read past EOF characters?

[iamtheky]

Problem being:  I used bash on windows to 'cat ntuser.pol' and pipe the output to a txt file on the desktop.  This file is chock full of EOF characters, so none of the file* nor _file* commands get past the first few characters.  Notepad.exe does not care, and as a bonus feature converts all those EOFs to squares when you save, but opening it, sending ctrl-s, and closing it, is clunky and am hoping there is a better alternative.

[water]

What does 'FileRead("filename", count)' return where count is the real size of the file?

[iamtheky]

a couple of characters (it also returns them as asian, when they are clearly not in the file), and they are not the entirety of the line.  Same characters are printed with -1 as are with 1-50ish.

read to array also returns the same.  1 line, only the few bytes up to the EOF.

Edited June 1, 2016 by iamtheky

[jchd]

Read the file verbatim (i.e. in binary mode).

[TheDcoder]

And convert the binary input to a string using BinaryToString

[iamtheky]

Nope, same.  The fileread function will not make it past the first EOF in any mode.  I will try and generate a file that doesnt have pii so you can play along, but I cant edit one because any save operation in notepad fixes it.

[MuffinMan]

Notepad++ ( https://notepad-plus-plus.org/ ) would probably allow you to remove/replace the PII while preserving the EOF characters that are causing your issues.

[jchd]

  On 6/2/2016 at 11:47 AM, iamtheky said:

 The fileread function will not make it past the first EOF in any mode.

Expand  

I sincerely doubt that. Binary mode will hopefully read anything. First use FileOpen in binary mode and use the handle returned to FileRead it. The binary result is the actual content of the file, verbatim.

[iamtheky]

I cannot post the screenshots, because apparently you can not delete attachments from the attachment menu (and i have reached my limit).

However, opening as bin is fine.  Fileread as Bin is also fine, that was incorrect.  BinaryToString or any attempts at converting that back to text (including stringreplace so far), result in the same 3 characters and the EOF.  

Edited June 2, 2016 by iamtheky

[TheDcoder]

  On 6/2/2016 at 11:47 AM, iamtheky said:

but I cant edit one because any save operation in notepad fixes it.

Expand  

Use a Hex Editor (like HxD), Hex Editors don't mess up and only save the changes you made, no extra changes like Notepad or Notepad++ .

[iamtheky]

I need the changes, so normal string ops work

once my attachments are freed up i will post all the goods.

[TheDcoder]

  On 6/2/2016 at 1:38 PM, iamtheky said:

I cannot post the screenshots, because apparently you can not delete attachments from the attachment menu (and i have reached my limit).

Expand  

Use an external image host like PIXHOSTER

[TheDcoder]

And you can use WeTransfer to transport your goods, the files are stored for 7 days .

[iamtheky]

https://drive.google.com/open?id=0B68V1P1HtXeOT3l0aXBDTnd4eDA

https://drive.google.com/open?id=0B68V1P1HtXeObnY1QUNnRE9OcW8

[TheDcoder]

I want NTUserPol.txt to study the file myself.

[iamtheky]

 

You can always gen your own while I am waiting, though i think i may be the only forum member who has added the linux subsystem to Win10.,,

[TheDcoder]

I am running Windows 8.1, not sure if I have bash.exe

[water]

The Windows Substystem for Linux only comes for Windows 10.
https://en.wikipedia.org/wiki/Windows_Subsystem_for_Linux

[iamtheky]

nope, you have to download the Win10 insider build, and then add the linux subsystem feature.  I dont know that you can even get to Ntuser.Pol from a session, nor if there is a utility like CAT that would pipe the contents to a txt.

But I am sure there are similar files to test with, maybe someone smart will offer suggestions for a reproducer

Edited June 2, 2016 by iamtheky

[jchd]

I wonder what you're trying to do.

ntuser.pol is a binary file with a ASCII header("PREG") then a sequence of a UNIT dword folowed by a series of UTF16-LE strings. You can check by yourself by using a hex editor. So this is a mixed content file which you can't easily and reliably modify with a text editor.