|
| 1 | +# coding: utf-8 |
| 2 | + |
| 3 | +# This script autogenerates `IPython.core.latex_symbols.py`, which contains a |
| 4 | +# single dict , named `latex_symbols`. The keys in this dict are latex symbols, |
| 5 | +# such as `\\alpha` and the values in the dict are the unicode equivalents for |
| 6 | +# those. Most importantly, only unicode symbols that are valid identifers in |
| 7 | +# Python 3 are included. |
| 8 | + |
| 9 | +# |
| 10 | +# The original mapping of latex symbols to unicode comes from the `latex_symbols.jl` files from Julia. |
| 11 | + |
| 12 | +from __future__ import print_function |
| 13 | + |
| 14 | +# Import the Julia LaTeX symbols |
| 15 | +print('Importing latex_symbols.js from Julia...') |
| 16 | +import requests |
| 17 | +url = 'https://raw.githubusercontent.com/JuliaLang/julia/master/base/latex_symbols.jl' |
| 18 | +r = requests.get(url) |
| 19 | + |
| 20 | + |
| 21 | +# Build a list of key, value pairs |
| 22 | +print('Building a list of (latex, unicode) key-vaule pairs...') |
| 23 | +lines = r.text.splitlines()[60:] |
| 24 | +lines = [line for line in lines if '=>' in line] |
| 25 | +lines = [line.replace('=>',':') for line in lines] |
| 26 | + |
| 27 | +def line_to_tuple(line): |
| 28 | + """Convert a single line of the .jl file to a 2-tuple of strings like ("\\alpha", "α")""" |
| 29 | + kv = line.split(',')[0].split(':') |
| 30 | +# kv = tuple(line.strip(', ').split(':')) |
| 31 | + k, v = kv[0].strip(' "'), kv[1].strip(' "') |
| 32 | +# if not test_ident(v): |
| 33 | +# print(line) |
| 34 | + return k, v |
| 35 | + |
| 36 | +assert line_to_tuple(' "\\sqrt" : "\u221A",') == ('\\sqrt', '\u221A') |
| 37 | +lines = [line_to_tuple(line) for line in lines] |
| 38 | + |
| 39 | + |
| 40 | +# Filter out non-valid identifiers |
| 41 | +print('Filtering out characters that are not valid Python 3 identifiers') |
| 42 | + |
| 43 | +def test_ident(i): |
| 44 | + """Is the unicode string a valid Python 3 identifer.""" |
| 45 | + try: |
| 46 | + exec('a%s = 10' % i, {}, {}) |
| 47 | + except SyntaxError: |
| 48 | + return False |
| 49 | + else: |
| 50 | + return True |
| 51 | + |
| 52 | +assert test_ident("α") |
| 53 | +assert not test_ident('‴') |
| 54 | + |
| 55 | +valid_idents = [line for line in lines if test_ident(line[1])] |
| 56 | + |
| 57 | + |
| 58 | +# Write the `latex_symbols.py` module in the cwd |
| 59 | + |
| 60 | +s = """# encoding: utf-8 |
| 61 | +
|
| 62 | +# This file is autogenerated from the file: |
| 63 | +# https://raw.githubusercontent.com/JuliaLang/julia/master/base/latex_symbols.jl |
| 64 | +# This original list is filtered to remove any unicode characters that are not valid |
| 65 | +# Python identifiers. |
| 66 | +
|
| 67 | +latex_symbols = {\n |
| 68 | +""" |
| 69 | +for line in valid_idents: |
| 70 | + s += ' "%s" : "%s",\n' % (line[0], line[1]) |
| 71 | +s += "}\n" |
| 72 | + |
| 73 | +with open('latex_symbols.py', 'w', encoding='utf-8') as f: |
| 74 | + f.write(s) |
| 75 | + |
| 76 | + |
0 commit comments