cuqueclicker 1.1.0

A TUI idle clicker where you finger an ASCII ass instead of clicking a cookie.
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
//! `Mag` — a non-negative real represented as `log10(value)`.
//!
//! The whole point: gameplay-side quantities (cuques, FPS, tree multipliers,
//! per-fingerer modifier mults) compound multiplicatively across thousands
//! of bought nodes and powerup catches. Stored as `f64`, the product
//! reaches `Infinity` after a long-haul session — the headline bug from
//! the wild "broken save" triage. Stored as `log10`, the same product
//! fits comfortably inside an `f64` for any value the player could
//! conceivably observe; the dynamic range is `10^(±1.8e308)`, i.e.
//! "truly infinite" for any practical gameplay.
//!
//! ## Representation
//!
//! `Mag { log10: f64 }`, where `log10` is the base-10 logarithm of the
//! value. Zero is the sentinel `log10 == f64::NEG_INFINITY`; everything
//! else is finite. Negatives are not representable (the only place the
//! game produced one was an underflow that signalled a bug anyway —
//! `try_sub` returns `Option`, callers handle the deficit explicitly).
//!
//! ## Operations
//!
//! - `mul` / `div`: addition / subtraction in log space — cheap.
//! - `add`: requires un-logging the smaller addend; uses
//!   `log10(a + b) = log10(a) + log10(1 + 10^(b - a))` for `a >= b`.
//! - `try_sub`: same trick. Returns `None` if the result would be
//!   negative.
//! - `cmp`: direct comparison of the log values.
//! - `to_f64`: useful at display sites (for the legacy `format::big`
//!   path) and inside formulas that need a real `f64` (e.g. RNG range
//!   bounds, multiplier caps from the procgen).

use serde::{Deserialize, Deserializer, Serialize, Serializer};

/// Non-negative log-magnitude. `Mag::ZERO == Mag { log10: -inf }`,
/// `Mag::ONE == Mag { log10: 0.0 }`. All other instances carry a
/// finite `log10`.
///
/// ## Save compatibility
///
/// Serializes as a plain `f64` when the value fits inside one
/// (`log10 < 300`, well clear of overflow); otherwise as
/// `{"log10": <f64>}`. Deserialization accepts BOTH shapes — so existing
/// V4 saves with `"cuques": 1.234e10` and `"MulFactor": 7.0` continue
/// to load correctly, no schema bump required. New saves use the
/// struct form only when the value exceeds the `f64` finite range.
#[derive(Clone, Copy, Debug)]
pub struct Mag {
    pub log10: f64,
}

impl Serialize for Mag {
    fn serialize<S: Serializer>(&self, s: S) -> Result<S::Ok, S::Error> {
        // Below the `f64` overflow horizon, emit a plain number so saves
        // stay human-readable and old-format-compatible. Above it, fall
        // back to the explicit struct so the log-magnitude survives the
        // round-trip exactly.
        if self.log10 == f64::NEG_INFINITY {
            return 0.0_f64.serialize(s);
        }
        if self.log10 < 300.0 {
            return self.to_f64().serialize(s);
        }
        // Struct form for truly-huge values.
        #[derive(Serialize)]
        struct Wide {
            log10: f64,
        }
        Wide { log10: self.log10 }.serialize(s)
    }
}

impl<'de> Deserialize<'de> for Mag {
    fn deserialize<D: Deserializer<'de>>(d: D) -> Result<Self, D::Error> {
        // Accept either a plain JSON number (legacy + small new saves)
        // or the explicit `{"log10": ...}` struct (large new saves). The
        // untagged enum gives serde the freedom to try both without us
        // hand-writing a visitor.
        #[derive(Deserialize)]
        #[serde(untagged)]
        enum Either {
            Number(f64),
            Struct { log10: f64 },
        }
        match Either::deserialize(d)? {
            Either::Number(v) => Ok(Mag::from_f64(v)),
            Either::Struct { log10 } => Ok(Mag { log10 }),
        }
    }
}

impl Mag {
    pub const ZERO: Mag = Mag {
        log10: f64::NEG_INFINITY,
    };
    pub const ONE: Mag = Mag { log10: 0.0 };

