diff options
Diffstat (limited to 'crates/ra_syntax/src/string_lexing')
-rw-r--r-- | crates/ra_syntax/src/string_lexing/byte.rs | 51 | ||||
-rw-r--r-- | crates/ra_syntax/src/string_lexing/byte_string.rs | 51 | ||||
-rw-r--r-- | crates/ra_syntax/src/string_lexing/char.rs | 176 | ||||
-rw-r--r-- | crates/ra_syntax/src/string_lexing/parser.rs | 92 | ||||
-rw-r--r-- | crates/ra_syntax/src/string_lexing/string.rs | 197 |
5 files changed, 221 insertions, 346 deletions
diff --git a/crates/ra_syntax/src/string_lexing/byte.rs b/crates/ra_syntax/src/string_lexing/byte.rs deleted file mode 100644 index b3228d6ca..000000000 --- a/crates/ra_syntax/src/string_lexing/byte.rs +++ /dev/null | |||
@@ -1,51 +0,0 @@ | |||
1 | use super::parser::Parser; | ||
2 | use super::CharComponent; | ||
3 | |||
4 | pub fn parse_byte_literal(src: &str) -> ByteComponentIterator { | ||
5 | ByteComponentIterator { | ||
6 | parser: Parser::new(src), | ||
7 | has_closing_quote: false, | ||
8 | } | ||
9 | } | ||
10 | |||
11 | pub struct ByteComponentIterator<'a> { | ||
12 | parser: Parser<'a>, | ||
13 | pub has_closing_quote: bool, | ||
14 | } | ||
15 | |||
16 | impl<'a> Iterator for ByteComponentIterator<'a> { | ||
17 | type Item = CharComponent; | ||
18 | fn next(&mut self) -> Option<CharComponent> { | ||
19 | if self.parser.pos == 0 { | ||
20 | assert!( | ||
21 | self.parser.advance() == 'b', | ||
22 | "Byte literal should start with a `b`" | ||
23 | ); | ||
24 | |||
25 | assert!( | ||
26 | self.parser.advance() == '\'', | ||
27 | "Byte literal should start with a `b`, followed by a quote" | ||
28 | ); | ||
29 | } | ||
30 | |||
31 | if let Some(component) = self.parser.parse_char_component() { | ||
32 | return Some(component); | ||
33 | } | ||
34 | |||
35 | // We get here when there are no char components left to parse | ||
36 | if self.parser.peek() == Some('\'') { | ||
37 | self.parser.advance(); | ||
38 | self.has_closing_quote = true; | ||
39 | } | ||
40 | |||
41 | assert!( | ||
42 | self.parser.peek() == None, | ||
43 | "byte literal should leave no unparsed input: src = {:?}, pos = {}, length = {}", | ||
44 | self.parser.src, | ||
45 | self.parser.pos, | ||
46 | self.parser.src.len() | ||
47 | ); | ||
48 | |||
49 | None | ||
50 | } | ||
51 | } | ||
diff --git a/crates/ra_syntax/src/string_lexing/byte_string.rs b/crates/ra_syntax/src/string_lexing/byte_string.rs deleted file mode 100644 index a6056159b..000000000 --- a/crates/ra_syntax/src/string_lexing/byte_string.rs +++ /dev/null | |||
@@ -1,51 +0,0 @@ | |||
1 | use super::parser::Parser; | ||
2 | use super::StringComponent; | ||
3 | |||
4 | pub fn parse_byte_string_literal(src: &str) -> ByteStringComponentIterator { | ||
5 | ByteStringComponentIterator { | ||
6 | parser: Parser::new(src), | ||
7 | has_closing_quote: false, | ||
8 | } | ||
9 | } | ||
10 | |||
11 | pub struct ByteStringComponentIterator<'a> { | ||
12 | parser: Parser<'a>, | ||
13 | pub has_closing_quote: bool, | ||
14 | } | ||
15 | |||
16 | impl<'a> Iterator for ByteStringComponentIterator<'a> { | ||
17 | type Item = StringComponent; | ||
18 | fn next(&mut self) -> Option<StringComponent> { | ||
19 | if self.parser.pos == 0 { | ||
20 | assert!( | ||
21 | self.parser.advance() == 'b', | ||
22 | "byte string literal should start with a `b`" | ||
23 | ); | ||
24 | |||
25 | assert!( | ||
26 | self.parser.advance() == '"', | ||
27 | "byte string literal should start with a `b`, followed by double quotes" | ||
28 | ); | ||
29 | } | ||
30 | |||
