sketchucation logo sketchucation
    • Login
    πŸ€‘ SketchPlus 1.3 | 44 Tools for $15 until June 20th Buy Now

    [Code] PCFileTools

    Scheduled Pinned Locked Moved Developers' Forum
    27 Posts 3 Posters 3.9k Views 3 Watching
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • TIGT Offline
      TIG Moderator
      last edited by

      This is my attempt to get over the issue of UTF-8 characters on PCs [only] when filepaths etc return incorrect results in some of Ruby's native File and Dir methods.
      It uses win32ole.so, which you need to have an appropriate version of in Plugins
      [ http://forums.sketchucation.com/viewtopic.php?p=380121#p380121 ]
      It returns correct results as far as I've tested it...
      Can some of you try testing it too...
      e.g. replace File.exist?() with PCFile.exist?()
      and see if filepaths with accented characters etc are then returned ' true' [properly]...
      These methods mimic the native tools fairly well and there are also some 'better' copy/move/rename methods; but there's no ' binmode' and the read/write syntax etc is slightly different [but still logical]...

      Put into Plugins...

      (c) TIG 2012
      Permission to use, copy, modify, and distribute this software for
      any purpose and without fee is hereby granted, provided that the above
      copyright notice appear in all copies.
      THIS SOFTWARE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR
      IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
      WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

      PCFileTools.rb

      Adds new PC ONLY tools similar to the Ruby 'File' and 'Dir' tools,
      BUT which avoid issues with UTF-8 characters in filepaths etc...
      [NOTE: MACs have no equivalent issues so they don't need this!]

      PCFile... & PCDir...

      They generally work like the Ruby equivalent
      So
      File.exist?("C:/Temp/bouΓ©.txt")
      which will WRONGLY return 'false,' is replaced by
      PCFile.exist?("C:/Temp/bouΓ©.txt")
      which correctly returns 'true'...

      PCFile.basename( fileName [, aSuffix] )
      returns the basename of the file or folder, e.g. "C:/Temp/file.txt"
      -> "file.txt" ; if the 'aSuffix' is used then any matching extension
      is stripped, using ".*" removes ANY file extension - e.g. -> "file".

      PCFile.delete( fileName )
      deletes the given file; if it's not an 'absolute path' the file is
      sought in the pwd.

      PCFile.dirname( fileName)
      returns the directory [path] of the passed filepath, stripping the
      terminal file/folder; e.g. "C:/Temp/file.txt" -> "C:/Temp"

      PCFile.directory?( fileName )
      returns true or false, depending on whether or not the given
      'fileName' refers to a 'directory'.

      PCFile.exist?( fileName )
      returns true or false, depending on whether or not the given file
      [OR folder] exists.

      PCFile.extname( fileName )
      returns the extension-name [if any] of the passed file;
      e.g. "C:/Temp/file.txt" > ".txt"

      PCFile.ftype( fileName )
      returns the type of the object as 'directory' or 'file'.
      [Note that it is more limited that native File.ftype()]

      PCFile.join( string, string* )
      joins strings to return a file-path;
      e.g. PCFile.join("C:/Temp", "bouΓ©.txt") -> "C:/Temp/bouΓ©.txt"

      PCFile.rename( fileName, newName )
      renames the file 'fileName' as 'newName'; unlike File.rename(...) it
      can 'move' a file to a new folder, if the specified folder exists,
      if no folder path is given it assumes the same folder as 'fileName'.

      PCFile.move( fileName, nname )
      equivalent to PCFilename.rename(...)

      PCFile.copy( fileName, newName, [overwrite=false] )
      makes a copy of the file 'fileName' as 'newName'; it can 'copy' a file
      to a new folder, if the specified folder exists, if no folder path is
      given it assumes the same folder as 'fileName'.
      If a 'true' 3rd argument is given then any existing file is
      overwritten, if a 'false' or no 3rd argument is given then any
      existing file is NOT overwritten.

