#!/usr/bin/env python
# -*- coding: UTF-8 -*-
#
# Copyright 2016-2019 European Commission (JRC);
# Licensed under the EUPL (the 'Licence');
# You may not use this work except in compliance with the Licence.
# You may obtain a copy of the Licence at: http://ec.europa.eu/idabc/eupl
"""
It provides Ranges class.
"""
import itertools
import numpy as np
from .tokens.operand import (
_re_range, range2parts, _index2col, maxrow, maxcol, Error
)
from .errors import RangeValueError
from .functions import Array
import schedula as sh
def _has_same_sheet(x, y):
try:
return x['excel'] == y['excel'] and x['sheet'] == y['sheet']
except KeyError:
return False
def _have_intersect(x, y):
if _has_same_sheet(x, y):
n1, n2 = max(y['n1'], x['n1']), min(y['n2'], x['n2'])
if n1 <= n2:
r1 = max(int(y['r1']), int(x['r1']))
r2 = min(int(y['r2']), int(x['r2']))
if r1 <= r2:
return {
'excel': x['excel'], 'sheet': x['sheet'], 'n1': n1,
'r1': str(r1), 'n2': n2, 'r2': str(r2)
}
return {}
def _split(base, rng, intersect=None, format_range=range2parts):
z = _have_intersect(base, rng)
if not z:
return rng,
if intersect is not None:
intersect.update(z)
ranges = []
rng = sh.selector(('excel', 'sheet', 'n1', 'n2', 'r1', 'r2'), rng)
it = ('n1', 'n2', 1), ('n2', 'n1', -1), ('r1', 'r2', 1), ('r2', 'r1', -1)
for i, j, n in it:
if z[i] != rng[i]:
r = rng.copy()
r[j] = str(int(z[i]) - n) if j[0] == 'r' else z[i] - n
r = dict(format_range(('name', 'n1', 'n2'), **r))
ranges.append(r)
rng[i] = z[i]
return tuple(ranges)
def _intersect(rng, ranges):
for r in ranges:
z = _have_intersect(rng, r)
if z:
yield z
def _merge_raw_update(base, rng):
if _has_same_sheet(base, rng):
if base['n1'] == rng['n2'] and int(base['r2']) + 1 >= int(rng['r1']):
base['r2'] = rng['r2']
return True
def _merge_col_update(base, rng):
if _has_same_sheet(base, rng):
if (base['n2'] + 1) == rng['n1']:
if base['r1'] == rng['r1'] and base['r2'] == rng['r2']:
base['n2'] = rng['n2']
return True
def _get_indices_intersection(base, i):
r, c = int(base['r1']), base['n1']
r = slice(int(i['r1']) - r, int(i['r2']) - r + 1)
c = slice(i['n1'] - c, i['n2'] - c + 1)
return r, c
def _assemble_values(base, values, empty=''):
res = np.empty(_shape(**base), object)
res[:, :] = empty
for k, (rng, value) in sorted(values.items()):
ist = _have_intersect(base, rng)
if ist:
br, bc = _get_indices_intersection(base, ist)
rr, rc = _get_indices_intersection(rng, ist)
res[br, bc] = value[rr, rc]
return res
# noinspection PyUnusedLocal
def _shape(n1, n2, r1, r2, **kw):
r1, r2 = int(r1), int(r2)
r = maxrow if r1 == 0 or r2 == maxrow else (r2 - r1 + 1)
c = maxcol if n1 == 0 or n2 == maxcol else (n2 - n1 + 1)
return r, c
def _reshape_array_as_excel(value, base_shape):
try:
return np.reshape(value, base_shape)
except ValueError:
res, (r, c) = np.empty(base_shape, object), value.shape
res[:, :] = getattr(value, '_default', Error.errors['#N/A'])
r = None if r == 1 else r
c = None if c == 1 else c
try:
res[:r, :c] = value
except ValueError:
res[:, :] = Error.errors['#VALUE!']