    /// Construct from an `f64`. Non-finite / negative inputs collapse to
    /// `ZERO` — there is no honest log-magnitude for them and silently
    /// turning a `NaN` into a zero is friendlier than panicking on a
    /// codepath the player never asked for.
    pub fn from_f64(v: f64) -> Mag {
        if !v.is_finite() || v <= 0.0 {
            Mag::ZERO
        } else {
            Mag { log10: v.log10() }
        }
    }

    /// Decode back to an `f64`. Clamps very-large `log10` values to
    /// `f64::MAX` — losing precision in the high end is the price of
    /// asking for an `f64`; the `Mag` representation itself stays
    /// faithful. Use this at the display boundary or wherever a downstream
    /// API truly demands `f64`. Internal math should chain `Mag` ops.
    pub fn to_f64(self) -> f64 {
        if self.log10.is_nan() {
            return 0.0;
        }
        if self.log10 == f64::NEG_INFINITY {
            return 0.0;
        }
        if self.log10 >= f64::MAX.log10() {
            return f64::MAX;
        }
        10f64.powf(self.log10)
    }

    pub fn is_zero(self) -> bool {
        self.log10 == f64::NEG_INFINITY
    }

    // `clippy::should_implement_trait` would have us drop the inherent
    // method in favor of the `std::ops::Mul` impl below. We keep both
    // because the call sites read more naturally as `a.mul(b)` than
    // `Mag::Mul::mul(a, b)` and the `*` operator alone obscures that
    // Mag isn't a primitive `f64`. The trait impl exists so `*` works
    // and so the inherent method is provably aligned with std semantics.
    #[allow(clippy::should_implement_trait)]
    pub fn mul(self, rhs: Mag) -> Mag {
        if self.is_zero() || rhs.is_zero() {
            return Mag::ZERO;
        }
        Mag {
            log10: self.log10 + rhs.log10,
        }
    }

    #[allow(clippy::should_implement_trait)]
    pub fn div(self, rhs: Mag) -> Mag {
        if self.is_zero() {
            return Mag::ZERO;
        }
        if rhs.is_zero() {
            // Defining 1/0 as ZERO here is a deliberate, defensive
            // choice: nothing in gameplay should ever divide by zero,
            // but if it does we'd rather collapse to "no contribution"
            // than panic or produce an `Infinity` that re-introduces
            // the bug class this whole module exists to kill.
            return Mag::ZERO;
        }
        Mag {
            log10: self.log10 - rhs.log10,
        }
    }

    #[allow(clippy::should_implement_trait)]
    pub fn add(self, rhs: Mag) -> Mag {
        if self.is_zero() {
            return rhs;
        }
        if rhs.is_zero() {
            return self;
        }
        let (hi, lo) = if self.log10 >= rhs.log10 {
            (self.log10, rhs.log10)
        } else {
            (rhs.log10, self.log10)
        };
        let diff = lo - hi;
        if diff < -16.0 {
            // 10^-16 is past f64 precision relative to 1.0 — the smaller
            // addend's contribution rounds away. Skip the (1.0 + 10^diff)
            // dance and return the larger.
            return Mag { log10: hi };
        }
        Mag {
            log10: hi + (1.0 + 10f64.powf(diff)).log10(),
        }
    }

    /// Saturating subtraction: returns `Some(self - rhs)` when
    /// `self >= rhs`, otherwise `None`. Cuque spending uses this on
    /// purchase — the caller already gated on `affordable`, so `None`
    /// means a logic bug, not a regular game state.
    pub fn try_sub(self, rhs: Mag) -> Option<Mag> {
        if rhs.is_zero() {
            return Some(self);
        }
        if self.log10 < rhs.log10 {
            return None;
        }
        if self.is_zero() {
            return Some(Mag::ZERO);
        }
        let diff = rhs.log10 - self.log10;
        if diff < -16.0 {
            return Some(self);
        }
        let factor = 1.0 - 10f64.powf(diff);
        if factor <= 0.0 {
            return Some(Mag::ZERO);
        }
        Some(Mag {
            log10: self.log10 + factor.log10(),
        })
    }

