| 1 | // pest. The Elegant Parser |
| 2 | // Copyright (c) 2018 Dragoș Tiselice |
| 3 | // |
| 4 | // Licensed under the Apache License, Version 2.0 |
| 5 | // <LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0> or the MIT |
| 6 | // license <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your |
| 7 | // option. All files in the project carrying such notice may not be copied, |
| 8 | // modified, or distributed except according to those terms. |
| 9 | |
| 10 | use alloc::rc::Rc; |
| 11 | use alloc::vec::Vec; |
| 12 | use core::fmt; |
| 13 | use core::str; |
| 14 | |
| 15 | use super::queueable_token::QueueableToken; |
| 16 | use crate::position; |
| 17 | use crate::token::Token; |
| 18 | use crate::RuleType; |
| 19 | |
| 20 | /// An iterator over [`Token`]s. It is created by [`Pair::tokens`] and [`Pairs::tokens`]. |
| 21 | /// |
| 22 | /// [`Token`]: ../enum.Token.html |
| 23 | /// [`Pair::tokens`]: struct.Pair.html#method.tokens |
| 24 | /// [`Pairs::tokens`]: struct.Pairs.html#method.tokens |
| 25 | #[derive (Clone)] |
| 26 | pub struct Tokens<'i, R> { |
| 27 | queue: Rc<Vec<QueueableToken<'i, R>>>, |
| 28 | input: &'i str, |
| 29 | start: usize, |
| 30 | end: usize, |
| 31 | } |
| 32 | |
| 33 | pub fn new<'i, R: RuleType>( |
| 34 | queue: Rc<Vec<QueueableToken<'i, R>>>, |
| 35 | input: &'i str, |
| 36 | start: usize, |
| 37 | end: usize, |
| 38 | ) -> Tokens<'i, R> { |
| 39 | if cfg!(debug_assertions) { |
| 40 | for tok: &QueueableToken<'_, R> in queue.iter() { |
| 41 | match *tok { |
| 42 | QueueableToken::Start { input_pos: usize, .. } | QueueableToken::End { input_pos: usize, .. } => { |
| 43 | assert!( |
| 44 | input.get(input_pos..).is_some(), |
| 45 | "💥 INVALID `Tokens` CREATED 💥" |
| 46 | ) |
| 47 | } |
| 48 | } |
| 49 | } |
| 50 | } |
| 51 | |
| 52 | Tokens { |
| 53 | queue, |
| 54 | input, |
| 55 | start, |
| 56 | end, |
| 57 | } |
| 58 | } |
| 59 | |
| 60 | impl<'i, R: RuleType> Tokens<'i, R> { |
| 61 | fn create_token(&self, index: usize) -> Token<'i, R> { |
| 62 | match self.queue[index] { |
| 63 | QueueableToken::Start { |
| 64 | end_token_index, |
| 65 | input_pos, |
| 66 | } => { |
| 67 | let rule = match self.queue[end_token_index] { |
| 68 | QueueableToken::End { rule, .. } => rule, |
| 69 | _ => unreachable!(), |
| 70 | }; |
| 71 | |
| 72 | Token::Start { |
| 73 | rule, |
| 74 | pos: position::Position::new_internal(self.input, input_pos), |
| 75 | } |
| 76 | } |
| 77 | QueueableToken::End { |
| 78 | rule, input_pos, .. |
| 79 | } => Token::End { |
| 80 | rule, |
| 81 | pos: position::Position::new_internal(self.input, input_pos), |
| 82 | }, |
| 83 | } |
| 84 | } |
| 85 | } |
| 86 | |
| 87 | impl<'i, R: RuleType> ExactSizeIterator for Tokens<'i, R> { |
| 88 | fn len(&self) -> usize { |
| 89 | self.end - self.start |
| 90 | } |
| 91 | } |
| 92 | |
| 93 | impl<'i, R: RuleType> Iterator for Tokens<'i, R> { |
| 94 | type Item = Token<'i, R>; |
| 95 | |
| 96 | fn next(&mut self) -> Option<Self::Item> { |
| 97 | if self.start >= self.end { |
| 98 | return None; |
| 99 | } |
| 100 | |
| 101 | let token: Token<'_, R> = self.create_token(self.start); |
| 102 | |
| 103 | self.start += 1; |
| 104 | |
| 105 | Some(token) |
| 106 | } |
| 107 | |
| 108 | fn size_hint(&self) -> (usize, Option<usize>) { |
| 109 | let len: usize = <Self as ExactSizeIterator>::len(self); |
| 110 | (len, Some(len)) |
| 111 | } |
| 112 | } |
| 113 | |
| 114 | impl<'i, R: RuleType> DoubleEndedIterator for Tokens<'i, R> { |
| 115 | fn next_back(&mut self) -> Option<Self::Item> { |
| 116 | if self.end <= self.start { |
| 117 | return None; |
| 118 | } |
| 119 | |
| 120 | let token: Token<'_, R> = self.create_token(self.end - 1); |
| 121 | |
| 122 | self.end -= 1; |
| 123 | |
| 124 | Some(token) |
| 125 | } |
| 126 | } |
| 127 | |
| 128 | impl<'i, R: RuleType> fmt::Debug for Tokens<'i, R> { |
| 129 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
| 130 | f.debug_list().entries(self.clone()).finish() |
| 131 | } |
| 132 | } |
| 133 | |
| 134 | #[cfg (test)] |
| 135 | mod tests { |
| 136 | use super::super::super::macros::tests::*; |
| 137 | use super::super::super::Parser; |
| 138 | use super::Token; |
| 139 | use alloc::vec::Vec; |
| 140 | |
| 141 | #[test ] |
| 142 | fn double_ended_iter_for_tokens() { |
| 143 | let pairs = AbcParser::parse(Rule::a, "abcde" ).unwrap(); |
| 144 | let mut tokens = pairs.clone().tokens().collect::<Vec<Token<'_, Rule>>>(); |
| 145 | tokens.reverse(); |
| 146 | let reverse_tokens = pairs.tokens().rev().collect::<Vec<Token<'_, Rule>>>(); |
| 147 | assert_eq!(tokens, reverse_tokens); |
| 148 | } |
| 149 | |
| 150 | #[test ] |
| 151 | fn exact_size_iter_for_tokens() { |
| 152 | let tokens = AbcParser::parse(Rule::a, "abcde" ).unwrap().tokens(); |
| 153 | assert_eq!(tokens.len(), tokens.count()); |
| 154 | |
| 155 | let tokens = AbcParser::parse(Rule::a, "我很漂亮e" ).unwrap().tokens(); |
| 156 | assert_eq!(tokens.len(), tokens.count()); |
| 157 | |
| 158 | let tokens = AbcParser::parse(Rule::a, "abcde" ).unwrap().tokens().rev(); |
| 159 | assert_eq!(tokens.len(), tokens.count()); |
| 160 | |
| 161 | let mut tokens = AbcParser::parse(Rule::a, "abcde" ).unwrap().tokens(); |
| 162 | let tokens_len = tokens.len(); |
| 163 | let _ = tokens.next().unwrap(); |
| 164 | assert_eq!(tokens.count() + 1, tokens_len); |
| 165 | } |
| 166 | } |
| 167 | |