aboutsummaryrefslogtreecommitdiffstats
path: root/askama_parser/src/lib.rs
blob: e7c387bfa6dc527be6c92d3fd8ef1ab939eff2cb (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
#![deny(unreachable_pub)]
#![deny(elided_lifetimes_in_paths)]

use std::cell::Cell;
use std::{fmt, str};

use nom::branch::alt;
use nom::bytes::complete::{escaped, is_not, tag, take_till};
use nom::character::complete::char;
use nom::character::complete::{anychar, digit1};
use nom::combinator::{eof, map, not, opt, recognize, value};
use nom::error::ErrorKind;
use nom::multi::separated_list1;
use nom::sequence::{delimited, pair, tuple};
use nom::{error_position, AsChar, IResult, InputTakeAtPosition};

pub use self::expr::Expr;
pub use self::node::{Cond, CondTest, Loop, Macro, Node, Target, When, Whitespace, Ws};

mod expr;
mod node;
#[cfg(test)]
mod tests;

struct State<'a> {
    syntax: &'a Syntax<'a>,
    loop_depth: Cell<usize>,
}

impl<'a> State<'a> {
    fn new(syntax: &'a Syntax<'a>) -> State<'a> {
        State {
            syntax,
            loop_depth: Cell::new(0),
        }
    }

    fn enter_loop(&self) {
        self.loop_depth.set(self.loop_depth.get() + 1);
    }

    fn leave_loop(&self) {
        self.loop_depth.set(self.loop_depth.get() - 1);
    }

    fn is_in_loop(&self) -> bool {
        self.loop_depth.get() > 0
    }
}

impl From<char> for Whitespace {
    fn from(c: char) -> Self {
        match c {
            '+' => Self::Preserve,
            '-' => Self::Suppress,
            '~' => Self::Minimize,
            _ => panic!("unsupported `Whitespace` conversion"),
        }
    }
}

mod _parsed {
    use std::mem;

    use super::{parse, Node, ParseError, Syntax};

    pub struct Parsed {
        #[allow(dead_code)]
        source: String,
        nodes: Vec<Node<'static>>,
    }

    impl Parsed {
        pub fn new(source: String, syntax: &Syntax<'_>) -> Result<Self, ParseError> {
            // Self-referential borrowing: `self` will keep the source alive as `String`,
            // internally we will transmute it to `&'static str` to satisfy the compiler.
            // However, we only expose the nodes with a lifetime limited to `self`.
            let src = unsafe { mem::transmute::<&str, &'static str>(source.as_str()) };
            let nodes = match parse(src, syntax) {
                Ok(nodes) => nodes,
                Err(e) => return Err(e),
            };

            Ok(Self { source, nodes })
        }

        // The return value's lifetime must be limited to `self` to uphold the unsafe invariant.
        pub fn nodes(&self) -> &[Node<'_>] {
            &self.nodes
        }
    }
}

pub use _parsed::Parsed;

pub fn parse<'a>(src: &'a str, syntax: &Syntax<'_>) -> Result<Vec<Node<'a>>, ParseError> {
    match Node::parse(src, &State::new(syntax)) {
        Ok((left, res)) => {
            if !left.is_empty() {
                Err(ParseError(format!("unable to parse template:\n\n{left:?}")))
            } else {
                Ok(res)
            }
        }

        Err(nom::Err::Error(err)) | Err(nom::Err::Failure(err)) => {
            let nom::error::Error { input, .. } = err;
            let offset = src.len() - input.len();
            let (source_before, source_after) = src.split_at(offset);

            let source_after = match source_after.char_indices().enumerate().take(41).last() {
                Some((40, (i, _))) => format!("{:?}...", &source_after[..i]),
                _ => format!("{source_after:?}"),
            };

            let (row, last_line) = source_before.lines().enumerate().last().unwrap();
            let column = last_line.chars().count();

            let msg = format!(
                "problems parsing template source at row {}, column {} near:\n{}",
                row + 1,
                column,
                source_after,
            );

            Err(ParseError(msg))
        }

        Err(nom::Err::Incomplete(_)) => Err(ParseError("parsing incomplete".into())),
    }
}

#[derive(Debug, Clone, PartialEq, Eq)]
pub struct ParseError(String);

impl std::error::Error for ParseError {}

impl fmt::Display for ParseError {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        self.0.fmt(f)
    }
}

fn is_ws(c: char) -> bool {
    matches!(c, ' ' | '\t' | '\r' | '\n')
}

fn not_ws(c: char) -> bool {
    !is_ws(c)
}

fn ws<'a, O>(
    inner: impl FnMut(&'a str) -> IResult<&'a str, O>,
) -> impl FnMut(&'a str) -> IResult<&'a str, O> {
    delimited(take_till(not_ws), inner, take_till(not_ws))
}

fn split_ws_parts(s: &str) -> Node<'_> {
    let trimmed_start = s.trim_start_matches(is_ws);
    let len_start = s.len() - trimmed_start.len();
    let trimmed = trimmed_start.trim_end_matches(is_ws);
    Node::Lit(&s[..len_start], trimmed, &trimmed_start[trimmed.len()..])
}