return res
[docs]class Ranges:
input_fields = ('excel', 'sheet', 'n1', 'n2', 'r1', 'r2')
[docs] def __init__(self, ranges=(), values=None):
self.ranges = ranges
self.values = values or {}
self._value = sh.NONE
[docs] def pushes(self, refs, values=(), context=None):
for r, v in itertools.zip_longest(refs, values, fillvalue=sh.EMPTY):
self.push(r, value=v, context=context)
return self
[docs] def set_value(self, rng, value=sh.EMPTY):
self._value = sh.NONE
self.ranges += rng,
if value is not sh.EMPTY:
if not isinstance(value, Array):
if not np.ndim(value):
value = [[value]]
value = np.asarray(value, object)
shape = _shape(**rng)
value = _reshape_array_as_excel(value, shape)
self.values[rng['name']] = (rng, value)
return self
@property
def is_set(self):
return len(self.ranges) > 1
[docs] @staticmethod
def get_range(format_range, ref, context=None):
ctx = (context or {}).copy()
for k, v in _re_range.match(ref).groupdict().items():
if v is not None:
if k == 'ref':
raise ValueError
ctx[k] = v
return dict(format_range(('name', 'n1', 'n2'), **ctx))
[docs] def push(self, ref, value=sh.EMPTY, context=None):
rng = self.get_range(self.format_range, ref, context)
return self.set_value(rng, value)
def __add__(self, other): # Expand.
ranges = self.ranges[1:] + other.ranges
rng = sh.selector(self.input_fields, self.ranges[0])
for k in ('r1', 'r2'):
rng[k] = int(rng[k])
for r in ranges:
if not _has_same_sheet(rng, r):
raise RangeValueError('{}:{}'.format(self, other))
else:
rng['r1'] = min(rng['r1'], int(r['r1']))
rng['n1'] = min(rng['n1'], r['n1'])
rng['r2'] = max(rng['r2'], int(r['r2']))
rng['n2'] = max(rng['n2'], r['n2'])
rng = dict(self.format_range(('name', 'n1', 'n2'), **rng))
if self.values and other.values:
values = self.values.copy()
values.update(other.values)
value = _assemble_values(rng, values)
return Ranges().push(rng['name'], value)
return Ranges((rng,))
def __or__(self, other): # Union.
values = self.values.copy()
values.update(other.values)
return Ranges(self.ranges + other.ranges, values)
[docs] def intersect(self, other):
if self.ranges:
for rng in other.ranges:
yield from _intersect(rng, self.ranges)
def __and__(self, other): # Intersection.
r = tuple(dict(self.format_range(('name', 'n1', 'n2'), **i))
for i in self.intersect(other))
values = self.values.copy()
values.update(other.values)
return Ranges(r, values)
def __sub__(self, other):
base = other.ranges
for r0 in self.ranges:
stack = [r0]
for b in base:
s = stack.copy()
stack = []
for r in s:
stack.extend(_split(b, r, format_range=self.format_range))
base += tuple(stack)
base, values = base[len(other.ranges):], self.values
return Ranges(base, values)
[docs] def simplify(self):
rng = self.ranges
if len(rng) <= 1:
return self
it = range(min(r['n1'] for r in rng), max(r['n2'] for r in rng) + 1)
it = ['{0}:{0}'.format(_index2col(c)) for c in it]
spl = (self & Ranges().pushes(it))._merge()
return spl
def _merge(self):
# noinspection PyPep8
key = lambda x: (x['n1'], int(x['r1']), -x['n2'], -int(x['r2']))
rng = self.ranges
for merge, select in ((_merge_raw_update, 1), (_merge_col_update, 0)):
it, rng = sorted(rng, key=key), []
for r in it:
if not (rng and merge(rng[-1], r)):
if rng:
rng.append(rng.pop())
if select:
r = sh.selector(self.input_fields, r)
rng.append(r)
rng = [dict(self.format_range(['name'], **r)) for r in rng]
return Ranges(tuple(rng), self.values)
def __repr__(self):
ranges = ', '.join(r['name'] for r in self.ranges)
value = '={}'.format(self.value) if ranges and self.values else ''
return '<%s>(%s)%s' % (self.__class__.__name__, ranges, value)
@property
def value(self):
if self._value is not sh.NONE:
return self._value
if self.ranges and not self.values:
raise RangeValueError(str(self))
stack, values = list(self.ranges), []
while stack:
update = False
for k, (rng, value) in sorted(self.values.items()):
if not stack:
break
i = {}
new_rng = _split(
rng, stack[-1], intersect=i, format_range=self.format_range
)
if i:
update = True
stack.pop()
stack.extend(new_rng)
r, c = _get_indices_intersection(rng, i)
values.append(value[:, c][r])
else:
if not update:
break
if self.is_set:
self._value = np.concatenate([v.ravel() for v in values])
elif values:
self._value = values[0]
else:
self._value = np.asarray([[Error.errors['#NULL!']]], object)
return self._value