diff options
Diffstat (limited to 'crates/ra_syntax/src/string_lexing/char.rs')
-rw-r--r-- | crates/ra_syntax/src/string_lexing/char.rs | 176 |
1 files changed, 0 insertions, 176 deletions
diff --git a/crates/ra_syntax/src/string_lexing/char.rs b/crates/ra_syntax/src/string_lexing/char.rs deleted file mode 100644 index e01813176..000000000 --- a/crates/ra_syntax/src/string_lexing/char.rs +++ /dev/null | |||
@@ -1,176 +0,0 @@ | |||
1 | use super::parser::Parser; | ||
2 | use super::CharComponent; | ||
3 | |||
4 | pub fn parse_char_literal(src: &str) -> CharComponentIterator { | ||
5 | CharComponentIterator { | ||
6 | parser: Parser::new(src), | ||
7 | has_closing_quote: false, | ||
8 | } | ||
9 | } | ||
10 | |||
11 | pub struct CharComponentIterator<'a> { | ||
12 | parser: Parser<'a>, | ||
13 | pub has_closing_quote: bool, | ||
14 | } | ||
15 | |||
16 | impl<'a> Iterator for CharComponentIterator<'a> { | ||
17 | type Item = CharComponent; | ||
18 | fn next(&mut self) -> Option<CharComponent> { | ||
19 | if self.parser.pos == 0 { | ||
20 | assert!( | ||
21 | self.parser.advance() == '\'', | ||
22 | "char literal should start with a quote" | ||
23 | ); | ||
24 | } | ||
25 | |||
26 | if let Some(component) = self.parser.parse_char_component() { | ||
27 | return Some(component); | ||
28 | } | ||
29 | |||
30 | // We get here when there are no char components left to parse | ||
31 | if self.parser.peek() == Some('\'') { | ||
32 | self.parser.advance(); | ||
33 | self.has_closing_quote = true; | ||
34 | } | ||
35 | |||
36 | assert!( | ||
37 | self.parser.peek() == None, | ||
38 | "char literal should leave no unparsed input: src = {:?}, pos = {}, length = {}", | ||
39 | self.parser.src, | ||
40 | self.parser.pos, | ||
41 | self.parser.src.len() | ||
42 | ); | ||
43 | |||
44 | None | ||
45 | } | ||
46 | } | ||
47 | |||
48 | #[cfg(test)] | ||
49 | mod tests { | ||
50 | use rowan::TextRange; | ||
51 | use crate::string_lexing::{ | ||
52 | CharComponent, | ||
53 | CharComponentKind::*, | ||
54 | }; | ||
55 | |||
56 | fn parse(src: &str) -> (bool, Vec<CharComponent>) { | ||
57 | let component_iterator = &mut super::parse_char_literal(src); | ||
58 | let components: Vec<_> = component_iterator.collect(); | ||
59 | (component_iterator.has_closing_quote, components) | ||
60 | } | ||
61 | |||
62 | fn unclosed_char_component(src: &str) -> CharComponent { | ||
63 | let (has_closing_quote, components) = parse(src); | ||
64 | assert!(!has_closing_quote, "char should not have closing quote"); | ||
65 | assert!(components.len() == 1); | ||
66 | components[0].clone() | ||
67 | } | ||
68 | |||
69 | fn closed_char_component(src: &str) -> CharComponent { | ||
70 | let (has_closing_quote, components) = parse(src); | ||
71 | assert!(has_closing_quote, "char should have closing quote"); | ||
72 | assert!( | ||
73 | components.len() == 1, | ||
74 | "Literal: {}\nComponents: {:#?}", | ||
75 | src, | ||
76 | components | ||
77 | ); | ||
78 | components[0].clone() | ||
79 | } | ||
80 | |||
81 | fn closed_char_components(src: &str) -> Vec<CharComponent> { | ||
82 | let (has_closing_quote, components) = parse(src); | ||
83 | assert!(has_closing_quote, "char should have closing quote"); | ||
84 | components | ||
85 | } | ||
86 | |||
87 | fn range_closed(src: &str) -> TextRange { | ||