/// Skips input until `end` was found, but does not consume it.
/// Returns tuple that would be returned when parsing `end`.
fn skip_till<'a, O>(
    end: impl FnMut(&'a str) -> IResult<&'a str, O>,
) -> impl FnMut(&'a str) -> IResult<&'a str, (&'a str, O)> {
    enum Next<O> {
        IsEnd(O),
        NotEnd(char),
    }
    let mut next = alt((map(end, Next::IsEnd), map(anychar, Next::NotEnd)));
    move |start: &'a str| {
        let mut i = start;
        loop {
            let (j, is_end) = next(i)?;
            match is_end {
                Next::IsEnd(lookahead) => return Ok((i, (j, lookahead))),
                Next::NotEnd(_) => i = j,
            }
        }
    }
}

fn keyword<'a>(k: &'a str) -> impl FnMut(&'a str) -> IResult<&'a str, &'a str> {
    move |i: &'a str| -> IResult<&'a str, &'a str> {
        let (j, v) = identifier(i)?;
        if k == v {
            Ok((j, v))
        } else {
            Err(nom::Err::Error(error_position!(i, ErrorKind::Tag)))
        }
    }
}

fn identifier(input: &str) -> IResult<&str, &str> {
    fn start(s: &str) -> IResult<&str, &str> {
        s.split_at_position1_complete(
            |c| !(c.is_alpha() || c == '_' || c >= '\u{0080}'),
            nom::error::ErrorKind::Alpha,
        )
    }

    fn tail(s: &str) -> IResult<&str, &str> {
        s.split_at_position1_complete(
            |c| !(c.is_alphanum() || c == '_' || c >= '\u{0080}'),
            nom::error::ErrorKind::Alpha,
        )
    }

    recognize(pair(start, opt(tail)))(input)
}

fn bool_lit(i: &str) -> IResult<&str, &str> {
    alt((keyword("false"), keyword("true")))(i)
}

fn num_lit(i: &str) -> IResult<&str, &str> {
    recognize(pair(digit1, opt(pair(char('.'), digit1))))(i)
}

fn str_lit(i: &str) -> IResult<&str, &str> {
    let (i, s) = delimited(
        char('"'),
        opt(escaped(is_not("\\\""), '\\', anychar)),
        char('"'),
    )(i)?;
    Ok((i, s.unwrap_or_default()))
}

fn char_lit(i: &str) -> IResult<&str, &str> {
    let (i, s) = delimited(
        char('\''),
        opt(escaped(is_not("\\\'"), '\\', anychar)),
        char('\''),
    )(i)?;
    Ok((i, s.unwrap_or_default()))
}

fn nested_parenthesis(i: &str) -> IResult<&str, ()> {
    let mut nested = 0;
    let mut last = 0;
    let mut in_str = false;
    let mut escaped = false;

    for (i, b) in i.chars().enumerate() {
        if !(b == '(' || b == ')') || !in_str {
            match b {
                '(' => nested += 1,
                ')' => {
                    if nested == 0 {
                        last = i;
                        break;
                    }
                    nested -= 1;
                }
                '"' => {
                    if in_str {
                        if !escaped {
                            in_str = false;
                        }
                    } else {
                        in_str = true;
                    }
                }
                '\\' => {
                    escaped = !escaped;
                }
                _ => (),
            }
        }

        if escaped && b != '\\' {
            escaped = false;
        }
    }

    if nested == 0 {
        Ok((&i[last..], ()))
    } else {
        Err(nom::Err::Error(error_position!(
            i,
            ErrorKind::SeparatedNonEmptyList
        )))
    }
}

fn path(i: &str) -> IResult<&str, Vec<&str>> {
    let root = opt(value("", ws(tag("::"))));
    let tail = separated_list1(ws(tag("::")), identifier);

    match tuple((root, identifier, ws(tag("::")), tail))(i) {
        Ok((i, (root, start, _, rest))) => {
            let mut path = Vec::new();
            path.extend(root);
            path.push(start);
            path.extend(rest);
            Ok((i, path))
        }
        Err(err) => {
            if let Ok((i, name)) = identifier(i) {
                // The returned identifier can be assumed to be path if:
                // - Contains both a lowercase and uppercase character, i.e. a type name like `None`
                // - Doesn't contain any lowercase characters, i.e. it's a constant
                // In short, if it contains any uppercase characters it's a path.
                if name.contains(char::is_uppercase) {
                    return Ok((i, vec![name]));
                }
            }

            // If `identifier()` fails then just return the original error
            Err(err)
        }
    }
}

fn take_content<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, Node<'a>> {
    let p_start = alt((
        tag(s.syntax.block_start),
        tag(s.syntax.comment_start),
        tag(s.syntax.expr_start),
    ));

    let (i, _) = not(eof)(i)?;
    let (i, content) = opt(recognize(skip_till(p_start)))(i)?;
    let (i, content) = match content {
        Some("") => {
            // {block,comment,expr}_start follows immediately.
            return Err(nom::Err::Error(error_position!(i, ErrorKind::TakeUntil)));
        }
        Some(content) => (i, content),
        None => ("", i), // there is no {block,comment,expr}_start: take everything
    };
    Ok((i, split_ws_parts(content)))
}

fn tag_block_start<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, &'a str> {
    tag(s.syntax.block_start)(i)
}

fn tag_block_end<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, &'a str> {
    tag(s.syntax.block_end)(i)
}

fn tag_comment_start<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, &'a str> {
    tag(s.syntax.comment_start)(i)
}

fn tag_comment_end<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, &'a str> {
    tag(s.syntax.comment_end)(i)
}

fn tag_expr_start<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, &'a str> {
    tag(s.syntax.expr_start)(i)
}

fn tag_expr_end<'a>(i: &'a str, s: &State<'_>) -> IResult<&'a str, &'a str> {
    tag(s.syntax.expr_end)(i)
}

#[derive(Debug)]
pub struct Syntax<'a> {
    pub block_start: &'a str,
    pub block_end: &'a str,
    pub expr_start: &'a str,
    pub expr_end: &'a str,
    pub comment_start: &'a str,
    pub comment_end: &'a str,
}

impl Default for Syntax<'static> {
    fn default() -> Self {
        Self {
            block_start: "{%",
            block_end: "%}",
            expr_start: "{{",
            expr_end: "}}",
            comment_start: "{#",
            comment_end: "#}",
        }
    }
}