rustling 0.8.0

A blazingly fast library for computational linguistics
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

//! Convert CHAT data to SRT (SubRip Subtitle) format.

use super::reader::ChatFile;
use crate::srt::format_srt_time;

/// A single subtitle entry collected from CHAT utterances.
struct SrtEntry {
    text: String,
    start_ms: i64,
    end_ms: i64,
}

/// Convert a single [`ChatFile`] to an SRT format string.
///
/// If `participants` is `Some`, only those participant codes are included.
/// If `None`, all participants are included.
/// Utterances without time marks are skipped (SRT requires time ranges).
/// When multiple participants are present, the subtitle text is prefixed
/// with the participant code (e.g., `"CHI: more cookie ."`).
pub(crate) fn chat_file_to_srt_str(file: &ChatFile, participants: Option<&[String]>) -> String {
    // Collect participant codes that appear.
    let mut seen_participants: Vec<String> = Vec::new();

    // Collect entries.
    let mut entries: Vec<SrtEntry> = Vec::new();

    for utt in file.real_utterances() {
        let participant = match &utt.participant {
            Some(p) => p,
            None => continue,
        };

        // Filter by requested participants.
        if let Some(codes) = participants
            && !codes.iter().any(|c| c == participant)
        {
            continue;
        }

        // Skip utterances without time marks.
        let (start, end) = match utt.time_marks {
            Some((s, e)) => (s, e),
            None => continue,
        };

        if !seen_participants.contains(participant) {
            seen_participants.push(participant.clone());
        }

        // Get the main text from tiers.
        let text = utt
            .tiers
            .as_ref()
            .and_then(|t| t.get(participant))
            .cloned()
            .unwrap_or_default();

        // Strip CHAT bullet markers from text: \x15start_end\x15
        let text = strip_bullet_markers(&text);

        entries.push(SrtEntry {
            text: format_subtitle_text(&text, participant, seen_participants.len() > 1),
            start_ms: start,
            end_ms: end,
        });
    }

    // Sort by start time, then end time.
    entries.sort_by(|a, b| {
        a.start_ms
            .cmp(&b.start_ms)
            .then_with(|| a.end_ms.cmp(&b.end_ms))
    });

    // Check if we need participant prefixes (re-evaluate after filtering).
    let multi_participant = seen_participants.len() > 1;
    // If multi_participant changed after collecting, rebuild text. But we already
    // formatted above with the knowledge of > 1 participant; the entries are already correct
    // because we push to seen_participants before creating entries for subsequent participants,
    // BUT the first participant's entries won't have the prefix if they were added
    // before seeing a second participant. Re-format if needed.
    if multi_participant {
        // Already correctly formatted since we check `seen_participants.len() > 1`
        // at entry creation time. However, the first entries for the first participant
        // were created when `seen_participants.len() == 1`. Let's just regenerate.
        return generate_srt_from_chat_file(file, participants);
    }

    // Generate SRT output.
    let mut output = String::with_capacity(4096);
    for (i, entry) in entries.iter().enumerate() {
        if i > 0 {
            output.push('\n');
        }
        output.push_str(&format!("{}\n", i + 1));
        output.push_str(&format!(
            "{} --> {}\n",
            format_srt_time(entry.start_ms),
            format_srt_time(entry.end_ms),
        ));
        output.push_str(&entry.text);
        output.push('\n');
    }
    output
}

/// Full generation that knows participant count upfront.
fn generate_srt_from_chat_file(file: &ChatFile, participants: Option<&[String]>) -> String {
    // First pass: determine which participants are present.
    let mut seen_participants: Vec<String> = Vec::new();
    for utt in file.real_utterances() {
        if let Some(participant) = &utt.participant {
            if let Some(codes) = participants
                && !codes.iter().any(|c| c == participant)
            {
                continue;
            }
            if utt.time_marks.is_some() && !seen_participants.contains(participant) {
                seen_participants.push(participant.clone());
            }
        }
    }

    let multi = seen_participants.len() > 1;

    // Second pass: collect entries.
    let mut entries: Vec<SrtEntry> = Vec::new();
    for utt in file.real_utterances() {
        let participant = match &utt.participant {
            Some(p) => p,
            None => continue,
        };
        if let Some(codes) = participants
            && !codes.iter().any(|c| c == participant)
        {
            continue;
        }
        let (start, end) = match utt.time_marks {
            Some((s, e)) => (s, e),
            None => continue,
        };

        let text = utt
            .tiers
            .as_ref()
            .and_then(|t| t.get(participant))
            .cloned()
            .unwrap_or_default();
        let text = strip_bullet_markers(&text);

        entries.push(SrtEntry {
            text: format_subtitle_text(&text, participant, multi),
            start_ms: start,
            end_ms: end,
        });
    }

    entries.sort_by(|a, b| {
        a.start_ms
            .cmp(&b.start_ms)
            .then_with(|| a.end_ms.cmp(&b.end_ms))
    });

    let mut output = String::with_capacity(4096);
    for (i, entry) in entries.iter().enumerate() {
        if i > 0 {
            output.push('\n');
        }
        output.push_str(&format!("{}\n", i + 1));
        output.push_str(&format!(
            "{} --> {}\n",
            format_srt_time(entry.start_ms),
            format_srt_time(entry.end_ms),
        ));
        output.push_str(&entry.text);
        output.push('\n');
    }
    output
}

