c2nim tool added

doc/manual.txt

@@ -155,7 +155,7 @@ underscores ``__`` are not allowed::
 
   letter ::= 'A'..'Z' | 'a'..'z' | '\x80'..'\xff'
   digit ::= '0'..'9'
-  IDENTIFIER ::= letter ( ['_'] letter | digit )*
+  IDENTIFIER ::= letter ( ['_'] (letter | digit) )*
 
 The following `keywords`:idx: are reserved and cannot be used as identifiers:
 

lib/impure/osinfo_posix.nim

@@ -0,0 +1,68 @@
+import posix, strutils, os
+
+type
+  Tstatfs {.importc: "struct statfs64", 
+            header: "<sys/statfs.h>", final, pure.} = object
+    f_type: int
+    f_bsize: int
+    f_blocks: int
+    f_bfree: int
+    f_bavail: int
+    f_files: int
+    f_ffree: int
+    f_fsid: int
+    f_namelen: int
+
+proc statfs(path: string, buf: var Tstatfs): int {.
+  importc, header: "<sys/vfs.h>".}
+
+
+proc getSystemVersion*(): string =
+  result = ""
+  
+  var unix_info: TUtsname
+  
+  if uname(unix_info) != 0:
+    os.OSError()
+  
+  if $unix_info.sysname == "Linux":
+    # Linux
+    result.add("Linux ")
+
+    result.add($unix_info.release & " ")
+    result.add($unix_info.machine)
+  elif $unix_info.sysname == "Darwin":
+    # Darwin
+    result.add("Mac OS X ")
+    if "10" in $unix_info.release:
+      result.add("v10.6 Snow Leopard")
+    elif "9" in $unix_info.release:
+      result.add("v10.5 Leopard")
+    elif "8" in $unix_info.release:
+      result.add("v10.4 Tiger")
+    elif "7" in $unix_info.release:
+      result.add("v10.3 Panther")
+    elif "6" in $unix_info.release:
+      result.add("v10.2 Jaguar")
+    elif "1.4" in $unix_info.release:
+      result.add("v10.1 Puma")
+    elif "1.3" in $unix_info.release:
+      result.add("v10.0 Cheetah")
+    elif "0" in $unix_info.release:
+      result.add("Server 1.0 Hera")
+  else:
+    result.add($unix_info.sysname & " " & $unix_info.release)
+    
+    
+when false:
+  var unix_info: TUtsname
+  echo(uname(unix_info))
+  echo(unix_info.sysname)
+  echo("8" in $unix_info.release)
+
+  echo(getSystemVersion())
+
+  var stfs: TStatfs
+  echo(statfs("sysinfo_posix.nim", stfs))
+  echo(stfs.f_files)
+  

