Replace module - update string comparison method from bytes to unicode (#85785)

changelogs/fragments/replace-update-string-comparison-method-to-unicode.yml

@@ -0,0 +1,3 @@
+minor_changes:
+  - replace - read/write files in text-mode as unicode chars instead of as bytes
+    and switch regex matching to unicode chars instead of bytes. (https://github.com/ansible/ansible/pull/85785).

lib/ansible/modules/replace.py

@@ -183,14 +183,14 @@ import os
 import re
 import tempfile
 
-from ansible.module_utils.common.text.converters import to_text, to_bytes
+from ansible.module_utils.common.text.converters import to_text
 from ansible.module_utils.basic import AnsibleModule
 
 
-def write_changes(module, contents, path):
+def write_changes(module, contents, path, encoding='utf-8'):
 
     tmpfd, tmpfile = tempfile.mkstemp(dir=module.tmpdir)
-    with os.fdopen(tmpfd, 'wb') as f:
+    with os.fdopen(tmpfd, 'w', encoding=encoding) as f:
         f.write(contents)
 
     validate = module.params.get('validate', None)
@@ -254,8 +254,8 @@ def main():
         module.fail_json(rc=257, msg='Path %s does not exist !' % path)
     else:
         try:
-            with open(path, 'rb') as f:
-                contents = to_text(f.read(), errors='surrogate_or_strict', encoding=encoding)
+            with open(path, 'r', encoding=encoding) as f:
+                contents = f.read()
         except OSError as ex:
             raise Exception(f"Unable to read the contents of {path!r}.") from ex
 
@@ -307,7 +307,7 @@ def main():
             res_args['backup_file'] = module.backup_local(path)
         # We should always follow symlinks so that we change the real file
         path = os.path.realpath(path)
-        write_changes(module, to_bytes(result[0], encoding=encoding), path)
+        write_changes(module, result[0], path, encoding=encoding)
 
     res_args['msg'], res_args['changed'] = check_file_attrs(module, changed, msg)
     module.exit_json(**res_args)

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

@@ -280,3 +280,68 @@
     that:
       - replace_test9 is failure
       - replace_test9.msg.startswith("Unable to process replace")
+
+# test replace operation in non-utf8 encoded files
+
+- name: create test file encoded in cp273
+  lineinfile:
+    path: "{{ remote_tmp_dir }}/test-encoding-file.txt"
+    create: True
+    line: "Test line encoded in cp273"
+    state: present
+    encoding: cp273
+
+- name: stat the new test file
+  stat:
+    path: "{{ remote_tmp_dir }}/test-encoding-file.txt"
+  register: replace_test10
+
+- name: validate test file content
+  assert:
+    that:
+      - replace_test10.stat.exists
+      - replace_test10.stat.checksum == 'e0bf2bdca94727d03483a9adab66543bcec4b99a'
+
+- name: replace only the 'e' in 'line' with a '3' using both the before and after options
+  replace:
+    path: "{{ remote_tmp_dir }}/test-encoding-file.txt"
+    before: 'encoded'
+    after: 'Test'
+    regexp: 'e'
+    replace: '3'
+    encoding: cp273
+  register: replace_test11
+
+- name: stat the modified file
+  stat:
+    path: "{{ remote_tmp_dir }}/test-encoding-file.txt"
+  register: replace_test12
+
+- name: validate the test file content is modified as expected.
+  assert:
+    that:
+      - replace_test11.msg == "1 replacements made"
+      - replace_test12.stat.exists
+      - replace_test12.stat.checksum == '485e97f6e38e50680228b8d49e1eb23160f19fdc'
+
+- name: idempotence check - replace only the 'e' 'line' with a '3' again
+  replace:
+    path: "{{ remote_tmp_dir }}/test-encoding-file.txt"
+    before: 'encoded'
+    after: 'Test'
+    regexp: 'e'
+    replace: '3'
+    encoding: cp273
+  register: replace_test13
+
+- name: stat the un-modified file
+  stat:
+    path: "{{ remote_tmp_dir }}/test-encoding-file.txt"
+  register: replace_test14
+
+- name: validate the test content is unchanged
+  assert:
+    that:
+      - replace_test13.changed == False
+      - replace_test14.stat.exists
+      - replace_test14.stat.checksum == '485e97f6e38e50680228b8d49e1eb23160f19fdc'