hjkl-engine 0.6.3

Vim FSM, motion grammar, and ex commands. Pre-1.0 churn.
Documentation
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
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
//! Public substitute command parser and applicator.
//!
//! Exposes [`parse_substitute`] and [`apply_substitute`] for the
//! `:[range]s/pattern/replacement/[flags]` ex command.
//!
//! ## Vim compatibility notes (v1 limitations)
//!
//! - Delimiter is **always `/`**. Alternate delimiters (`s|x|y|`,
//!   `s#x#y#`) are not supported. The parser returns an error when the
//!   first character after the keyword is not `/`.
//! - The `c` (confirm) flag is **parsed but silently ignored**. No
//!   interactive replacement. See vim's `:help :s_c` for what a full
//!   implementation looks like.
//! - The `\v` very-magic mode is not supported. The regex crate uses
//!   ERE syntax by default. Most ERE patterns work, but vim-specific
//!   extensions (`\<`, `\>`, `\s`, `\+`) may not. Use POSIX ERE
//!   equivalents or the `regex` crate's syntax.
//! - Capture-group references use vim notation (`\1`…`\9`, `&`); the
//!   parser translates them to `$1`…`$9`, `$0` for the `regex` crate.
//!
//! See vim's `:help :substitute` for the full spec.

use regex::Regex;

use crate::Editor;

/// Error type returned by [`parse_substitute`] and [`apply_substitute`].
pub type SubstError = String;

/// Parsed `:s/pattern/replacement/flags` command.
///
/// Produced by [`parse_substitute`]. Pass to [`apply_substitute`].
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct SubstituteCmd {
    /// The literal pattern string. `None` means "reuse `last_search`
    /// from the editor" (the user typed `:s//replacement/`).
    pub pattern: Option<String>,
    /// The replacement string in vim notation (`&`, `\1`…`\9`).
    /// Empty string deletes the match.
    pub replacement: String,
    /// Parsed flags.
    pub flags: SubstFlags,
}

/// Flags for the substitute command.
#[derive(Debug, Clone, Copy, PartialEq, Eq, Default)]
pub struct SubstFlags {
    /// `g` — replace all occurrences on each line (default: first only).
    pub all: bool,
    /// `i` — case-insensitive (overrides editor `ignorecase`).
    pub ignore_case: bool,
    /// `I` — case-sensitive (overrides editor `ignorecase`).
    pub case_sensitive: bool,
    /// `c` — confirm mode. **Parsed but ignored in v1.** Behaves as if
    /// not set; all matches are replaced without prompting. This is a
    /// known divergence from vim. See vim's `:help :s_c`.
    pub confirm: bool,
}

/// Result of [`apply_substitute`].
#[derive(Debug, Clone, Copy, PartialEq, Eq, Default)]
pub struct SubstituteOutcome {
    /// Total number of individual replacements made across all lines.
    pub replacements: usize,
    /// Number of lines that had at least one replacement.
    pub lines_changed: usize,
}

/// Parse the tail of a substitute command (everything after the leading
/// `s` / `substitute` keyword).
///
/// # Examples
///
/// ```
/// use hjkl_engine::substitute::parse_substitute;
///
/// let cmd = parse_substitute("/foo/bar/gi").unwrap();
/// assert_eq!(cmd.pattern.as_deref(), Some("foo"));
/// assert_eq!(cmd.replacement, "bar");
/// assert!(cmd.flags.all);
/// assert!(cmd.flags.ignore_case);
///
/// // Empty pattern — reuse last_search.
/// let cmd = parse_substitute("//bar/").unwrap();
/// assert!(cmd.pattern.is_none());
/// assert_eq!(cmd.replacement, "bar");
/// ```
///
/// # Errors
///
/// Returns an error when:
/// - `s` is not followed by `/` (no delimiter or alternate delimiter).
/// - The flag string contains an unknown character.
/// - The separator `/` is absent (less than two fields).
pub fn parse_substitute(s: &str) -> Result<SubstituteCmd, SubstError> {
    // Require leading `/`. Alternate delimiters are out of scope for v1.
    let rest = s
        .strip_prefix('/')
        .ok_or_else(|| format!("substitute: expected '/' delimiter, got {s:?}"))?;

