get_url: add support for BSD-style digest (#84485)

* Added support for BSD-style digest file to test
  checksum of downloaded file.

Fixes: #84476

Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>

changelogs/fragments/get_url_bsd_style_digest.yml

@@ -0,0 +1,3 @@
+---
+bugfixes:
+  - get_url - add support for BSD-style checksum digest file (https://github.com/ansible/ansible/issues/84476).

lib/ansible/modules/get_url.py

@@ -460,6 +460,37 @@ def is_url(checksum):
     return urlsplit(checksum).scheme in supported_schemes
 
 
+def parse_digest_lines(filename, lines):
+    """Returns a list of tuple containing the filename and digest depending upon
+      the lines provided
+
+    Args:
+        filename (str): Name of the filename, used only when the digest is one-liner
+        lines (list): A list of lines containing filenames and checksums
+    """
+    checksum_map = []
+    BSD_DIGEST_LINE = re.compile(r'^(\w+) ?\((?P<path>.+)\) ?= (?P<digest>[\w.]+)$')
+    GNU_DIGEST_LINE = re.compile(r'^(?P<digest>[\w.]+) ([ *])(?P<path>.+)$')
+
+    if len(lines) == 1 and len(lines[0].split()) == 1:
+        # Only a single line with a single string
+        # treat it as a checksum only file
+        checksum_map.append((lines[0], filename))
+        return checksum_map
+    # The assumption here is the file is in the format of
+    # checksum filename
+    for line in lines:
+        match = BSD_DIGEST_LINE.match(line)
+        if match:
+            checksum_map.append((match.group('digest'), match.group('path')))
+        else:
+            match = GNU_DIGEST_LINE.match(line)
+            if match:
+                checksum_map.append((match.group('digest'), match.group('path').lstrip("./")))
+
+    return checksum_map
+
+
 # ==============================================================
 # main
 
@@ -527,31 +558,13 @@ def main():
         if is_url(checksum):
             checksum_url = checksum
             # download checksum file to checksum_tmpsrc
-            checksum_tmpsrc, checksum_info = url_get(module, checksum_url, dest, use_proxy, last_mod_time, force, timeout, headers, tmp_dest,
+            checksum_tmpsrc, _dummy = url_get(module, checksum_url, dest, use_proxy, last_mod_time, force, timeout, headers, tmp_dest,
-                                                     unredirected_headers=unredirected_headers, ciphers=ciphers, use_netrc=use_netrc)
+                                              unredirected_headers=unredirected_headers, ciphers=ciphers, use_netrc=use_netrc)
             with open(checksum_tmpsrc) as f:
                 lines = [line.rstrip('\n') for line in f]
             os.remove(checksum_tmpsrc)
-            checksum_map = []
             filename = url_filename(url)
-            if len(lines) == 1 and len(lines[0].split()) == 1:
+            checksum_map = parse_digest_lines(filename=filename, lines=lines)
-                # Only a single line with a single string
-                # treat it as a checksum only file
-                checksum_map.append((lines[0], filename))
-            else:
-                # The assumption here is the file is in the format of
-                # checksum filename
-                for line in lines:
-                    # Split by one whitespace to keep the leading type char ' ' (whitespace) for text and '*' for binary
-                    parts = line.split(" ", 1)
-                    if len(parts) == 2:
-                        # Remove the leading type char, we expect
-                        if parts[1].startswith((" ", "*",)):
-                            parts[1] = parts[1][1:]
-
-                        # Append checksum and path without potential leading './'
-                        checksum_map.append((parts[0], parts[1].lstrip("./")))
-
             # Look through each line in the checksum file for a hash corresponding to
             # the filename in the url, returning the first hash that is found.
             for cksum in (s for (s, f) in checksum_map if f == filename):

test/integration/targets/get_url/tasks/main.yml

@@ -376,6 +376,15 @@
       30949cc401e30ac494d695ab8764a9f76aae17c5d73c67f65e9b558f47eff892 *not_target1.txt
       d0dbfc1945bc83bf6606b770e442035f2c4e15c886ee0c22fb3901ba19900b5b *not_target2.txt
 