lib/impure/osinfo_win.nim

@@ -0,0 +1,399 @@
+# XXX clean up this mess!
+
+import winlean
+
+const
+  INVALID_HANDLE_VALUE = int(- 1) # GetStockObject
+
+type
+  TMEMORYSTATUSEX {.final, pure.} = object
+    dwLength: int32
+    dwMemoryLoad: int32
+    ullTotalPhys: int64
+    ullAvailPhys: int64
+    ullTotalPageFile: int64
+    ullAvailPageFile: int64
+    ullTotalVirtual: int64
+    ullAvailVirtual: int64
+    ullAvailExtendedVirtual: int64
+    
+  SYSTEM_INFO* {.final, pure.} = object
+    wProcessorArchitecture*: int16
+    wReserved*: int16
+    dwPageSize*: int32
+    lpMinimumApplicationAddress*: pointer
+    lpMaximumApplicationAddress*: pointer
+    dwActiveProcessorMask*: int32
+    dwNumberOfProcessors*: int32
+    dwProcessorType*: int32
+    dwAllocationGranularity*: int32
+    wProcessorLevel*: int16
+    wProcessorRevision*: int16
+
+  LPSYSTEM_INFO* = ptr SYSTEM_INFO
+  TSYSTEMINFO* = SYSTEM_INFO
+
+  TMemoryInfo* = object
+    MemoryLoad*: int ## occupied memory, in percent
+    TotalPhysMem*: int64 ## Total Physical memory, in bytes
+    AvailablePhysMem*: int64 ## Available physical memory, in bytes
+    TotalPageFile*: int64 ## The current committed memory limit 
+                          ## for the system or the current process, whichever is smaller, in bytes.
+    AvailablePageFile*: int64 ## The maximum amount of memory the current process can commit, in bytes.
+    TotalVirtualMem*: int64 ## Total virtual memory, in bytes
+    AvailableVirtualMem*: int64 ## Available virtual memory, in bytes
+    
+  TOSVERSIONINFOEX {.final, pure.} = object
+    dwOSVersionInfoSize: int32
+    dwMajorVersion: int32
+    dwMinorVersion: int32
+    dwBuildNumber: int32
+    dwPlatformId: int32
+    szCSDVersion: array[0..127, char]
+    wServicePackMajor: int16
+    wServicePackMinor: int16
+    wSuiteMask: int16
+    wProductType: int8
+    wReserved: char
+    
+  TVersionInfo* = object
+    majorVersion*: int
+    minorVersion*: int
+    buildNumber*: int
+    platformID*: int
+    SPVersion*: string ## Full Service pack version string
+    SPMajor*: int ## Major service pack version
+    SPMinor*: int ## Minor service pack version
+    SuiteMask*: int
+    ProductType*: int
+    
+  TPartitionInfo* = tuple[FreeSpace, TotalSpace: filetime]
+  
+const
+  # SuiteMask - VersionInfo.SuiteMask
+  VER_SUITE_BACKOFFICE* = 0x00000004
+  VER_SUITE_BLADE* = 0x00000400
+  VER_SUITE_COMPUTE_SERVER* = 0x00004000
+  VER_SUITE_DATACENTER* = 0x00000080
+  VER_SUITE_ENTERPRISE* = 0x00000002
+  VER_SUITE_EMBEDDEDNT* = 0x00000040
+  VER_SUITE_PERSONAL* = 0x00000200
+  VER_SUITE_SINGLEUSERTS* = 0x00000100
+  VER_SUITE_SMALLBUSINESS* = 0x00000001
+  VER_SUITE_SMALLBUSINESS_RESTRICTED* = 0x00000020
+  VER_SUITE_STORAGE_SERVER* = 0x00002000
+  VER_SUITE_TERMINAL* = 0x00000010
+  VER_SUITE_WH_SERVER* = 0x00008000
+
+  # ProductType - VersionInfo.ProductType
+  VER_NT_DOMAIN_CONTROLLER* = 0x0000002
+  VER_NT_SERVER* = 0x0000003
+  VER_NT_WORKSTATION* = 0x0000001
+  
+  VER_PLATFORM_WIN32_NT* = 2
+  
+  # Product Info - getProductInfo() - (Remove unused ones ?)
+  PRODUCT_BUSINESS* = 0x00000006
+  PRODUCT_BUSINESS_N* = 0x00000010
+  PRODUCT_CLUSTER_SERVER* = 0x00000012
+  PRODUCT_DATACENTER_SERVER* = 0x00000008
+  PRODUCT_DATACENTER_SERVER_CORE* = 0x0000000C
+  PRODUCT_DATACENTER_SERVER_CORE_V* = 0x00000027
+  PRODUCT_DATACENTER_SERVER_V* = 0x00000025
+  PRODUCT_ENTERPRISE* = 0x00000004
+  PRODUCT_ENTERPRISE_E* = 0x00000046
+  PRODUCT_ENTERPRISE_N* = 0x0000001B
+  PRODUCT_ENTERPRISE_SERVER* = 0x0000000A
+  PRODUCT_ENTERPRISE_SERVER_CORE* = 0x0000000E
+  PRODUCT_ENTERPRISE_SERVER_CORE_V* = 0x00000029
+  PRODUCT_ENTERPRISE_SERVER_IA64* = 0x0000000F
+  PRODUCT_ENTERPRISE_SERVER_V* = 0x00000026
+  PRODUCT_HOME_BASIC* = 0x00000002
+  PRODUCT_HOME_BASIC_E* = 0x00000043
+  PRODUCT_HOME_BASIC_N* = 0x00000005
+  PRODUCT_HOME_PREMIUM* = 0x00000003
+  PRODUCT_HOME_PREMIUM_E* = 0x00000044
+  PRODUCT_HOME_PREMIUM_N* = 0x0000001A
+  PRODUCT_HYPERV* = 0x0000002A
+  PRODUCT_MEDIUMBUSINESS_SERVER_MANAGEMENT* = 0x0000001E
+  PRODUCT_MEDIUMBUSINESS_SERVER_MESSAGING* = 0x00000020
+  PRODUCT_MEDIUMBUSINESS_SERVER_SECURITY* = 0x0000001F
+  PRODUCT_PROFESSIONAL* = 0x00000030
+  PRODUCT_PROFESSIONAL_E* = 0x00000045
+  PRODUCT_PROFESSIONAL_N* = 0x00000031
+  PRODUCT_SERVER_FOR_SMALLBUSINESS* = 0x00000018
+  PRODUCT_SERVER_FOR_SMALLBUSINESS_V* = 0x00000023
+  PRODUCT_SERVER_FOUNDATION* = 0x00000021
+  PRODUCT_SMALLBUSINESS_SERVER* = 0x00000009
+  PRODUCT_STANDARD_SERVER* = 0x00000007
+  PRODUCT_STANDARD_SERVER_CORE * = 0x0000000D
+  PRODUCT_STANDARD_SERVER_CORE_V* = 0x00000028
+  PRODUCT_STANDARD_SERVER_V* = 0x00000024
+  PRODUCT_STARTER* = 0x0000000B
+  PRODUCT_STARTER_E* = 0x00000042
+  PRODUCT_STARTER_N* = 0x0000002F
+  PRODUCT_STORAGE_ENTERPRISE_SERVER* = 0x00000017
+  PRODUCT_STORAGE_EXPRESS_SERVER* = 0x00000014
+  PRODUCT_STORAGE_STANDARD_SERVER* = 0x00000015
+  PRODUCT_STORAGE_WORKGROUP_SERVER* = 0x00000016
+  PRODUCT_UNDEFINED* = 0x00000000
+  PRODUCT_ULTIMATE* = 0x00000001
+  PRODUCT_ULTIMATE_E* = 0x00000047
+  PRODUCT_ULTIMATE_N* = 0x0000001C
+  PRODUCT_WEB_SERVER* = 0x00000011
+  PRODUCT_WEB_SERVER_CORE* = 0x0000001D
+  
+  PROCESSOR_ARCHITECTURE_AMD64* = 9 ## x64 (AMD or Intel)
+  PROCESSOR_ARCHITECTURE_IA64* = 6 ## Intel Itanium Processor Family (IPF)
+  PROCESSOR_ARCHITECTURE_INTEL* = 0 ## x86
+  PROCESSOR_ARCHITECTURE_UNKNOWN* = 0xffff ## Unknown architecture.
+  
+  # GetSystemMetrics
+  SM_SERVERR2 = 89 
+  
+proc GlobalMemoryStatusEx*(lpBuffer: var TMEMORYSTATUSEX){.stdcall, dynlib: "kernel32",
+    importc: "GlobalMemoryStatusEx".}
+    
+proc getMemoryInfo*(): TMemoryInfo =
+  ## Retrieves memory info
+  var statex: TMEMORYSTATUSEX
+  statex.dwLength = sizeof(statex)
+
+  GlobalMemoryStatusEx(statex)
+  result.MemoryLoad = statex.dwMemoryLoad
+  result.TotalPhysMem = statex.ullTotalPhys
+  result.AvailablePhysMem = statex.ullAvailPhys
+  result.TotalPageFile = statex.ullTotalPageFile
+  result.AvailablePageFile = statex.ullAvailPageFile
+  result.TotalVirtualMem = statex.ullTotalVirtual
+  result.AvailableVirtualMem = statex.ullAvailExtendedVirtual
+
+proc GetVersionEx*(lpVersionInformation: var TOSVERSIONINFOEX): WINBOOL{.stdcall,
+    dynlib: "kernel32", importc: "GetVersionExA".}
+
+proc GetProcAddress*(hModule: int, lpProcName: cstring): pointer{.stdcall,
+    dynlib: "kernel32", importc: "GetProcAddress".}
+
+proc GetModuleHandleA*(lpModuleName: cstring): int{.stdcall,
+     dynlib: "kernel32", importc.}
+
+proc getVersionInfo*(): TVersionInfo =
+  ## Retrieves operating system info
+  var osvi: TOSVERSIONINFOEX
+  osvi.dwOSVersionInfoSize = sizeof(osvi)
+  discard GetVersionEx(osvi)
+  result.majorVersion = osvi.dwMajorVersion
+  result.minorVersion = osvi.dwMinorVersion
+  result.buildNumber = osvi.dwBuildNumber
+  result.platformID = osvi.dwPlatformId
+  result.SPVersion = $osvi.szCSDVersion
+  result.SPMajor = osvi.wServicePackMajor
+  result.SPMinor = osvi.wServicePackMinor
+  result.SuiteMask = osvi.wSuiteMask
+  result.ProductType = osvi.wProductType
+
+proc getProductInfo*(majorVersion, minorVersion, SPMajorVersion, 
+                     SPMinorVersion: int): int =
+  ## Retrieves Windows' ProductInfo, this function only works in Vista and 7
+
+  var pGPI = cast[proc (dwOSMajorVersion, dwOSMinorVersion, 
+              dwSpMajorVersion, dwSpMinorVersion: int32, outValue: Pint32)](GetProcAddress(
+                GetModuleHandleA("kernel32.dll"), "GetProductInfo"))
+                
+  if pGPI != nil:
+    var dwType: int32
+    pGPI(int32(majorVersion), int32(minorVersion), int32(SPMajorVersion), int32(SPMinorVersion), addr(dwType))
+    result = int(dwType)
+  else:
+    return PRODUCT_UNDEFINED
+
+proc GetSystemInfo*(lpSystemInfo: LPSYSTEM_INFO){.stdcall, dynlib: "kernel32",
+    importc: "GetSystemInfo".}
+    
+proc getSystemInfo*(): TSYSTEM_INFO =
+  ## Returns the SystemInfo
+
+  # Use GetNativeSystemInfo if it's available
+  var pGNSI = cast[proc (lpSystemInfo: LPSYSTEM_INFO)](GetProcAddress(
+                GetModuleHandleA("kernel32.dll"), "GetNativeSystemInfo"))
+                
+  var systemi: TSYSTEM_INFO              
+  if pGNSI != nil:
+    pGNSI(addr(systemi))
+  else:
+    GetSystemInfo(addr(systemi))
+
+  return systemi
+
+proc GetSystemMetrics*(nIndex: int32): int32{.stdcall, dynlib: "user32",
+    importc: "GetSystemMetrics".}
+
+proc `$`*(osvi: TVersionInfo): string =
+  ## Turns a VersionInfo object, into a string
+
+  if osvi.platformID == VER_PLATFORM_WIN32_NT and osvi.majorVersion > 4:
+    result = "Microsoft "
+    
+    var si = getSystemInfo()
+    # Test for the specific product
+    if osvi.majorVersion == 6:
+      if osvi.minorVersion == 0:
+        if osvi.ProductType == VER_NT_WORKSTATION:
+          result.add("Windows Vista ")
+        else: result.add("Windows Server 2008 ")
+      elif osvi.minorVersion == 1:
+        if osvi.ProductType == VER_NT_WORKSTATION:
+          result.add("Windows 7 ")
+        else: result.add("Windows Server 2008 R2 ")
+    
+      var dwType = getProductInfo(osvi.majorVersion, osvi.minorVersion, 0, 0)
+      case dwType
+      of PRODUCT_ULTIMATE:
+        result.add("Ultimate Edition")
+      of PRODUCT_PROFESSIONAL:
+        result.add("Professional")
+      of PRODUCT_HOME_PREMIUM:
+        result.add("Home Premium Edition")
+      of PRODUCT_HOME_BASIC:
+        result.add("Home Basic Edition")
+      of PRODUCT_ENTERPRISE:
+        result.add("Enterprise Edition")
+      of PRODUCT_BUSINESS:
+        result.add("Business Edition")
+      of PRODUCT_STARTER:
+        result.add("Starter Edition")
+      of PRODUCT_CLUSTER_SERVER:
+        result.add("Cluster Server Edition")
+      of PRODUCT_DATACENTER_SERVER:
+        result.add("Datacenter Edition")
+      of PRODUCT_DATACENTER_SERVER_CORE:
+        result.add("Datacenter Edition (core installation)")
+      of PRODUCT_ENTERPRISE_SERVER:
+        result.add("Enterprise Edition")
+      of PRODUCT_ENTERPRISE_SERVER_CORE:
+        result.add("Enterprise Edition (core installation)")
+      of PRODUCT_ENTERPRISE_SERVER_IA64:
+        result.add("Enterprise Edition for Itanium-based Systems")
+      of PRODUCT_SMALLBUSINESS_SERVER:
+        result.add("Small Business Server")
+      of PRODUCT_STANDARD_SERVER:
+        result.add("Standard Edition")
+      of PRODUCT_STANDARD_SERVER_CORE:
+        result.add("Standard Edition (core installation)")
+      of PRODUCT_WEB_SERVER:
+        result.add("Web Server Edition")
+      else:
+        nil
+    # End of Windows 6.*
+
+    if osvi.majorVersion == 5 and osvi.minorVersion == 2:
+      if GetSystemMetrics(SM_SERVERR2) != 0:
+        result.add("Windows Server 2003 R2, ")
+      elif (osvi.SuiteMask and VER_SUITE_PERSONAL) != 0: # Not sure if this will work
+        result.add("Windows Storage Server 2003")
+      elif (osvi.SuiteMask and VER_SUITE_WH_SERVER) != 0:
+        result.add("Windows Home Server")
+      elif osvi.ProductType == VER_NT_WORKSTATION and 
+          si.wProcessorArchitecture==PROCESSOR_ARCHITECTURE_AMD64:
+        result.add("Windows XP Professional x64 Edition")
+      else:
+        result.add("Windows Server 2003, ")
+      
+      # Test for the specific product
+      if osvi.ProductType != VER_NT_WORKSTATION:
+        if ze(si.wProcessorArchitecture) == PROCESSOR_ARCHITECTURE_IA64:
+          if (osvi.SuiteMask and VER_SUITE_DATACENTER) != 0:
+            result.add("Datacenter Edition for Itanium-based Systems")
+          elif (osvi.SuiteMask and VER_SUITE_ENTERPRISE) != 0:
+            result.add("Enterprise Edition for Itanium-based Systems")
+        elif ze(si.wProcessorArchitecture) == PROCESSOR_ARCHITECTURE_AMD64:
+          if (osvi.SuiteMask and VER_SUITE_DATACENTER) != 0:
+            result.add("Datacenter x64 Edition")
+          elif (osvi.SuiteMask and VER_SUITE_ENTERPRISE) != 0:
+            result.add("Enterprise x64 Edition")
+          else:
+            result.add("Standard x64 Edition")
+        else:
+          if (osvi.SuiteMask and VER_SUITE_COMPUTE_SERVER) != 0:
+            result.add("Compute Cluster Edition")
+          elif (osvi.SuiteMask and VER_SUITE_DATACENTER) != 0:
+            result.add("Datacenter Edition")
+          elif (osvi.SuiteMask and VER_SUITE_ENTERPRISE) != 0:
+            result.add("Enterprise Edition")
+          elif (osvi.SuiteMask and VER_SUITE_BLADE) != 0:
+            result.add("Web Edition")
+          else:
+            result.add("Standard Edition")
+    # End of 5.2
+    
+    if osvi.majorVersion == 5 and osvi.minorVersion == 1:
+      result.add("Windows XP ")
+      if (osvi.SuiteMask and VER_SUITE_PERSONAL) != 0:
+        result.add("Home Edition")
+      else:
+        result.add("Professional")
+    # End of 5.1
+    
+    if osvi.majorVersion == 5 and osvi.minorVersion == 0:
+      result.add("Windows 2000 ")
+      if osvi.ProductType == VER_NT_WORKSTATION:
+        result.add("Professional")
+      else:
+        if (osvi.SuiteMask and VER_SUITE_DATACENTER) != 0:
+          result.add("Datacenter Server")
+        elif (osvi.SuiteMask and VER_SUITE_ENTERPRISE) != 0:
+          result.add("Advanced Server")
+        else:
+          result.add("Server")
+    # End of 5.0
+    
+    # Include service pack (if any) and build number.
+    if len(osvi.SPVersion) > 0:
+      result.add(" ")
+      result.add(osvi.SPVersion)
+    
+    result.add(" (build " & $osvi.buildNumber & ")")
+    
+    if osvi.majorVersion >= 6:
+      if ze(si.wProcessorArchitecture) == PROCESSOR_ARCHITECTURE_AMD64:
+        result.add(", 64-bit")
+      elif ze(si.wProcessorArchitecture) == PROCESSOR_ARCHITECTURE_INTEL:
+        result.add(", 32-bit")
+    
+  else:
+    # Windows 98 etc...
+    result = "Unknown version of windows[Kernel version <= 4]"
+    
+
+proc getFileSize*(file: string): biggestInt =
+  var fileData: TWIN32_FIND_DATA
+  var hFile = FindFirstFileA(file, fileData)
+  
+  if hFile == INVALID_HANDLE_VALUE:
+    raise newException(EIO, $GetLastError())
+  
+  return fileData.nFileSizeLow
+
+proc GetDiskFreeSpaceEx*(lpDirectoryName: cstring, lpFreeBytesAvailableToCaller,
+                         lpTotalNumberOfBytes,
+                         lpTotalNumberOfFreeBytes: var filetime): WINBOOL{.
+    stdcall, dynlib: "kernel32", importc: "GetDiskFreeSpaceExA".}
+
+proc getPartitionInfo*(partition: string): TPartitionInfo =
+  ## Retrieves partition info, for example ``partition`` may be ``"C:\"``
+  var FreeBytes, TotalBytes, TotalFreeBytes: filetime 
+  var res = GetDiskFreeSpaceEx(r"C:\", FreeBytes, TotalBytes, 
+                               TotalFreeBytes)
+  return (FreeBytes, TotalBytes)
+
+when isMainModule:
+  var r = getMemoryInfo()
+  echo("Memory load: ", r.MemoryLoad, "%")
+  
+  var osvi = getVersionInfo()
+  
+  echo($osvi)
+
+  echo(getFileSize(r"osinfo_win.nim") div 1024 div 1024)
+  
+  echo(rdFileTime(getPartitionInfo(r"C:\")[0]))

