parsevalue script
ActivePublic
Actions

Authored by matej_suchanek on Mar 14 2023, 8:19 AM.

Tags

None

Referenced Files

	F36910717: parsevalue script
	Mar 14 2023, 8:19 AM

Subscribers

None

	from collections import namedtuple
	from enum import IntEnum

	import matplotlib.colors as mcolors
	import matplotlib.patches as mpatches
	import matplotlib.pyplot as plt
	import numpy as np
	import pywikibot
	from pywikibot import WbTime
	from pywikibot.tools.itertools import itergroup


	class Status(IntEnum):
	UNDEFINED = 0
	ERROR = 1
	OK = 2
	MONTH_DAY = 3
	WRONG = 4


	colors = ['black', 'grey', 'green', 'yellow', 'red']
	labels = ['Undefined', 'Error', 'OK', 'MDY', 'Different date']
	cmap = mcolors.ListedColormap(colors)
	patches = [mpatches.Patch(color=color, label=label)
	for color, label in zip(colors, labels)]

	site = pywikibot.Site('wikidata', 'wikidata')

	matrix = np.full((12, 31), Status.UNDEFINED, dtype=np.uint8)
	months = [31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31]
	for m, d in enumerate(months):
	matrix[m, :d] = Status.ERROR

	language = 'cs'
	year = 2022
	template = '{0.day}.{0.month}.{0.year}'
	#template = '{0.day:02}.{0.month:02}.{0.year:02}'
	dates = []
	d_type = namedtuple('Date', ['year', 'month', 'day'])
	for m, days in enumerate(months, start=1):
	for d in range(1, days+1):
	dates.append(d_type(year, m, d))

	all_strings = [template.format(date) for date in dates]
	length = len(all_strings)
	batch = 16
	stack = []
	for i in range(0, length, batch):
	stack.append((i, min(i+batch, length)))
	stack = stack[::-1]

	while stack:
	start, end = stack.pop()
	strings = all_strings[start:end]
	try:
	parsed = site.parsevalue('time', strings, options={'lang': language}, language=language)
	except ValueError:
	mid = start + (end - start) // 2
	if start != mid and mid != end:
	stack.append((mid, end))
	stack.append((start, mid))
	continue

	for date, text, out in zip(dates[start:end], strings, parsed):
	year, m, d = date
	value = WbTime.fromWikibase(out, site)
	if value.year == year and value.month == m and value.day == d:
	matrix[m-1, d-1] = Status.OK
	elif value.year == year and value.month == d and value.day == m:
	matrix[m-1, d-1] = Status.MONTH_DAY
	else:
	matrix[m-1, d-1] = Status.WRONG

	plt.figure(tight_layout=True)
	plt.imshow(matrix.T, cmap=cmap, interpolation='nearest', vmin=0, vmax=4)
	plt.title(f'Format: "{all_strings[0]}"')
	plt.xlabel('Month')
	plt.yticks(range(12), map(str, range(1, 13)))
	plt.ylabel('Day')
	plt.yticks(range(31), map(str, range(1, 32)))
	plt.legend(handles=patches, bbox_to_anchor=(1.05, 1.0), loc='upper left')

	plt.savefig(f"../{template.format(d_type(year, 1, 1)).replace(' ', '_')}.png",
	format='png')
	plt.show()

Event Timeline

matej_suchanek created this paste.Mar 14 2023, 8:19 AM

matej_suchanek mentioned this in T221097: The most commonly used date format in the Czech Republic produces wrong date when used as a value in Wikidata.Mar 14 2023, 8:28 AM

parsevalue scriptActivePublicActions

Event Timeline

parsevalue script
ActivePublic
Actions