31 | if let Some(component) = self.parser.parse_string_component() { | ||
32 | return Some(component); | ||
33 | } | ||
34 | |||
35 | // We get here when there are no char components left to parse | ||
36 | if self.parser.peek() == Some('"') { | ||
37 | self.parser.advance(); | ||
38 | self.has_closing_quote = true; | ||
39 | } | ||
40 | |||
41 | assert!( | ||
42 | self.parser.peek() == None, | ||
43 | "byte string literal should leave no unparsed input: src = {:?}, pos = {}, length = {}", | ||
44 | self.parser.src, | ||
45 | self.parser.pos, | ||
46 | self.parser.src.len() | ||
47 | ); | ||
48 | |||
49 | None | ||
50 | } | ||
51 | } | ||
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 | } | ||
diff --git a/crates/ra_syntax/src/string_lexing/parser.rs b/crates/ra_syntax/src/string_lexing/parser.rs index 4a6d5bc93..14c6015c2 100644 --- a/crates/ra_syntax/src/string_lexing/parser.rs +++ b/crates/ra_syntax/src/string_lexing/parser.rs | |||
@@ -1,15 +1,16 @@ | |||
1 | use rowan::{TextRange, TextUnit}; | 1 | use rowan::{TextRange, TextUnit}; |
2 | 2 | ||
3 | use self::CharComponentKind::*; | 3 | use self::StringComponentKind::*; |
4 | 4 | ||
5 | pub struct Parser<'a> { | 5 | pub struct Parser<'a> { |
6 | pub(super) quote: u8, | ||
6 | pub(super) src: &'a str, | 7 | pub(super) src: &'a str, |
7 | pub(super) pos: usize, | 8 | pub(super) pos: usize, |
8 | } | 9 | } |
9 | 10 | ||
10 | impl<'a> Parser<'a> { | 11 | impl<'a> Parser<'a> { |
11 | pub fn new(src: &'a str) -> Parser<'a> { | 12 | pub fn new(src: &'a str, quote: u8) -> Parser<'a> { |
12 | Parser { src, pos: 0 } | 13 | Parser { quote, src, pos: 0 } |
13 | } | 14 | } |
14 | 15 | ||
15 | // Utility methods | 16 | // Utility methods |
@@ -42,7 +43,7 @@ impl<'a> Parser<'a> { | |||
42 | 43 | ||
43 | // Char parsing methods | 44 | // Char parsing methods |
44 | 45 | ||
45 | fn parse_unicode_escape(&mut self, start: TextUnit) -> CharComponent { | 46 | fn parse_unicode_escape(&mut self, start: TextUnit) -> StringComponent { |
46 | match self.peek() { | 47 | match self.peek() { |
47 | Some('{') => { | 48 | Some('{') => { |
48 | self.advance(); | 49 | self.advance(); |
@@ -56,16 +57,16 @@ impl<'a> Parser<'a> { | |||
56 | } | 57 | } |
57 | 58 | ||
58 | let end = self.get_pos(); | 59 | let end = self.get_pos(); |
59 | CharComponent::new(TextRange::from_to(start, end), UnicodeEscape) | 60 | StringComponent::new(TextRange::from_to(start, end), UnicodeEscape) |
60 | } | 61 | } |
61 | Some(_) | None => { | 62 | Some(_) | None => { |
62 | let end = self.get_pos(); | 63 | let end = self.get_pos(); |
63 | CharComponent::new(TextRange::from_to(start, end), UnicodeEscape) | 64 | StringComponent::new(TextRange::from_to(start, end), UnicodeEscape) |
64 | } | 65 | } |
65 | } | 66 | } |
66 | } | 67 | } |
67 | 68 | ||
68 | fn parse_ascii_code_escape(&mut self, start: TextUnit) -> CharComponent { | 69 | fn parse_ascii_code_escape(&mut self, start: TextUnit) -> StringComponent { |
69 | let code_start = self.get_pos(); | 70 | let code_start = self.get_pos(); |
70 | while let Some(next) = self.peek() { | 71 | while let Some(next) = self.peek() { |
71 | if next == '\'' || (self.get_pos() - code_start == 2.into()) { | 72 | if next == '\'' || (self.get_pos() - code_start == 2.into()) { |
@@ -76,12 +77,12 @@ impl<'a> Parser<'a> { | |||
76 | } | 77 | } |
77 | 78 | ||
78 | let end = self.get_pos(); | 79 | let end = self.get_pos(); |
79 | CharComponent::new(TextRange::from_to(start, end), AsciiCodeEscape) | 80 | StringComponent::new(TextRange::from_to(start, end), AsciiCodeEscape) |