      PCFile.new( fileName, [modeString="r"] )
      opens a file as f=PCFile.new(...), making it or using it if it exists,
      the mode defaults to "r" = 'read', the options are:
      "r" =read
      "r+" =read+write
      "w" =write
      "w+" =write+read
      "a" =append
      "a+" =append+read

      PCFile.open( fileName, [modeString="r"] )
      equivalent to PCFile.new(...)

      NOTE: the slightly different syntax for the read/write/close file 'f'...
      [there is no 'binmode']

      PCFile.read(f)
      reads all of the contents of the new/open file 'f'.

      PCFile.readline(f)
      reads the first line of the new/open file 'f'.

      PCFile.readlines(f)
      reads all of the lines of the new/open file 'f' as an array of lines.

      PCFile.write(f, string)
      writes the 'string' to the new/open file 'f', with no added \n

      PCFile.print(f, string)
      writes the 'string' to the new/open file 'f', with no added \n

      PCFile.puts(f, string|array)
      writes the 'string' to the new/open file 'f', adding a \n;
      if it's an array it writes it a line at a time adding \n to each line.

      PCFile.close(f)
      closes the new/open file 'f' made as above, after you've finished
      reading/writing it etc.

      PCDir.pwd
      returns the path to the current directory [folder].

      PCDir.chdir( dirPath )
      changes the current directory [folder] to 'dirPath', if it exists.

      PCDir.mkdir( dirPath )
      makes the specified new directory [folder].

      PCDir.rmdir( dirPath )
      deletes the specified directory - note that unlike the Dir. method it
      will remove a directory [folder] even if it contains files!

      PCDir.exist?( dirPath )
      returns true or false, depending on whether or not the folder exists.

      PCDir.copy( dirPath, dirNew )
      makes a copy of a folder at the new location/name.

      PCDir.move( dirPath, dirNew )
      moves a folder to the new location/name.

      PCDir.entries( dirPath )
      returns an array of the names all of the files/folders within the
      specified folder [including '.' and '..']

      NOTE:
      Requires a compatible version of 'win32ole.so' in the Plugins folder.
      Available from here:
      http://forums.sketchucation.com/viewtopic.php?p=380121#p380121

      Donations:
      info @ revitrev.org via PayPal.com

      Version:
      1.0 20120202 First release.

      TIG

      1 Reply Last reply Reply Quote 0
      • Dan RathbunD Offline
        Dan Rathbun
        last edited by

        Took a quick look.

        It's a good start.

        First thoughts:

        1. The methods should act the same as the native Ruby methods, so that they can be a replacement (via aliasing within an Author's namespace.)
        2. They need to be wrapped within some library namespace. ( TIG::Lib, SKX::WIN, etc.)
        3. The container's need to be classes, not modules.
        4. They need to be subclasses of the Ruby baseclasses, so that they inherit all the methods that you will not be overriding.

        So, as an example, let's say you put them in SKX::WIN

        They would need to be declared:

        class SKX;;WIN;;Dir < ;;Dir
          ### the overriden methods
        end#class
        
        class SKX;;WIN;;File < ;;File
          ### the overriden methods
        end#class
        

        Within an author's namespace, he could use a local constant File, that points at your subclasses:

        module Author
          class FancyTool
        
            File = SKX;;WIN;;File
            # now he can use it as he's always done.
        
            # he can always refer to the base class,
            # by using the toplevel scope prefix, thus;
            ;;File.basename(pathstring)
        
            # or if he's done using the PC specific class,
            # he can make the local constant to point at
            # the base class again, by reassignment;
            File = ;;File
            #
            # or by removing the local constant, thus;
            remove_const(;File)
         
          end#class
        end#module
        

        I'm not here much anymore.

