[Biopython-dev] Phylo: support for JSON trees

[Fields, Christopher J]

Agreed.  This is veering towards the deadly sin of reinvention.  Instead of that, maybe a good look should be taken at the current XML tree formats to see if they can reasonably be represented in or translated to JSON.  And only if a JSON-based format doesn't already exist.

Chris

> On Dec 9, 2015, at 4:12 PM, Peter Cock <p.j.a.cock at googlemail.com> wrote:
> 
> Hi Fabio,
> 
> You're right JSON has a lot of advantages (people said this about XML
> too), but inventing your own file format is not something to be taken
> lightly.
> 
> Are there really no pre-existing tree formats using JSON?
> 
> Peter
> 
>> On Wed, Dec 9, 2015 at 9:12 PM, Fabio Zanini <fabio.zanini at fastmail.fm> wrote:
>> Hi all,
>> 
>> I've been working with phylogenetic trees for a few years and found it
>> very convenient to convert Bio.Phylo trees to/from JSON, e.g.
>> 
>> {'name': 'root', 'children': [
>>  {'name': 'leaf1', 'branch_length': 0.01, children: []},
>>  {'name': 'leaf2', 'branch_length': 0.02, children: []}]
>> }
>> 
>> I would like to implement such import/export functions into Bio.Phylo if
>> there is interest. Because Python has a JSON parser already, it's like
>> 50 lines of code altogether.
>> 
>> The main arguments in favour of this are:
>> - *JSON is standard for the web*, and the web is where all is moving
>> (Jupyter, bokeh, but also specific tree-oriented web apps such as
>> http://nextflu.org)
>> - embedding metadata is trivial (just add more keys to those dicts!)
>> 
>> How's that sound, shall I make a github PR?
>> 
>> Cheers,
>> Fabio
>> 
>> 
>> _______________________________________________
>> Biopython-dev mailing list
>> Biopython-dev at mailman.open-bio.org
>> http://mailman.open-bio.org/mailman/listinfo/biopython-dev
> _______________________________________________
> Biopython-dev mailing list
> Biopython-dev at mailman.open-bio.org
> http://mailman.open-bio.org/mailman/listinfo/biopython-dev