Bug with IniReadSectionNames if file is coded with UTF-8-BOM

[anonymous]

I have an ini-file that is formatted with UTF-8-BOM. If the file does not start with a linebreak, IniReadSectionNames does not retrieve the first section.

Local $sPathA, $sPathB, $sPathC, $sPathD, $hFile

; File A - UTF-8-BOM - No Leading LineBreak
$sPathA = @DesktopDir & "\File_UTF-8-BOM_v1.ini"
$hFile = FileOpen($sPathA, 2 + 8 + 128)
FileWrite($hFile, "[SectionA]" & @CRLF & "Key=Value")
FileClose($hFile)

; File B - UTF-8-BOM - Leading LineBreak
$sPathB = @DesktopDir & "\File_UTF-8-BOM_v2.ini"
$hFile = FileOpen($sPathB, 2 + 8 + 128)
FileWrite($hFile, @CRLF & "[SectionB]" & @CRLF & "Key=Value")
FileClose($hFile)

; File C - UTF-8 - No Leading LineBreak
$sPathC = @DesktopDir & "\File_UTF-8_v1.ini"
$hFile = FileOpen($sPathC, 2 + 8 + 256)
FileWrite($hFile, "[SectionA]" & @CRLF & "Key=Value")
FileClose($hFile)

; File D - UTF-8 - Leading LineBreak
$sPathD = @DesktopDir & "\File_UTF-8_v2.ini"
$hFile = FileOpen($sPathD, 2 + 8 + 256)
FileWrite($hFile, @CRLF & "[SectionB]" & @CRLF & "Key=Value")
FileClose($hFile)

Local $aSectionA = IniReadSectionNames($sPathA)
Local $aSectionB = IniReadSectionNames($sPathB)
Local $aSectionC = IniReadSectionNames($sPathC)
Local $aSectionD = IniReadSectionNames($sPathD)
ConsoleWrite("$aSectionA" & @TAB & UBound($aSectionA) & @CRLF)
ConsoleWrite("$aSectionB" & @TAB & UBound($aSectionB) & @CRLF)
ConsoleWrite("$aSectionC" & @TAB & UBound($aSectionC) & @CRLF)
ConsoleWrite("$aSectionD" & @TAB & UBound($aSectionD) & @CRLF)
Exit

[anonymous]

see discussion at ​https://www.autoitscript.com/forum/topic/213616-bug-with-inireadsectionnames-if-file-is-coded-with-utf-8-bom/#comment-1551213

[mLipok]

@JPM please also read:

​https://www.autoitscript.com/forum/topic/206029-ini-functions-and-ini-file-encoding/#findComment-1483659

and:
​https://www.autoitscript.com/forum/topic/206029-ini-functions-and-ini-file-encoding/#findComment-1483660

also:
​https://www.autoitscript.com/forum/topic/174335-file-encoding-problem/

[Jpm]

It is a bug in MS API used by AutoIt
It is true also for IniReadSection and IniRead
Those API was cerlainly created before UTF8 with BOM support
The only think I can do is to make a warning in the doc
Cheers

Local $sPathA, $sPathB, $sPathC, $sPathD, $hFile
Local $sSectionA = "SectionA"
Local $sSectionA1 = "SectionA1"
Local $sSectionB = "SectionB"
; File A - UTF-8-BOM - No Leading LineBreak
$sPathA = @ScriptDir & "\File_UTF-8-BOM_v1.ini"
$hFile = FileOpen($sPathA, 2 + 8 + 128)
FileWrite($hFile, "[" & $sSectionA & "]" & @CRLF & "KeyA=ValueA")
FileWrite($hFile,  @CRLF & "[" & $sSectionA1 & "]" & @CRLF & "KeyA1=ValueA1")
FileClose($hFile)

;~ ; File B - UTF-8-BOM - Leading LineBreak
$sPathB = @ScriptDir & "\File_UTF-8-BOM_v2.ini"
$hFile = FileOpen($sPathB, 2 + 8 + 128)
FileWrite($hFile, @CRLF & "[" & $sSectionB & "]" & @CRLF & "KeyB=ValueB")
FileClose($hFile)