        1 Reply Last reply Reply Quote 0
        • Dan RathbunD Offline
          Dan Rathbun
          last edited by

          Here's a better conditional load wrapper:

          $LOAD_MSGS = '' unless defined($LOAD_MSGS)
          unless RUBY_PLATFORM =~ /(mswin|mingw)/
            msg = "Error; #{File.basename(__FILE__)}; loads PC only subclasses!\n"
            msg<< "This platform's #{RUBY_PLATFORM} is not recognized."
            puts(msg)
            $LOAD_MSGS << msg
          else
            # running on Windows
            begin
              require('win32ole.so')
            rescue LoadError => e
              msg = "Error; #<LoadError; in #{File.basename(__FILE__)}.>\n"
              msg<< "The shared library file; 'win32ole.so' could not "
              msg<< "be found using the $LOAD_PATH search array.\n"
              msg<< "Please install the 'win32ole.so' where it can be found.\n"
              puts(msg)
              $LOAD_MSGS << msg
            else
              # The 'win32ole.so' file WAS loaded...
              
              #
              ### define subclasses HERE
              #
              
              unless file_loaded?(File.basename(__FILE__))
                file_loaded(File.basename(__FILE__))
              end
          
            end# win32ole required
            #
          end#unless on PC
          
          

          The ### define subclasses HERE line can actually load the definition file (to save indents):
          require('skx/win/file_dir_defs.rb')
          and the file can have a exit clause at the top (just in case):

          exit(2) unless defined?(WIN32OLE)
          
          module SKX; end
          module SKX;;Win; end
          
          class SKX;;WIN;;Dir < ;;Dir
            ### method defs ...
          end#class
          
          class SKX;;WIN;;File < ;;File
            ### method defs ...
          end#class
          
          

          I'm not here much anymore.

          1 Reply Last reply Reply Quote 0
          • TIGT Offline
            TIG Moderator
            last edited by

            Dan

            I'm sure that the 'use' of the core of the code can be made better - although I'd be wary of overwriting base class ?
            A lot of this is 'above my pay grade'...

            I've kept everything separated for testing - so it still works provided you substitute PCFile.exist?(filepath) for File.exist?(filepath) and so on...

            What I'm asking initially is for developers to test these new/replacement PCFile methods and see if they are buggy and/or work as hoped.

            The biggest bind with 1.8~ Ruby File on PCs is that many operations like File.exist?() can return false when you know the file really does exist, these 'failures' are caused by UTF-8 characters in the filepath string [typically these are 'accented letters' in FR/DE/ES/PT/NO etc, but results from Chinese users etc would be great too...].

            I've tried to write simple code using pack/ unpack and equivalent Win32OLE methods that return ' true' correctly when the filepath exists and it contains such characters...

            There are many other failures with File that I hope I have now trapped with these equivalent PCFile methods. I've had to make slightly different read/write code to get around some Win32OLE vagaries etc... and I can't see currently how to do the equivalent of ' binmode' etc. Any suggestions/additions gratefully received...

            At this stage I really just want feedback on the efficacy of these methods...
            We can then debate later how the finalized 'fixed' methods are shoehorned into Ruby... πŸ˜‰

            TIG

            1 Reply Last reply Reply Quote 0
            • Dan RathbunD Offline
              Dan Rathbun
              last edited by

              It's UTF-8 (not UFT-8, BTW.)

              I'm not here much anymore.

              1 Reply Last reply Reply Quote 0
              • TIGT Offline
                TIG Moderator
                last edited by

                WTF... Doh! my fingers often type in the wrong order πŸ˜’

                TIG

                1 Reply Last reply Reply Quote 0
                • Dan RathbunD Offline
                  Dan Rathbun
                  last edited by

                  The main point, I'm trying to get across.. is future implementation.
                  Ie, how scripters will wish to use unicode extensions to File, Dir and String.

                  The goal is to write cross-platform plugins.

