tokrs 0.2.0

A CLI tool for counting tokens
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

use super::*;
use std::fs;
use std::path::PathBuf;
use std::time::{SystemTime, UNIX_EPOCH};

const TS: i64 = 1_788_256_800;

fn temp_dir() -> PathBuf {
    let dir = std::env::temp_dir().join(format!(
        "tokrs-grok-{}-{}",
        std::process::id(),
        SystemTime::now()
            .duration_since(UNIX_EPOCH)
            .unwrap()
            .as_nanos()
    ));
    fs::create_dir_all(&dir).unwrap();
    dir
}

fn write_updates(base: &Path, root: &str, session: &str, lines: &[String]) {
    let dir = base.join(root).join("enc-cwd").join(session);
    fs::create_dir_all(&dir).unwrap();
    let mut content = lines.join("\n");
    content.push('\n');
    fs::write(dir.join("updates.jsonl"), content).unwrap();
}

fn counters(input: u64, output: u64, cached: u64) -> String {
    format!(r#"{{"inputTokens":{input},"outputTokens":{output},"cachedReadTokens":{cached}}}"#)
}

fn model_usage(pairs: &[(&str, String)]) -> String {
    let inner = pairs
        .iter()
        .map(|(m, c)| format!(r#""{m}":{c}"#))
        .collect::<Vec<_>>()
        .join(",");
    format!(r#"{{"modelUsage":{{{inner}}}}}"#)
}

fn turn_line(ts: i64, prompt_id: Option<&str>, usage: &str) -> String {
    let pid = match prompt_id {
        Some(p) => format!(r#""prompt_id":"{p}","#),
        None => String::new(),
    };
    format!(
        r#"{{"timestamp":{ts},"method":"_x.ai/session/update","params":{{"update":{{"sessionUpdate":"turn_completed",{pid}"usage":{usage}}}}}}}"#
    )
}

