Skip to content

fix(codearts)：按需付费积分非零时可用积分仍显示翻倍（#261） #262

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
aimod-cc merged 1 commit into from
Oct 15, 2026
Merged

fix(codearts)：按需付费积分非零时可用积分仍显示翻倍（#261） #262

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
68 changes: 55 additions & 13 deletions desktop-tauri/src-tauri/server/src/server/core/providers/codearts/balance.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -486,24 +486,34 @@ pub async fn fetch_both(
/// `available` 只在**有积分数值**时给数：福利池是 token 不是积分，把两家的数
/// 加在一起是一个谁也对不上的值 —— 宁可 null，让界面按 wallets 逐项显示。
///
/// 「套餐积分」是**汇总口径**不是独立分账：`usageTotalPackageCredit` 的剩余/总量
/// 恰好等于「基础包 + 赠送」两行之和（2026-10-14 实测 trial 账号：套餐 8500 =
/// 基础包 500 + 赠送 8000，`available` 若三行全加得 17000 = 真值的 2 倍，与
/// issue #225 的 14956 = 7478×2 同一形状）。合计时识别这种包含关系，命中就只取
/// 分项行 —— 用容差比较，上游数值走浮点。分项缺失（比如哪天上游不再单独给
/// 赠送行）时不剔除：宁可照旧全加，也不凭猜丢掉一笔真实的钱。
/// 「套餐积分」是**汇总口径**不是独立分账：`usageTotalPackageCredit` 的剩余等于
/// 各分项行之和（2026-10-14 实测 trial 账号：5172.57 = 基础包 172.57 + 按需 0 +
/// 赠送 5000；issue #225 的 14956 = 7478×2、#261 的 11000 = 5500×2 同一形状）。
/// 合计时识别这种包含关系，命中就只取分项行 —— 用容差比较，上游数值走浮点。
///
/// 分项行**不点名**（#261 之后的推广）：按需付费行（`usageOnDemandPackageCredit`）
/// 非零的账号，「基础包 + 赠送」两行之和 ≠ 汇总行，点名判据认不出包含关系、
/// 翻倍照旧 —— 所以分项取「除汇总行外所有带剩余积分的行」。认不出包含关系
/// （数值对不上 / 分项缺席）时不剔除：宁可照旧全加，也不凭猜丢掉一笔真实的钱。
pub fn usage_document(statistics: Option<&Statistics>, benefit: Option<&BenefitBalance>) -> Value {
let mut wallets = Vec::new();
let mut credit_left: Option<f64> = None;
if let Some(statistics) = statistics {
// 「基础包 + 赠送」两行的剩余之和（任一行缺席则组合判据不成立）。
let component_sum = || -> Option<f64> {
// 分项行的剩余之和：**除汇总行外**所有带 `credit_remaining` 的行
// （有没有分项、有哪几族由上游的 metrics 决定，这里不点名）。
let parts_sum = || -> Option<f64> {
let mut sum = 0.0;
for wanted in ["usageBasicPackageCredit", "usageBonusPackageCredit"] {
let meter = statistics.meters.iter().find(|meter| meter.name == wanted)?;
sum += meter.credit_remaining?;
let mut any = false;
for meter in &statistics.meters {
if meter.name == "usageTotalPackageCredit" {
continue;
}
if let Some(left) = meter.credit_remaining {
sum += left;
any = true;
}
}
Some(sum)
any.then_some(sum)
}();
for meter in &statistics.meters {
wallets.push(json!({
Expand All @@ -519,7 +529,7 @@ pub fn usage_document(statistics: Option<&Statistics>, benefit: Option<&BenefitB
}));
let Some(left) = meter.credit_remaining else { continue };
let duplicated = meter.name == "usageTotalPackageCredit"
&& component_sum.is_some_and(|sum| (sum - left).abs() <= f64::EPSILON.max(left.abs() * 1e-6));
&& parts_sum.is_some_and(|sum| (sum - left).abs() <= f64::EPSILON.max(left.abs() * 1e-6));
if !duplicated {
credit_left = Some(credit_left.unwrap_or(0.0) + left);
}
Expand Down Expand Up @@ -713,6 +723,38 @@ mod tests {
assert_eq!(Some(8500.0), usage_document(Some(&parsed), None)["available"].as_f64());
}

/// 按需付费行**非零**的账号：点名「基础包 + 赠送」的旧判据下，两行之和 ≠ 汇总行，
/// 包含关系认不出来、翻倍照旧 —— #261 报的 11000 = 5500×2 就可能是这个形状。
/// 分项不点名后，按需行也进分项之和，包含关系照常命中。
#[test]
fn the_summary_is_deduped_even_when_the_on_demand_row_is_nonzero() {
let body = r#"{
"metrics": [
{"name":"usageTotalPackageCredit","value":0,"package_credit_amount":12000,"package_credit_remain":8800,"show":true},
{"name":"usageBasicPackageCredit","value":0,"package_credit_amount":500,"package_credit_remain":500,"show":true},
{"name":"usageOnDemandPackageCredit","value":0,"package_credit_amount":3500,"package_credit_remain":300,"show":true},
{"name":"usageBonusPackageCredit","value":0,"package_credit_amount":8000,"package_credit_remain":8000,"show":true}
]
}"#;
let parsed = parse_statistics(body).unwrap();
// 分项之和 500 + 300 + 8000 = 8800 = 汇总行 ⇒ 命中包含关系，只取分项
assert_eq!(Some(8800.0), usage_document(Some(&parsed), None)["available"].as_f64());

// 反对照：若汇总行**不含**按需（8500 = 500 + 8000）而按需另有一笔，包含关系
// 认不出来 ⇒ 照旧全加（17300）。这条钉住「宁加错不少算」的取舍没被顺手改掉 ——
// 两种语义哪种为真要靠上游实测，猜一个方向去夹都可能丢掉真实的钱。
let excluding = r#"{
"metrics": [
{"name":"usageTotalPackageCredit","value":0,"package_credit_amount":8500,"package_credit_remain":8500,"show":true},
{"name":"usageBasicPackageCredit","value":0,"package_credit_amount":500,"package_credit_remain":500,"show":true},
{"name":"usageOnDemandPackageCredit","value":0,"package_credit_amount":300,"package_credit_remain":300,"show":true},
{"name":"usageBonusPackageCredit","value":0,"package_credit_amount":8000,"package_credit_remain":8000,"show":true}
]
}"#;
let parsed = parse_statistics(excluding).unwrap();
assert_eq!(Some(17300.0), usage_document(Some(&parsed), None)["available"].as_f64());
}

/// 包含关系不成立（数值对不上 / 分项行缺席）时不得剔除：「套餐积分」万一
/// 哪天真是独立分账，剔除会把真实的钱丢掉一半。宁加错不可少算的方向与
/// 旧行为一致，只是不再加重复的那份。
Expand Down
Loading