                  Ex:

                  MAC =( RUBY_PLATFORM =~ /(darwin)/ ? true ; false ) unless defined?(MAC)
                  WIN =( not MAC ) unless defined?(WIN)
                  
                  module Author
                    class FancyTool
                   
                      if WIN
                        require('skx/win/File')
                        File = SKX;;WIN;;File
                      end
                  
                      # Now he can use File as he's always done.
                      #
                      # On Mac, there is no local constant "File", and the
                      # call is evaluated to the object, that is referenced,
                      # by the toplevel constant "File".
                      #
                      # But on PC, within this namespace ONLY, the call
                      # is evaluated, to the object pointed to, by the
                      # local (constant) reference "File", or fully
                      # qualified; "Author;;FancyTool;;File",
                      # which does not touch ;;File (aka; Object;;File)
                   
                    end#class
                  end#module
                  

                  There is no way, scripters will want to changed all calls to methods of class File or Dir, into platform conditional statements, like:

                  filename =( WIN ? PCFile.basename(pathstr) ; File.basename(pathstr) )
                  

                  or (worse):

                  filename = if WIN
                    PCFile.basename(pathstr)
                  else
                    File.basename(pathstr)
                  end
                  

                  πŸ€“

                  I'm not here much anymore.

                  1 Reply Last reply Reply Quote 0
                  • Dan RathbunD Offline
                    Dan Rathbun
                    last edited by

                    @tig said:

                    A lot of this is 'above my pay grade'...

                    No doubt that goes for any single ONE rubyist. Something like this needs to be a group project.
                    Others have done some work in this area already. ThomThom on wide strings (which also need to be addressed, because many of the base class String methods, will garble unicode strings.)
                    Dan Berger's "win32-api" toolkit plays a bit with creating a WideString class. (I think he considers it beta.)
                    Also, we should not ignore the extended class Pathname (which is actually a wrapper class, not a String subclass.)

                    @tig said:

                    We can then debate later how the finalized 'fixed' methods are shoehorned into Ruby... πŸ˜‰

                    OK, I made my point, on this issue.

                    The only thing to add is to state the obvious. There are really only 2 alternatives.

                    My suggestion is the least invasive, with the least responsibility. (Authors are left to decide to use the extension on a namespace by namespace basis.)

                    The other option, to actually redefine any Ruby base classes that will use unicode strings, on the Windows platform(s).
                    This would mean taking responsibility for string functionality of ALL plugins (running on Windows.)
                    Not what (I think,) you would wish to do. Nor any project group. (Not at least without compensation, of some kind,.... and a lot of time to devote to the "cause.")
                    Even the Ruby Core guys are taking forever to implement unicode support.

                    Enough said... πŸ˜’

                    I'm not here much anymore.

                    1 Reply Last reply Reply Quote 0
                    • Dan RathbunD Offline
                      Dan Rathbun
                      last edited by

                      @tig said:

                      ... although I'd be wary of overwriting base class ?

                      I didn't say anyone should. I said they should be subclasses of the base class. If they are, Ruby would not let anyone make the subclass overwrite the baseclass, because it would be a circular reference.

                      EDIT: I tested this at the console. Ruby does not check for circular references using the c-side **=** operator. Anyway, it does not actually create a circular reference. The reference to original baseclass becomes un-identified by any constant, but the object can still be got, via the superclass() method.

                      Making a LOCAL constant point at an object (in this case a Class definition,) does not overwrite anything.

                      Now obviously, if some stupid 'newb' types File = SKX::WIN::File in the Ruby Console, or in an unwrapped script, THAT global constant that references the base class definition object, is changed (and affects all scripts that are also using it.) So, in order to set it back [without restarting,] it might be a good idea to have a 'secret' reference to the base class definition, kept "out of sight."

                      module Ruby
                        
                        Refs = {}
                        # base classes
                        Object.constants.sort.each {|ref|
                          obj = Object.class_eval "#{ref}"
                          Refs[ref]= obj if obj.class==(Class)
                        }
                        # base modules
                        [Comparable,Enumerable,Errno,FileTest,GC,
                         Kernel,Marshal,Math,ObjectSpace,Process].each {|obj|
                          Refs[obj.name]= obj
                        }
                        Refs['TOPLEVEL_BINDING']= TOPLEVEL_BINDING
                        Refs.freeze
                      