+- name: create sha256 checksum file of src in BSD-style checksum (--tag)
+  copy:
+    dest: '{{ files_dir }}/sha256sum_bsd_style.txt'
+    content: |
+      SHA256 (27617.txt) = b1b6ce5073c8fac263a8fc5edfffdbd5dec1980c784e09c5bc69f8fb6056f006.
+      SHA256 (71420.txt) = b1b6ce5073c8fac263a8fc5edfffdbd5dec1980c784e09c5bc69f8fb6056f006.
+      SHA256 (not_target1.txt) = 30949cc401e30ac494d695ab8764a9f76aae17c5d73c67f65e9b558f47eff892
+      SHA256 (not_target2.txt) = d0dbfc1945bc83bf6606b770e442035f2c4e15c886ee0c22fb3901ba19900b5b
+
 # completing 27617 with bug 54390
 - name: create sha256 checksum only with no filename inside
   copy:
@@ -463,6 +472,17 @@
     path: "{{ remote_tmp_dir }}/27617sha256_with_dot.txt"
   register: stat_result_sha256_with_file_scheme
 
+- name: download src with sha256 checksum url with BSD style checksum
+  get_url:
+    url: 'http://localhost:{{ http_port }}/27617.txt'
+    dest: '{{ remote_tmp_dir }}/27617sha256_with_bsd_style.txt'
+    checksum: 'sha256:file://{{ files_dir }}/sha256sum_bsd_style.txt'
+  register: result_sha256_with_bsd_style
+
+- stat:
+    path: "{{ remote_tmp_dir }}/27617sha256_with_bsd_style.txt"
+  register: stat_result_sha256_with_bsd_style
+
 - name: download 71420.txt with sha1 checksum url
   get_url:
     url: 'http://localhost:{{ http_port }}/71420.txt'
@@ -538,11 +558,13 @@
       - result_sha256_with_dot is changed
       - result_sha256_with_asterisk is changed
       - result_sha256_with_file_scheme is changed
+      - result_sha256_with_bsd_style is changed
       - "stat_result_sha1.stat.exists == true"
       - "stat_result_sha256.stat.exists == true"
       - "stat_result_sha256_with_dot.stat.exists == true"
       - "stat_result_sha256_with_asterisk.stat.exists == true"
       - "stat_result_sha256_with_file_scheme.stat.exists == true"
+      - stat_result_sha256_with_bsd_style.stat.exists
       - result_sha1_71420 is changed
       - result_sha256_71420 is changed
       - result_sha256_with_dot_71420 is changed

test/units/modules/test_get_url.py

@@ -0,0 +1,45 @@
+# -*- coding: utf-8 -*-
+
+# Copyright: Contributors to the Ansible project
+# GNU General Public License v3.0+ (see COPYING or https://www.gnu.org/licenses/gpl-3.0.txt)
+from __future__ import annotations
+
+import pytest
+
+from ansible.modules.get_url import parse_digest_lines
+
+
+@pytest.mark.parametrize(
+    ("lines", "expected"),
+    [
+        pytest.param(
+            [
+                "a97e6837f60cec6da4491bab387296bbcd72bdba",
+            ],
+            [("a97e6837f60cec6da4491bab387296bbcd72bdba", "sample.txt")],
+            id="single-line-digest",
+        ),
+        pytest.param(
+            [
+                "a97e6837f60cec6da4491bab387296bbcd72bdba  sample.txt",
+            ],
+            [("a97e6837f60cec6da4491bab387296bbcd72bdba", "sample.txt")],
+            id="GNU-style-digest",
+        ),
+        pytest.param(
+            [
+                "SHA256 (sample.txt) = b1b6ce5073c8fac263a8fc5edfffdbd5dec1980c784e09c5bc69f8fb6056f006.",
+            ],
+            [
+                (
+                    "b1b6ce5073c8fac263a8fc5edfffdbd5dec1980c784e09c5bc69f8fb6056f006.",
+                    "sample.txt",
+                )
+            ],
+            id="BSD-style-digest",
+        ),
+    ],
+)
+def test_parse_digest_lines(lines, expected):
+    filename = "sample.txt"
+    assert parse_digest_lines(filename, lines) == expected