude_vcs, debug=debug)
     nar.serialize(directory)
-    print(convert_fn[format_output](h.hexdigest()))
+    print(nar.result())

swh/loader/core/tests/test_nar.py

+# Copyright (C) 2023  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+from pathlib import Path
+
+from swh.core.tarball import uncompress
+from swh.loader.core.nar import Nar
+
+
+def test_nar_tarball(tmpdir, tarball_with_nar_hashes):
+    tarball_path, nar_hashes = tarball_with_nar_hashes
+
+    directory_path = Path(tmpdir)
+    directory_path.mkdir(parents=True, exist_ok=True)
+    uncompress(str(tarball_path), dest=str(directory_path))
+    path_on_disk = next(directory_path.iterdir())
+
+    for hash_algo, expected_hash_value in nar_hashes.items():
+        nar = Nar(hash_algo=hash_algo)
+        nar.serialize(path_on_disk)
+        assert nar.result() == nar_hashes[hash_algo]
+
+
+def test_nar_content(content_with_nar_hashes):
+    content_path, nar_hashes = content_with_nar_hashes
+
+    for hash_algo, expected_hash_value in nar_hashes.items():
+        nar = Nar(hash_algo=hash_algo)
+        nar.serialize(content_path)
+        assert nar.result() == expected_hash_value