    /// Saturating subtraction with floor at zero. Convenience for paths
    /// that don't care about the underflow case (e.g. visual tween
    /// targets).
    pub fn saturating_sub(self, rhs: Mag) -> Mag {
        self.try_sub(rhs).unwrap_or(Mag::ZERO)
    }

    /// Raise to an integer power. `pow(0) == ONE`, `pow(n) = self * self * …`.
    /// Cheap: just multiplies the log by `n`.
    ///
    /// `0^n` for any `n != 0` returns `ZERO`. Mathematically `0^negative` is
    /// undefined, but returning ZERO is the conservative choice for this
    /// codebase: it propagates "no contribution" downstream rather than
    /// producing an `Infinity`, which would reintroduce the bug class
    /// the `Mag` type exists to kill.
    pub fn pow_i(self, n: i32) -> Mag {
        if n == 0 {
            return Mag::ONE;
        }
        if self.is_zero() {
            return Mag::ZERO;
        }
        Mag {
            log10: self.log10 * (n as f64),
        }
    }

    /// Floored `to_f64`, useful for "how many of these can the player
    /// afford" math. Clamps to `u64::MAX`.
    pub fn floor_u64(self) -> u64 {
        let v = self.to_f64().floor();
        if v <= 0.0 {
            0
        } else if v >= u64::MAX as f64 {
            u64::MAX
        } else {
            v as u64
        }
    }
}

impl Default for Mag {
    fn default() -> Self {
        Mag::ZERO
    }
}

impl PartialEq for Mag {
    fn eq(&self, other: &Self) -> bool {
        // NaN-safe: treat any NaN log as ZERO for equality purposes.
        let a = if self.log10.is_nan() {
            f64::NEG_INFINITY
        } else {
            self.log10
        };
        let b = if other.log10.is_nan() {
            f64::NEG_INFINITY
        } else {
            other.log10
        };
        a == b
    }
}

impl PartialOrd for Mag {
    fn partial_cmp(&self, other: &Self) -> Option<std::cmp::Ordering> {
        self.log10.partial_cmp(&other.log10)
    }
}

// Convenience: ergonomic conversion from common literal types so call
// sites can keep reading naturally ("Mag::from(0.05)" inside a test
// fixture).
impl From<f64> for Mag {
    fn from(v: f64) -> Self {
        Mag::from_f64(v)
    }
}

impl From<u32> for Mag {
    fn from(v: u32) -> Self {
        Mag::from_f64(v as f64)
    }
}

impl From<u64> for Mag {
    fn from(v: u64) -> Self {
        Mag::from_f64(v as f64)
    }
}

// Operator traits forward to the inherent methods so `a * b` / `a + b` /
// `a / b` work alongside the explicit `.mul(b)` / `.add(b)` / `.div(b)`
// the rest of the codebase uses. The trait impls also silence clippy's
// `method_can_be_confused_for_std_trait` warning — the methods *are*
// the standard operations now.
//
// Subtraction deliberately doesn't get a `Sub` impl: the only honest
// answer for `a - b` when `b > a` is "underflow", and `try_sub →
// Option` keeps every spend site forced to acknowledge that case.

impl std::ops::Mul for Mag {
    type Output = Mag;
    fn mul(self, rhs: Mag) -> Mag {
        Mag::mul(self, rhs)
    }
}

impl std::ops::Div for Mag {
    type Output = Mag;
    fn div(self, rhs: Mag) -> Mag {
        Mag::div(self, rhs)
    }
}

impl std::ops::Add for Mag {
    type Output = Mag;
    fn add(self, rhs: Mag) -> Mag {
        Mag::add(self, rhs)
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    fn approx_eq(a: f64, b: f64) -> bool {
        if a == b {
            return true;
        }
        (a - b).abs() / b.abs().max(1e-12) < 1e-10
    }