                        def self.reset_object_ref(ref)
                          if ref.is_a?(String) || ref.is_a?(Symbol)
                            ref = ref.to_s 
                          elsif ref.is_a?(Module) # includes Class
                            ref = ref.name
                          else
                            return nil
                          end
                          if Ref.has_key?(ref)
                            begin
                              eval( "#{ref} = ObjectSpace._id2ref(#{Ref[ref].object_id})", TOPLEVEL_BINDING )
                            rescue
                              return false
                            else
                              return true
                            end
                          end
                          return false
                        end
                      
                      end#module
                      ###
                      

                      EDIT: I added a reset method to the example, just for kicks.

                      @tig said:

                      I've kept everything separated for testing - so it still works provided you substitute PCFile.exist?(filepath) for File.exist?(filepath) and so on...

                      You won't find many who are willing to go through their code and search and replace, "File." with "PCFile.", besides you haven't provided aliases for the methods you did not implement, which makes a simple editor search and replace, into a tedious manual edit session.

                      You might consider, overriding the method_missing callback, whilst you still have them as custom modules:

                      def method_missing( sym, *args )
                        if File.respond_to?(sym)
                          File.method(sym).call(*args)
                        else
                          raise(NoMethodError,"undefined method `#{sym.to_s}' for #{self.name};#{self.class.name}",caller)
                        end
                      end#def
                      

                      and similar for PCDir:

                      def method_missing( sym, *args )
                        if Dir.respond_to?(sym)
                          Dir.method(sym).call(*args)
                        else
                          raise(NoMethodError,"undefined method `#{sym.to_s}' for #{self.name};#{self.class.name}",caller)
                        end
                      end#def
                      

                      I'm not here much anymore.

                      1 Reply Last reply Reply Quote 0
                      • Dan RathbunD Offline
                        Dan Rathbun
                        last edited by

                        @tig said:

                        ... and I can't see currently how to do the equivalent of ' binmode' etc. Any suggestions/additions gratefully received..

                        Ya know.. it's weird that the Core guys made this switch, without a way to test later IF the stream was IN binmode or not.

                        I checked the docs, and the Ruby 1.8.7 branch is still the same.

                        BUT... in the 1.9.x trunk, they have added a binmode?() boolean query method. (See the online docs for IO class. They also added methods binread and binwrite, special open methods.)
                        Of course, they have added a bunch of options, to read and write in several encodings.

                        I'm not here much anymore.

                        1 Reply Last reply Reply Quote 0
                        • Dan RathbunD Offline
                          Dan Rathbun
                          last edited by

                          What effect do you imagine WIN32OLE.codepage= WIN32OLE::CP_UTF8 would have over CP_ACP (ANSI?/ASCII?), which seems to be the default.

                          I'm not here much anymore.

                          1 Reply Last reply Reply Quote 0
                          • Dan RathbunD Offline
                            Dan Rathbun
                            last edited by

                            Ok TIG.. a general question:

                            most all of the methods are doing this:

                            arg = arg.unpack("U*").map{|c|c.chr}.join

                            which seems to convert a UTF-8 string (if it is one,) to an ANSI string, before passing it to Windows FSO methods that take and return Unicode strings ...

                            1) correct ??

                            2) and why ??

                            I'm not here much anymore.