    // Split on unescaped `/`, collecting at most 3 segments:
    // [pattern, replacement, flags?]
    let parts = split_on_slash(rest);

    if parts.len() < 2 {
        return Err("substitute needs /pattern/replacement/".into());
    }

    let raw_pattern = &parts[0];
    let raw_replacement = &parts[1];
    let raw_flags = parts.get(2).map(String::as_str).unwrap_or("");

    // Empty pattern → reuse last_search.
    let pattern = if raw_pattern.is_empty() {
        None
    } else {
        Some(raw_pattern.clone())
    };

    // Translate vim replacement notation to regex crate notation.
    let replacement = translate_replacement(raw_replacement);

    let mut flags = SubstFlags::default();
    for ch in raw_flags.chars() {
        match ch {
            'g' => flags.all = true,
            'i' => flags.ignore_case = true,
            'I' => flags.case_sensitive = true,
            'c' => flags.confirm = true, // parsed, silently ignored
            other => return Err(format!("unknown flag '{other}' in substitute")),
        }
    }

    Ok(SubstituteCmd {
        pattern,
        replacement,
        flags,
    })
}

/// Apply a parsed substitute command to `line_range` (0-based inclusive)
/// in the editor's buffer.
///
/// # Pattern resolution
///
/// If `cmd.pattern` is `None` (user typed `:s//rep/`), the editor's
/// `last_search()` is used. Returns an error with `"no previous regular
/// expression"` when both are empty.
///
/// # Case-sensitivity precedence
///
/// `flags.case_sensitive` wins over `flags.ignore_case`, which wins over
/// the editor's `settings().ignore_case`.
///
/// # Cursor
///
/// After a successful substitution the cursor is placed at column 0 of the
/// **last line that changed**, matching vim semantics. When no replacements
/// are made the cursor is left unchanged.
///
/// # Undo
///
/// One undo snapshot is pushed before the first edit. If no replacements
/// occur the snapshot is popped so the undo stack stays clean.
///
/// # Errors
///
/// Returns an error when pattern resolution fails or the regex is invalid.
pub fn apply_substitute<H: crate::types::Host>(
    ed: &mut Editor<hjkl_buffer::Buffer, H>,
    cmd: &SubstituteCmd,
    line_range: std::ops::RangeInclusive<u32>,
) -> Result<SubstituteOutcome, SubstError> {
    // Resolve pattern.
    let pattern_str: String = match &cmd.pattern {
        Some(p) => p.clone(),
        None => ed
            .last_search()
            .map(str::to_owned)
            .ok_or_else(|| "no previous regular expression".to_string())?,
    };

    // Case-sensitivity.
    let case_insensitive = if cmd.flags.case_sensitive {
        false
    } else if cmd.flags.ignore_case {
        true
    } else {
        ed.settings().ignore_case
    };

    let effective_pattern = if case_insensitive {
        format!("(?i){pattern_str}")
    } else {
        pattern_str.clone()
    };

    let regex = Regex::new(&effective_pattern).map_err(|e| format!("bad pattern: {e}"))?;

    ed.push_undo();

    let start = *line_range.start() as usize;
    let end = *line_range.end() as usize;
    let total = ed.buffer().lines().len();

    let clamp_end = end.min(total.saturating_sub(1));
    let mut new_lines: Vec<String> = ed.buffer().lines().to_vec();
    let mut replacements = 0usize;
    let mut lines_changed = 0usize;
    let mut last_changed_row = 0usize;

    if start <= clamp_end {
        for (row, line) in new_lines[start..=clamp_end].iter_mut().enumerate() {
            let (replaced, n) = do_replace(&regex, line, &cmd.replacement, cmd.flags.all);
            if n > 0 {
                *line = replaced;
                replacements += n;
                lines_changed += 1;
                last_changed_row = start + row;
            }
        }
    }

    if replacements == 0 {
        ed.pop_last_undo();
        return Ok(SubstituteOutcome {
            replacements: 0,
            lines_changed: 0,
        });
    }