80 | } | 81 | } |
81 | 82 | ||
82 | fn parse_escape(&mut self, start: TextUnit) -> CharComponent { | 83 | fn parse_escape(&mut self, start: TextUnit) -> StringComponent { |
83 | if self.peek().is_none() { | 84 | if self.peek().is_none() { |
84 | return CharComponent::new(TextRange::from_to(start, start), AsciiEscape); | 85 | return StringComponent::new(TextRange::from_to(start, start), AsciiEscape); |
85 | } | 86 | } |
86 | 87 | ||
87 | let next = self.advance(); | 88 | let next = self.advance(); |
@@ -90,29 +91,7 @@ impl<'a> Parser<'a> { | |||
90 | match next { | 91 | match next { |
91 | 'x' => self.parse_ascii_code_escape(start), | 92 | 'x' => self.parse_ascii_code_escape(start), |
92 | 'u' => self.parse_unicode_escape(start), | 93 | 'u' => self.parse_unicode_escape(start), |
93 | _ => CharComponent::new(range, AsciiEscape), | 94 | _ => StringComponent::new(range, AsciiEscape), |
94 | } | ||
95 | } | ||
96 | |||
97 | pub fn parse_char_component(&mut self) -> Option<CharComponent> { | ||
98 | let next = self.peek()?; | ||
99 | |||
100 | // Ignore character close | ||
101 | if next == '\'' { | ||
102 | return None; | ||
103 | } | ||
104 | |||
105 | let start = self.get_pos(); | ||
106 | self.advance(); | ||
107 | |||
108 | if next == '\\' { | ||
109 | Some(self.parse_escape(start)) | ||
110 | } else { | ||
111 | let end = self.get_pos(); | ||
112 | Some(CharComponent::new( | ||
113 | TextRange::from_to(start, end), | ||
114 | CodePoint, | ||
115 | )) | ||
116 | } | 95 | } |
117 | } | 96 | } |
118 | 97 | ||
@@ -131,11 +110,11 @@ impl<'a> Parser<'a> { | |||
131 | } | 110 | } |
132 | } | 111 | } |
133 | 112 | ||
134 | pub fn parse_string_component(&mut self) -> Option<StringComponent> { | 113 | pub fn parse_component(&mut self) -> Option<StringComponent> { |
135 | let next = self.peek()?; | 114 | let next = self.peek()?; |
136 | 115 | ||
137 | // Ignore string close | 116 | // Ignore string close |
138 | if next == '"' { | 117 | if next == self.quote as char { |
139 | return None; | 118 | return None; |
140 | } | 119 | } |
141 | 120 | ||
@@ -145,21 +124,31 @@ impl<'a> Parser<'a> { | |||
145 | if next == '\\' { | 124 | if next == '\\' { |
146 | // Strings can use `\` to ignore newlines, so we first try to parse one of those | 125 | // Strings can use `\` to ignore newlines, so we first try to parse one of those |
147 | // before falling back to parsing char escapes | 126 | // before falling back to parsing char escapes |
148 | self.parse_ignore_newline(start).or_else(|| { | 127 | if self.quote == b'"' { |
149 | let char_component = self.parse_escape(start); | 128 | if let Some(component) = self.parse_ignore_newline(start) { |
150 | Some(StringComponent::new( | 129 | return Some(component); |
151 | char_component.range, | 130 | } |
152 | StringComponentKind::Char(char_component.kind), | 131 | } |
153 | )) | 132 | |
154 | }) | 133 | Some(self.parse_escape(start)) |
155 | } else { | 134 | } else { |
156 | let end = self.get_pos(); | 135 | let end = self.get_pos(); |
157 | Some(StringComponent::new( | 136 | Some(StringComponent::new( |
158 | TextRange::from_to(start, end), | 137 | TextRange::from_to(start, end), |
159 | StringComponentKind::Char(CodePoint), | 138 | CodePoint, |
160 | )) | 139 | )) |
161 | } | 140 | } |
162 | } | 141 | } |
142 | |||
143 | pub fn parse_suffix(&mut self) -> Option<TextRange> { | ||
144 | let start = self.get_pos(); | ||
145 | let _ = self.peek()?; | ||
