From e78f508997f9384518c2dd1a005c5306ccd24783 Mon Sep 17 00:00:00 2001
From: Winfried Donkers <winfrieddonkers@libreoffice.org>
Date: Tue, 21 Nov 2017 16:08:32 +0100
Subject: tdf#97198 Make calc function MID work with Unicode non-BMP
 characters.

Change-Id: Ic86344495490d6ca942f9fd2752874da22ef531f
Reviewed-on: https://gerrit.libreoffice.org/45040
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
---
 sc/qa/unit/data/functions/text/fods/mid.fods | 175 ++++++++++++++++++++++-----
 sc/source/core/tool/interpr1.cxx             |  28 +++--
 2 files changed, 166 insertions(+), 37 deletions(-)

diff --git a/sc/qa/unit/data/functions/text/fods/mid.fods b/sc/qa/unit/data/functions/text/fods/mid.fods
index 56fcd55d3f27..68c7feb10299 100644
--- a/sc/qa/unit/data/functions/text/fods/mid.fods
+++ b/sc/qa/unit/data/functions/text/fods/mid.fods
@@ -1054,41 +1054,158 @@
      </table:table-cell>
      <table:table-cell table:number-columns-repeated="5"/>
     </table:table-row>
-    <table:table-row table:style-name="ro5">
-     <table:table-cell/>
-     <table:table-cell table:style-name="ce17"/>
-     <table:table-cell table:style-name="ce27"/>
-     <table:table-cell table:style-name="ce11"/>
-     <table:table-cell table:number-columns-repeated="5"/>
+    <table:table-row table:style-name="ro2">
+     <table:table-cell table:formula="of:=MID([.I2];12;8)">
+      <text:p/>
+     </table:table-cell>
+     <table:table-cell table:formula="of:=MID([.J2];12;8)">
+      <text:p/>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce38" table:formula="of:=[.A6]=[.B6]" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A6])" office:value-type="string" office:string-value="=MID(I2;12;8)" calcext:value-type="string">
+      <text:p>=MID(I2;12;8)</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="7"/>
     </table:table-row>
     <table:table-row table:style-name="ro2">
-     <table:table-cell table:style-name="ce9" table:number-columns-repeated="2"/>
-     <table:table-cell table:style-name="ce27"/>
-     <table:table-cell table:style-name="ce11"/>
-     <table:table-cell table:number-columns-repeated="5"/>
+     <table:table-cell table:formula="of:=MID([.I2];0;3)" office:value-type="string" office:string-value="" calcext:value-type="error">
+      <text:p>Err:502</text:p>
+     </table:table-cell>
+     <table:table-cell table:formula="of:#ERR502!" office:value-type="string" office:string-value="" calcext:value-type="error">
+      <text:p>Err:502</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce39" table:formula="of:=ISERROR([.A7])" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A7])" office:value-type="string" office:string-value="=MID(I2;0;3)" calcext:value-type="string">
+      <text:p>=MID(I2;0;3)</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="7"/>
     </table:table-row>
     <table:table-row table:style-name="ro2">
-     <table:table-cell table:number-columns-repeated="2"/>
-     <table:table-cell table:style-name="ce27"/>
-     <table:table-cell table:style-name="ce11"/>
-     <table:table-cell table:number-columns-repeated="5"/>
+     <table:table-cell table:formula="of:=MID([.I8];[.J8];[.K8])" office:value-type="string" office:string-value="b😂" calcext:value-type="string">
+      <text:p>b😂</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce35" office:value-type="string" calcext:value-type="string">
+      <text:p>b😂</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce40" table:formula="of:=[.A8]=[.B8]" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A8])" office:value-type="string" office:string-value="=MID(I8;J8;K8)" calcext:value-type="string">
+      <text:p>=MID(I8;J8;K8)</text:p>
+     </table:table-cell>
+     <table:table-cell office:value-type="string" calcext:value-type="string">
+      <text:p>result Excel 2016 is b😂</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="3"/>
+     <table:table-cell table:style-name="ce46" office:value-type="string" calcext:value-type="string"><text:p>ab<text:span text:style-name="T1">😂de𝔖gh𝕬𝖈𝖍𝖙𝖚𝖓𝖌!</text:span></text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="2" table:style-name="ce46" office:value-type="float" office:value="2" calcext:value-type="float">