    // Apply the new content in one shot.
    ed.buffer_mut().replace_all(&new_lines.join("\n"));

    // Cursor lands on the start of the last changed line.
    ed.buffer_mut()
        .set_cursor(hjkl_buffer::Position::new(last_changed_row, 0));

    ed.mark_content_dirty();

    // Update last_search so n/N can repeat the same pattern.
    ed.set_last_search(Some(pattern_str), true);

    Ok(SubstituteOutcome {
        replacements,
        lines_changed,
    })
}

/// Split `s` on unescaped `/`. Each `\/` in `s` becomes a literal `/`
/// in the output segment. Other `\x` sequences pass through unchanged
/// (so regex escape syntax survives).
///
/// Returns at most 3 segments: `[pattern, replacement, flags]`. Anything
/// after the third `/` is absorbed into the flags segment.
fn split_on_slash(s: &str) -> Vec<String> {
    let mut out: Vec<String> = Vec::new();
    let mut cur = String::new();
    let mut chars = s.chars().peekable();
    while let Some(c) = chars.next() {
        if c == '\\' {
            match chars.peek() {
                Some(&'/') => {
                    // Escaped delimiter → literal slash in this segment.
                    cur.push('/');
                    chars.next();
                }
                Some(_) => {
                    // Any other escape: preserve both chars so regex
                    // syntax (\d, \s, \1, \n …) survives.
                    let next = chars.next().unwrap();
                    cur.push('\\');
                    cur.push(next);
                }
                None => cur.push('\\'),
            }
        } else if c == '/' {
            if out.len() < 2 {
                out.push(std::mem::take(&mut cur));
            } else {
                // Third delimiter found: treat rest as flags.
                // Everything up to this point was the replacement;
                // collect the flags into `cur` and break.
                cur.push(c);
                // Keep going to collect remaining chars as flags.
                // (Actually we already consumed the `/`, so just let
                // the outer loop continue accumulating into cur.)
            }
        } else {
            cur.push(c);
        }
    }
    out.push(cur);
    out
}

/// Translate vim-style replacement tokens to regex-crate syntax.
///
/// - `&` → `$0` (whole match)
/// - `\&` → literal `&`
/// - `\1`…`\9` → `$1`…`$9` (capture groups)
/// - `\\` → `\` (literal backslash)
/// - Any other `\x` → `x` (drop the backslash)
fn translate_replacement(s: &str) -> String {
    let mut out = String::with_capacity(s.len() + 4);
    let mut chars = s.chars().peekable();
    while let Some(c) = chars.next() {
        if c == '&' {
            out.push_str("$0");
        } else if c == '\\' {
            match chars.next() {
                Some('&') => out.push('&'),   // \& → literal &
                Some('\\') => out.push('\\'), // \\ → literal \
                Some(d @ '1'..='9') => {
                    out.push('$');
                    out.push(d);
                }
                Some(other) => out.push(other), // drop backslash
                None => {}                      // trailing \ ignored
            }
        } else {
            out.push(c);
        }
    }
    out
}

/// Replace first or all occurrences of `regex` in `text` using the
/// already-translated `replacement` string. Returns `(new_text, count)`.
fn do_replace(regex: &Regex, text: &str, replacement: &str, all: bool) -> (String, usize) {
    let matches = regex.find_iter(text).count();
    if matches == 0 {
        return (text.to_string(), 0);
    }
    let replaced = if all {
        regex.replace_all(text, replacement).into_owned()
    } else {
        regex.replace(text, replacement).into_owned()
    };
    let count = if all { matches } else { 1 };
    (replaced, count)
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::types::{DefaultHost, Options};
    use hjkl_buffer::Buffer;

    fn editor_with(content: &str) -> Editor<Buffer, DefaultHost> {
        let mut e = Editor::new(Buffer::new(), DefaultHost::new(), Options::default());
        e.set_content(content);
        e
    }