146 | while let Some(_) = self.peek() { | ||
147 | self.advance(); | ||
148 | } | ||
149 | let end = self.get_pos(); | ||
150 | Some(TextRange::from_to(start, end)) | ||
151 | } | ||
163 | } | 152 | } |
164 | 153 | ||
165 | #[derive(Debug, Eq, PartialEq, Clone)] | 154 | #[derive(Debug, Eq, PartialEq, Clone)] |
@@ -177,23 +166,6 @@ impl StringComponent { | |||
177 | #[derive(Debug, Eq, PartialEq, Clone)] | 166 | #[derive(Debug, Eq, PartialEq, Clone)] |
178 | pub enum StringComponentKind { | 167 | pub enum StringComponentKind { |
179 | IgnoreNewline, | 168 | IgnoreNewline, |
180 | Char(CharComponentKind), | ||
181 | } | ||
182 | |||
183 | #[derive(Debug, Eq, PartialEq, Clone)] | ||
184 | pub struct CharComponent { | ||
185 | pub range: TextRange, | ||
186 | pub kind: CharComponentKind, | ||
187 | } | ||
188 | |||
189 | impl CharComponent { | ||
190 | fn new(range: TextRange, kind: CharComponentKind) -> CharComponent { | ||
191 | CharComponent { range, kind } | ||
192 | } | ||
193 | } | ||
194 | |||
195 | #[derive(Debug, Eq, PartialEq, Clone)] | ||
196 | pub enum CharComponentKind { | ||
197 | CodePoint, | 169 | CodePoint, |
198 | AsciiEscape, | 170 | AsciiEscape, |
199 | AsciiCodeEscape, | 171 | AsciiCodeEscape, |
diff --git a/crates/ra_syntax/src/string_lexing/string.rs b/crates/ra_syntax/src/string_lexing/string.rs index d8351e9af..064f08544 100644 --- a/crates/ra_syntax/src/string_lexing/string.rs +++ b/crates/ra_syntax/src/string_lexing/string.rs | |||
@@ -1,41 +1,92 @@ | |||
1 | use super::parser::Parser; | 1 | use crate::{ |
2 | use super::StringComponent; | 2 | TextRange, |
3 | string_lexing::{ | ||
4 | parser::Parser, | ||
5 | StringComponent, | ||
6 | }}; | ||
3 | 7 | ||
4 | pub fn parse_string_literal(src: &str) -> StringComponentIterator { | 8 | pub fn parse_string_literal(src: &str) -> StringComponentIterator { |
5 | StringComponentIterator { | 9 | StringComponentIterator { |
6 | parser: Parser::new(src), | 10 | parser: Parser::new(src, b'"'), |
7 | has_closing_quote: false, | 11 | has_closing_quote: false, |
12 | suffix: None, | ||
13 | prefix: None, | ||
14 | quote: b'"', | ||
15 | } | ||
16 | } | ||
17 | |||
18 | pub fn parse_byte_string_literal(src: &str) -> StringComponentIterator { | ||
19 | StringComponentIterator { | ||
20 | parser: Parser::new(src, b'"'), | ||
21 | has_closing_quote: false, | ||
22 | suffix: None, | ||
23 | prefix: Some(b'b'), | ||
24 | quote: b'"', | ||
25 | } | ||
26 | } | ||
27 | |||
28 | pub fn parse_char_literal(src: &str) -> StringComponentIterator { | ||
29 | StringComponentIterator { | ||
30 | parser: Parser::new(src, b'\''), | ||
31 | has_closing_quote: false, | ||
32 | suffix: None, | ||
33 | prefix: None, | ||
34 | quote: b'\'', | ||
35 | } | ||
36 | } | ||
37 | |||
38 | pub fn parse_byte_literal(src: &str) -> StringComponentIterator { | ||
39 | StringComponentIterator { | ||
40 | parser: Parser::new(src, b'\''), | ||
41 | has_closing_quote: false, | ||
42 | suffix: None, | ||
43 | prefix: Some(b'b'), | ||
44 | quote: b'\'', | ||
8 | } | 45 | } |
9 | } | 46 | } |
10 | 47 | ||
11 | pub struct StringComponentIterator<'a> { | 48 | pub struct StringComponentIterator<'a> { |
12 | parser: Parser<'a>, | 49 | parser: Parser<'a>, |
13 | pub has_closing_quote: bool, | 50 | pub has_closing_quote: bool, |
51 | pub suffix: Option<TextRange>, | ||
52 | prefix: Option<u8>, | ||
53 | quote: u8, | ||
14 | } | 54 | } |
15 | 55 | ||
16 | impl<'a> Iterator for StringComponentIterator<'a> { | 56 | impl<'a> Iterator for StringComponentIterator<'a> { |
