Trim text in a file without storing the contents of anywhere

[TheDcoder]

Hello, I have a simple task today, I am sure that I made a similar post a long time ago... I want trim (delete or remove or wipe) the contents of a file (text or binary) WITHOUT storing the contents of the file anywhere in the script. Here is a text file for the purpose of experimenting: (contents of text.txt)

1234567890

The task is simple, remove "456" from the contents of text.txt WITHOUT storing the contents anywhere! I wanted to post some code but it seems impossible to provide any relevant code this time...

 

Good luck with the challenge! TD

Edited November 6, 2015 by TheDcoder
removed "(whole)" from the question

[UEZ]

What is the purpose of this "challenge"? Why should I use the slow file I/O only?

[TheDcoder]

@UEZ Autually it was my problem which I am facing right now...

Why should I use the slow file I/O only?

Because I process *big* binary files (archives to be precise), I need to trim some stuff in it

[UEZ]

Define "big".

[JohnOne]

Based on the exact criteria for your request.

$f = FileOpen("file.txt", 2)
FileWrite($f, "1237890")
FileClose($f)

 

[JohnOne]

But I always thought this was not possible without direct access to the physical disk and knowing where the file is, on it.

[water]

J1, helpful as always

[water]

Use FileRead in a loop and read the data in chunks by setting parameter "count" to e.g. 64K. Process the variable and write the result to a new file. When finished delete the original file and rename the new one.

[TheDcoder]

Define "big".

10 - 100 MB

[Danyfirex]

You can do this...

Local $sEndData = ""
Local $hFile = FileOpen("1.txt", 17)
FileSetPos($hFile, 3, 0)
FileSetPos($hFile, 3 + 3, 0)
$sEndData = FileRead($hFile)
FileSetPos($hFile, 3, 0)
FileSetEnd($hFile)
FileWrite($hFile, $sEndData)
FileClose($hFile)

Saludos

[water]

100MB isn't too big. You should be able to read and process the whole file in one cycle.

What have you tried so far?

[water]

Because I process *big* binary files (archives to be precise), I need to trim some stuff in it

What kind of binary archives are you talking about?
Wouldn't changing the content/length destroy the internal structure of the archive?

[TheDcoder]

What have you tried so far?

Tried some crazy solutions like passing empty string, Null, Binary 0x000000 to trim the contents, currently I just finished my dinner...

Wouldn't changing the content/length destroy the internal structure of the archive?

I am writing additional information in the header, I tried it with a hex editor, It worked!

Use FileRead in a loop and read the data in chunks...

I will consider it using if its the most reliable

 

@Danyfirex Please wait, I am experimenting

[JohnOne]

danyfirex code in form of function

_FileEdit("file.txt", "456")


Func _FileEdit($sFilePath, $sText)
    $iPos = StringInStr(FileRead($sFilePath), $sText) -1
    $iLen = StringLen($sText)
    Local $sEndData = ""
    Local $hFile = FileOpen($sFilePath, 17)
    FileSetPos($hFile, $iPos, 0)
    FileSetPos($hFile, $iPos + $iLen, 0)
    $sEndData = FileRead($hFile)
    FileSetPos($hFile, $iPos, 0)
    FileSetEnd($hFile)
    FileWrite($hFile, $sEndData)
    FileClose($hFile)
EndFunc   ;==>_FileEdit

 

[TheDcoder]

@Danyfirex Did you forgot that you are not allowed to store the contents of the file in a variable?

[BrewManNH]

Unless you're planning on hex editing the disk drive directly, your requirement is never going to be possible.

[TheDcoder]

@BrewManNH Hmmm.... .... I find water's solution to read the data in chucks now the best. I might reconsider some things tomorrow

[Bowmore]

When working with big files reading the file in chunks is the way to go. I wrote an AutoIt script at work to manipulate very large (15 to 20 GigaByte) XML files and it works like a charm. It's not difficult and relatively fast. The one case you need to consider that some forget, is where the section of the file that you are interested in falls across the boundary of 2 chunks. The way around this is to save some data from the end of the previous chunck, that is 1 byte shorter than what you are looking for and append the next chunk to this. This will ensure that you have the complete section of interest in one chunk of data. It's also worth experimenting with different chunk  sizees to find the optimal size for your particular set up.

[guinness]

Hello, I have a simple task today, I am sure that I made a similar post a long time ago... I want trim (delete or remove or wipe) the contents of a file (text or binary) WITHOUT storing the contents of the file anywhere in the script.

When you read data into a variable, that is stored in memory, not the script.

Good luck with the challenge! TD 

If I'm not mistaken this is not a challenge as you're asking for help.

[TheDcoder]

Quote (TheDcoder said): ...in the script.

When you read data into a variable, that is stored in memory, not the script.

I was referring to that, what is it called?.... uhhhh... .... Ah! Synthetic Sugar!, "in the script" is synthetic sugar for "in the memory"

If I'm not mistaken this is not a challenge as you're asking for help.

Busted , Yep, I wanted some help