    // ── Parser tests ─────────────────────────────────────────────────

    #[test]
    fn parse_basic() {
        let cmd = parse_substitute("/foo/bar/").unwrap();
        assert_eq!(cmd.pattern.as_deref(), Some("foo"));
        assert_eq!(cmd.replacement, "bar");
        assert!(!cmd.flags.all);
    }

    #[test]
    fn parse_trailing_slash_optional() {
        let cmd = parse_substitute("/foo/bar").unwrap();
        assert_eq!(cmd.pattern.as_deref(), Some("foo"));
        assert_eq!(cmd.replacement, "bar");
    }

    #[test]
    fn parse_global_flag() {
        let cmd = parse_substitute("/x/y/g").unwrap();
        assert!(cmd.flags.all);
    }

    #[test]
    fn parse_ignore_case_flag() {
        let cmd = parse_substitute("/x/y/i").unwrap();
        assert!(cmd.flags.ignore_case);
    }

    #[test]
    fn parse_case_sensitive_flag() {
        let cmd = parse_substitute("/x/y/I").unwrap();
        assert!(cmd.flags.case_sensitive);
    }

    #[test]
    fn parse_confirm_flag_accepted() {
        let cmd = parse_substitute("/x/y/c").unwrap();
        assert!(cmd.flags.confirm);
    }

    #[test]
    fn parse_multi_flags() {
        let cmd = parse_substitute("/x/y/gi").unwrap();
        assert!(cmd.flags.all);
        assert!(cmd.flags.ignore_case);
    }

    #[test]
    fn parse_unknown_flag_errors() {
        let err = parse_substitute("/x/y/z").unwrap_err();
        assert!(err.to_string().contains("unknown flag 'z'"), "{err}");
    }

    #[test]
    fn parse_empty_pattern_is_none() {
        let cmd = parse_substitute("//bar/").unwrap();
        assert!(cmd.pattern.is_none());
        assert_eq!(cmd.replacement, "bar");
    }

    #[test]
    fn parse_empty_replacement_ok() {
        let cmd = parse_substitute("/foo//").unwrap();
        assert_eq!(cmd.pattern.as_deref(), Some("foo"));
        assert_eq!(cmd.replacement, "");
    }

    #[test]
    fn parse_escaped_slash_in_pattern() {
        let cmd = parse_substitute("/a\\/b/c/").unwrap();
        assert_eq!(cmd.pattern.as_deref(), Some("a/b"));
    }

    #[test]
    fn parse_escaped_slash_in_replacement() {
        let cmd = parse_substitute("/a/b\\/c/").unwrap();
        // Replacement is already translated; literal / survives.
        assert_eq!(cmd.replacement, "b/c");
    }

    #[test]
    fn parse_ampersand_becomes_dollar_zero() {
        let cmd = parse_substitute("/foo/[&]/").unwrap();
        assert_eq!(cmd.replacement, "[$0]");
    }

    #[test]
    fn parse_escaped_ampersand_is_literal() {
        let cmd = parse_substitute("/foo/\\&/").unwrap();
        assert_eq!(cmd.replacement, "&");
    }

    #[test]
    fn parse_group_ref_translates() {
        let cmd = parse_substitute("/(foo)/\\1/").unwrap();
        assert_eq!(cmd.replacement, "$1");
    }

    #[test]
    fn parse_group_ref_nine() {
        let cmd = parse_substitute("/(x)/\\9/").unwrap();
        assert_eq!(cmd.replacement, "$9");
    }

    #[test]
    fn parse_wrong_delimiter_errors() {
        let err = parse_substitute("|foo|bar|").unwrap_err();
        assert!(err.to_string().contains("'/'"), "{err}");
    }

    #[test]
    fn parse_too_few_fields_errors() {
        let err = parse_substitute("/foo").unwrap_err();
        assert!(
            err.to_string().contains("needs /pattern/replacement"),
            "{err}"
        );
    }

    // ── Apply tests ──────────────────────────────────────────────────

    #[test]
    fn apply_single_line_first_only() {
        let mut e = editor_with("foo foo");
        let cmd = parse_substitute("/foo/bar/").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(out.replacements, 1);
        assert_eq!(out.lines_changed, 1);
        assert_eq!(e.buffer().lines()[0], "bar foo");
    }

