[with patch, positive review] subsets and subwords bug fix + improvements.

[hivert]

This patches improves subwords and subsets and deals with several issues:

1. It implements subsets for finite multisets (sets with repetitions).

   sage: Subsets([2,2,3], multiset=True).list() 
   [[], [2], [3], [2, 2], [2, 3], [2, 2, 3]] 
   

2. It implement __contains__ which was missing for subsets and subwords: Before:

   sage: st = Subsets([1,2,2,3]); Set([1,2]) in st 
   --------------------------------------------------------------------------- 
   NotImplementedError                       Traceback (most recent call last) 
   

   After:

   sage: st = Subsets([1,2,2,3]); Set([1,2]) in st 
   True 
   

3. It fixes a bug in smallest_positions: Before:

   sage: sage.combinat.subword.smallest_positions([2,4,3,3,1,2],[1,3,3]) 
   [4, 4, 4] 
   

   After:

   sage.combinat.subword.smallest_positions([2,4,3,3,1,2],[1,3,3]) 
   False 
   

   which means that 113 is not a subword of 243312.
4. It finally improves the doc and the tests.

Cheers,

Florent

[hivert]

It has been decided (direct talk with Nicolas + irssi with Mike) that the user has to explicitely set that he want multisets. Therefore, on the contrary that is anounced, the following is not working:

sage: Subsets([2,2,3]).list()
[[], [2], [3], [2, 2], [2, 3], [2, 2, 3]]

Instead one should write

sage: Subsets([2,2,3], multiset=True).list()
[[], [2], [3], [2, 2], [2, 3], [2, 2, 3]]

[hivert]

Sorry for the mess with the description... I just realize because of my mistake that it is possible to change the description. :-)

I've uploaded a new patch according to remark of Nicolas and Mike. It should be ready for review and hopefully integration.

Cheers,

Florent

[hivert]

New patch after Nicolas review.

[hivert]

Nicolas sent me a review on sage-combinat-devel. The new subsets_subwords-5200-submitted.2.patch patch fixes the various problems pointed out by the review (typos + code improvements).

Florent

[nthiery]

Florent: feel free to switch the title to with review after the following micro updates:

• iterate -> iterates
• of all -> ""

[hivert]

Last review patch.

[hivert]

The review patch I just submitted address Nicolas last comment. Sorry for the mess with several version of the patch. The correct patch are:

• attachment:subsets_subwords-5200-submitted.2.patch​
• attachment:subsets_subwords-5200-review.patch​

According to Nicolas la remarks I allow myself to change the review title. Please tell me if it's not allowed by the review rules.

Florent

[mabshoff]

"with review" is meaningless :)

Nicolas: Can you give this a formal positive review?

Cheers,

Michael

[mabshoff]

After re-reading Nicola's comment I am giving this patch a positive review in his name. The two patches also pass doctests.

Cheers,

Michael

[mabshoff]

Merged both patches in Sage 3.4.1.alpha0.

Cheers,

Michael