#[test]
fn test_face_value_and_snapshot_filtering() {
    let base = temp_dir();
    write_updates(
        &base,
        "sessions",
        "sess-7",
        &[
            r#"{"method":"other","params":{"update":{}}}"#.to_string(),
            // 显式 usage_snapshot 带 usage: 不得导入(防中途快照双算)
            turn_line(TS, Some("px"), &model_usage(&[("m", counters(9999, 9, 0))]))
                .replace("turn_completed", "usage_snapshot"),
            turn_line(
                TS,
                Some("p1"),
                &model_usage(&[("grok-4.5-build", counters(100, 10, 5))]),
            ),
            // 下一轮面值与前一轮相同: 仍是两笔真实用量
            turn_line(
                TS + 60,
                Some("p2"),
                &model_usage(&[("grok-4.5-build", counters(100, 10, 5))]),
            ),
        ],
    );
    let entries = collect_from(&base).unwrap();
    assert_eq!(entries.len(), 2);
    for e in &entries {
        assert_eq!(e.model, "grok-4.5-build");
        // inputTokens 含 cachedRead: fresh = 100-5 = 95
        assert_eq!(e.input_tokens, 95);
        assert_eq!(e.output_tokens, 10);
        assert_eq!(e.cache_read_tokens, 5);
        assert_eq!(e.cache_creation_tokens, 0);
        assert_eq!(e.session_id.as_deref(), Some("sess-7"));
    }
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_missing_session_update_and_prompt_id_accepted() {
    let base = temp_dir();
    // 缺 sessionUpdate 字段: 向后兼容放行; 缺 prompt_id: 用事件序号
    let line = format!(
        r#"{{"timestamp":{TS},"method":"_x.ai/session/update","params":{{"update":{{"usage":{}}}}}}}"#,
        counters(5, 1, 0)
    );
    write_updates(&base, "sessions", "s", &[line.clone(), line]);
    let entries = collect_from(&base).unwrap();
    // 两条同内容事件序号不同(idx0/idx1), 不互相吞并
    assert_eq!(entries.len(), 2);
    assert_eq!(entries[0].model, "unknown");
    assert_eq!(entries[0].created_at, TS);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_multi_model_and_top_level_fallback() {
    let base = temp_dir();
    write_updates(
        &base,
        "sessions",
        "s",
        &[
            turn_line(
                TS,
                Some("p1"),
                &model_usage(&[
                    ("model-b", counters(2, 2, 0)),
                    ("model-a", counters(1, 1, 0)),
                ]),
            ),
            // 缺 modelUsage: 回退顶层 usage, 模型名 unknown
            turn_line(TS, Some("p2"), &counters(50, 5, 3)),
        ],
    );
    let mut entries = collect_from(&base).unwrap();
    entries.sort_by_key(|e| e.model.clone());
    assert_eq!(entries.len(), 3);
    assert_eq!(entries[0].model, "model-a");
    assert_eq!(entries[1].model, "model-b");
    assert_eq!(entries[2].model, "unknown");
    assert_eq!(entries[2].input_tokens, 47);
    assert_eq!(entries[2].cache_read_tokens, 3);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_same_prompt_id_last_wins() {
    let base = temp_dir();
    write_updates(
        &base,
        "sessions",
        "s",
        &[
            turn_line(TS, Some("p1"), &model_usage(&[("m", counters(100, 10, 0))])),
            turn_line(TS, Some("p1"), &model_usage(&[("m", counters(200, 20, 0))])),
        ],
    );
    let entries = collect_from(&base).unwrap();
    assert_eq!(entries.len(), 1);
    assert_eq!(entries[0].input_tokens, 200);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_zero_and_timestampless_skipped() {
    let base = temp_dir();
    let no_ts = format!(
        r#"{{"method":"_x.ai/session/update","params":{{"update":{{"sessionUpdate":"turn_completed","prompt_id":"p9","usage":{}}}}}}}"#,
        model_usage(&[("m", counters(77, 7, 0))])
    );
    write_updates(
        &base,
        "sessions",
        "s",
        &[
            turn_line(TS, Some("p1"), &model_usage(&[("m", counters(0, 0, 0))])),
            no_ts,
        ],
    );
    assert!(collect_from(&base).unwrap().is_empty());
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_pure_cache_hit_kept_with_zero_fresh_input() {
    let base = temp_dir();
    write_updates(
        &base,
        "sessions",
        "s",
        &[turn_line(
            TS,
            Some("p1"),
            &model_usage(&[("m", counters(40, 0, 40))]),
        )],
    );
    let entries = collect_from(&base).unwrap();
    // input 全部来自缓存: fresh 归 0, 但缓存命中仍是真实用量, 必须保留
    assert_eq!(entries.len(), 1);
    assert_eq!(entries[0].input_tokens, 0);
    assert_eq!(entries[0].cache_read_tokens, 40);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_cost_usd_ticks_captured() {
    let base = temp_dir();
    // 1 tick = 1e-10 USD: 338880000 -> $0.033888
    let usage = r#"{"modelUsage":{"m":{"inputTokens":10,"outputTokens":1,"cachedReadTokens":0,"costUsdTicks":338880000}}}"#;
    write_updates(&base, "sessions", "s", &[turn_line(TS, Some("p1"), usage)]);
    let entries = collect_from(&base).unwrap();
    assert_eq!(entries.len(), 1);
    let cost = entries[0].self_cost_usd.unwrap();
    assert!((cost - 0.033_888).abs() < 1e-12);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_zero_tokens_with_cost_kept_and_partial_ignored() {
    let base = temp_dir();
    // p1: token 全零但有可信自报(0.001 USD) -> 保留
    let cost_only = r#"{"modelUsage":{"m":{"inputTokens":0,"outputTokens":0,"cachedReadTokens":0,"costUsdTicks":10000000}}}"#;
    // p2: costIsPartial 标记自报仅为下界 -> 不采自报
    let partial = r#"{"modelUsage":{"m":{"inputTokens":10,"outputTokens":1,"cachedReadTokens":0,"costUsdTicks":10000000,"costIsPartial":true}}}"#;
    write_updates(
        &base,
        "sessions",
        "s",
        &[
            turn_line(TS, Some("p1"), cost_only),
            turn_line(TS, Some("p2"), partial),
        ],
    );
    let mut entries = collect_from(&base).unwrap();
    assert_eq!(entries.len(), 2);
    entries.sort_by_key(|e| e.input_tokens);
    assert_eq!(entries[0].self_cost_usd, Some(0.001));
    assert_eq!(entries[1].self_cost_usd, None);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_archived_converges_and_other_files_ignored() {
    let base = temp_dir();
    let lines = vec![turn_line(
        TS,
        Some("p1"),
        &model_usage(&[("m", counters(9, 1, 0))]),
    )];
    write_updates(&base, "sessions", "s1", &lines);
    write_updates(&base, "archived_sessions", "s1", &lines);
    // 非 updates.jsonl 不采集
    let other = base
        .join("sessions")
        .join("enc-cwd")
        .join("s1")
        .join("summary.json");
    fs::write(&other, lines[0].clone()).unwrap();
    let entries = collect_from(&base).unwrap();
    // 同 session_id 归档副本与活跃副本经去重收敛
    assert_eq!(entries.len(), 1);
    fs::remove_dir_all(&base).ok();
}

#[test]
fn test_missing_base_returns_empty() {
    let base = temp_dir();
    fs::remove_dir_all(&base).unwrap();
    assert!(collect_from(&base).unwrap().is_empty());
}