Last active
December 30, 2021 05:04
-
-
Save wasi-master/c68a8065fd2234b196cfe2a8c1723afc to your computer and use it in GitHub Desktop.
NKFC Normalization mapping
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
{ | |
' ': ['\xa0', '\u2000', '\u2001', '\u2002', '\u2003', '\u2004', '\u2005', '\u2006', '\u2007', '\u2008', '\u2009', '\u200a', '\u202f', '\u205f', '\u3000'], | |
'!': ['︕', '﹗', '!'], | |
'!?': ['⁉'], | |
'"': ['"'], | |
'#': ['﹟', '#'], | |
'$': ['﹩', '$'], | |
'%': ['﹪', '%'], | |
'&': ['﹠', '&'], | |
"'": ['''], | |
'(': ['⁽', '₍', '︵', '﹙', '('], | |
')': ['⁾', '₎', '︶', '﹚', ')'], | |
'*': ['﹡', '*'], | |
'+': ['⁺', '₊', '﬩', '﹢', '+'], | |
',': ['︐', '﹐', ','], | |
'-': ['﹣', '-'], | |
'.': ['․', '﹒', '.'], | |
'..': ['‥', '︰'], | |
'...': ['…', '︙'], | |
'/': ['/'], | |
'0': ['⁰', '₀', '⓪', '0'], | |
'1': ['¹', '₁', '①', '1'], | |
'1.': ['⒈'], | |
'10.': ['⒑'], | |
'11.': ['⒒'], | |
'12.': ['⒓'], | |
'13.': ['⒔'], | |
'14.': ['⒕'], | |
'15.': ['⒖'], | |
'16.': ['⒗'], | |
'17.': ['⒘'], | |
'18.': ['⒙'], | |
'19.': ['⒚'], | |
'2': ['²', '₂', '②', '2'], | |
'2.': ['⒉'], | |
'20.': ['⒛'], | |
'3': ['³', '₃', '③', '3'], | |
'3.': ['⒊'], | |
'4': ['⁴', '₄', '④', '4'], | |
'4.': ['⒋'], | |
'5': ['⁵', '₅', '⑤', '5'], | |
'5.': ['⒌'], | |
'6': ['⁶', '₆', '⑥', '6'], | |
'6.': ['⒍'], | |
'7': ['⁷', '₇', '⑦', '7'], | |
'7.': ['⒎'], | |
'8': ['⁸', '₈', '⑧', '8'], | |
'8.': ['⒏'], | |
'9': ['⁹', '₉', '⑨', '9'], | |
'9.': ['⒐'], | |
':': ['︓', '﹕', ':'], | |
'::=': ['⩴'], | |
';': [';', '︔', '﹔', ';'], | |
'<': ['﹤', '<'], | |
'=': ['⁼', '₌', '﹦', '='], | |
'==': ['⩵'], | |
'===': ['⩶'], | |
'>': ['﹥', '>'], | |
'?': ['︖', '﹖', '?'], | |
'?!': ['⁈'], | |
'??': ['⁇'], | |
'@': ['﹫', '@'], | |
'A': ['ᴬ', 'Ⓐ', 'A'], | |
'B': ['ᴮ', 'ℬ', 'Ⓑ', 'B'], | |
'C': ['ℂ', 'ℭ', 'Ⅽ', 'Ⓒ', 'C'], | |
'Co.': ['㏇'], | |
'D': ['ᴰ', 'ⅅ', 'Ⅾ', 'Ⓓ', 'D'], | |
'E': ['ᴱ', 'ℰ', 'Ⓔ', 'E'], | |
'F': ['ℱ', 'Ⓕ', 'F'], | |
'G': ['ᴳ', 'Ⓖ', 'G'], | |
'H': ['ᴴ', 'ℋ', 'ℌ', 'ℍ', 'Ⓗ', 'H'], | |
'I': ['ᴵ', 'ℐ', 'ℑ', 'Ⅰ', 'Ⓘ', 'I'], | |
'J': ['ᴶ', 'Ⓙ', 'J'], | |
'K': ['ᴷ', 'K', 'Ⓚ', 'K'], | |
'L': ['ᴸ', 'ℒ', 'Ⅼ', 'Ⓛ', 'L'], | |
'M': ['ᴹ', 'ℳ', 'Ⅿ', 'Ⓜ', 'M'], | |
'N': ['ᴺ', 'ℕ', 'Ⓝ', 'N'], | |
'O': ['ᴼ', 'Ⓞ', 'O'], | |
'P': ['ᴾ', 'ℙ', 'Ⓟ', 'P'], | |
'Q': ['ℚ', 'Ⓠ', 'Q'], | |