88 | TextRange::from_to(1.into(), (src.len() as u32 - 1).into()) | ||
89 | } | ||
90 | |||
91 | fn range_unclosed(src: &str) -> TextRange { | ||
92 | TextRange::from_to(1.into(), (src.len() as u32).into()) | ||
93 | } | ||
94 | |||
95 | #[test] | ||
96 | fn test_unicode_escapes() { | ||
97 | let unicode_escapes = &[r"{DEAD}", "{BEEF}", "{FF}", "{}", ""]; | ||
98 | for escape in unicode_escapes { | ||
99 | let escape_sequence = format!(r"'\u{}'", escape); | ||
100 | let component = closed_char_component(&escape_sequence); | ||
101 | let expected_range = range_closed(&escape_sequence); | ||
102 | assert_eq!(component.kind, UnicodeEscape); | ||
103 | assert_eq!(component.range, expected_range); | ||
104 | } | ||
105 | } | ||
106 | |||
107 | #[test] | ||
108 | fn test_unicode_escapes_unclosed() { | ||
109 | let unicode_escapes = &["{DEAD", "{BEEF", "{FF"]; | ||
110 | for escape in unicode_escapes { | ||
111 | let escape_sequence = format!(r"'\u{}'", escape); | ||
112 | let component = unclosed_char_component(&escape_sequence); | ||
113 | let expected_range = range_unclosed(&escape_sequence); | ||
114 | assert_eq!(component.kind, UnicodeEscape); | ||
115 | assert_eq!(component.range, expected_range); | ||
116 | } | ||
117 | } | ||
118 | |||
119 | #[test] | ||
120 | fn test_empty_char() { | ||
121 | let (has_closing_quote, components) = parse("''"); | ||
122 | assert!(has_closing_quote, "char should have closing quote"); | ||
123 | assert!(components.len() == 0); | ||
124 | } | ||
125 | |||
126 | #[test] | ||
127 | fn test_unclosed_char() { | ||
128 | let component = unclosed_char_component("'a"); | ||
129 | assert!(component.kind == CodePoint); | ||
130 | assert!(component.range == TextRange::from_to(1.into(), 2.into())); | ||
131 | } | ||
132 | |||
133 | #[test] | ||
134 | fn test_digit_escapes() { | ||
135 | let literals = &[r"", r"5", r"55"]; | ||
136 | |||
137 | for literal in literals { | ||
138 | let lit_text = format!(r"'\x{}'", literal); | ||
139 | let component = closed_char_component(&lit_text); | ||
140 | assert!(component.kind == AsciiCodeEscape); | ||
141 | assert!(component.range == range_closed(&lit_text)); | ||
142 | } | ||
143 | |||
144 | // More than 2 digits starts a new codepoint | ||
145 | let components = closed_char_components(r"'\x555'"); | ||
146 | assert!(components.len() == 2); | ||
147 | assert!(components[1].kind == CodePoint); | ||
148 | } | ||
149 | |||
150 | #[test] | ||
151 | fn test_ascii_escapes() { | ||
152 | let literals = &[ | ||
153 | r"\'", "\\\"", // equivalent to \" | ||
154 | r"\n", r"\r", r"\t", r"\\", r"\0", | ||
155 | ]; | ||
156 | |||
157 | for literal in literals { | ||
158 | let lit_text = format!("'{}'", literal); | ||
159 | let component = closed_char_component(&lit_text); | ||
160 | assert!(component.kind == AsciiEscape); | ||
161 | assert!(component.range == range_closed(&lit_text)); | ||
162 | } | ||
163 | } | ||
164 | |||
165 | #[test] | ||
166 | fn test_no_escapes() { | ||
167 | let literals = &['"', 'n', 'r', 't', '0', 'x', 'u']; | ||
168 | |||
169 | for &literal in literals { | ||
170 | let lit_text = format!("'{}'", literal); | ||
171 | let component = closed_char_component(&lit_text); | ||
172 | assert!(component.kind == CodePoint); | ||
173 | assert!(component.range == range_closed(&lit_text)); | ||
174 | } | ||
175 | } | ||
176 | } | ||