                            1 Reply Last reply Reply Quote 0
                            • TIGT Offline
                              TIG Moderator
                              last edited by

                              I found that if I didn't do that change to the string then any tests of UTF-8 strings, like PCFile.exist?(path) do not work properly and return ' false' when it should be ' true'... just like the File.exist?(path) version; BUT making that change to the string before testing it seems to return correct results - consistently ' true' when it should be ' true' and ' false' when it should be ' false'. For a simple ANSI character string it works fine either way [the unpack/join has no affect], but if you test with a UTF-8 string with accented characters [perhaps obtained from a UI.openpanel()], that is unpack/joined etc then you can see the difference between what the File.. and PCFile.. versions return...
                              There's probably a more elegant way to do this... BUT it seems to work the way I've bodged it together, so now perhaps we can think of better ways of achieving the same difference... πŸ˜•

                              TIG

                              1 Reply Last reply Reply Quote 0
                              • Dan RathbunD Offline
                                Dan Rathbun
                                last edited by

                                That seems to indicate that the FSO methods are doing ANSI comparisons (perhaps by default.)

                                I'm not here much anymore.

                                1 Reply Last reply Reply Quote 0
                                • thomthomT Offline
                                  thomthom
                                  last edited by

                                  So, does PCFile.exist?() return true for a file with, for example, Japanese characters?

                                  Thomas Thomassen β€” SketchUp Monkey & Coding addict
                                  List of my plugins and link to the CookieWare fund

                                  1 Reply Last reply Reply Quote 0
                                  • Dan RathbunD Offline
                                    Dan Rathbun
                                    last edited by

                                    Would the Japanese Kanji chars be in the UTF-16 set?

                                    Ya know we are all back to the ol' String encoding problem, really.

                                    I thot about using Dan Berger's String subclass(es) WideString or whatever he called them, but it seem like it would be combersome. Unless they converted themselves automatically similar to how Numerics use coerce().
                                    Currently the interpreter always makes ANSI strings from **" "** and **' '** literals. (and their **%**dilimeter equivs.)

                                    I wonder if possible to create a %u function that creates UTF8 strings. And maybe a %U that creates UTF16 ?
                                    (Are these defined in Kernel, or are they C-side interpreter functions?

                                    (Just throwing issues in the air, "musing out load.")

                                    I'm not here much anymore.

                                    1 Reply Last reply Reply Quote 0
                                    • thomthomT Offline
                                      thomthom
                                      last edited by

                                      The underlying problem in Ruby 1.8 under windows is that it calls the A version of the file functions instead of the W versions. If a function is called FileFunction is used in C/C++ - when compiled it will translate to FileFunctionA or FileFunctionW depending on whether UNICODE is defined.
                                      I was thinking that a C Extension that would forcefully call the FileFunctionW variants would be sure to work as it would be the system doing all the work. Meddling with the string in Ruby is quite likely to cause data to be lost or corrupted.

                                      Thomas Thomassen β€” SketchUp Monkey & Coding addict
                                      List of my plugins and link to the CookieWare fund

                                      1 Reply Last reply Reply Quote 0
                                      • TIGT Offline
                                        TIG Moderator
                                        last edited by

                                        I only works for UTF-8 [i.e. 'European' accented-characters etc] - the more complex Chinese/Japanese return false when it should be true πŸ˜’
                                        However, if we have a way of resolving one hopefully the other will follow...

                                        TIG

                                        1 Reply Last reply Reply Quote 0
                                        • TIGT Offline
                                          TIG Moderator
                                          last edited by

                                          http://www.danielstutzman.com/2011/04/how-to-write-unicode-filenames-in-ruby-1-8-6 uses Win32API and iconv ...

                                          TIG

                                          1 Reply Last reply Reply Quote 0
                                          • Dan RathbunD Offline
                                            Dan Rathbun
                                            last edited by

                                            BTW.. if interested:

                                            This is the Extended lib module FileUtils from Ruby v1.8.6-p287
                                            module FileUtils (Ruby v1.8.6-p287)

                                            I'm not here much anymore.

                                            1 Reply Last reply Reply Quote 0
                                            • 1
                                            • 2
                                            • 1 / 2
                                            • First post
                                              Last post
                                            Buy SketchPlus
                                            Buy SUbD
                                            Buy WrapR
                                            Buy eBook
                                            Buy Modelur
                                            Buy Vertex Tools
                                            Buy SketchCuisine
                                            Buy FormFonts

                                            Advertisement