+      <text:p>2</text:p>
+     </table:table-cell>
     </table:table-row>
-    <table:table-row table:style-name="ro5">
-     <table:table-cell table:style-name="ce14"/>
-     <table:table-cell table:style-name="ce17"/>
-     <table:table-cell table:style-name="ce27"/>
-     <table:table-cell table:style-name="ce11"/>
-     <table:table-cell table:number-columns-repeated="5"/>
+    <table:table-row table:style-name="ro2">
+     <table:table-cell table:formula="of:=MID([.I9];[.J9];[.K9])" office:value-type="string" office:string-value="b" calcext:value-type="string">
+      <text:p>b</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce35" office:value-type="string" calcext:value-type="string">
+      <text:p>b</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce41" table:formula="of:=[.A9]=[.B9]" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A9])" office:value-type="string" office:string-value="=MID(I9;J9;K9)" calcext:value-type="string">
+      <text:p>=MID(I9;J9;K9)</text:p>
+     </table:table-cell>
+     <table:table-cell office:value-type="string" calcext:value-type="string">
+      <text:p>result Excel 2016 is b</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="3"/>
+     <table:table-cell table:style-name="ce46" office:value-type="string" calcext:value-type="string">
+      <text:p>ab😂</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce46" office:value-type="float" office:value="2" calcext:value-type="float">
+      <text:p>2</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce46" office:value-type="float" office:value="1" calcext:value-type="float">
+      <text:p>1</text:p>
+     </table:table-cell>
     </table:table-row>
-    <table:table-row table:style-name="ro5">
-     <table:table-cell table:style-name="ce16"/>
-     <table:table-cell table:style-name="ce17"/>
-     <table:table-cell table:style-name="ce27"/>
-     <table:table-cell table:style-name="ce11"/>
-     <table:table-cell table:number-columns-repeated="4"/>
-     <table:table-cell table:style-name="ce17"/>
+    <table:table-row table:style-name="ro6">
+     <table:table-cell table:formula="of:=MID([.I10];[.J10];[.K10])" office:value-type="string" office:string-value="𧌒𧀬" calcext:value-type="string">
+      <text:p>𧌒𧀬</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce35" office:value-type="string" calcext:value-type="string">
+      <text:p>𧌒𧀬</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce42" table:formula="of:=[.A10]=[.B10]" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A10])" office:value-type="string" office:string-value="=MID(I10;J10;K10)" calcext:value-type="string">
+      <text:p>=MID(I10;J10;K10)</text:p>
+     </table:table-cell>
+     <table:table-cell office:value-type="string" calcext:value-type="string">
+      <text:p>result Excel 2016 is 𧌒</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="3"/>
+     <table:table-cell table:style-name="ce47" office:value-type="string" calcext:value-type="string"><text:p>&quot;<text:span text:style-name="T2">𧌒𧀬</text:span>&quot;</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="2" table:style-name="ce46" office:value-type="float" office:value="2" calcext:value-type="float">
+      <text:p>2</text:p>
+     </table:table-cell>
+    </table:table-row>
+    <table:table-row table:style-name="ro3">
+     <table:table-cell table:formula="of:=MID([.I11];[.J11];[.K11])" office:value-type="string" office:string-value="ü" calcext:value-type="string">
+      <text:p>ü</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce35" office:value-type="string" calcext:value-type="string">
+      <text:p>ü</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce43" table:formula="of:=[.A11]=[.B11]" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A11])" office:value-type="string" office:string-value="=MID(I11;J11;K11)" calcext:value-type="string">
+      <text:p>=MID(I11;J11;K11)</text:p>
+     </table:table-cell>
