hagfish/src/types.rs

use serde::{Deserialize, Serialize};
use std::collections::HashMap;
use std::path::Path;

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct Config {
    pub duckdb_path: String,
    pub bind_address: String,
    pub data_source_url: String,
    pub log_file_path: Option<String>,
    #[serde(default = "default_allowed_origins")]
    pub allowed_origins: Vec<String>,
}

fn default_allowed_origins() -> Vec<String> {
    vec![]
}

impl Default for Config {
    fn default() -> Self {
        Self {
            duckdb_path: "hagfish.db".to_string(),
            bind_address: "127.0.0.1:8090".to_string(),
            data_source_url: "https://statbank.hagstova.fo/api/v1/fo/H2/VV/VV01/fisknv_md.px"
                .to_string(),
            log_file_path: Some("hagfish.log".to_string()),
            allowed_origins: default_allowed_origins(),
        }
    }
}

impl Config {
    pub fn load(path: &Path) -> anyhow::Result<Self> {
        let contents = std::fs::read_to_string(path)
            .map_err(|e| anyhow::anyhow!("Failed to read config file {}: {e}", path.display()))?;
        let config: Config = serde_json::from_str(&contents)
            .map_err(|e| anyhow::anyhow!("Failed to parse config file {}: {e}", path.display()))?;
        tracing::info!("Loaded config from {}", path.display());
        Ok(config)
    }
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct MetadataResponse {
    #[serde(default)]
    pub title: Option<String>,
    pub variables: Vec<VariableMeta>,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct AvailableFiltersDto {
    pub species: Vec<String>,
    pub zones: Vec<String>,
    pub gear: Vec<String>,
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct VariableMeta {
    pub code: String,
    #[serde(default)]
    pub text: String,
    #[serde(default)]
    pub role: Option<String>,
    #[serde(default)]
    pub values: Vec<String>,
    #[serde(default, rename = "valueTexts")]
    pub value_texts: Vec<String>,
    #[serde(flatten)]
    pub extra: HashMap<String, serde_json::Value>,
}

impl VariableMeta {
    pub fn lookup_map(&self) -> HashMap<String, String> {
        self.values
            .iter()
            .zip(self.value_texts.iter())
            .map(|(code, label)| (code.clone(), label.clone()))
            .collect()
    }
}

#[derive(Debug, Clone, Serialize)]
pub struct Query {
    pub query: Vec<QueryItem>,
    pub response: QueryResponse,
}

#[derive(Debug, Clone, Serialize)]
pub struct QueryItem {
    pub code: String,
    pub selection: Selection,
}

#[derive(Debug, Clone, Serialize)]
pub struct Selection {
    pub filter: String,
    pub values: Vec<String>,
}

#[derive(Debug, Clone, Serialize)]
pub struct QueryResponse {
    pub format: String,
}

impl Default for QueryResponse {
    fn default() -> Self {
        Self {
            format: "json-stat2".to_string(),
        }
    }
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct DataResponse {
    #[serde(flatten)]
    pub dataset: Dataset,
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct Dataset {
    #[serde(default)]
    pub id: Vec<String>,
    #[serde(default)]
    pub size: Vec<usize>,
    #[serde(flatten)]
    pub extra_fields: HashMap<String, serde_json::Value>,
    pub dimension: HashMap<String, Dimension>,
    pub value: Vec<Option<f64>>,
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct Dimension {
    #[serde(default)]
    pub label: String,
    pub category: CategoryInfo,
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct CategoryInfo {
    pub index: HashMap<String, usize>,
    #[serde(default)]
    pub label: HashMap<String, String>,
}

#[allow(dead_code)]
#[derive(Debug, Clone)]
pub struct DataRow {
    pub month: String,
    pub species_code: String,
    pub species_label: String,
    pub gear_code: String,
    pub gear_label: String,
    pub zone_code: String,
    pub zone_label: String,
    pub processing_code: String,
    pub processing_label: String,
    pub preservation_code: String,
    pub preservation_label: String,
    pub shipsize_code: String,
    pub shipsize_label: String,
    pub measure_code: String,
    pub measure_label: String,
    pub value: Option<f64>,
}

#[derive(Debug, Clone)]
pub struct Landing {
    pub month: String,
    pub species_code: String,
    pub species_label: String,
    pub gear_code: String,
    pub zone_code: String,
    pub processing_code: String,
    pub preservation_code: String,
    pub shipsize_code: String,
    pub measure_code: String,
    pub value: Option<f64>,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SpeciesDto {
    pub code: String,
    pub label: String,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct LookupDto {
    pub code: String,
    pub label: String,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct LandingDto {
    pub month: String,
    pub species_code: String,
    pub species_label: String,
    pub gear_code: String,
    pub zone_code: String,
    pub processing_code: String,
    pub preservation_code: String,
    pub shipsize_code: String,
    pub measure_code: String,
    pub value: Option<f64>,
}

impl From<Landing> for LandingDto {
    fn from(l: Landing) -> Self {
        Self {
            month: l.month,
            species_code: l.species_code,
            species_label: l.species_label,
            gear_code: l.gear_code,
            zone_code: l.zone_code,
            processing_code: l.processing_code,
            preservation_code: l.preservation_code,
            shipsize_code: l.shipsize_code,
            measure_code: l.measure_code,
            value: l.value,
        }
    }
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct MonthlyAggregate {
    pub month: String,
    pub total_mass: Option<f64>,
    pub total_value: Option<f64>,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct TopSpecies {
    pub species_code: String,
    pub species_label: String,
    pub total_value: Option<f64>,
    pub total_mass: Option<f64>,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct PriceTrend {
    pub month: String,
    pub price_per_kg: Option<f64>,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SummaryDto {
    pub monthly: Vec<MonthlyAggregate>,
    pub top_species: Vec<TopSpecies>,
    pub price_trend: Vec<PriceTrend>,
}

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct MonthlySpeciesBreakdown {
    pub month: String,
    pub species_code: String,
    pub species_label: String,
    pub total_value: Option<f64>,
    pub total_mass: Option<f64>,
}

#[derive(Debug, thiserror::Error)]
pub enum IngestError {
    #[error("HTTP request failed: {0}")]
    HttpError(#[from] reqwest::Error),

    #[error("JSON parsing failed: {0}")]
    JsonError(#[from] serde_json::Error),

    #[error("API returned status {status}: {body}")]
    ApiStatus { status: u16, body: String },

    #[error("Missing dimension in response: {0}")]
    MissingDimension(String),

    #[error("Invalid value code: {0}")]
    InvalidValueCode(String),

    #[error("API returned empty data set")]
    EmptyDataset,
}

pub type LookupMap = HashMap<String, HashMap<String, String>>;

pub type Result<T> = std::result::Result<T, IngestError>;

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

    #[test]
    fn test_config_load_missing_file_hard_error() {
        let result = Config::load(Path::new("/nonexistent/path/config.json"));
        assert!(result.is_err());
        let err_msg = result.unwrap_err().to_string();
        assert!(err_msg.contains("Failed to read config file"));
    }

    #[test]
    fn test_config_load_invalid_json_error() {
        let tmp = tempfile::NamedTempFile::new().unwrap();
        std::fs::write(tmp.path(), "not valid json {{{").unwrap();

        let result = Config::load(tmp.path());
        assert!(result.is_err());
        let err_msg = result.unwrap_err().to_string();
        assert!(err_msg.contains("Failed to parse config file"));
    }

    #[test]
    fn test_config_load_valid_file() {
        let tmp = tempfile::NamedTempFile::new().unwrap();
        let content = r#"{
            "duckdb_path": "test.db",
            "bind_address": "127.0.0.1:9090",
            "data_source_url": "http://example.com",
            "log_file_path": null,
            "allowed_origins": ["https://example.com"]
        }"#;
        std::fs::write(tmp.path(), content).unwrap();

        let config = Config::load(tmp.path()).unwrap();
        assert_eq!(config.duckdb_path, "test.db");
        assert_eq!(config.bind_address, "127.0.0.1:9090");
        assert_eq!(config.data_source_url, "http://example.com");
        assert!(config.log_file_path.is_none());
        assert_eq!(config.allowed_origins, vec!["https://example.com"]);
    }

    #[test]
    fn test_config_load_defaults_allowed_origins() {
        let tmp = tempfile::NamedTempFile::new().unwrap();
        let content = r#"{
            "duckdb_path": "test.db",
            "bind_address": "127.0.0.1:9090",
            "data_source_url": "http://example.com"
        }"#;
        std::fs::write(tmp.path(), content).unwrap();

        let config = Config::load(tmp.path()).unwrap();
        assert!(config.allowed_origins.is_empty());
    }

    #[test]
    fn test_config_default_values() {
        let config = Config::default();
        assert_eq!(config.duckdb_path, "hagfish.db");
        assert_eq!(config.bind_address, "127.0.0.1:8090");
        assert!(config.data_source_url.contains("statbank.hagstova.fo"));
        assert!(config.log_file_path.is_some());
        assert!(config.allowed_origins.is_empty());
    }
}