From 3287fdba28530671dfac3144983a790b4cc1eba4 Mon Sep 17 00:00:00 2001
From: HanishKVC <hanishkvc@gmail.com>
Date: Sat, 4 May 2024 11:43:15 +0530
Subject: [PATCH] SimpCfg:Fix/cleanup trim related test samples and flow

Use the commonality between Indian languages to show mixup issue
with the simple minded trim_dump logic and how trim_oversmart
could potentially avoid that.

Given that I am using valid strings to show the pitfalls of fixed
native char size driven logic, so no need to keep the dump and
oversmart flows seperate, so merge into a common loop.
---
 common/simpcfg.hpp | 19 ++++++++-----------
 1 file changed, 8 insertions(+), 11 deletions(-)

diff --git a/common/simpcfg.hpp b/common/simpcfg.hpp
index 2a7f96d6b..27fb9c56d 100644
--- a/common/simpcfg.hpp
+++ b/common/simpcfg.hpp
@@ -541,18 +541,15 @@ void check_nonenglish() {
         std::string sGotOSmart = str_trim_oversmart(sTest, {" \n\t"});
         std::cout << std::format("{}: Test1[{}] Dumb[{}] OverSmart[{}]", __func__, sTest, sGotDumb, sGotOSmart) << std::endl;
     }
-    std::vector<std::string> vTest2 = { "\n\t this र remove 0s at end 000 ", "\n\tthis र remove 0s and अs at end 000रअ0अ ", "\n\tthis र remove 0s and अs at end 000रअ0\xa4अ "};
+    // The string "\n\tthis र remove 0s and अs at end 000रअ0\xa4अ ",
+    // * will mess up str_trim_dumb,
+    // * but will rightly trigger a exception with oversmart.
+    std::vector<std::string> vTest2 = { "\n\t this र remove 0s at end 000 ", "\n\tthis र remove 0s, अs, ഇs at end 000रअ0अ ", "\n\tthis र remove 0s, अs, ഇs at end 000रअ0इअ "};
+    std::string trimChars = {" \n\tഇ0अ"};
     for (auto sTest: vTest2) {
-        std::string sGotDumb = str_trim_dumb(sTest, {" \n\t0अ"});
-        std::cout << std::format("{}: Test2[{}] Dumb[{}]", __func__, sTest, sGotDumb) << std::endl;
-    }
-    // This partly invalid utf8 string will mess up str_trim_dumb "\n\tthis र remove 0s and अs at end 000रअ0\xa4अ "
-    // but will trigger a exception with oversmart.
-    // std::vector<std::string> vTest3 = { "\n\t this र remove 0s at end 000 ", "\n\tthis र remove 0s and अs at end 000रअ0अ ", "\n\tthis र remove 0s and अs at end 000रअ0\xa4अ "};
-    std::vector<std::string> vTest3 = { "\n\t this र remove 0s at end 000 ", "\n\tthis र remove 0s and अs at end 000रअ0अ ", "\n\tthis र remove 0s and अs at end 000रअ0\xe0\xa4\x30अ "}; // \xe0\xa4
-    for (auto sTest: vTest3) {
-        std::string sGotOSmart = str_trim_oversmart(sTest, {" \n\t0अ"});
-        std::cout << std::format("{}: Test3[{}] OverSmart[{}]", __func__, sTest, sGotOSmart) << std::endl;
+        std::string sGotDumb = str_trim_dumb(sTest, trimChars);
+        std::string sGotOSmart = str_trim_oversmart(sTest, trimChars);
+        std::cout << std::format("{}: Test2 [{}]\n\tDumb[{}]\n\tOverSmart[{}]", __func__, sTest, sGotDumb, sGotOSmart) << std::endl;
     }
 }