17 | type Item = StringComponent; | 57 | type Item = StringComponent; |
18 | fn next(&mut self) -> Option<StringComponent> { | 58 | fn next(&mut self) -> Option<StringComponent> { |
19 | if self.parser.pos == 0 { | 59 | if self.parser.pos == 0 { |
60 | if let Some(prefix) = self.prefix { | ||
61 | assert!( | ||
62 | self.parser.advance() == prefix as char, | ||
63 | "literal should start with a {:?}", | ||
64 | prefix as char, | ||
65 | ); | ||
66 | } | ||
20 | assert!( | 67 | assert!( |
21 | self.parser.advance() == '"', | 68 | self.parser.advance() == self.quote as char, |
22 | "string literal should start with double quotes" | 69 | "literal should start with a {:?}", |
70 | self.quote as char, | ||
23 | ); | 71 | ); |
24 | } | 72 | } |
25 | 73 | ||
26 | if let Some(component) = self.parser.parse_string_component() { | 74 | if let Some(component) = self.parser.parse_component() { |
27 | return Some(component); | 75 | return Some(component); |
28 | } | 76 | } |
29 | 77 | ||
30 | // We get here when there are no char components left to parse | 78 | // We get here when there are no char components left to parse |
31 | if self.parser.peek() == Some('"') { | 79 | if self.parser.peek() == Some(self.quote as char) { |
32 | self.parser.advance(); | 80 | self.parser.advance(); |
33 | self.has_closing_quote = true; | 81 | self.has_closing_quote = true; |
82 | if let Some(range) = self.parser.parse_suffix() { | ||
83 | self.suffix = Some(range); | ||
84 | } | ||
34 | } | 85 | } |
35 | 86 | ||
36 | assert!( | 87 | assert!( |
37 | self.parser.peek() == None, | 88 | self.parser.peek() == None, |
38 | "string literal should leave no unparsed input: src = {:?}, pos = {}, length = {}", | 89 | "literal should leave no unparsed input: src = {:?}, pos = {}, length = {}", |
39 | self.parser.src, | 90 | self.parser.src, |
40 | self.parser.pos, | 91 | self.parser.pos, |
41 | self.parser.src.len() | 92 | self.parser.src.len() |
@@ -44,3 +95,133 @@ impl<'a> Iterator for StringComponentIterator<'a> { | |||
44 | None | 95 | None |
45 | } | 96 | } |
46 | } | 97 | } |
98 | |||
99 | #[cfg(test)] | ||
100 | mod tests { | ||
101 | use rowan::TextRange; | ||
102 | use crate::string_lexing::{ | ||
103 | StringComponent, | ||
104 | StringComponentKind::*, | ||
105 | }; | ||
106 | |||
107 | fn parse(src: &str) -> (bool, Vec<StringComponent>) { | ||
108 | let component_iterator = &mut super::parse_char_literal(src); | ||
109 | let components: Vec<_> = component_iterator.collect(); | ||
110 | (component_iterator.has_closing_quote, components) | ||
111 | } | ||
112 | |||
113 | fn unclosed_char_component(src: &str) -> StringComponent { | ||
114 | let (has_closing_quote, components) = parse(src); | ||
115 | assert!(!has_closing_quote, "char should not have closing quote"); | ||
116 | assert!(components.len() == 1); | ||
117 | components[0].clone() | ||
118 | } | ||
119 | |||
120 | fn closed_char_component(src: &str) -> StringComponent { | ||
121 | let (has_closing_quote, components) = parse(src); | ||
122 | assert!(has_closing_quote, "char should have closing quote"); | ||
123 | assert!( | ||
124 | components.len() == 1, | ||
125 | "Literal: {}\nComponents: {:#?}", | ||
126 | src, | ||
127 | components | ||
128 | ); | ||
129 | components[0].clone() | ||
130 | } | ||
131 | |||
132 | fn closed_char_components(src: &str) -> Vec<StringComponent> { | ||
133 | let (has_closing_quote, components) = parse(src); | ||