lib/nimbase.h

@@ -80,7 +80,7 @@ __TINYC__
 
 /* ---------------- casting without correct aliasing rules ----------- */
 
-#if defined(__GNUCC__)
+#if defined(__GNUC__)
 #  define NIM_CAST(type, ptr) (((union{type __x__;}*)(ptr))->__x__)
 #else
 #  define NIM_CAST(type, ptr) ((type)(ptr))

lib/pure/osproc.nim

@@ -117,7 +117,7 @@ when defined(macosx) or defined(bsd):
 
 proc countProcessors*(): int =
   ## returns the numer of the processors/cores the machine has.
-  ## Returns 0 if it cannot be determined.
+  ## Returns 0 if it cannot be detected.
   when defined(windows):
     var x = getenv("NUMBER_OF_PROCESSORS")
     if x.len > 0: result = parseInt(x)

lib/pure/strutils.nim

@@ -391,10 +391,13 @@ proc endsWith(s, suffix: string): bool =
   var
     i = 0
     j = len(s) - len(suffix)
-  while true:
+  while i+j <% s.len:
-    if suffix[i] == '\0': return true
     if s[i+j] != suffix[i]: return false
     inc(i)
+  if suffix[i] == '\0': return true
+
+# 012345
+#    345
 
 when false:
   proc abbrev(s: string, possibilities: openarray[string]): int = 

rod/c2nim/c2nim.nim

@@ -0,0 +1,74 @@
+#
+#
+#      c2nim - C to Nimrod source converter
+#        (c) Copyright 2010 Andreas Rumpf
+#
+#    See the file "copying.txt", included in this
+#    distribution, for details about the copyright.
+#
+
+import 
+  strutils, os, times, parseopt, llstream, ast, rnimsyn, options, msgs,
+  clex, cparse
+
+const
+  Version = "0.8.10"
+  Usage = """
+c2nim - C to Nimrod source converter
+  (c) 2010 Andreas Rumpf
+Usage: c2nim [options] inputfile [options]
+Options: 
+  -o, --out:FILE         set output filename
+  --dynlib:SYMBOL        import from dynlib: SYMBOL will be used for the import
+  --header:HEADER_FILE   import from a HEADER_FILE (discouraged!)
+  --cdecl                annotate procs with ``{.cdecl.}`` 
+  --stdcall              annotate procs with ``{.stdcall.}``
+  --ref                  convert typ* to ref typ (default: ptr typ)
+  --prefix:PREFIX        strip prefix for the generated Nimrod identifiers 
+                         (multiple --prefix options are supported)
+  --suffix:SUFFIX        strip suffix for the generated Nimrod identifiers 
+                         (multiple --suffix options are supported)
+  --skip:IDENT           skip IDENT in the input file
+  -v, --version          write c2nim's version
+  -h, --help             show this help
+"""
+
+proc main(infile, outfile: string, options: PParserOptions) = 
+  var start = getTime()
+  var stream = LLStreamOpen(infile, fmRead)
+  if stream == nil: rawMessage(errCannotOpenFile, infile)
+  var p: TParser
+  openParser(p, infile, stream, options)
+  var module = parseUnit(p)
+  closeParser(p)
+  renderModule(module, outfile)
+  rawMessage(hintSuccessX, [$gLinesCompiled, $(getTime() - start)])
+
+var
+  infile = ""
+  outfile = ""
+  parserOptions = newParserOptions()
+for kind, key, val in getopt():
+  case kind
+  of cmdArgument: infile = key
+  of cmdLongOption, cmdShortOption:
+    case key.toLower
+    of "help", "h":
+      stdout.write(Usage)
+      quit(0)
+    of "version", "v":
+      stdout.write(Version & "\n")
+      quit(0)
+    of "o", "out": outfile = key
+    else:
+      if not parserOptions.setOption(key, val):
+        stdout.write("[Error] unknown option: " & key)
+  of cmdEnd: assert(false)
+if infile.len == 0:
+  # no filename has been given, so we show the help:
+  stdout.write(Usage)
+else:
+  if outfile.len == 0:
+    outfile = changeFileExt(infile, "nim")
+  infile = addFileExt(infile, "h")
+  main(infile, outfile, parserOptions)