/// Format subtitle text, optionally prefixing with participant code.
fn format_subtitle_text(text: &str, participant: &str, multi_participant: bool) -> String {
    if multi_participant {
        format!("{participant}: {text}")
    } else {
        text.to_string()
    }
}

/// Strip CHAT bullet markers (`\x15start_end\x15`) from text.
fn strip_bullet_markers(text: &str) -> String {
    let mut result = String::with_capacity(text.len());
    let mut in_bullet = false;
    for ch in text.chars() {
        if ch == '\x15' {
            in_bullet = !in_bullet;
        } else if !in_bullet {
            result.push(ch);
        }
    }
    result.trim().to_string()
}

// ---------------------------------------------------------------------------
// Tests
// ---------------------------------------------------------------------------

#[cfg(test)]
mod tests {
    use super::*;
    use crate::chat::header::Headers;
    use crate::chat::reader::ChatFile;
    use crate::chat::utterance::Utterance;
    use std::collections::HashMap;

    fn make_utterance(
        participant: &str,
        main_text: &str,
        time_marks: Option<(i64, i64)>,
    ) -> Utterance {
        let mut tiers = HashMap::new();
        tiers.insert(participant.to_string(), main_text.to_string());
        Utterance {
            participant: Some(participant.to_string()),
            tokens: None,
            time_marks,
            tiers: Some(tiers),
            changeable_header: None,
            mor_tier_name: Some("%mor".to_string()),
            gra_tier_name: Some("%gra".to_string()),
        }
    }

    fn make_chat_file(events: Vec<Utterance>) -> ChatFile {
        ChatFile::new("test.cha".to_string(), Headers::default(), events, vec![])
    }

    #[test]
    fn test_single_participant() {
        let file = make_chat_file(vec![
            make_utterance("CHI", "hello world .", Some((0, 1500))),
            make_utterance("CHI", "goodbye .", Some((2000, 3500))),
        ]);
        let srt = chat_file_to_srt_str(&file, None);

        assert!(srt.contains("1\n00:00:00,000 --> 00:00:01,500\nhello world ."));
        assert!(srt.contains("2\n00:00:02,000 --> 00:00:03,500\ngoodbye ."));
        // No participant prefix for single participant.
        assert!(!srt.contains("CHI:"));
    }

    #[test]
    fn test_multi_participant() {
        let file = make_chat_file(vec![
            make_utterance("CHI", "more cookie .", Some((0, 2000))),
            make_utterance("MOT", "want more ?", Some((2500, 4000))),
        ]);
        let srt = chat_file_to_srt_str(&file, None);

        // Participant prefix for multiple participants.
        assert!(srt.contains("CHI: more cookie ."));
        assert!(srt.contains("MOT: want more ?"));
    }

    #[test]
    fn test_skip_no_time_marks() {
        let file = make_chat_file(vec![
            make_utterance("CHI", "hello .", Some((0, 1500))),
            make_utterance("CHI", "no time marks .", None),
            make_utterance("CHI", "goodbye .", Some((2000, 3000))),
        ]);
        let srt = chat_file_to_srt_str(&file, None);

        assert!(srt.contains("hello ."));
        assert!(!srt.contains("no time marks ."));
        assert!(srt.contains("goodbye ."));
        // Only 2 entries.
        assert!(srt.contains("1\n"));
        assert!(srt.contains("2\n"));
        assert!(!srt.contains("3\n"));
    }

    #[test]
    fn test_filter_participants() {
        let file = make_chat_file(vec![
            make_utterance("CHI", "more cookie .", Some((0, 2000))),
            make_utterance("MOT", "want more ?", Some((2500, 4000))),
        ]);
        let participants = vec!["CHI".to_string()];
        let srt = chat_file_to_srt_str(&file, Some(&participants));

        assert!(srt.contains("more cookie ."));
        assert!(!srt.contains("want more ?"));
        // Single participant after filtering, no prefix.
        assert!(!srt.contains("CHI:"));
    }

    #[test]
    fn test_strip_bullet_markers() {
        let text = "hello world . \x150_1500\x15";
        let result = strip_bullet_markers(text);
        assert_eq!(result, "hello world .");
    }

    #[test]
    fn test_empty_file() {
        let file = make_chat_file(vec![]);
        let srt = chat_file_to_srt_str(&file, None);
        assert!(srt.is_empty());
    }

    #[test]
    fn test_sorted_by_time() {
        let file = make_chat_file(vec![
            make_utterance("CHI", "second .", Some((2000, 3000))),
            make_utterance("CHI", "first .", Some((0, 1000))),
        ]);
        let srt = chat_file_to_srt_str(&file, None);

        let first_pos = srt.find("first .").unwrap();
        let second_pos = srt.find("second .").unwrap();
        assert!(first_pos < second_pos);
    }
}