'R': ['ᴿ', 'ℛ', 'ℜ', 'ℝ', 'Ⓡ', 'R'], | |
'S': ['Ⓢ', 'S'], | |
'T': ['ᵀ', 'Ⓣ', 'T'], | |
'U': ['ᵁ', 'Ⓤ', 'U'], | |
'V': ['Ⅴ', 'Ⓥ', 'ⱽ', 'V'], | |
'W': ['ᵂ', 'Ⓦ', 'W'], | |
'X': ['Ⅹ', 'Ⓧ', 'X'], | |
'Y': ['Ⓨ', 'Y'], | |
'Z': ['ℤ', 'ℨ', 'Ⓩ', 'Z'], | |
'[': ['﹇', '['], | |
'\\': ['﹨', '\'], | |
']': ['﹈', ']'], | |
'^': ['^'], | |
'_': ['︳', '︴', '﹍', '﹎', '﹏', '_'], | |
'`': ['`', '`'], | |
'a': ['ª', 'ᵃ', 'ₐ', 'ⓐ', 'a'], | |
'a.m.': ['㏂'], | |
'a/c': ['℀'], | |
'a/s': ['℁'], | |
'b': ['ᵇ', 'ⓑ', 'b'], | |
'c': ['ᶜ', 'ⅽ', 'ⓒ', 'c'], | |
'c/o': ['℅'], | |
'c/u': ['℆'], | |
'd': ['ᵈ', 'ⅆ', 'ⅾ', 'ⓓ', 'd'], | |
'e': ['ᵉ', 'ₑ', 'ℯ', 'ⅇ', 'ⓔ', 'e'], | |
'f': ['ᶠ', 'ⓕ', 'f'], | |
'g': ['ᵍ', 'ℊ', 'ⓖ', 'g'], | |
'h': ['ʰ', 'ℎ', 'ⓗ', 'h'], | |
'i': ['ᵢ', 'ⁱ', 'ℹ', 'ⅈ', 'ⅰ', 'ⓘ', 'i'], | |
'j': ['ʲ', 'ⅉ', 'ⓙ', 'ⱼ', 'j'], | |
'k': ['ᵏ', 'ⓚ', 'k'], | |
'l': ['ˡ', 'ℓ', 'ⅼ', 'ⓛ', 'l'], | |
'm': ['ᵐ', 'ⅿ', 'ⓜ', 'm'], | |
'n': ['ⁿ', 'ⓝ', 'n'], | |
'o': ['º', 'ᵒ', 'ₒ', 'ℴ', 'ⓞ', 'o'], | |
'p': ['ᵖ', 'ⓟ', 'p'], | |
'p.m.': ['㏘'], | |
'q': ['ⓠ', 'q'], | |
'r': ['ʳ', 'ᵣ', 'ⓡ', 'r'], | |
's': ['ſ', 'ˢ', 'ⓢ', 's'], | |
't': ['ᵗ', 'ⓣ', 't'], | |
'u': ['ᵘ', 'ᵤ', 'ⓤ', 'u'], | |
'v': ['ᵛ', 'ᵥ', 'ⅴ', 'ⓥ', 'v'], | |
'w': ['ʷ', 'ⓦ', 'w'], | |
'x': ['ˣ', 'ₓ', 'ⅹ', 'ⓧ', 'x'], | |
'y': ['ʸ', 'ⓨ', 'y'], | |
'z': ['ᶻ', 'ⓩ', 'z'], | |
'{': ['︷', '﹛', '{'], | |
'|': ['|'], | |
'}': ['︸', '﹜', '}'], | |
'~': ['~'] | |
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import requests | |
import rich | |
from bs4 import BeautifulSoup | |
from rich.syntax import Syntax | |
url = "https://appcheck-ng.com/wp-content/uploads/unicode_normalization.html" | |
soup = BeautifulSoup(requests.get(url).text, "lxml") | |
characters = soup.find("table").find_all("tr")[1:] | |
output = {} | |
for character in characters: | |
original = character.find("td").text | |
alternatives = character.find_all("td")[2:] | |
for alternative in alternatives: | |
if not "NFKC" in alternative.text: | |
print(alternative.text.split(";")[1], '"is not NFKC') | |
continue | |
if output.get(original): | |
output[original].append(alternative.find("span").text) | |
else: | |
output[original] = [alternative.find("span").text] | |
rich.print(output) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment