liter-llm 1.2.0

use std::collections::BTreeMap;

use serde::{Deserialize, Serialize};

use super::common::{
    AssistantMessage, ChatCompletionTool, Message, ResponseFormat, StopSequence, ToolChoice, ToolType, Usage,
};
use crate::cost;

// ─── Finish Reason ────────────────────────────────────────────────────────────

/// Why a choice stopped generating tokens.
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "snake_case")]
pub enum FinishReason {
    Stop,
    Length,
    ToolCalls,
    ContentFilter,
    /// Deprecated legacy finish reason; retained for API compatibility.
    #[serde(rename = "function_call")]
    FunctionCall,
    /// Catch-all for unknown finish reasons returned by non-OpenAI providers.
    ///
    /// Note: this intentionally does **not** carry the original string (e.g.
    /// `Other(String)`).  Using `#[serde(other)]` requires a unit variant, and
    /// switching to `#[serde(untagged)]` would change deserialization semantics
    /// for all variants.  The original value can be recovered by inspecting the
    /// raw JSON if needed.
    #[serde(other)]
    Other,
}

// ─── Reasoning Effort ────────────────────────────────────────────────────────

/// Controls how much reasoning effort the model should use.
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "lowercase")]
pub enum ReasoningEffort {
    Low,
    Medium,
    High,
}

// ─── Request ─────────────────────────────────────────────────────────────────

#[derive(Debug, Clone, Default, PartialEq, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
pub struct ChatCompletionRequest {
    pub model: String,
    pub messages: Vec<Message>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub temperature: Option<f64>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub top_p: Option<f64>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub n: Option<u32>,
    /// Whether to stream the response.
    ///
    /// This field is managed by the client layer (`prepare_request`) and should
    /// not be set directly by callers — use `chat` for non-streaming and
    /// `chat_stream` for streaming.  Making it `pub(crate)` prevents callers
    /// from setting it explicitly, which would conflict with the client's own
    /// stream flag injection.
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub(crate) stream: Option<bool>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub stop: Option<StopSequence>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub max_tokens: Option<u64>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub presence_penalty: Option<f64>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub frequency_penalty: Option<f64>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    /// Token bias map.  Uses `BTreeMap` (sorted keys) for deterministic
    /// serialization order — important when hashing or signing requests.
    pub logit_bias: Option<BTreeMap<String, f64>>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub user: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub tools: Option<Vec<ChatCompletionTool>>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub tool_choice: Option<ToolChoice>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub parallel_tool_calls: Option<bool>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub response_format: Option<ResponseFormat>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub stream_options: Option<StreamOptions>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub seed: Option<i64>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub reasoning_effort: Option<ReasoningEffort>,
    /// Provider-specific extra parameters merged into the request body.
    /// Use for guardrails, safety settings, grounding config, etc.
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub extra_body: Option<serde_json::Value>,
}

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
pub struct StreamOptions {
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub include_usage: Option<bool>,
}

// ─── Response ────────────────────────────────────────────────────────────────

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct ChatCompletionResponse {
    pub id: String,
    /// Always `"chat.completion"` from OpenAI-compatible APIs.  Stored as a
    /// plain `String` so non-standard provider values do not break deserialization.
    pub object: String,
    pub created: u64,
    pub model: String,
    pub choices: Vec<Choice>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub usage: Option<Usage>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub system_fingerprint: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub service_tier: Option<String>,
}

impl ChatCompletionResponse {
    /// Estimate the cost of this response based on embedded pricing data.
    ///
    /// Returns `None` if:
    /// - the `model` field is not present in the embedded pricing registry, or
    /// - the `usage` field is absent from the response.
    ///
    /// # Example
    ///
    /// ```rust,ignore
    /// let cost = response.estimated_cost();
    /// if let Some(usd) = cost {
    ///     println!("Request cost: ${usd:.6}");
    /// }
    /// ```
    #[must_use]
    pub fn estimated_cost(&self) -> Option<f64> {
        let usage = self.usage.as_ref()?;
        cost::completion_cost(&self.model, usage.prompt_tokens, usage.completion_tokens)
    }
}

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct Choice {
    pub index: u32,
    pub message: AssistantMessage,
    pub finish_reason: Option<FinishReason>,
}

// ─── Stream Chunk ────────────────────────────────────────────────────────────

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct ChatCompletionChunk {
    pub id: String,
    /// Always `"chat.completion.chunk"` from OpenAI-compatible APIs.  Stored
    /// as a plain `String` so non-standard provider values do not fail parsing.
    pub object: String,
    pub created: u64,
    pub model: String,
    pub choices: Vec<StreamChoice>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub usage: Option<Usage>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub system_fingerprint: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub service_tier: Option<String>,
}

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct StreamChoice {
    pub index: u32,
    pub delta: StreamDelta,
    pub finish_reason: Option<FinishReason>,
}

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct StreamDelta {
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub role: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub content: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub tool_calls: Option<Vec<StreamToolCall>>,
    /// Deprecated legacy function_call delta; retained for API compatibility.
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub function_call: Option<StreamFunctionCall>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub refusal: Option<String>,
}

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct StreamToolCall {
    pub index: u32,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub id: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none", rename = "type")]
    pub call_type: Option<ToolType>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub function: Option<StreamFunctionCall>,
}

#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct StreamFunctionCall {
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub name: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub arguments: Option<String>,
}