rod/c2nim/clex.nim

@@ -0,0 +1,751 @@
+#
+#
+#      c2nim - C to Nimrod source converter
+#        (c) Copyright 2010 Andreas Rumpf
+#
+#    See the file "copying.txt", included in this
+#    distribution, for details about the copyright.
+#
+
+# This module implements an Ansi C scanner. This is an adaption from
+# the scanner module. Keywords are not handled here, but in the parser to make
+# it more flexible.
+
+
+import 
+  options, msgs, strutils, platform, lexbase, llstream
+
+const 
+  MaxLineLength* = 80         # lines longer than this lead to a warning
+  numChars*: TCharSet = {'0'..'9', 'a'..'z', 'A'..'Z'} 
+  SymChars*: TCharSet = {'a'..'z', 'A'..'Z', '0'..'9', '_', '\x80'..'\xFF'}
+  SymStartChars*: TCharSet = {'a'..'z', 'A'..'Z', '_', '\x80'..'\xFF'}
+
+type
+  TTokKind* = enum 
+    pxInvalid, pxEof,         
+    pxStarComment,            # /* */ comment
+    pxLineComment,            # // comment
+    pxDirective,              # #define, etc.
+    pxDirectiveParLe,         # #define m( with parle (yes, C is that ugly!)
+    pxDirConc,                # ##
+    pxNewLine,                # newline: end of directive
+    pxAmp,                    # &
+    pxAmpAmp,                 # && 
+    pxAmpAsgn,                # &=
+    pxAmpAmpAsgn,             # &&=
+    pxBar,                    # |
+    pxBarBar,                 # ||
+    pxBarAsgn,                # |=
+    pxBarBarAsgn,             # ||=
+    pxNot,                    # !
+    pxPlusPlus,               # ++
+    pxMinusMinus,             # --
+    pxPlus,                   # +
+    pxPlusAsgn,               # +=
+    pxMinus,                  # -
+    pxMinusAsgn,              # -=
+    pxMod,                    # %
+    pxModAsgn,                # %=
+    pxSlash,                  # /
+    pxSlashAsgn,              # /=
+    pxStar,                   # *
+    pxStarAsgn,               # *=
+    pxHat,                    # ^
+    pxHatAsgn,                # ^=
+    pxAsgn,                   # =
+    pxEquals,                 # ==
+    pxDot,                    # .
+    pxDotDotDot,              # ...
+    pxLe,                     # <=
+    pxLt,                     # <
+    pxGe,                     # >=
+    pxGt,                     # >
+    pxNeq,                    # != 
+    pxConditional,            # ?
+    pxShl,                    # <<
+    pxShlAsgn,                # <<=
+    pxShr,                    # >>
+    pxShrAsgn,                # >>=
+    pxTilde,                  # ~
+    pxTildeAsgn,              # ~=
+    pxArrow,                  # ->
+    pxScope,                  # ::
+
+    pxStrLit, 
+    pxCharLit,
+    pxSymbol,                 # a symbol
+    pxIntLit, 
+    pxInt64Lit, # long constant like 0x70fffffff or out of int range
+    pxFloatLit, 
+    pxParLe, pxParRi, 
+    pxBracketLe, pxBracketRi, 
+    pxComma, pxSemiColon, pxColon,
+    pxCurlyLe, pxCurlyRi
+  TTokKinds* = set[TTokKind]
+
+type
+  TNumericalBase* = enum base10, base2, base8, base16
+  TToken* = object
+    xkind*: TTokKind          # the type of the token
+    s*: string                # parsed symbol, char or string literal
+    iNumber*: BiggestInt      # the parsed integer literal
+    fNumber*: BiggestFloat    # the parsed floating point literal
+    base*: TNumericalBase     # the numerical base; only valid for int
+                              # or float literals
+    next*: ref TToken         # for C we need arbitrary look-ahead :-(
+  
+  TLexer* = object of TBaseLexer
+    filename*: string
+    inDirective: bool
+  
+
+proc getTok*(L: var TLexer, tok: var TToken)
+proc PrintTok*(tok: TToken)
+proc `$`*(tok: TToken): string
+# implementation
+
+var
+  gLinesCompiled*: int
+
+proc fillToken(L: var TToken) = 
+  L.xkind = pxInvalid
+  L.iNumber = 0
+  L.s = ""
+  L.fNumber = 0.0
+  L.base = base10
+  
+proc openLexer*(lex: var TLexer, filename: string, inputstream: PLLStream) = 
+  openBaseLexer(lex, inputstream)
+  lex.filename = filename
+
+proc closeLexer*(lex: var TLexer) = 
+  inc(gLinesCompiled, lex.LineNumber)
+  closeBaseLexer(lex)
+
+proc getColumn*(L: TLexer): int = 
+  result = getColNumber(L, L.bufPos)
+
+proc getLineInfo*(L: TLexer): TLineInfo = 
+  result = newLineInfo(L.filename, L.linenumber, getColNumber(L, L.bufpos))
+
+proc lexMessage*(L: TLexer, msg: TMsgKind, arg = "") = 
+  msgs.liMessage(getLineInfo(L), msg, arg)
+
+proc lexMessagePos(L: var TLexer, msg: TMsgKind, pos: int, arg = "") = 
+  var info = newLineInfo(L.filename, L.linenumber, pos - L.lineStart)
+  msgs.liMessage(info, msg, arg)
+
+proc TokKindToStr*(k: TTokKind): string =
+  case k
+  of pxEof: result = "[EOF]"
+  of pxInvalid: result = "[invalid]"
+  of pxStarComment, pxLineComment: result = "[comment]" 
+  of pxStrLit: result = "[string literal]"
+  of pxCharLit: result = "[char literal]"
+
+  of pxDirective, pxDirectiveParLe: result = "#"             # #define, etc.
+  of pxDirConc: result = "##"
+  of pxNewLine: result = "[NewLine]"
+  of pxAmp: result = "&"                   # &
+  of pxAmpAmp: result = "&&"                # && 
+  of pxAmpAsgn: result = "&="                # &=
+  of pxAmpAmpAsgn: result = "&&="            # &&=
+  of pxBar: result = "|"                   # |
+  of pxBarBar: result = "||"                # ||
+  of pxBarAsgn: result = "|="               # |=
+  of pxBarBarAsgn: result = "||="            # ||=
+  of pxNot: result = "!"                   # !
+  of pxPlusPlus: result = "++"             # ++
+  of pxMinusMinus: result = "--"            # --
+  of pxPlus: result = "+"                  # +
+  of pxPlusAsgn: result = "+="              # +=
+  of pxMinus: result = "-"                 # -
+  of pxMinusAsgn: result = "-="             # -=
+  of pxMod: result = "%"                   # %
+  of pxModAsgn: result = "%="               # %=
+  of pxSlash: result = "/"                 # /
+  of pxSlashAsgn: result = "/="             # /=
+  of pxStar: result = "*"                  # *
+  of pxStarAsgn: result = "*="              # *=
+  of pxHat: result = "^"                   # ^
+  of pxHatAsgn: result = "^="               # ^=
+  of pxAsgn: result = "="                  # =
+  of pxEquals: result = "=="                # ==
+  of pxDot: result = "."                   # .
+  of pxDotDotDot: result = "..."             # ...
+  of pxLe: result = "<="                    # <=
+  of pxLt: result = "<"                    # <
+  of pxGe: result = ">="                    # >=
+  of pxGt: result = ">"                    # >
+  of pxNeq: result = "!="                   # != 
+  of pxConditional: result = "?"
+  of pxShl: result = "<<"
+  of pxShlAsgn: result = "<<="
+  of pxShr: result = ">>"
+  of pxShrAsgn: result = ">>="
+  of pxTilde: result = "~"
+  of pxTildeAsgn: result = "~="
+  of pxArrow: result = "->"
+  of pxScope: result = "::"
+
+  of pxSymbol: result = "[identifier]"
+  of pxIntLit, pxInt64Lit: result = "[integer literal]"
+  of pxFloatLit: result = "[floating point literal]"
+  of pxParLe: result = "("
+  of pxParRi: result = ")"
+  of pxBracketLe: result = "["
+  of pxBracketRi: result = "]"
+  of pxComma: result = ","
+  of pxSemiColon: result = ";"
+  of pxColon: result = ":"
+  of pxCurlyLe: result = "{"
+  of pxCurlyRi: result = "}"
+
+proc `$`(tok: TToken): string = 
+  case tok.xkind
+  of pxSymbol, pxInvalid, pxStarComment, pxLineComment, pxStrLit: result = tok.s
+  of pxIntLit, pxInt64Lit: result = $tok.iNumber
+  of pxFloatLit: result = $tok.fNumber
+  else: result = TokKindToStr(tok.xkind)
+  
+proc PrintTok(tok: TToken) = 
+  writeln(stdout, $tok)
+  
+proc matchUnderscoreChars(L: var TLexer, tok: var TToken, chars: TCharSet) = 
+  # matches ([chars]_)*
+  var pos = L.bufpos              # use registers for pos, buf
+  var buf = L.buf
+  while true: 
+    if buf[pos] in chars: 
+      add(tok.s, buf[pos])
+      Inc(pos)
+    else: 
+      break 
+    if buf[pos] == '_': 
+      add(tok.s, '_')
+      Inc(pos)
+  L.bufPos = pos
+
+proc isFloatLiteral(s: string): bool = 
+  for i in countup(0, len(s)-1): 
+    if s[i] in {'.', 'e', 'E'}: 
+      return true
+
+proc getNumber2(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos + 2 # skip 0b
+  tok.base = base2
+  var xi: biggestInt = 0
+  var bits = 0
+  while true: 
+    case L.buf[pos]
+    of 'A'..'Z', 'a'..'z':
+      # ignore type suffix:
+      inc(pos)
+    of '2'..'9', '.': 
+      lexMessage(L, errInvalidNumber)
+      inc(pos)
+    of '_': 
+      inc(pos)
+    of '0', '1': 
+      xi = `shl`(xi, 1) or (ord(L.buf[pos]) - ord('0'))
+      inc(pos)
+      inc(bits)
+    else: break 
+  tok.iNumber = xi
+  if (bits > 32): tok.xkind = pxInt64Lit
+  else: tok.xkind = pxIntLit
+  L.bufpos = pos
+
+proc getNumber8(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos + 2 # skip 0b
+  tok.base = base8
+  var xi: biggestInt = 0
+  var bits = 0
+  while true: 
+    case L.buf[pos]
+    of 'A'..'Z', 'a'..'z':
+      # ignore type suffix:
+      inc(pos)
+    of '8'..'9', '.': 
+      lexMessage(L, errInvalidNumber)
+      inc(pos)
+    of '_': 
+      inc(pos)
+    of '0'..'7': 
+      xi = `shl`(xi, 3) or (ord(L.buf[pos]) - ord('0'))
+      inc(pos)
+      inc(bits)
+    else: break 
+  tok.iNumber = xi
+  if (bits > 12): tok.xkind = pxInt64Lit
+  else: tok.xkind = pxIntLit
+  L.bufpos = pos
+
+proc getNumber16(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos + 2          # skip 0x
+  tok.base = base16
+  var xi: biggestInt = 0
+  var bits = 0
+  while true: 
+    case L.buf[pos]
+    of 'G'..'Z', 'g'..'z': 
+      # ignore type suffix:
+      inc(pos)
+    of '_': inc(pos)
+    of '0'..'9': 
+      xi = `shl`(xi, 4) or (ord(L.buf[pos]) - ord('0'))
+      inc(pos)
+      inc(bits, 4)
+    of 'a'..'f': 
+      xi = `shl`(xi, 4) or (ord(L.buf[pos]) - ord('a') + 10)
+      inc(pos)
+      inc(bits, 4)
+    of 'A'..'F': 
+      xi = `shl`(xi, 4) or (ord(L.buf[pos]) - ord('A') + 10)
+      inc(pos)
+      inc(bits, 4)
+    else: break 
+  tok.iNumber = xi
+  if bits > 32: tok.xkind = pxInt64Lit
+  else: tok.xkind = pxIntLit
+  L.bufpos = pos
+
+proc getNumber(L: var TLexer, tok: var TToken) = 
+  tok.base = base10
+  matchUnderscoreChars(L, tok, {'0'..'9'})
+  if (L.buf[L.bufpos] == '.') and (L.buf[L.bufpos + 1] in {'0'..'9'}): 
+    add(tok.s, '.')
+    inc(L.bufpos)
+    matchUnderscoreChars(L, tok, {'e', 'E', '+', '-', '0'..'9'})
+  try: 
+    if isFloatLiteral(tok.s): 
+      tok.fnumber = parseFloat(tok.s)
+      tok.xkind = pxFloatLit
+    else: 
+      tok.iNumber = ParseInt(tok.s)
+      if (tok.iNumber < low(int32)) or (tok.iNumber > high(int32)): 
+        tok.xkind = pxInt64Lit
+      else: 
+        tok.xkind = pxIntLit
