From 3df91b4d7a7e2c35b4093767f9c7c2a618c0045c Mon Sep 17 00:00:00 2001
From: Nguyen Thanh Dat <ntdat812@gmail.com>
Date: Mon, 31 Aug 2026 15:31:28 +0700
Subject: [PATCH] Keep the startup banners from crashing on a non-UTF-8 stdout
 (#3050)

---
 maigret/notify.py    | 25 ++++++++++++++++++++--
 tests/test_notify.py | 49 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 72 insertions(+), 2 deletions(-)

diff --git a/maigret/notify.py b/maigret/notify.py
index 43bd3315..e1672474 100644
--- a/maigret/notify.py
+++ b/maigret/notify.py
@@ -271,6 +271,27 @@ PATREON_URL = "https://www.patreon.com/soxoj"
 INTRO_TEXT = "MAIGRET - collect a dossier by username from 3000+ sites"
 
 
+def _print_encodable(text: str) -> None:
+    """Print text the active stdout encoding may not be able to represent.
+
+    On Windows, Python takes stdout's encoding from the process ANSI
+    codepage, which is cp1252 on a default install and has no U+2665. The
+    banners below carry one, so printing them raised UnicodeEncodeError
+    before a single site was checked, and --no-color did not help because
+    the character sits in that branch too. PYTHONIOENCODING cannot rescue
+    the PyInstaller build, which ignores PYTHON* environment variables.
+
+    Catching the write rather than reconfiguring the stream is deliberate:
+    colorama's init() replaces sys.stdout with a wrapper that has no
+    reconfigure(), so a stream-level fix would depend on running first.
+    """
+    try:
+        print(text)
+    except UnicodeEncodeError:
+        encoding = getattr(sys.stdout, "encoding", None) or "ascii"
+        print(text.encode(encoding, errors="replace").decode(encoding, errors="replace"))
+
+
 def _format_intro(use_color: bool) -> str:
     if not use_color:
         return f"[+] {INTRO_TEXT}"
@@ -283,7 +304,7 @@ def print_intro_banner(no_color: bool = False, silent: bool = False) -> None:
     """Print the Maigret intro tagline. Skipped only in silent (--ai) mode."""
     if silent:
         return
-    print(_format_intro(use_color=not no_color))
+    _print_encodable(_format_intro(use_color=not no_color))
 
 
 def _format_donate_banner(use_color: bool) -> str:
@@ -304,4 +325,4 @@ def print_donate_banner(no_color: bool = False, silent: bool = False) -> None:
     """Print a colored donation banner. Skipped only in silent (--ai) mode."""
     if silent:
         return
-    print(_format_donate_banner(use_color=not no_color))
+    _print_encodable(_format_donate_banner(use_color=not no_color))
diff --git a/tests/test_notify.py b/tests/test_notify.py
index a4c6b993..8824895f 100644
--- a/tests/test_notify.py
+++ b/tests/test_notify.py
@@ -204,3 +204,52 @@ def test_update_shows_extractor_field_when_verbose(monkeypatch):
     out = "\n".join(captured)
     assert "uid: 42" in out
     assert "_extractor: SomeSchemeAPI" in out
+
+
+def test_banners_survive_a_stdout_that_cannot_encode_them(tmp_path):
+    """The banners carry '♥', which the Windows ANSI codepage (cp1252 on a
+    default install) cannot encode, so printing them raised UnicodeEncodeError
+    before a single site was checked. --no-color did not help: the character is
+    in that branch too, and the PyInstaller build ignores PYTHONIOENCODING, so
+    there was no way around it from outside.
+
+    Run the banners in a child process whose stdout really is cp1252 rather
+    than mocking the failure, so this fails on any machine if the guard goes.
+    """
+    import os
+    import subprocess
+    import sys
+    import textwrap
+
+    probe = tmp_path / "banner_probe.py"
+    probe.write_text(
+        textwrap.dedent(
+            """
+            from maigret.notify import print_intro_banner, print_donate_banner
+
+            print_intro_banner(no_color=True)
+            print_donate_banner(no_color=True)
+            print_intro_banner(no_color=False)
+            print_donate_banner(no_color=False)
+            """
+        ),
+        encoding="utf-8",
+    )
+
+    env = dict(os.environ)
+    env["PYTHONIOENCODING"] = "cp1252"
+    result = subprocess.run(
+        [sys.executable, str(probe)],
+        capture_output=True,
+        encoding="cp1252",
+        errors="replace",
+        env=env,
+        cwd=str(tmp_path),
+    )
+
+    assert result.returncode == 0, (
+        f"the banner must not crash the run: stderr={result.stderr!r}"
+    )
+    assert "Support Maigret" in result.stdout, (
+        f"the banner must still be printed: stdout={result.stdout!r}"
+    )