From 86e1ec8e32e1656cdb57bb13cf26e52bb2cc1fe7 Mon Sep 17 00:00:00 2001
From: m1ngsama <contact@m1ng.space>
Date: Sun, 24 May 2026 08:58:51 +0800
Subject: [PATCH] i18n: tolerate whitespace in language parsing

---
 docs/CHANGELOG.md      |  2 ++
 src/i18n.c             | 22 +++++++++++++++++++++-
 tests/unit/test_i18n.c | 15 +++++++++++++++
 3 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/docs/CHANGELOG.md b/docs/CHANGELOG.md
index 9c9218f..b1d6683 100644
--- a/docs/CHANGELOG.md
+++ b/docs/CHANGELOG.md
@@ -64,6 +64,8 @@
   concurrent-session and connection-rate scenarios.
 - CI memory-leak smoke checks now use an isolated state directory, wait for
   real SSH readiness, and clean up the exact server PID instead of `pkill`.
+- Language parsing now tolerates surrounding whitespace and accepts the
+  `english` alias, improving `TNT_LANG` and `:lang` ergonomics.
 - Refreshed README and quick-reference module maps to match the current
   `cli_text`, `help_text`, `support_text`, i18n, exec, and rate-limit modules.
 - NORMAL mode now opens at the latest visible messages instead of the oldest
diff --git a/src/i18n.c b/src/i18n.c
index c721a51..844053c 100644
--- a/src/i18n.c
+++ b/src/i18n.c
@@ -2,9 +2,28 @@
 
 #include <ctype.h>
 
+static const char *skip_space(const char *value) {
+    while (value && *value &&
+           isspace((unsigned char)*value)) {
+        value++;
+    }
+    return value;
+}
+
+static bool is_lang_boundary(const char *value) {
+    if (*value == '\0' || *value == '_' || *value == '-' || *value == '.') {
+        return true;
+    }
+    if (!isspace((unsigned char)*value)) {
+        return false;
+    }
+    return *skip_space(value) == '\0';
+}
+
 static bool starts_with_lang(const char *value, const char *prefix) {
     if (!value || !prefix) return false;
 
+    value = skip_space(value);
     while (*prefix) {
         if (tolower((unsigned char)*value) !=
             tolower((unsigned char)*prefix)) {
@@ -14,7 +33,7 @@ static bool starts_with_lang(const char *value, const char *prefix) {
         prefix++;
     }
 
-    return *value == '\0' || *value == '_' || *value == '-' || *value == '.';
+    return is_lang_boundary(value);
 }
 
 bool i18n_try_parse_lang(const char *value, help_lang_t *lang) {
@@ -30,6 +49,7 @@ bool i18n_try_parse_lang(const char *value, help_lang_t *lang) {
     }
 
     if (starts_with_lang(value, "en") ||
+        starts_with_lang(value, "english") ||
         starts_with_lang(value, "c") ||
         starts_with_lang(value, "posix")) {
         if (lang) *lang = LANG_EN;
diff --git a/tests/unit/test_i18n.c b/tests/unit/test_i18n.c
index ef98c60..d1ab38f 100644
--- a/tests/unit/test_i18n.c
+++ b/tests/unit/test_i18n.c
@@ -38,6 +38,20 @@ TEST(parse_unknown_uses_fallback) {
     assert(i18n_parse_lang("fr_FR.UTF-8", LANG_ZH) == LANG_ZH);
 }
 
+TEST(parse_ignores_surrounding_whitespace) {
+    help_lang_t lang;
+
+    assert(i18n_try_parse_lang("  zh  ", &lang) == true);
+    assert(lang == LANG_ZH);
+    assert(i18n_parse_lang("\ten_US.UTF-8\n", LANG_ZH) == LANG_EN);
+    assert(i18n_parse_lang(" english ", LANG_ZH) == LANG_EN);
+    assert(i18n_try_parse_lang("zh CN", &lang) == false);
+
+    setenv("TNT_LANG", " zh ", 1);
+    setenv("LC_ALL", "en_US.UTF-8", 1);
+    assert(i18n_default_lang() == LANG_ZH);
+}
+
 TEST(default_prefers_tnt_lang) {
     setenv("TNT_LANG", "zh_CN.UTF-8", 1);
     setenv("LC_ALL", "en_US.UTF-8", 1);
@@ -123,6 +137,7 @@ int main(void) {
 
     RUN_TEST(parse_explicit_languages);
     RUN_TEST(parse_unknown_uses_fallback);
+    RUN_TEST(parse_ignores_surrounding_whitespace);
     RUN_TEST(default_prefers_tnt_lang);
     RUN_TEST(default_uses_locale_when_no_tnt_lang);
     RUN_TEST(text_lookup_matches_language);