+  except EInvalidValue: 
+    lexMessage(L, errInvalidNumber, tok.s)
+  except EOverflow: 
+    lexMessage(L, errNumberOutOfRange, tok.s)
+  # ignore type suffix:
+  while L.buf[L.bufpos] in {'A'..'Z', 'a'..'z'}: inc(L.bufpos)
+  
+proc HandleCRLF(L: var TLexer, pos: int): int = 
+  case L.buf[pos]
+  of CR: result = lexbase.HandleCR(L, pos)
+  of LF: result = lexbase.HandleLF(L, pos)
+  else: result = pos
+  
+proc escape(L: var TLexer, tok: var TToken, allowEmpty=false) = 
+  inc(L.bufpos) # skip \ 
+  case L.buf[L.bufpos]
+  of 'b', 'B':
+    add(tok.s, '\b')
+    inc(L.bufpos)
+  of 't', 'T':
+    add(tok.s, '\t')
+    inc(L.bufpos)
+  of 'n', 'N':
+    add(tok.s, '\L')
+    inc(L.bufpos)
+  of 'f', 'F':
+    add(tok.s, '\f')
+    inc(L.bufpos)
+  of 'r', 'R':
+    add(tok.s, '\r')
+    inc(L.bufpos)
+  of '\'':
+    add(tok.s, '\'')
+    inc(L.bufpos)
+  of '"':
+    add(tok.s, '"')
+    inc(L.bufpos)
+  of '\\':
+    add(tok.s, '\b')
+    inc(L.bufpos)  
+  of '0'..'7':
+    var xi = ord(L.buf[L.bufpos]) - ord('0')
+    inc(L.bufpos)
+    if L.buf[L.bufpos] in {'0'..'7'}:
+      xi = (xi shl 3) or (ord(L.buf[L.bufpos]) - ord('0'))
+      inc(L.bufpos)
+      if L.buf[L.bufpos] in {'0'..'7'}:
+        xi = (xi shl 3) or (ord(L.buf[L.bufpos]) - ord('0'))
+        inc(L.bufpos)
+    add(tok.s, chr(xi))
+  elif not allowEmpty:
+    lexMessage(L, errInvalidCharacterConstant)
+  
+proc getCharLit(L: var TLexer, tok: var TToken) = 
+  inc(L.bufpos) # skip '
+  if L.buf[L.bufpos] == '\\':
+    escape(L, tok)
+  else:
+    add(tok.s, L.buf[L.bufpos])
+    inc(L.bufpos)
+  if L.buf[L.bufpos] == '\'':
+    inc(L.bufpos)
+  else:
+    lexMessage(L, errMissingFinalQuote)
+  tok.xkind = pxCharLit
+
+proc getString(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufPos + 1          # skip "
+  var buf = L.buf                 # put `buf` in a register
+  var line = L.linenumber         # save linenumber for better error message
+  while true: 
+    case buf[pos]
+    of '\"': 
+      Inc(pos)
+      break
+    of CR: 
+      pos = lexbase.HandleCR(L, pos)
+      buf = L.buf
+    of LF: 
+      pos = lexbase.HandleLF(L, pos)
+      buf = L.buf
+    of lexbase.EndOfFile: 
+      var line2 = L.linenumber
+      L.LineNumber = line
+      lexMessagePos(L, errClosingQuoteExpected, L.lineStart)
+      L.LineNumber = line2
+      break 
+    of '\\': 
+      # we allow an empty \ for line concatenation, but we don't require it
+      # for line concatenation
+      L.bufpos = pos
+      escape(L, tok, allowEmpty=true)
+      pos = L.bufpos
+    else: 
+      add(tok.s, buf[pos])
+      Inc(pos)
+  L.bufpos = pos
+  tok.xkind = pxStrLit
+
+proc getSymbol(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos
+  var buf = L.buf
+  while true: 
+    var c = buf[pos]
+    if c notin SymChars: break
+    add(tok.s, c)
+    Inc(pos)
+  L.bufpos = pos
+  tok.xkind = pxSymbol
+
+proc scanLineComment(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos
+  var buf = L.buf 
+  # a comment ends if the next line does not start with the // on the same
+  # column after only whitespace
+  tok.xkind = pxLineComment
+  var col = getColNumber(L, pos)
+  while true: 
+    inc(pos, 2)               # skip //
+    add(tok.s, '#')
+    while not (buf[pos] in {CR, LF, lexbase.EndOfFile}): 
+      add(tok.s, buf[pos])
+      inc(pos)
+    pos = handleCRLF(L, pos)
+    buf = L.buf
+    var indent = 0
+    while buf[pos] == ' ': 
+      inc(pos)
+      inc(indent)
+    if (col == indent) and (buf[pos] == '/') and (buf[pos + 1] == '/'): 
+      add(tok.s, "\n")
+    else: 
+      break 
+  L.bufpos = pos
+
+proc scanStarComment(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos
+  var buf = L.buf
+  tok.s = "#"
+  tok.xkind = pxStarComment
+  while true: 
+    case buf[pos]
+    of CR, LF: 
+      pos = HandleCRLF(L, pos)
+      buf = L.buf
+      add(tok.s, "\n#")
+      # skip annoying stars as line prefix: (eg.
+      # /*
+      #  * ugly comment <-- this star
+      #  */
+      while buf[pos] in {' ', '\t'}:
+        add(tok.s, ' ')
+        inc(pos)
+      if buf[pos] == '*' and buf[pos+1] != '/': inc(pos)
+    of '*': 
+      inc(pos)
+      if buf[pos] == '/': 
+        inc(pos)
+        break 
+      else: 
+        add(tok.s, '*')
+    of lexbase.EndOfFile: 
+      lexMessage(L, errTokenExpected, "*/")
+    else: 
+      add(tok.s, buf[pos])
+      inc(pos)
+  L.bufpos = pos
+
+proc skip(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos
+  var buf = L.buf
+  while true: 
+    case buf[pos]
+    of '\\': 
+      # Ignore \ line continuation characters when not inDirective
+      inc(pos)
+      if L.inDirective:
+        while buf[pos] in {' ', '\t'}: inc(pos)
+        if buf[pos] in {CR, LF}:
+          pos = HandleCRLF(L, pos)
+          buf = L.buf
+    of ' ', Tabulator: 
+      Inc(pos)                # newline is special:
+    of CR, LF: 
+      pos = HandleCRLF(L, pos)
+      buf = L.buf
+      if L.inDirective:
+        tok.xkind = pxNewLine
+        L.inDirective = false
+    else: 
+      break                   # EndOfFile also leaves the loop
+  L.bufpos = pos
+
+proc getDirective(L: var TLexer, tok: var TToken) = 
+  var pos = L.bufpos + 1
+  var buf = L.buf
+  while buf[pos] in {' ', '\t'}: inc(pos)
+  while buf[pos] in SymChars: 
+    add(tok.s, buf[pos])
+    inc(pos)
+  # a HACK: we need to distinguish 
+  # #define x (...)
+  # from:
+  # #define x(...)
+  # 
+  L.bufpos = pos
+  # look ahead:
+  while buf[pos] in {' ', '\t'}: inc(pos)
+  while buf[pos] in SymChars: inc(pos)
+  if buf[pos] == '(': tok.xkind = pxDirectiveParLe
+  else: tok.xkind = pxDirective
+  L.inDirective = true
+
+proc getTok(L: var TLexer, tok: var TToken) = 
+  tok.xkind = pxInvalid
+  fillToken(tok)
+  skip(L, tok)
+  if tok.xkind == pxNewLine: return
+  var c = L.buf[L.bufpos]
+  if c in SymStartChars: 
+    getSymbol(L, tok)
+  elif c == '0':
+    case L.buf[L.bufpos+1]
+    of 'x', 'X': getNumber16(L, tok)
+    of 'b', 'B': getNumber2(L, tok)
+    of '1'..'7': getNumber8(L, tok)
+    else: getNumber(L, tok)
+  elif c in {'1'..'9'}: 
+    getNumber(L, tok)
+  else: 
+    case c
+    of ';': 
+      tok.xkind = pxSemicolon
+      Inc(L.bufpos)
+    of '/': 
+      if L.buf[L.bufpos + 1] == '/': 
+        scanLineComment(L, tok)
+      elif L.buf[L.bufpos+1] == '*':
+        inc(L.bufpos, 2)
+        scanStarComment(L, tok)
+      elif L.buf[L.bufpos+1] == '=':
+        inc(L.bufpos, 2)
+        tok.xkind = pxSlashAsgn
+      else: 
+        tok.xkind = pxSlash
+        inc(L.bufpos)
+    of ',': 
+      tok.xkind = pxComma
+      Inc(L.bufpos)
+    of '(': 
+      Inc(L.bufpos)
+      tok.xkind = pxParLe
+    of '*': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        inc(L.bufpos)
+        tok.xkind = pxStarAsgn
+      else:
+        tok.xkind = pxStar
+    of ')': 
+      Inc(L.bufpos)
+      tok.xkind = pxParRi
+    of '[': 
+      Inc(L.bufpos)
+      tok.xkind = pxBracketLe
+    of ']': 
+      Inc(L.bufpos)
+      tok.xkind = pxBracketRi
+    of '.': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '.' and L.buf[L.bufpos+1] == '.': 
+        tok.xkind = pxDotDotDot
+        inc(L.bufpos, 2)
+      else: 
+        tok.xkind = pxDot
+    of '{': 
+      Inc(L.bufpos)
+      tok.xkind = pxCurlyLe
+    of '}': 
+      Inc(L.bufpos)
+      tok.xkind = pxCurlyRi
+    of '+': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxPlusAsgn
+        inc(L.bufpos)
+      elif L.buf[L.bufpos] == '+': 
+        tok.xkind = pxPlusPlus
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxPlus
+    of '-': 
+      inc(L.bufpos)
+      case L.buf[L.bufpos] 
+      of '>':
+        tok.xkind = pxArrow
+        inc(L.bufpos)
+      of '=':
+        tok.xkind = pxMinusAsgn
+        inc(L.bufpos)
+      of '-': 
+        tok.xkind = pxMinusMinus
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxMinus
+    of '?':
+      inc(L.bufpos)
+      tok.xkind = pxConditional
+    of ':': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == ':':
+        tok.xkind = pxScope
+        inc(L.bufpos)
+      else:
+        tok.xkind = pxColon
+    of '!':
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxNeq
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxNot
+    of '<': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=': 
+        inc(L.bufpos)
+        tok.xkind = pxLe
+      elif L.buf[L.bufpos] == '<':
+        inc(L.bufpos)
+        if L.buf[L.bufpos] == '=': 
+          inc(L.bufpos)
+          tok.xkind = pxShlAsgn
+        else:
+          tok.xkind = pxShl
+      else: 
+        tok.xkind = pxLt
+    of '>': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=': 
+        inc(L.bufpos)
+        tok.xkind = pxGe
+      elif L.buf[L.bufpos] == '>':
+        inc(L.bufpos)
+        if L.buf[L.bufpos] == '=': 
+          inc(L.bufpos)
+          tok.xkind = pxShrAsgn
+        else:
+          tok.xkind = pxShr
+      else: 
+        tok.xkind = pxGt
+    of '=': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxEquals
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxAsgn
+    of '&': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxAmpAsgn
+        inc(L.bufpos)
+      elif L.buf[L.bufpos] == '&':
+        inc(L.bufpos)
+        if L.buf[L.bufpos] == '=':
+          inc(L.bufpos)
+          tok.xkind = pxAmpAmpAsgn
+        else:
+          tok.xkind = pxAmpAmp
+      else: 
+        tok.xkind = pxAmp
+    of '|': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxBarAsgn
+        inc(L.bufpos)
+      elif L.buf[L.bufpos] == '|':
+        inc(L.bufpos)
+        if L.buf[L.bufpos] == '=':
+          inc(L.bufpos)
+          tok.xkind = pxBarBarAsgn
+        else:
+          tok.xkind = pxBarBar
+      else: 
+        tok.xkind = pxBar
+    of '^': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxHatAsgn
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxHat
+    of '%': 
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxModAsgn
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxMod
+    of '~':
+      inc(L.bufpos)
+      if L.buf[L.bufpos] == '=':
+        tok.xkind = pxTildeAsgn
+        inc(L.bufpos)
+      else: 
+        tok.xkind = pxTilde
+    of '#': 
+      if L.buf[L.bufpos+1] == '#':
+        inc(L.bufpos, 2)
+        tok.xkind = pxDirConc
+      else:
+        getDirective(L, tok)
+    of '"': getString(L, tok)
+    of '\'': getCharLit(L, tok)
+    of lexbase.EndOfFile: 
+      tok.xkind = pxEof
+    else: 
+      tok.s = $c
+      tok.xkind = pxInvalid
+      lexMessage(L, errInvalidToken, c & " (\\" & $(ord(c)) & ')')
+      Inc(L.bufpos)

