Extapolate text from software?

[MaxV]

Good morning Rebolers,
I need an help to get a simple way to translate to english the text of software.
I'm italian, and obviously my software is first written with italian language (buttons, texts, and so on ...) then I rewrite it with in English.
Is there a way to extrapolate text form rebol file?
Usually all the texts are between " " (examle: "hello world!"), is there a way with parse or find or something else?
Any idea is welcome

[Graham]

Well, Rebol code has values such as string!, date!, function! etc.

So, all you need to do is load your program and examine those values. If you find string! values, they are a good candidate for translation ...

Have a look at Carl's color-code script on rebol.org which examines the value of each word in a script and assigns a color based on its datatype!

[notchent]

MaxV, here's a quickie that may be useful:

rebol []
code: {
    view layout [
        btn "some text"
        btn "some more text"
    ] 
}
strings: copy []
parse code [any [thru {"} copy a-string to {"} (append strings a-string)] to end]
foreach str strings [
    if true = request rejoin [{Change "} str {"?}] [
        replace/all code str request-text/title rejoin [{Change "} str {" to:}]
    ]
]
editor code

This will obviously work better if you use opening and closing curly braces instead of quotes to represent strings in your code.

[MaxV]

Very good, I added a check for corrected ", so I obtain only the real strings:

code: { 
    view layout [
        btn "some text"
        btn "some more text"
    ] 
}
strings: copy []
test: 1
parse code [
   any [
      thru {"} (test: test + 1)   copy a-string
      to {"}  (if ((remainder test  2 ) = 0) [append strings a-string]) 
      ] 
    to end]
parse code [any [thru "{"  copy a-string to "}" (append strings a-string) ] to end]
probe strings
alert "Done!"

Now may next step is: how to use google translator with these strings?
I can read the page of the translation simply with my browser, example: I want tor translate "Ciao mondo"

http://translate.google.com/?hl=it#auto|en|Ciao mondo

but I can't read anywhere "Hello world" in source code of HTML, any suggestion?
Thank you and best regards
Max

[Graham]

Must be being generated in Javascript so that's why you can't see it.

[MaxV]

I found that to substitute the strings, it's fundamental use mold.
Without mold, "Hello" could be:

• "Hello"
• {Hello}

with mold i obtain: {"Hello"} or {{Hello}}

so the correct script is:

theurl: to-file request-file
code: read/string  theurl
strings: copy []
test: 1
parse code [
   any [
      thru {"} (test: test + 1)   copy a-string
      to {"}  (if ((remainder test  2 ) = 0) [append strings mold a-string]) 
      ] 
    to end]
parse code [any [thru "{"  copy a-string to "}" (append strings mold a-string) ] to end]
append  theurl  ".txt"
sort strings
strings: unique/case strings
write theurl ""  ; clean file
foreach temp strings [
	write/append theurl {[ }	
	write/append theurl (mold temp)
	write/append theurl { ]}
	write/append theurl "^/"
	]
alert "Done!"

[swhite]

I once wrote a small program where I tried out the idea of putting ALL text (that is, button values, messages, EVERYTHING) into one file. I still have the code if you would like to examine it, but it is too big to paste into this message. The idea behind this was that if I wanted to translate the program to another language, I would have to translate just that one file. I never thought of using something like google to do the translating.

Because all the text strings were identified by words, it would be possible to write a program to get each text string, although what one would DO with each text string, I have no idea.

[Graham]

If you build your application with RebGUI, you get language localization built in.