Sergey Chelnokov

It's annoying, but okay. I continue to use text file generators and run DocFetcher on them.

And there, let the user at least run OCR on *.png if he wants (for example, what user can add to table FILTER-COMMAND)

And there, let the user at least run OCR on *.png if he wants

No, I don't want for version that support djvu. I want version with suport the external user-defined converters (that converts document to plain text, and then DocFetcher can use they stdout for indexing) I.e. user just setup a table: Filter Command Stdout encoding *.djvu;*.djv c:\tooling\convert_djvu_to_txt.bat {file} cp1251 *.mcdx c:\tooling\convert_mathcad_to_txt.bat {file} utf-8 May be it will be more clear to convert documents with intermediate file: Filter Command *.djvu;*.djv " C:\Program...

No, I don't want for version that support djvu. I want version with suport external user-defined converters (that converts document to plain text, and then DocFetcher can they stdout for indexing) I.e. user just setup a table: Filter Command Stdout encoding *.djvu;*.djv c:\tooling\convert_djvu_to_txt.bat {file} cp1251 *.mcdx c:\tooling\convert_mathcad_to_txt.bat {file} utf-8 May be it will be more clear to convert documents with intermediate file: Filter Command *.djvu;*.djv " C:\Program Files (x86)\DjvuLibre\djvutxt"...

Now I use batch files to generate .meta-files with text from djvu/mcdx(Mathcad). For djvu uses DjvuLibre\djvutxt.exe, for matchad - 7-zip. And I use special batch file associated with .meta, that find and open origin document - it's usable, but not easy and effective. Is it possible to customize user-defined document (djvu, mathcad, e.t.c.) to text converter?

I don't know how I use it since 2 years - alternative found already... But it's cool :)