rod/c2nim/cpp.nim

@@ -0,0 +1,231 @@
+# Preprocessor support
+
+const
+  c2nimSymbol = "C2NIM"
+
+proc eatNewLine(p: var TParser, n: PNode) = 
+  if p.tok.xkind == pxLineComment:
+    skipCom(p, n)
+    if p.tok.xkind == pxNewLine: getTok(p)
+  else:
+    eat(p, pxNewLine)
+
+proc parseDefineBody(p: var TParser, tmplDef: PNode): string = 
+  if p.tok.xkind == pxCurlyLe or 
+    (p.tok.xkind == pxSymbol and (declKeyword(p.tok.s) or stmtKeyword(p.tok.s))):
+    addSon(tmplDef, statement(p))
+    result = "stmt"
+  elif p.tok.xkind in {pxLineComment, pxNewLine}:
+    addSon(tmplDef, buildStmtList(newNodeP(nkNilLit, p)))
+    result = "stmt"
+  else:
+    addSon(tmplDef, buildStmtList(expression(p)))
+    result = "expr"
+
+proc parseDefine(p: var TParser): PNode = 
+  if p.tok.xkind == pxDirectiveParLe: 
+    # a macro with parameters:
+    result = newNodeP(nkTemplateDef, p)
+    getTok(p)
+    addSon(result, skipIdent(p))
+    eat(p, pxParLe)
+    var params = newNodeP(nkFormalParams, p)
+    # return type; not known yet:
+    addSon(params, nil)  
+    var identDefs = newNodeP(nkIdentDefs, p)
+    while p.tok.xkind != pxParRi: 
+      addSon(identDefs, skipIdent(p))
+      skipStarCom(p, nil)
+      if p.tok.xkind != pxComma: break
+      getTok(p)
+    addSon(identDefs, newIdentNodeP("expr", p))
+    addSon(identDefs, nil)
+    addSon(params, identDefs)
+    eat(p, pxParRi)
+    
+    addSon(result, nil) # no generic parameters
+    addSon(result, params)
+    addSon(result, nil) # no pragmas
+    var kind = parseDefineBody(p, result)
+    params.sons[0] = newIdentNodeP(kind, p)
+    eatNewLine(p, result)
+  else:
+    # a macro without parameters:
+    result = newNodeP(nkConstSection, p)
+    while p.tok.xkind == pxDirective and p.tok.s == "define":
+      getTok(p) # skip #define
+      var c = newNodeP(nkConstDef, p)
+      addSon(c, skipIdent(p))
+      addSon(c, nil)
+      skipStarCom(p, c)
+      if p.tok.xkind in {pxLineComment, pxNewLine, pxEof}:
+        addSon(c, newIdentNodeP("true", p))
+      else:
+        addSon(c, expression(p))
+      addSon(result, c)
+      eatNewLine(p, c)
+  
+proc isDir(p: TParser, dir: string): bool = 
+  result = p.tok.xkind in {pxDirectiveParLe, pxDirective} and p.tok.s == dir
+
+proc parseInclude(p: var TParser): PNode = 
+  result = newNodeP(nkImportStmt, p)
+  while isDir(p, "include"):
+    getTok(p) # skip "include"
+    if p.tok.xkind == pxStrLit:
+      var file = newStrNodeP(nkStrLit, p.tok.s, p)
+      addSon(result, file)
+      getTok(p)
+      skipStarCom(p, file)
+    elif p.tok.xkind == pxLt: 
+      while p.tok.xkind notin {pxEof, pxNewLine, pxLineComment}: getTok(p)
+    else:
+      parMessage(p, errXExpected, "string literal")
+    eatNewLine(p, nil)
+  if sonsLen(result) == 0: 
+    # we only parsed includes that we chose to ignore:
+    result = nil
+
+proc definedExprAux(p: var TParser): PNode = 
+  result = newNodeP(nkCall, p)
+  addSon(result, newIdentNodeP("defined", p))
+  addSon(result, skipIdent(p))
+
+proc parseStmtList(p: var TParser): PNode = 
+  result = newNodeP(nkStmtList, p)
+  while true: 
+    case p.tok.xkind
+    of pxEof: break 
+    of pxDirectiveParLe, pxDirective: 
+      case p.tok.s
+      of "else", "endif", "elif": break
+    else: nil
+    addSon(result, statement(p))
+  
+proc parseIfDirAux(p: var TParser, result: PNode) = 
+  addSon(result.sons[0], parseStmtList(p))
+  while isDir(p, "elif"): 
+    var b = newNodeP(nkElifBranch, p)
+    getTok(p)
+    addSon(b, expression(p))
+    eatNewLine(p, nil)
+    addSon(b, parseStmtList(p))
+    addSon(result, b)
+  if isDir(p, "else"): 
+    var s = newNodeP(nkElse, p)
+    while p.tok.xkind notin {pxEof, pxNewLine, pxLineComment}: getTok(p)
+    eatNewLine(p, nil)
+    addSon(s, parseStmtList(p))
+    addSon(result, s)
+  if isDir(p, "endif"): 
+    while p.tok.xkind notin {pxEof, pxNewLine, pxLineComment}: getTok(p)
+    eatNewLine(p, nil)
+  else: 
+    parMessage(p, errXExpected, "#endif")
+  
+proc specialIf(p: TParser): bool = 
+  ExpectIdent(p)
+  result = p.tok.s == c2nimSymbol
+  
+proc chooseBranch(whenStmt: PNode, branch: int): PNode = 
+  var L = sonsLen(whenStmt)
+  if branch < L: 
+    if L == 2 and whenStmt[1].kind == nkElse or branch == 0: 
+      result = lastSon(whenStmt[branch])
+    else:
+      var b = whenStmt[branch]
+      assert(b.kind == nkElifBranch)
+      result = newNodeI(nkWhenStmt, whenStmt.info)
+      for i in branch .. L-1:
+        addSon(result, whenStmt[i])
+  
+proc skipIfdefCPlusPlus(p: var TParser): PNode =
+  while p.tok.xkind != pxEof:
+    if isDir(p, "endif"): 
+      while p.tok.xkind notin {pxEof, pxNewLine, pxLineComment}: getTok(p)
+      eatNewLine(p, nil)
+      return
+    getTok(p)
+  parMessage(p, errXExpected, "#endif")
+  
+proc parseIfdefDir(p: var TParser): PNode = 
+  result = newNodeP(nkWhenStmt, p)
+  addSon(result, newNodeP(nkElifBranch, p))
+  getTok(p)
+  var special = specialIf(p)
+  if p.tok.s == "__cplusplus": 
+    return skipIfdefCPlusPlus(p)
+  addSon(result.sons[0], definedExprAux(p))
+  eatNewLine(p, nil)
+  parseIfDirAux(p, result)
+  if special: 
+    result = chooseBranch(result, 0)
+
+proc parseIfndefDir(p: var TParser): PNode = 
+  result = newNodeP(nkWhenStmt, p)
+  addSon(result, newNodeP(nkElifBranch, p))
+  getTok(p)
+  var special = specialIf(p)
+  var e = newNodeP(nkCall, p)
+  addSon(e, newIdentNodeP("not", p))
+  addSon(e, definedExprAux(p))
+  eatNewLine(p, nil)
+  addSon(result.sons[0], e)
+  parseIfDirAux(p, result)
+  if special:
+    result = chooseBranch(result, 1)
+
+proc parseIfDir(p: var TParser): PNode = 
+  result = newNodeP(nkWhenStmt, p)
+  addSon(result, newNodeP(nkElifBranch, p))
+  getTok(p)
+  addSon(result.sons[0], expression(p))
+  eatNewLine(p, nil)
+  parseIfDirAux(p, result)
+
+proc parseMangleDir(p: var TParser) = 
+  var col = getColumn(p.lex) + 2
+  getTok(p)
+  if p.tok.xkind != pxStrLit: ExpectIdent(p)
+  try:
+    var pattern = parsePeg(
+      input = p.tok.s, 
+      filename = p.lex.filename, 
+      line = p.lex.linenumber, 
+      col = col)
+    getTok(p)
+    if p.tok.xkind != pxStrLit: ExpectIdent(p)
+    p.options.mangleRules.add((pattern, p.tok.s))
+    getTok(p)
+  except EInvalidPeg:
+    parMessage(p, errUser, getCurrentExceptionMsg())
+  eatNewLine(p, nil)
+
+proc parseDir(p: var TParser): PNode = 
+  assert(p.tok.xkind in {pxDirective, pxDirectiveParLe})
+  case p.tok.s
+  of "define": result = parseDefine(p)
+  of "include": result = parseInclude(p)
+  of "ifdef": result = parseIfdefDir(p)
+  of "ifndef": result = parseIfndefDir(p)
+  of "if": result = parseIfDir(p)
+  of "cdecl", "stdcall", "ref": 
+    discard setOption(p.options, p.tok.s)
+    getTok(p)
+    eatNewLine(p, nil)
+  of "dynlib", "header", "prefix", "suffix", "skip": 
+    var key = p.tok.s
+    getTok(p)
+    if p.tok.xkind != pxStrLit: ExpectIdent(p)
+    discard setOption(p.options, key, p.tok.s)
+    getTok(p)
+    eatNewLine(p, nil)
+  of "mangle":
+    parseMangleDir(p)
+  else: 
+    # ignore unimportant/unknown directive ("undef", "pragma", "error")
+    while true:
+      getTok(p)
+      if p.tok.xkind in {pxEof, pxNewLine, pxLineComment}: break
+    eatNewLine(p, nil)
+

rod/c2nim/manual.txt

@@ -0,0 +1,235 @@
+=================================
+  c2nim  User's manual
+=================================
+
+:Author: Andreas Rumpf
+:Version: 0.8.10
+
+Introduction
+============
+
+c2nim is a tool to translate Ansi C code to Nimrod. The output is 
+human-readable Nimrod code that is meant to be tweaked by hand after the 
+translation process. c2nim is no real compiler!
+
+c2nim is preliminary meant to translate C header files. Because of this, the 
+preprocessor is part of the parser. For example:
+
+.. code-block:: C
+  #define abc 123
+  #define xyz 789
+  
+Is translated into:
+
+.. code-block:: Nimrod
+  const
+    abc* = 123
+    xyz* = 789
+    
+
+c2nim is meant to translate fragments of C code and thus does not follow 
+include files. c2nim cannot parse all of Ansi C and many constructs cannot
+be represented in Nimrod: for example `duff's device`:idx: cannot be translated
+to Nimrod. 
+
+
+Preprocessor support
+====================
+
+Even though the translation process is not perfect, it is often the case that
+the translated Nimrod code does not need any tweaking by hand. In other cases
+it may be preferable to modify the input file instead of the generated Nimrod
+code so that c2nim can parse it properly. c2nim's preprocessor defines the 
+symbol ``C2NIM`` that can be used to mark code sections: 
+
+.. code-block:: C
+  #ifndef C2NIM
+    // C2NIM should ignore this prototype:
+    int fprintf(FILE* f, const char* frmt, ...);
+  #endif
+  
+The ``C2NIM`` symbol is only recognized in ``#ifdef`` and ``#ifndef`` 
+constructs! ``#if defined(C2NIM)`` does **not** work. 
+
+c2nim *processes* ``#ifdef C2NIM`` and ``#ifndef C2NIM`` directives, but other
+``#if[def]`` directives are *translated* into Nimrod's ``when`` construct: 
+
+.. code-block:: C
+  #ifdef DEBUG
+  #  define OUT(x) printf("%s\n", x)
+  #else
+  #  define OUT(x)
+  #endif
+  
+Is translated into:
+
+.. code-block:: Nimrod
+  when defined(debug):
+    template OUT(x: expr): expr = 
+      printf("%s\x0A", x)
+  else:
+    template OUT(x: expr): stmt = 
+      nil
+  
+As can been seen from the example, C's macros with parameters are mapped
+to Nimrod's templates. This mapping is the best one can do, but it is of course
+not accurate: Nimrod's templates operate on syntax trees whereas C's 
+macros work on the token level. c2nim cannot translate any macro that contains
+the ``##`` token concatenation operator.
+
+c2nim's preprocessor supports special directives that affect how the output
+is generated. They should be put into a ``#ifdef C2NIM`` section so that 
+ordinary C compilers ignore them.
+
+
+``#stdcall`` and ``#cdecl`` directives
+--------------------------------------
+**Note**: There are also ``--stdcall`` and ``--cdecl`` command line options
+that can be used for the same purpose.
+
+These directives tell c2nim that it should annotate every proc (or proc type)
+with the ``stdcall`` / ``cdecl`` calling convention.
+
+
+``#dynlib`` directive
+---------------------
+**Note**: There is also a ``--dynlib`` command line option that can be used for
+the same purpose.
+
+This directive tells c2nim that it should annotate every proc that resulted
+from a C function prototype with the ``dynlib`` pragma:
+ 
+.. code-block:: C
+  
+  #ifdef C2NIM
+  #  dynlib iupdll
+  #  cdecl
+  #  if defined(windows)
+  #    define iupdll "iup.dll"
+  #  elif defined(macosx)
+  #    define iupdll "libiup.dynlib"
+  #  else
+  #    define iupdll "libiup.so"
+  #  endif
+  #endif
+  
+  int IupConvertXYToPos(PIhandle ih, int x, int y);
+
+Is translated to:
+  
+.. code-block:: Nimrod
+  when defined(windows):
+    const iupdll* = "iup.dll"
+  elif defined(macosx):
+    const iupdll* = "libiup.dynlib"
+  else:
+    const iupdll* = "libiup.so"
+  
+  proc IupConvertXYToPos*(ih: PIhandle, x: cint, y: cint): cint {.
+    importc: "IupConvertXYToPos", cdecl, dynlib: iupdll.}
+
+Note how the example contains extra C code to declare the ``iupdll`` symbol
+in the generated Nimrod code.
+
+
+``#header`` directive
+---------------------
+**Note**: There is also a ``--header`` command line option that can be used for
+the same purpose.
+
+The ``#header`` directive tells c2nim that it should annotate every proc that
+resulted from a C function prototype and every exported variable and type with
+the ``header`` pragma:
+
+.. code-block:: C
+  
+  #ifdef C2NIM
+  #  header "iup.h"
+  #endif
+  
+  int IupConvertXYToPos(PIhandle ih, int x, int y);
+
+Is translated to:
+  
+.. code-block:: Nimrod
+  proc IupConvertXYToPos*(ih: PIhandle, x: cint, y: cint): cint {.
+    importc: "IupConvertXYToPos", header: "iup.h".}
+
+The ``#header`` and the ``#dynlib`` directives are mutually exclusive. 
+A binding that uses ``dynlib`` is much more preferable over one that uses
+``header``! The Nimrod compiler might drop support for the ``header`` pragma
+in the future as it cannot work for backends that do not generate C code.
+
+
+``#prefix`` and ``#suffix`` directives
+--------------------------------------
+
+**Note**: There are also ``--prefix`` and ``--suffix`` command line options
+that can be used for the same purpose.
+
+c2nim does not do any name mangling by default. However the 
+ ``#prefix`` and ``#suffix`` directives can be used to strip prefixes and
+suffixes from the identifiers in the C code: 
+ 
+.. code-block:: C
+  
+  #ifdef C2NIM
+  #  prefix Iup
+  #  dynlib dllname
+  #  cdecl
+  #endif
+  
+  int IupConvertXYToPos(PIhandle ih, int x, int y);
+
+Is translated to:
+  
+.. code-block:: Nimrod
+
+  proc ConvertXYToPos*(ih: PIhandle, x: cint, y: cint): cint {.
+    importc: "IupConvertXYToPos", cdecl, dynlib: dllname.}
+
+
+``#mangle`` directive
+---------------------
+
+Even more sophisticated name mangling can be achieved by the ``#mangle`` 
+directive: It takes a PEG pattern and format string that specify how the
+identifier should be converted: 
+
+.. code-block:: C
+  #mangle "'GTK_'{.*}" "TGtk$1"
+
+
+``#skip`` directive
+-------------------
+**Note**: There is also ``--skip`` command line option that can be used for the
+same purpose.
+
+Often C code contains special macros that affect the declaration of a function
+prototype but confuse c2nim's parser:
+
+.. code-block:: C
+  // does not parse!
+  EXPORT int f(void); 
+  EXPORT int g(void); 
+
+Instead of to remove ``EXPORT`` from the input source file, one can tell c2nim
+to skip special identifiers:
+
+.. code-block:: C
+  #skip EXPORT
+  // does parse now!
+  EXPORT int f(void);
+  EXPORT int g(void); 
+
+
+Limitations
+===========
+
+* C's ``,`` operator (comma operator) is not supported.
+* C's ``union`` has no equivalent in Nimrod.
+* Standalone ``struct x {}`` declarations are not implemented. Put them into
+  a ``typedef``.
+* The condition in a ``do while(condition)`` statement must be ``0``.
+* Lots of other small issues...
+

