phyloschema

PhyloSchema

This class (importable from peyotl.nexson_syntax) provides a simple container for holding the set of variables needed to convert from one format to another (with error checking). The primary motivation for this class is to:

1. generate type conversion errors up front when some one requests a particular coercion. For example, this allows the phylesystem api to raise an error before it fetches the data in cases in which the user is requesting a format/content combination is not currently supported (or not possible)
2. allow that agreed-upon coercion to be done later with a simple call to convert or serialize. So the class acts like a closure that can transform any nexson to the desired format (if NexSON has the necessary content)

This class is used internally by peyotl. Client code may have to create instances of PhyloSchema objects, but should not need to call methods of the objects.

Initialization

PhyloSchema.__init__(schema, **kwargs) creates an object where the arguments are:

• schema None one of nexson | newick | nexml|nexus`
• content one of file | meta | otu | otumap | otus | study | subtree | tree
• content_id = None if content == study or meta)
  a string if content is file, otu, otumap, otus, tree, or
  a tuple of strings (TREE_ID, NODE_ID) if content==tree
• type_ext (only used if schema is None to infer schema using the rule
  '.nexson' -> 'nexson',
  '.nexml'-> 'nexml',
  '.nex'-> 'nexus',
  '.tre'-> 'newick',
  '.nwk'-> 'newick',
  
• output_nexml2json used if the schema is NexSON should be 0.0.0, 1.0.0, 1.2.1 specifies the version of NexSON
• format_str acts like schema (but has lower priority than type_ext and output_nexml2json)
• version acts like output_nexml2json
• otu_label (see below)

If OTUs are requested, the otu_label specifies what field will be used to represent the otu. This should be ot:originalLabel, ot:ottId, or ot:ottTaxonName (but the value is not case sensitive).

create_content_spec factory

create_content_spec returns a PhyloSchema object. As seen above, the PhyloSchema(...) initializer is quite complex and its generic nature leads to some cryptic names (e.g. content_id can hold a tree ID, or an OTU ID depending on the content argument). The create_content_spec(**kwargs) supports some more specific arguments that lead to clearer calls on the client side. The arguments for this function support:

• format as an alias for format_str
• nexson_version as an alias of version
• tip_label as an alias for otu_label
• tree_id=TREEID, subtree_id=NODEID or tree_id=TREEID, node_id=NODEID as aliases for content="subtree", content_id=(TREEID, NODEID)
• tree_id=TREEID as an alias for content="tree", content=TREEID
• otus_id=OTUSID as an alias for content="otus", content=OTUSID
• otu_id=OTUID as an alias for content="otu", content=OTUID