From 5ec283a3592dfdef7aff34ab00ca8685c4d37470 Mon Sep 17 00:00:00 2001
From: Kurt McKee <contactme@kurtmckee.org>
Date: Sun, 6 Sep 2020 07:53:35 -0500
Subject: Overhaul Javascript numeric literals (#1534)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

* Rename the "Javascript" tests to reflect that they are for CoffeeScript

This change also modifies the module docstring to reflect the file's purpose.

* Overhaul the Javascript numeric literal parsing

Fixes #307

This patch contains the following changes:

* Adds 50+ unit tests for Javascript numeric literals
* Forces ASCII numbers for float literals (so, now reject `.୪`)
* Adds support for Javascript's BigInt notation (`100n`)
* Adds support for leading-zero-only octal notation (`0777`)
* Adds support for scientific notation with no significand (`1e10`)

Numeric literal parsing is based on information at:

* https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Grammar_and_types
* https://developer.mozilla.org/en-US/docs/Web/JavaScript/Data_structures
---
 tests/test_coffeescript.py | 85 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 85 insertions(+)
 create mode 100644 tests/test_coffeescript.py

(limited to 'tests/test_coffeescript.py')

diff --git a/tests/test_coffeescript.py b/tests/test_coffeescript.py
new file mode 100644
index 00000000..41ca8e0d
--- /dev/null
+++ b/tests/test_coffeescript.py
@@ -0,0 +1,85 @@
+# -*- coding: utf-8 -*-
+"""
+    CoffeeScript tests
+    ~~~~~~~~~~~~~~~~~~
+
+    :copyright: Copyright 2006-2020 by the Pygments team, see AUTHORS.
+    :license: BSD, see LICENSE for details.
+"""
+
+import pytest
+
+from pygments.lexers import CoffeeScriptLexer
+from pygments.token import Token
+
+COFFEE_SLASH_GOLDEN = [
+    # input_str, slashes_are_regex_here
+    (r'/\\/', True),
+    (r'/\\/i', True),
+    (r'/\//', True),
+    (r'/(\s)/', True),
+    ('/a{2,8}/', True),
+    ('/b*c?d+/', True),
+    ('/(capture-match)/', True),
+    ('/(?:do-not-capture-match)/', True),
+    ('/this|or|that/', True),
+    ('/[char-set]/', True),
+    ('/[^neg-char_st]/', True),
+    ('/^.*$/', True),
+    (r'/\n(\f)\0\1\d\b\cm\u1234/', True),
+    (r'/^.?([^/\\\n\w]*)a\1+$/.something(or_other) # something more complex', True),
+    ("foo = (str) ->\n  /'|\"/.test str", True),
+    ('a = a / b / c', False),
+    ('a = a/b/c', False),
+    ('a = a/b/ c', False),
+    ('a = a /b/c', False),
+    ('a = 1 + /d/.test(a)', True),
+]
+
+
+@pytest.fixture(scope='module')
+def lexer():
+    yield CoffeeScriptLexer()
+
+
+@pytest.mark.parametrize('golden', COFFEE_SLASH_GOLDEN)
+def test_coffee_slashes(lexer, golden):
+    input_str, slashes_are_regex_here = golden
+    output = list(lexer.get_tokens(input_str))
+    print(output)
+    for t, s in output:
+        if '/' in s:
+            is_regex = t is Token.String.Regex
+            assert is_regex == slashes_are_regex_here, (t, s)
+
+
+def test_mixed_slashes(lexer):
+    fragment = u'a?/foo/:1/2;\n'
+    tokens = [
+        (Token.Name.Other, u'a'),
+        (Token.Operator, u'?'),
+        (Token.Literal.String.Regex, u'/foo/'),
+        (Token.Operator, u':'),
+        (Token.Literal.Number.Integer, u'1'),
+        (Token.Operator, u'/'),
+        (Token.Literal.Number.Integer, u'2'),
+        (Token.Punctuation, u';'),
+        (Token.Text, u'\n'),
+    ]
+    assert list(lexer.get_tokens(fragment)) == tokens
+
+
+def test_beware_infinite_loop(lexer):
+    # This demonstrates the case that "This isn't really guarding" comment
+    # refers to.
+    fragment = '/a/x;\n'
+    tokens = [
+        (Token.Text, ''),
+        (Token.Operator, '/'),
+        (Token.Name.Other, 'a'),
+        (Token.Operator, '/'),
+        (Token.Name.Other, 'x'),
+        (Token.Punctuation, ';'),
+        (Token.Text, '\n'),
+    ]
+    assert list(lexer.get_tokens(fragment)) == tokens
-- 
cgit v1.2.1