rod/ccgstmts.nim

@@ -581,16 +581,13 @@ proc genTryStmt(p: BProc, t: PNode) =
   #      sp.status = RangeError; /* if raise; else 0 */
   #    }
   #  }
+  #  excHandler = excHandler->prev; /* deactivate this safe point */ 
   #  /* finally: */
   #  printf('fin!\n');
   #  if (sp.status != 0)
   #    longjmp(excHandler->context, sp.status);
-  #  excHandler = excHandler->prev; /* deactivate this safe point */ 
-  var 
-    i, length, blen: int
-    safePoint, orExpr: PRope
   genLineDir(p, t)
-  safePoint = getTempName()
+  var safePoint = getTempName()
   useMagic(p.module, "TSafePoint")
   useMagic(p.module, "E_Base")
   useMagic(p.module, "excHandler")
@@ -600,13 +597,13 @@ proc genTryStmt(p: BProc, t: PNode) =
   if optStackTrace in p.Options: 
     app(p.s[cpsStmts], "framePtr = (TFrame*)&F;" & tnl)
   appf(p.s[cpsStmts], "if ($1.status == 0) {$n", [safePoint])
-  length = sonsLen(t)
+  var length = sonsLen(t)
   inc(p.nestedTryStmts)
   genStmts(p, t.sons[0])
   app(p.s[cpsStmts], "} else {" & tnl)
-  i = 1
+  var i = 1
   while (i < length) and (t.sons[i].kind == nkExceptBranch): 
-    blen = sonsLen(t.sons[i])
+    var blen = sonsLen(t.sons[i])
     if blen == 1: 
       # general except section:
       if i > 1: app(p.s[cpsStmts], "else {" & tnl)
@@ -614,7 +611,7 @@ proc genTryStmt(p: BProc, t: PNode) =
       appf(p.s[cpsStmts], "$1.status = 0;$n", [safePoint])
       if i > 1: app(p.s[cpsStmts], '}' & tnl)
     else: 
-      orExpr = nil
+      var orExpr: PRope = nil
       for j in countup(0, blen - 2): 
         assert(t.sons[i].sons[j].kind == nkType)
         if orExpr != nil: app(orExpr, "||")

rod/extccomp.nim

@@ -50,7 +50,7 @@ const
      compilerExe: "gcc", 
      compileTmpl: "-c $options $include -o $objfile $file", 
      buildGui: " -mwindows", 
-     buildDll: " -mdll", 
+     buildDll: " -shared", 
      linkerExe: "gcc", 
      linkTmpl: "$options $buildgui $builddll -o $exefile $objfiles", 
      includeCmd: " -I", 
@@ -65,7 +65,7 @@ const
      compilerExe: "llvm-gcc", 
      compileTmpl: "-c $options $include -o $objfile $file", 
      buildGui: " -mwindows", 
-     buildDll: " -mdll", 
+     buildDll: " -shared", 
      linkerExe: "llvm-gcc", 
      linkTmpl: "$options $buildgui $builddll -o $exefile $objfiles", 
      includeCmd: " -I", 
@@ -225,7 +225,7 @@ const
 
 var ccompiler*: TSystemCC = ccGcc
 
-const                         # the used compiler
+const                   # the used compiler
   hExt* = "h"
 
 var cExt*: string = "c" # extension of generated C/C++ files
@@ -490,7 +490,6 @@ proc CallCCompiler(projectfile: string) =
     generateScript(projectFile, script)
 
 proc genMappingFiles(list: TLinkedList): PRope = 
-  result = nil
   var it = PStrEntry(list.head)
   while it != nil: 
     appf(result, "--file:r\"$1\"$n", [toRope(AddFileExt(it.data, cExt))])

rod/pragmas.nim

@@ -1,7 +1,7 @@
 #
 #
 #           The Nimrod Compiler
-#        (c) Copyright 2009 Andreas Rumpf
+#        (c) Copyright 2010 Andreas Rumpf
 #
 #    See the file "copying.txt", included in this
 #    distribution, for details about the copyright.
@@ -37,7 +37,7 @@ const
   lambdaPragmas* = {FirstCallConv..LastCallConv, wImportc, wExportc, wNodecl, 
     wNosideEffect, wSideEffect, wNoreturn, wDynLib, wHeader, wPure, wDeprecated}
   typePragmas* = {wImportc, wExportc, wDeprecated, wMagic, wAcyclic, wNodecl, 
-    wPure, wHeader, wCompilerProc, wFinal}
+    wPure, wHeader, wCompilerProc, wFinal, wSize}
   fieldPragmas* = {wImportc, wExportc, wDeprecated}
   varPragmas* = {wImportc, wExportc, wVolatile, wRegister, wThreadVar, wNodecl, 
     wMagic, wHeader, wDeprecated, wCompilerProc, wDynLib}
@@ -344,6 +344,13 @@ proc pragma(c: PContext, sym: PSym, n: PNode, validPragmas: TSpecialWords) =
           sym.typ.align = expectIntLit(c, it)
           if not IsPowerOfTwo(sym.typ.align) and (sym.typ.align != 0): 
             liMessage(it.info, errPowerOfTwoExpected)
+        of wSize: 
+          if sym.typ == nil: invalidPragma(it)
+          var size = expectIntLit(c, it)
+          if not IsPowerOfTwo(size) or size <= 0 or size > 8: 
+            liMessage(it.info, errPowerOfTwoExpected)
+          else:
+            sym.typ.size = size
         of wNodecl: 
           noVal(it)
           incl(sym.loc.Flags, lfNoDecl)

rod/rnimsyn.nim

@@ -615,13 +615,19 @@ proc gproc(g: var TSrcGen, n: PNode) =
 proc gblock(g: var TSrcGen, n: PNode) = 
   var c: TContext
   initContext(c)
-  putWithSpace(g, tkBlock, "block")
+  if n.sons[0] != nil:
-  gsub(g, n.sons[0])
+    putWithSpace(g, tkBlock, "block")
+    gsub(g, n.sons[0])
+  else:
+    put(g, tkBlock, "block")
   putWithSpace(g, tkColon, ":")
   if longMode(n) or (lsub(n.sons[1]) + g.lineLen > maxLineLen): 
     incl(c.flags, rfLongMode)
   gcoms(g)
+  # XXX I don't get why this is needed here! gstmts should already handle this!
+  indentNL(g)
   gstmts(g, n.sons[1], c)
+  dedent(g)
 
 proc gasm(g: var TSrcGen, n: PNode) = 
   putWithSpace(g, tkAsm, "asm")

rod/scanner.nim

@@ -427,7 +427,7 @@ proc getEscapedChar(L: var TLexer, tok: var TToken) =
   case L.buf[L.bufpos]
   of 'n', 'N': 
     if tok.toktype == tkCharLit: lexMessage(L, errNnotAllowedInCharacter)
-    tok.literal = tok.literal & tnl
+    add(tok.literal, tnl)
     Inc(L.bufpos)
   of 'r', 'R', 'c', 'C': 
     add(tok.literal, CR)

rod/semgnrc.nim

@@ -1,11 +1,12 @@
 #
 #
 #           The Nimrod Compiler
-#        (c) Copyright 2009 Andreas Rumpf
+#        (c) Copyright 2010 Andreas Rumpf
 #
 #    See the file "copying.txt", included in this
 #    distribution, for details about the copyright.
 #
+
 # This implements the first pass over the generic body; it resolves some
 # symbols. Thus for generics there is a two-phase symbol lookup just like
 # in C++.
@@ -155,8 +156,8 @@ proc semGenericStmt(c: PContext, n: PNode, flags: TSemGenericFlags = {}): PNode
       if (a.kind != nkIdentDefs): IllFormedAst(a)
       checkMinSonsLen(a, 3)
       L = sonsLen(a)
-      a.sons[L - 2] = semGenericStmt(c, a.sons[L - 2], {withinTypeDesc}) # do not perform symbol lookup for default 
+      a.sons[L - 2] = semGenericStmt(c, a.sons[L - 2], {withinTypeDesc}) 
-                                                                         # expressions 
+      # do not perform symbol lookup for default expressions 
       for j in countup(0, L - 3): 
         addDecl(c, newSymS(skUnknown, getIdentNode(a.sons[j]), c))
   of nkConstSection: 
@@ -196,8 +197,7 @@ proc semGenericStmt(c: PContext, n: PNode, flags: TSemGenericFlags = {}): PNode
       of nkEnumFieldDef: a = n.sons[i].sons[0]
       of nkIdent: a = n.sons[i]
       else: illFormedAst(n)
-      addDeclAt(c, newSymS(skUnknown, getIdentNode(a.sons[i]), c), c.tab.tos -
+      addDeclAt(c, newSymS(skUnknown, getIdentNode(a.sons[i]), c), c.tab.tos-1)
-          1)
   of nkObjectTy, nkTupleTy: 
     nil
   of nkFormalParams: 

rod/sigmatch.nim

@@ -92,18 +92,15 @@ proc getNotFoundError(c: PContext, n: PNode): string =
   # Gives a detailed error message; this is seperated from semDirectCall,
   # as semDirectCall is already pretty slow (and we need this information only
   # in case of an error).
-  var 
-    sym: PSym
-    o: TOverloadIter
-    candidates: string
   result = msgKindToString(errTypeMismatch)
   for i in countup(1, sonsLen(n) - 1): 
     #debug(n.sons[i].typ);
     add(result, typeToString(n.sons[i].typ))
     if i != sonsLen(n) - 1: add(result, ", ")
   add(result, ')')
-  candidates = ""
+  var candidates = ""
-  sym = initOverloadIter(o, c, n.sons[0])
+  var o: TOverloadIter  
+  var sym = initOverloadIter(o, c, n.sons[0])
   while sym != nil: 
     if sym.kind in {skProc, skMethod, skIterator, skConverter}: 
       add(candidates, getProcHeader(sym))

rod/syntaxes.nim

@@ -37,6 +37,7 @@ proc parseAll*(p: var TParsers): PNode
 proc parseTopLevelStmt*(p: var TParsers): PNode
   # implements an iterator. Returns the next top-level statement or nil if end
   # of stream.
+
 # implementation
 
 proc ParseFile(filename: string): PNode = 
@@ -57,7 +58,8 @@ proc parseAll(p: var TParsers): PNode =
   of skinBraces: 
     result = pbraces.parseAll(p.parser)
   of skinEndX: 
-    InternalError("parser to implement") # skinEndX: result := pendx.parseAll(p.parser);
+    InternalError("parser to implement") 
+    # skinEndX: result := pendx.parseAll(p.parser);
   
