Source code for glypy.utils.base

import warnings
import os
import random
import sys
import gzip

from collections import defaultdict
try:  # pragma: no cover
    import cPickle as pickle
except:  # pragma: no cover
    import pickle
try:  # pragma: no cover
    from lxml import etree as ET
except ImportError:  # pragma: no cover
    try:
        from xml.etree import cElementTree as ET
    except:
        from xml.etree import ElementTree as ET

try:   # pragma: no cover
    from cStringIO import StringIO
except:  # pragma: no cover
    try:
        from StringIO import StringIO
    except:
        from io import StringIO

try:  # pragma: no cover
    i128 = long
    basestring = basestring
except:  # pragma: no cover
    i128 = int
    basestring = (bytes, str)


def cyclewarning():
    '''
    Used to warn users about the presence of cylcical glycans, which are harder to
    reason about for crossring_cleavages.
    '''
    warnings.warn("A cyclic glycan may be present. They may not cross-ring fragment correctly.", stacklevel=3)


def opener(obj, mode='r'):
    '''
    Try to use `obj` to access a file-object. If `obj` is a string, assume
    it denotes a path to a file, and open that file in the specified mode.
    If `obj` has an attribute `read`, assume it
    itself is a file-like object and return it.

    Parameters
    ----------
    obj: basestring or file-like object
        If `obj` is a base string it is treated like a file path, else if it supports
        the file-like operation `read`, return the object unchanged.
    mode: str, optional
        The mode, if any, to open `obj` with if it is a file path. Defaults to 'r', `read`

    '''
    if isinstance(obj, basestring):
        if obj[-2:] == 'gz':  # pragma: no cover
            return gzip.open(obj, mode)
        return open(obj, mode)
    elif hasattr(obj, "read"):
        return obj
    else:  # pragma: no cover
        raise IOError("Can't find a way to open {}".format(obj))


def invert_dict(d):
    return {v: k for k, v in d.items()}


def make_counter(start=1):
    '''
    Create a functor whose only internal piece of data is a mutable container
    with a reference to an integer, `start`. When the functor is called, it returns
    current `int` value of `start` and increments the mutable value by one.

    Parameters
    ----------
    start: int, optional
        The number to start counting from. Defaults to `1`.

    Returns
    -------
    int:
        The next number in the count progression.
    '''
    start = [start]

    def count_up():
        ret_val = start[0]
        start[0] += 1
        return ret_val
    return count_up


def identity(x):   # pragma: no cover
    return x


def nullop(*args, **kwargs):   # pragma: no cover
    pass


def chrinc(a='a', i=1):
    return chr(ord(a) + i)


def make_struct(name, fields, debug=False, docstring=None):
    '''
    A convenience function for defining plain-old-data (POD) objects that are optimized
    for named accessor lookup, unlike `namedtuple`. If the named container does not
    require any special logic and won't be extended, the resulting structure is best for
    storing and accessing the data.

    Parameters
    ----------
    name: str
        The name of the new class structure
    fields: iterable of str
    debug: bool
        Should the generated code be printed
    docstring: str, optional
        The docstring to apply to the generated class
    '''
    # Generate a totally uninformative docstring
    if docstring is None:
        docstring = """{name} is a plain-old-data holder.

Attributes
----------
""".format(name=name)
        for field in fields:
            docstring += "    {field}: object\n".format(field=field)
        docstring = '\n'.join(["    " + line for line in docstring.splitlines()]).strip()
    template = ('''
class {name}(object):
    r"""{docstring!s}
    """

    __slots__ = {fields!r}
    def __init__(self, {args}):
        {self_fields} = {args}
    def __getitem__(self, idx):
        return getattr(self, fields[idx])
    def __setitem__(self, idx, value):
        return setattr(self, fields[idx], value)
    def __getstate__(self):
        return ({self_fields})
    def __setstate__(self, state):
        {self_fields} = state
    def __repr__(self):
        rep = "{name}("
        i = 0
        for f in {fields!r}:
            if i != 0:
                rep += ", "
            i += 1
            rep += f + "=" + str(getattr(self, f))
        rep += ")"
        return rep
    def __eq__(self, other):
        for f in {fields!r}:
            if getattr(self, f) != getattr(other, f):
                return False
        return True
    def __ne__(self, other):
        return not self == other

    @property
    def __dict__(self):
        d = dict()
        for f in {fields!r}:
            d[f] = getattr(self, f)
        return d
    ''').format(
        name=name,
        fields=fields,
        args=','.join(fields),
        self_fields=','.join('self.' + f for f in fields),
        docstring=docstring)
    d = {'fields': fields}
    if debug:  # pragma: no cover
        print(template)
    exec(template, d)
    result = d[name]
    # Patch the class to support pickling, as is done for namedtuple
    try:
        result.__module__ = sys._getframe(1).f_globals.get('__name__', '__main__')
    except (AttributeError, ValueError):  # pragma: no cover
        pass
    return result


class ClassPropertyDescriptor(object):
    '''
    Standard Class Property Descriptor Implementation
    '''
    def __init__(self, fget, fset=None):
        self.fget = fget
        self.fset = fset

    def __get__(self, obj, klass=None):
        if klass is None:  # pragma: no cover
            klass = type(obj)
        return self.fget.__get__(obj, klass)()

    def __set__(self, obj, value):
        if not self.fset:  # pragma: no cover
            raise AttributeError("can't set attribute")
        type_ = type(obj)
        return self.fset.__get__(obj, type_)(value)

    def setter(self, func):
        if not isinstance(func, (classmethod, staticmethod)):
            func = classmethod(func)
        self.fset = func
        return self


def classproperty(func):
    '''
    Applies ClassPropertyDescriptor as you would a normal
    @property descriptor
    '''
    if not isinstance(func, (classmethod, staticmethod)):
        func = classmethod(func)

    return ClassPropertyDescriptor(func)


[docs]class RootProtocolNotSupportedError(TypeError): """A subclass of :class:`TypeError` that signals that the provided type does not support the `__root__` protocol. """ pass
[docs]def root(structure): """A generic method for obtaining the root of a structure representing or containing a glycan graph with a single distinct root. Implementing objects should provide a `__root__` method that returns a :class:`~.Monosaccharide` object. Parameters ---------- structure : any An object that implements the `root` protocol, containing a tree structure somewhere inside it. Returns ------- Monosaccharide : The root of the |Glycan| tree Raises ------ RootProtocolNotSupportedError """ try: root = structure.__root__() return root except AttributeError: raise RootProtocolNotSupportedError( "{} does not support the `root` protocol".format(structure.__class__.__name__))
[docs]class TreeProtocolNotSupportedError(TypeError): """A subclass of :class:`TypeError` that signals that the provided type does not support the `__tree__` protocol. """ pass
[docs]def tree(structure): """A generic method for obtaining the :class:`Glycan` of a structure representing or containing a glycan graph. Implementing objects should provide a `__tree__` method that returns a :class:`~.Glycan` object. Parameters ---------- structure : any An object that implements the `tree` protocol, containing a tree structure somewhere inside it. Returns ------- Glycan Raises ------ TreeProtocolNotSupportedError """ try: tree = structure.__tree__() except AttributeError: # pragma: no cover raise TreeProtocolNotSupportedError( "{} does not support the `tree` protocol.".format(structure.__class__.__name__)) return tree
def groupby(ungrouped_list, key_fn=identity): groups = defaultdict(list) for item in ungrouped_list: key_value = key_fn(item) groups[key_value].append(item) return groups def where(iterable, fn): return [i for i, k in enumerate(iterable) if fn(k)] def uid(n=128): int_ = random.getrandbits(n) return int_