Allow slicing of screed records

[ctb]

This pattern occurs several times in the khmer code base:

new_read.name = read.name
new_read.sequence = read.sequence[:trim_at]
if hasattr(read, 'accuracy'):
    accuracy = read.accuracy
    new_read.accuracy = read.accuracy[:trim_at]

and should really just be new_read = read[:trim_at]...

This could be done at the same time as #705, too.

[bocajnotnef]

I'm only seeing this in scripts/trim-low-abund. Are there others?

[bocajnotnef]

(Also, starting work on this)

[ctb]

On Fri, Feb 27, 2015 at 08:44:46AM -0800, bocajnotnef wrote:

I'm only seeing this in scripts/trim-low-abund. Are there others?

filter-abund and filter-abund-single, too. And maybe more down the road :)

[bocajnotnef]

Alright. Is there an obvious place I should look to fix this? Otherwise I'm just gonna try to do it and see what breaks.

[ctb]

On Fri, Feb 27, 2015 at 08:47:43AM -0800, bocajnotnef wrote:

Alright. Is there an obvious place I should look to fix this? Otherwise I'm just gonna try to do it and see what breaks.

Let's talk briefly at noon.

[bocajnotnef]

• see _screed_recorcd_dict (rename to RecordDict and make public?)
• rewrite getitem to handle slice objects

[themangoemoji]

@bocajnotnef are you still working on this? Thinking of picking it up.

[camillescott]

Relevant docs: https://docs.python.org/2/reference/datamodel.html#object.__getitem__; http://www.pythoncentral.io/how-to-slice-custom-objects-classes-in-python/

[bocajnotnef]

@wrightmhw Hadn't started--Go for it.

[bocajnotnef]

Closing--resolved by dib-lab/screed#29