 proc parseTopLevelStmt(p: var TParsers): PNode = 
   case p.skin
@@ -66,39 +68,33 @@ proc parseTopLevelStmt(p: var TParsers): PNode =
   of skinBraces: 
     result = pbraces.parseTopLevelStmt(p.parser)
   of skinEndX: 
-    InternalError("parser to implement") #skinEndX: result := pendx.parseTopLevelStmt(p.parser);
+    InternalError("parser to implement") 
+    #skinEndX: result := pendx.parseTopLevelStmt(p.parser);
   
 proc UTF8_BOM(s: string): int = 
-  if (s[0] == '\xEF') and (s[0 + 1] == '\xBB') and (s[0 + 2] == '\xBF'): 
+  if (s[0] == '\xEF') and (s[1] == '\xBB') and (s[2] == '\xBF'): 
     result = 3
   else: 
     result = 0
   
 proc containsShebang(s: string, i: int): bool = 
-  var j: int
-  result = false
   if (s[i] == '#') and (s[i + 1] == '!'): 
-    j = i + 2
+    var j = i + 2
     while s[j] in WhiteSpace: inc(j)
     result = s[j] == '/'
 
 proc parsePipe(filename: string, inputStream: PLLStream): PNode = 
-  var 
+  var s = LLStreamOpen(filename, fmRead)
-    line: string
-    s: PLLStream
-    i: int
-    q: TParser
-  result = nil
-  s = LLStreamOpen(filename, fmRead)
   if s != nil: 
-    line = LLStreamReadLine(s)
+    var line = LLStreamReadLine(s)
-    i = UTF8_Bom(line) + 0
+    var i = UTF8_Bom(line)
     if containsShebang(line, i): 
       line = LLStreamReadLine(s)
       i = 0
     if (line[i] == '#') and (line[i + 1] == '!'): 
       inc(i, 2)
       while line[i] in WhiteSpace: inc(i)
+      var q: TParser
       OpenParser(q, filename, LLStreamOpen(copy(line, i)))
       result = pnimsyn.parseAll(q)
       CloseParser(q)
@@ -124,12 +120,10 @@ proc getCallee(n: PNode): PIdent =
   else: 
     rawMessage(errXNotAllowedHere, renderTree(n))
   
-proc applyFilter(p: var TParsers, n: PNode, filename: string, stdin: PLLStream): PLLStream = 
+proc applyFilter(p: var TParsers, n: PNode, filename: string, 
-  var 
+                 stdin: PLLStream): PLLStream = 
-    ident: PIdent
+  var ident = getCallee(n)
-    f: TFilterKind
+  var f = getFilter(ident)
-  ident = getCallee(n)
-  f = getFilter(ident)
   case f
   of filtNone: 
     p.skin = getParser(ident)
@@ -146,7 +140,8 @@ proc applyFilter(p: var TParsers, n: PNode, filename: string, stdin: PLLStream):
       messageOut(result.s)
       rawMessage(hintCodeEnd, [])
 
-proc evalPipe(p: var TParsers, n: PNode, filename: string, start: PLLStream): PLLStream = 
+proc evalPipe(p: var TParsers, n: PNode, filename: string, 
+              start: PLLStream): PLLStream = 
   result = start
   if n == nil: return 
   if (n.kind == nkInfix) and (n.sons[0].kind == nkIdent) and
@@ -162,16 +157,14 @@ proc evalPipe(p: var TParsers, n: PNode, filename: string, start: PLLStream): PL
     result = applyFilter(p, n, filename, result)
   
 proc openParsers(p: var TParsers, filename: string, inputstream: PLLStream) = 
-  var 
+  var s: PLLStream
-    pipe: PNode
-    s: PLLStream
   p.skin = skinStandard
-  pipe = parsePipe(filename, inputStream)
+  var pipe = parsePipe(filename, inputStream)
   if pipe != nil: s = evalPipe(p, pipe, filename, inputStream)
   else: s = inputStream
   case p.skin
-  of skinStandard, skinBraces, skinEndX: pnimsyn.openParser(p.parser, filename, 
+  of skinStandard, skinBraces, skinEndX: 
-      s)
+    pnimsyn.openParser(p.parser, filename, s)
   
 proc closeParsers(p: var TParsers) = 
   pnimsyn.closeParser(p.parser)

rod/types.nim

@@ -502,7 +502,7 @@ proc base(t: PType): PType =
 
 proc firstOrd(t: PType): biggestInt = 
   case t.kind
-  of tyBool, tyChar, tySequence, tyOpenArray: 
+  of tyBool, tyChar, tySequence, tyOpenArray, tyString: 
     result = 0
   of tySet, tyVar: 
     result = firstOrd(t.sons[0])

tests/accept/compile/tenum3.nim

@@ -0,0 +1,16 @@
+# Test enum with explicit size
+
+type
+  TEnumHole {.size: sizeof(int).} = enum 
+    eA = 0,
+    eB = 4,
+    eC = 5
+    
+var
+  e: TEnumHole = eB
+  
+case e
+of eA: echo "A"
+of eB: echo "B"
+of eC: echo "C"
+

tests/accept/run/spec.csv

@@ -3,6 +3,7 @@ tambsym2.nim;7
 tambsys.nim;
 tarray.nim;10012
 tarray2.nim;[16, 25, 36]
+tarray3.nim;3
 tassert.nim;assertion failure!this shall be always written
 tbind1.nim;3
 tbind3.nim;1

tests/accept/run/tarray3.nim

@@ -0,0 +1,7 @@
+# simple check for two dimensional arrays
+
+const  
+  myData = [[1,2,3], [4, 5, 6]]
+
+echo myData[0][2] #OUT 3
+

todo.txt

@@ -0,0 +1,155 @@
+For version 0.8.10
+==================
+
+- fix exception handling
+- fix implicit generic routines
+- fix the streams implementation so that they use methods
+
+
+TO IMPLEMENT
+============
+
+* fix overloading resolution
+* wrong co-/contravariance
+* startsWith `=^`
+* endsWith `=$`
+* ignore case `=?` --> `=$?`
+
+Bugs
+----
+- proc (x: int) is passable to proc (x: var int)  !?
+- detected by pegs module 64bit: p(result, result) should use a temporary!
+- seq[TLoc] in C code generator always failed --> probably some reference
+  counting is wrong; try to reproduce this in the GC test
+- the parser allows empty object case branches
+- SDL event handling
+- accurate file/line information
+
+
+To implement
+------------
+
+* icon installation for the Windows installer
+* sort routine
+* hash tables and sets
+* the two other parsers
+* distinct types for array/seq indexes
+* constant sequences
+* IMPLEMENT GENERIC TYPES!
+  - implement expr parameters
+  - document generic types better
+
+* implement closures for the C code generator
+
+
+
+Low priority
+------------
+- resizing of strings/sequences could take into account the memory that
+  is allocated
+- implicit conversions from ``ptr/ref T`` to ``var T`` (from
+  ``ptr/ref T`` to ``T``)?
+- documentation: type converters
+- typeAllowed() for parameters...
+- find a way to reintroduce the cleanup() pass for C code generation: this
+  is hard because of partial evaluation --> symbol files will fix this as
+  a side effect
+- implement better error handling: errornous top level statements are ignored
+  and not part of the symbal table --> for REPL
+  --> this needs deletion operation for symbol table!
+- floating point checks for EcmaScript
+- enhance `` notation for identifier concatenation: `concat` a `these`
+- prefer proc in current module over other procs with same overloading result?
+- real types for template results
+- generalized case statement (requires better transf)
+- normalize for the DOM
+- script layer!
+- tlastmod returns wrong results on BSD (Linux, MacOS X: works)
+- nested tuple unpacking
+
+
+Library
+-------
+
+- float formatting
+- locale support
+- conversion between character sets
+- bignums
+- ftp, smtp (and other internet protocols)
+
+- finish json module: use coroutines for this!
+
+- pdcurses bindings
+- automate module: expect-like module for Nimrod
+
+- for system:
+  proc `@` [T](a: openArray[T]): seq[T] = 
+    newSeq(result, a.len)
+    for i in 0..a.len-1: result[i] = a[i]
+    
+  --> ensure @[] calls the array version!
+  
+
+For the next versions
+=====================
+
+- support for generation of dynamic libraries
+
+
+Further ideas/nice to have
+==========================
+
+- queues additional to streams: have two positions (read/write) instead of one
+
+
+Version 2
+---------
+
+- language change: inheritance should only work with reference types, so that
+  the ``type`` field is not needed for objects! --> zero overhead aggregation
+  BETTER: ``is`` and safe object conversions only work with ref objects. Same
+  for multi methods.
+  
+- explicit nil types?
+  * nil seq[int]
+  * nil string
+  * nil ref int
+  * nil ptr THallo
+  * nil proc 
+
+.. code-block:: nimrod
+  var
+    x: string = nil # initialized with invalid value!
+  if not isNil(x):
+    # now x 
+  
+- better for backwards compatibility: default nilable, but ``not nil``
+  notation::
+  
+  type
+    PWindow = ref TWindow not nil
+
+
+Low priority
+------------
+
+- ``when T is int`` for generic code
+- ``when validCode( proc ()  )`` for generic code
+  
+  
+  when compiles:
+    
+  elif compiles: 
+    
+
+- macros: ``typecheck`` pragma; this allows transformations based on types!
+- find a way for easy constructors and destructors; (destructors are much more
+  important than constructors)
+- code generated for type information is wasteful
+
+
+RST
+---
+- footnotes; prefix :i: whitespace before :i:, _reference, `reference`__
+  __ anonymous: www.nimrod.org
+

tools/build.tmpl

@@ -55,7 +55,7 @@ case $ucpu in
     mycpu="amd64" ;;
   *sparc*|*sun* ) 
     mycpu="sparc" ;;
-  *power* ) 
+  *power*|*Power* ) 
     mycpu="powerpc" ;;
   *mips* ) 
     mycpu="mips" ;;

web/news.txt

@@ -16,6 +16,8 @@ Bugfixes
 - Bugfix: ``dialogs.ChooseFileToSave`` uses ``STOCK_SAVE`` instead of 
   ``STOCK_OPEN`` for the GTK backend.
 - Bugfix: ``raise`` within an exception handler did not work.
+- Bugfix: ``low(somestring)`` crashed the compiler.
+- Bugfix: ``strutils.endsWith`` lacked range checking.
 
 
 Changes affecting backwards compatibility
@@ -37,6 +39,8 @@ Additions
 - Added ``times.epochTime`` and ``times.cpuTime``. 
 - Implemented explicit type arguments for generics.
 - Implemented implicit type arguments for generics.
+- Implemented ``{.size: sizeof(cint).}`` pragma for enum types. This is useful
+  for interfacing with C.
 
 
 2010-03-14 Version 0.8.8 released

web/question.txt

@@ -58,9 +58,29 @@ to write as low-level Nimrod code as C requires you to do. That said the only
 overhead Nimrod has over C is the GC which has been tuned for years.
 
 
+What about JVM/CLR backends?
+----------------------------
+
+A JVM backend is almost impossible. The JVM is not expressive enough. It has
+never been designed as a general purpose VM anyway. A CLR backend is possible
+but would require much work. 
+
+
 Compilation
 ===========
 
+Which option to use for the fastest executable?
+-----------------------------------------------
+
+For the standard configuration file, ``-d:release`` does the trick.
+
+
+Which option to use for the smallest executable?
+------------------------------------------------
+
+For the standard configuration file, ``-d:quick --opt:size`` does the trick.
+
+
 Execution of GCC fails (Windows)
 --------------------------------