+     <table:table-cell office:value-type="string" calcext:value-type="string">
+      <text:p>result Excel 2016 is identical</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="3"/>
+     <table:table-cell table:style-name="ce48" office:value-type="string" calcext:value-type="string">
+      <text:p>&quot;üë&quot;</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="2" table:style-name="ce46" office:value-type="float" office:value="2" calcext:value-type="float">
+      <text:p>2</text:p>
+     </table:table-cell>
+    </table:table-row>
+    <table:table-row table:style-name="ro2">
+     <table:table-cell table:formula="of:=MID([.I12];[.J12];[.K12])" office:value-type="string" office:string-value="𝖍𝖙𝖚" calcext:value-type="string">
+      <text:p>𝖍𝖙𝖚</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce35" office:value-type="string" calcext:value-type="string">
+      <text:p>𝖍𝖙𝖚</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce44" table:formula="of:=[.A12]=[.B12]" office:value-type="boolean" office:boolean-value="true" calcext:value-type="boolean">
+      <text:p>TRUE</text:p>
+     </table:table-cell>
+     <table:table-cell table:style-name="ce11" table:formula="of:=FORMULA([.A12])" office:value-type="string" office:string-value="=MID(I12;J12;K12)" calcext:value-type="string">
+      <text:p>=MID(I12;J12;K12)</text:p>
+     </table:table-cell>
+     <table:table-cell office:value-type="string" calcext:value-type="string">
+      <text:p>result Excel 2016 is 𝖈</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="3"/>
+     <table:table-cell table:style-name="ce49" office:value-type="string" calcext:value-type="string">
+      <text:p>𝕬𝖈𝖍𝖙𝖚𝖓𝖌!</text:p>
+     </table:table-cell>
+     <table:table-cell table:number-columns-repeated="2" table:style-name="ce46" office:value-type="float" office:value="3" calcext:value-type="float">
+      <text:p>3</text:p>
+     </table:table-cell>
     </table:table-row>
-    <table:table-row table:style-name="ro5" table:number-rows-repeated="27">
+    <table:table-row table:style-name="ro5" table:number-rows-repeated="25">
      <table:table-cell table:style-name="ce14"/>
      <table:table-cell table:style-name="ce17"/>
      <table:table-cell table:style-name="ce27"/>
@@ -1130,4 +1247,4 @@
    </table:named-expressions>
   </office:spreadsheet>
  </office:body>
-</office:document>
\ No newline at end of file
+</office:document>
diff --git a/sc/source/core/tool/interpr1.cxx b/sc/source/core/tool/interpr1.cxx
index 5bb02e6e363d..d53c00f9fcd7 100644
--- a/sc/source/core/tool/interpr1.cxx
+++ b/sc/source/core/tool/interpr1.cxx
@@ -8914,18 +8914,30 @@ void ScInterpreter::ScMid()
 {
     if ( MustHaveParamCount( GetByte(), 3 ) )
     {
-        double fCnt    = GetStringPositionArgument();
-        double fAnfang = GetStringPositionArgument();
+        sal_Int32 nSubLen = ( sal_Int32 )GetStringPositionArgument();
+        sal_Int32 nStart  = ( sal_Int32 )GetStringPositionArgument();
         OUString aStr = GetString().getString();
-        if (fAnfang < 1.0 || fCnt < 0.0)
+        if ( nStart < 1 || nSubLen < 0 )
             PushIllegalArgument();
         else
         {
-            sal_Int32 nCharacters = std::min<sal_Int32>(static_cast<sal_Int32>(fCnt), aStr.getLength() - fAnfang + 1);
-            OUString sRes;
-            if (nCharacters > 0)
-                sRes = aStr.copy(static_cast<sal_Int32>(fAnfang-1), nCharacters);
-            PushString(sRes);
+            sal_Int32 nLen = aStr.getLength();
+            sal_Int32 nIdx = 0;
+            sal_Int32 nCnt = 0;
+            while ( nIdx < nLen && nStart - 1 > nCnt )
+            {
+                aStr.iterateCodePoints( &nIdx );
+                ++nCnt;
+            }
+            sal_Int32 nIdx0 = nIdx;  //start position
+
+            while ( nIdx < nLen && nStart + nSubLen - 1 > nCnt )
+            {
+                aStr.iterateCodePoints( &nIdx );
+                ++nCnt;
+            }
+            aStr = aStr.copy( nIdx0, nIdx - nIdx0 );
+            PushString( aStr );
         }
     }
 }
-- 
cgit v1.2.3