;~ ; File C - UTF-8 - No Leading LineBreak
$sPathC = @ScriptDir & "\File_UTF-8_v1.ini"
$hFile = FileOpen($sPathC, 2 + 8 + 256)
FileWrite($hFile, "[" & $sSectionA & "]" & @CRLF & "KeyC=ValueC")
FileClose($hFile)

;~ ; File D - UTF-8 - Leading LineBreak
$sPathD = @ScriptDir & "\File_UTF-8_v2.ini"
$hFile = FileOpen($sPathD, 2 + 8 + 256)
FileWrite($hFile, @CRLF & "[" & $sSectionB & "]" & @CRLF & "KeyD=ValueD")
FileClose($hFile)

Local $aSectionA = IniReadSectionNames($sPathA)
ConsoleWrite('@@ Debug(' & @ScriptLineNumber & ') : $aSectionA = ' & $aSectionA & @CRLF & '>Error code: ' & @error & '    Extended code: ' & @extended & ' (0x' & Hex(@extended) & ')' & @CRLF) ;### Debug Console
Local $aSectionB = IniReadSectionNames($sPathB)
Local $aSectionC = IniReadSectionNames($sPathC)
Local $aSectionD = IniReadSectionNames($sPathD)
ConsoleWrite("$aSectionA" & @TAB & UBound($aSectionA) & @CRLF)
ConsoleWrite("$aSectionB" & @TAB & UBound($aSectionB) & @CRLF)
ConsoleWrite("$aSectionC" & @TAB & UBound($aSectionC) & @CRLF)
ConsoleWrite("$aSectionD" & @TAB & UBound($aSectionD) & @CRLF)

Local $aKeyA = IniReadSection($sPathA, $sSectionA)
ConsoleWrite('@@ Debug(' & @ScriptLineNumber & ') : $aKeyA = ' & $aKeyA & @CRLF & '>Error code: ' & @error & '    Extended code: ' & @extended & ' (0x' & Hex(@extended) & ')' & @CRLF) ;### Debug Console
Local $aKeyA1 = IniReadSection($sPathA, $sSectionA1)
Local $aKeyB = IniReadSection($sPathB, $sSectionB)
Local $aKeyC = IniReadSection($sPathC, $sSectionA)
Local $aKeyD = IniReadSection($sPathD, $sSectionB)
ConsoleWrite("$aKeyA" & @TAB & UBound($aKeyA) & @CRLF)
ConsoleWrite("$aKeyA1" & @TAB & UBound($aKeyA1) & @CRLF)
ConsoleWrite("$aKeyB" & @TAB & UBound($aKeyB) & @CRLF)
ConsoleWrite("$aKeyC" & @TAB & UBound($aKeyC) & @CRLF)
ConsoleWrite("$aKeyD" & @TAB & UBound($aKeyD) & @CRLF)

Local $sValueA = IniRead($sPathA, $sSectionA, "KeyA", "???")
ConsoleWrite('@@ Debug(' & @ScriptLineNumber & ') : $sValueA = ' & $sValueA & @CRLF & '>Error code: ' & @error & '    Extended code: ' & @extended & ' (0x' & Hex(@extended) & ')' & @CRLF) ;### Debug Console
Local $sValueA1 = IniRead($sPathA, $sSectionA1, "KeyA1", "???")
Local $sValueB = IniRead($sPathB, $sSectionB, "KeyB", "???")
Local $sValueC = IniRead($sPathC, $sSectionA, "KeyC", "???")
Local $sValueD = IniRead($sPathD, $sSectionB, "KeyD", "???")
ConsoleWrite("$sValueA" & @TAB & $sValueA & @CRLF)
ConsoleWrite("$sValueA1" & @TAB & $sValueA1 & @CRLF)
ConsoleWrite("$sValueB" & @TAB & $sValueB & @CRLF)
ConsoleWrite("$sValueC" & @TAB & $sValueC & @CRLF)
ConsoleWrite("$sValueD" & @TAB & $sValueD & @CRLF)
Exit

[Jpm]

Fixed by revision [13369]

Added #4086: Some precision about UTF8 with BOM ini file for ini functions.