134 | assert!(has_closing_quote, "char should have closing quote"); | ||
135 | components | ||
136 | } | ||
137 | |||
138 | fn range_closed(src: &str) -> TextRange { | ||
139 | TextRange::from_to(1.into(), (src.len() as u32 - 1).into()) | ||
140 | } | ||
141 | |||
142 | fn range_unclosed(src: &str) -> TextRange { | ||
143 | TextRange::from_to(1.into(), (src.len() as u32).into()) | ||
144 | } | ||
145 | |||
146 | #[test] | ||
147 | fn test_unicode_escapes() { | ||
148 | let unicode_escapes = &[r"{DEAD}", "{BEEF}", "{FF}", "{}", ""]; | ||
149 | for escape in unicode_escapes { | ||
150 | let escape_sequence = format!(r"'\u{}'", escape); | ||
151 | let component = closed_char_component(&escape_sequence); | ||
152 | let expected_range = range_closed(&escape_sequence); | ||
153 | assert_eq!(component.kind, UnicodeEscape); | ||
154 | assert_eq!(component.range, expected_range); | ||
155 | } | ||
156 | } | ||
157 | |||
158 | #[test] | ||
159 | fn test_unicode_escapes_unclosed() { | ||
160 | let unicode_escapes = &["{DEAD", "{BEEF", "{FF"]; | ||
161 | for escape in unicode_escapes { | ||
162 | let escape_sequence = format!(r"'\u{}'", escape); | ||
163 | let component = unclosed_char_component(&escape_sequence); | ||
164 | let expected_range = range_unclosed(&escape_sequence); | ||
165 | assert_eq!(component.kind, UnicodeEscape); | ||
166 | assert_eq!(component.range, expected_range); | ||
167 | } | ||
168 | } | ||
169 | |||
170 | #[test] | ||
171 | fn test_empty_char() { | ||
172 | let (has_closing_quote, components) = parse("''"); | ||
173 | assert!(has_closing_quote, "char should have closing quote"); | ||
174 | assert!(components.len() == 0); | ||
175 | } | ||
176 | |||
177 | #[test] | ||
178 | fn test_unclosed_char() { | ||
179 | let component = unclosed_char_component("'a"); | ||
180 | assert!(component.kind == CodePoint); | ||
181 | assert!(component.range == TextRange::from_to(1.into(), 2.into())); | ||
182 | } | ||
183 | |||
184 | #[test] | ||
185 | fn test_digit_escapes() { | ||
186 | let literals = &[r"", r"5", r"55"]; | ||
187 | |||
188 | for literal in literals { | ||
189 | let lit_text = format!(r"'\x{}'", literal); | ||
190 | let component = closed_char_component(&lit_text); | ||
191 | assert!(component.kind == AsciiCodeEscape); | ||
192 | assert!(component.range == range_closed(&lit_text)); | ||
193 | } | ||
194 | |||
195 | // More than 2 digits starts a new codepoint | ||
196 | let components = closed_char_components(r"'\x555'"); | ||
197 | assert!(components.len() == 2); | ||
198 | assert!(components[1].kind == CodePoint); | ||
199 | } | ||
200 | |||
201 | #[test] | ||
202 | fn test_ascii_escapes() { | ||
203 | let literals = &[ | ||
204 | r"\'", "\\\"", // equivalent to \" | ||
205 | r"\n", r"\r", r"\t", r"\\", r"\0", | ||
206 | ]; | ||
207 | |||
208 | for literal in literals { | ||
209 | let lit_text = format!("'{}'", literal); | ||
210 | let component = closed_char_component(&lit_text); | ||
211 | assert!(component.kind == AsciiEscape); | ||
212 | assert!(component.range == range_closed(&lit_text)); | ||
213 | } | ||
214 | } | ||
215 | |||
216 | #[test] | ||
217 | fn test_no_escapes() { | ||
218 | let literals = &['"', 'n', 'r', 't', '0', 'x', 'u']; | ||
219 | |||
220 | for &literal in literals { | ||
221 | let lit_text = format!("'{}'", literal); | ||
222 | let component = closed_char_component(&lit_text); | ||
223 | assert!(component.kind == CodePoint); | ||
224 | assert!(component.range == range_closed(&lit_text)); | ||
225 | } | ||
226 | } | ||
227 | } | ||