    #[test]
    fn apply_single_line_global() {
        let mut e = editor_with("foo foo foo");
        let cmd = parse_substitute("/foo/bar/g").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(out.replacements, 3);
        assert_eq!(out.lines_changed, 1);
        assert_eq!(e.buffer().lines()[0], "bar bar bar");
    }

    #[test]
    fn apply_multi_line_range() {
        let mut e = editor_with("foo\nfoo foo\nbar");
        let cmd = parse_substitute("/foo/xyz/g").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=2).unwrap();
        assert_eq!(out.replacements, 3);
        assert_eq!(out.lines_changed, 2);
        assert_eq!(e.buffer().lines()[0], "xyz");
        assert_eq!(e.buffer().lines()[1], "xyz xyz");
        assert_eq!(e.buffer().lines()[2], "bar");
    }

    #[test]
    fn apply_no_match_returns_zero() {
        let mut e = editor_with("hello");
        let original = e.buffer().lines()[0].to_string();
        let cmd = parse_substitute("/xyz/abc/").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(out.replacements, 0);
        assert_eq!(out.lines_changed, 0);
        assert_eq!(e.buffer().lines()[0], original);
    }

    #[test]
    fn apply_case_insensitive_flag() {
        let mut e = editor_with("Foo FOO foo");
        let cmd = parse_substitute("/foo/bar/gi").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(out.replacements, 3);
        assert_eq!(e.buffer().lines()[0], "bar bar bar");
    }

    #[test]
    fn apply_case_sensitive_flag_overrides_editor_setting() {
        let mut e = editor_with("Foo foo");
        // Enable ignorecase on the editor.
        e.settings_mut().ignore_case = true;
        // `I` (capital) forces case-sensitive.
        let cmd = parse_substitute("/foo/bar/I").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        // Only the lowercase "foo" matches.
        assert_eq!(out.replacements, 1);
        assert_eq!(e.buffer().lines()[0], "Foo bar");
    }

    #[test]
    fn apply_empty_pattern_reuses_last_search() {
        let mut e = editor_with("hello world");
        e.set_last_search(Some("world".to_string()), true);
        let cmd = parse_substitute("//planet/").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(out.replacements, 1);
        assert_eq!(e.buffer().lines()[0], "hello planet");
    }

    #[test]
    fn apply_empty_pattern_no_last_search_errors() {
        let mut e = editor_with("hello");
        let cmd = parse_substitute("//bar/").unwrap();
        let err = apply_substitute(&mut e, &cmd, 0..=0).unwrap_err();
        assert!(
            err.to_string().contains("no previous regular expression"),
            "{err}"
        );
    }

    #[test]
    fn apply_updates_last_search() {
        let mut e = editor_with("foo");
        let cmd = parse_substitute("/foo/bar/").unwrap();
        apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(e.last_search(), Some("foo"));
    }

    #[test]
    fn apply_empty_replacement_deletes_match() {
        let mut e = editor_with("hello world");
        let cmd = parse_substitute("/world//").unwrap();
        let out = apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(out.replacements, 1);
        assert_eq!(e.buffer().lines()[0], "hello ");
    }

    #[test]
    fn apply_undo_reverts_in_one_step() {
        let mut e = editor_with("foo");
        let cmd = parse_substitute("/foo/bar/").unwrap();
        apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(e.buffer().lines()[0], "bar");
        e.undo();
        assert_eq!(e.buffer().lines()[0], "foo");
    }

    #[test]
    fn apply_ampersand_in_replacement() {
        let mut e = editor_with("foo");
        let cmd = parse_substitute("/foo/[&]/").unwrap();
        apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(e.buffer().lines()[0], "[foo]");
    }

    #[test]
    fn apply_capture_group_reference() {
        let mut e = editor_with("hello world");
        let cmd = parse_substitute("/(\\w+)/<<\\1>>/g").unwrap();
        apply_substitute(&mut e, &cmd, 0..=0).unwrap();
        assert_eq!(e.buffer().lines()[0], "<<hello>> <<world>>");
    }
}