Feature request : slicing with lists

[Goutte]

Great lib ! I really like the elegance of it. That's good cheese. 🍷

I hacked support for the following tests :

    def test_slice_with_list_read(self):
        sl = sparse_list.SparseList([1, 2, 3, 4, 5])
        ip = [0, 2, 4]
        self.assertEquals([1, 3, 5], sl[ip])

    def test_slice_with_list_write(self):
        sl = sparse_list.SparseList([1, 2, 3, 4, 5])
        ip = [0, 2, 4]
        sl[ip] = [6, 7, 8]
        self.assertEquals([6, 2, 7, 4, 8], list(sl))

I wrote tests for numpy too (just to be sure) but decided against including them in a PR, unless you want them :

    def test_slice_with_numpy_array_read(self):
        import numpy
        sl = sparse_list.SparseList([1, 2, 3, 4, 5])
        ip = numpy.array([0, 2, 4])
        self.assertEquals([1, 3, 5], sl[ip])

    def test_slice_with_numpy_array_write(self):
        import numpy
        sl = sparse_list.SparseList([1, 2, 3, 4, 5])
        ip = numpy.array([0, 2, 4])
        sl[ip] = [6, 7, 8]
        self.assertEquals([6, 2, 7, 4, 8], list(sl))

---

I have a bunch of questions and remarks, while I'm at it (you don't have to answer, but given your SO score I guess explaining does not bother you too much :3) :

1. Why triple single quotes ''' and not double quotes """ for doctrings ?
2. There is a SyntaxError in setup.py with python 2.7, because of the positional argument install_requires('future', 'six'),
3. There's a fromage called benchmark whose simplicity and absence of code in multiline strings you might enjoy, if you don't already know about it.
4. I rather enjoyed your sense of humour, thank you for that too !
5. I know that comments are a failure but I took the liberty of adding two of them, you'll see. (and possibly remove them, I won't mind)

[johnsyweb]

I don't recall now, but I think I copied the language from the official documentation for list...

> pydoc list | fgrep -A1 count
 |  count(...)
 |      L.count(value) -> integer -- return number of occurrences of value
 |  

> pydoc sparse_list | fgrep -A1 count
     |  count(self, value)
     |      return number of occurrences of value

[Goutte]

Urgh. I hesitated for a while before changing the text, but then my inner grammarian assumed control, i guess. I can revert it back, I don't know how to choose between consistency with the existing doc and literacy. You choose. :)

[johnsyweb]

Hi @Goutte!

Thank you very much for this PR! It has been nice to revisit this project after having neglected it for so long. I don't think I've written any Python code since I wrote this module about three years ago, so my memory around this is a little faded. Apologies for my vagueness.

To address your questions and remarks:

Why triple single quotes ''' and not double quotes """ for doctrings ?

I think the convention at my previous employer was to use single quotes for strings wherever possible. I've been doing Ruby for so long now, I can't recall whether there was any difference in Python (like there is in Ruby). Were I to start again, I would probably adhere to http://legacy.python.org/dev/peps/pep-0257/ 😄

There is a SyntaxError in setup.py with python 2.7, because of the positional argument install_requires('future', 'six'),

Interesting. I wonder how I didn't pick that up. Perhaps I was using an older version of Python 2.7 than we have now.

There's a fromage called benchmark whose simplicity and absence of code in multiline strings you might enjoy, if you don't already know about it.

Thanks for this. I would certainly use benchmark instead of my dodgy time_sparse_list script.

I rather enjoyed your sense of humour, thank you for that too !

😊

I know that comments are a failure but I took the liberty of adding two of them, you'll see. (and possibly remove them, I won't mind)

He he. Thanks for reading that article!

[johnsyweb]

Looks like https://github.com/Goutte/python_sparse_vector is the bomb when it comes to numpy! 💥