Question

I am after a general purpose unspecialised plain text file extractor.

Firstly before people shout look at Apache Tika – my response is that it only supports some popular binary file formats like Office, BMPs etc.

Back to the problem – Many binary files have text strings embedded in them, which i would like to extract without the binary byte noise. THis would mean it could find simple text string sequences in exes and so on with the result only holding ascii words. I tried googling but could not find anything that did this. My basic idea is if a file is not handled by TIKA this simple binary file handler would try its best to find these text strings.

Answer 1

I ended up writing my code class to solve my problem.

Important features/considerations.

• Only accepts cr, nl, tab, space – char127 –
  • Ignores all characters that are no ascii
  • bad luck if file includes unicode.
• ignores character sequences that are leass than a few chars (configurable).
  • this means that bytes that are a single letter surrounded by other non ASCII values are ignored.
• a space is inserted between character sequences
  • this means that one string, some bytes and then other string appears in the result as two words separated by a string rather than a single long word.