Implement an option for shallow extraction

[chinyeungli]

Currect extractcode does not have an option for shallow extraction.

[pombredanne]

This makes a lot of sense. Especially since this is already supported internally.
As you can see in https://github.com/nexB/scancode-toolkit/blob/e45f8f401ca22a15fda6a78dff317e8acf94038e/src/extractcode/extract.py#L97 the function has a recurse argument which defaults to False but is set to True when called from the extractcode CLI

[steven-esser]

Ive implemented this feature on 262-shallow-extraction and its working as intended. Any feedback is appreciated.

[pombredanne]

Thanks! The code looks fine. Would you mind adding some tests? Especially for the case where you ran a shallow extraction once and run it again: are archives extracted from top level archives extracted further one level down?
You might want to refine the help text for the CLi option too. Extract top level archives only. may not be super self explanatory?

[pombredanne]

Also I wonder if recurse may be better than shallow ? though I am not too sure.

[steven-esser]

I was thinking about that too, but did not necessarily want to change the default behavior as it has been up to now.

[steven-esser]

Currently, archives are not extracted after running extractcode --shallow twice on the same directory. It has something to do with the way the -extract directory is created.

It appears that the archive.tgz-extract directory gets created again or overwritten when extractcode --shallow is run a second time, and the previously nested archives do not get extracted.

If I were to change the name of the extracted directory to something else, the command works as expected.

My guess is the problem lies in the order that pathnames/files are processed during extract events