    #[test]
    fn from_f64_roundtrip() {
        for v in [1.0, 1000.0, 1e20, 1e100, 1e300] {
            let m = Mag::from_f64(v);
            assert!(approx_eq(m.to_f64(), v), "roundtrip {v} got {}", m.to_f64());
        }
    }

    #[test]
    fn zero_and_negative_collapse_to_zero() {
        assert_eq!(Mag::from_f64(0.0), Mag::ZERO);
        assert_eq!(Mag::from_f64(-5.0), Mag::ZERO);
        assert_eq!(Mag::from_f64(f64::NAN), Mag::ZERO);
        assert_eq!(Mag::from_f64(f64::INFINITY), Mag::ZERO);
    }

    #[test]
    fn mul_compounds_huge() {
        let a = Mag::from_f64(1e200);
        let b = Mag::from_f64(1e200);
        let c = a.mul(b);
        // 10^400 — well past f64::MAX.
        assert!(approx_eq(c.log10, 400.0));
        // to_f64 saturates rather than producing infinity.
        assert_eq!(c.to_f64(), f64::MAX);
    }

    #[test]
    fn pow_i_handles_grindy_compounding() {
        // 1.005^100_000 — a stress test for late-late-game stacks.
        // Original f64 path would overflow long before this.
        let m = Mag::from_f64(1.005);
        let p = m.pow_i(100_000);
        let expected_log = (1.005_f64).log10() * 100_000.0;
        assert!(approx_eq(p.log10, expected_log));
    }

    #[test]
    fn add_handles_disparate_magnitudes() {
        // 10^100 + 10^50 ≈ 10^100 (within f64 precision).
        let a = Mag::from_f64(1e100);
        let b = Mag::from_f64(1e50);
        let s = a.add(b);
        assert!(approx_eq(s.log10, a.log10));
    }

    #[test]
    fn add_handles_equal_magnitudes() {
        let a = Mag::from_f64(1e100);
        let s = a.add(a);
        // 2 * 10^100 -> log10 = 100 + log10(2)
        assert!(approx_eq(s.log10, 100.0 + 2.0_f64.log10()));
    }

    #[test]
    fn try_sub_returns_none_on_underflow() {
        let a = Mag::from_f64(100.0);
        let b = Mag::from_f64(200.0);
        assert!(a.try_sub(b).is_none());
    }

    #[test]
    fn try_sub_handles_close_values() {
        let a = Mag::from_f64(100.0);
        let b = Mag::from_f64(99.0);
        let s = a.try_sub(b).expect("should fit");
        assert!(approx_eq(s.to_f64(), 1.0));
    }

    #[test]
    fn try_sub_zero_is_identity() {
        let a = Mag::from_f64(50.0);
        assert_eq!(a.try_sub(Mag::ZERO).unwrap(), a);
    }

    #[test]
    fn comparison_order_matches_value_order() {
        assert!(Mag::from_f64(10.0) < Mag::from_f64(100.0));
        assert!(Mag::ZERO < Mag::from_f64(1.0));
        // Truly huge values still order correctly.
        assert!(Mag::from_f64(1e200).mul(Mag::from_f64(1e200)) > Mag::from_f64(1e300));
    }

    #[test]
    fn mul_by_zero_is_zero() {
        let a = Mag::from_f64(1e100);
        assert_eq!(a.mul(Mag::ZERO), Mag::ZERO);
    }

    #[test]
    fn one_is_multiplicative_identity() {
        let a = Mag::from_f64(42.0);
        assert!(approx_eq(a.mul(Mag::ONE).to_f64(), a.to_f64()));
        assert!(approx_eq(Mag::ONE.mul(a).to_f64(), a.to_f64()));
    }

    #[test]
    fn floor_u64_clamps_at_u64_max() {
        let huge = Mag::from_f64(1e25);
        // 10^25 > u64::MAX (~1.8e19), should clamp.
        assert_eq!(huge.floor_u64(), u64::MAX);
        let small = Mag::from_f64(42.7);
        assert_eq!(small.floor_u64(), 42);
        assert_eq!(Mag::ZERO.floor_u64(), 0);
    }
}