vibestats 2.4.1

//! Claude Code harness: parses `~/.claude/projects/**/*.jsonl` session files.

use crate::harnesses::{DailyActivity, Harness};
use serde::Deserialize;
use std::collections::{BTreeMap, HashMap, HashSet};
use std::io::{BufRead, BufReader};
use std::path::{Path, PathBuf};

pub struct Claude;

impl Harness for Claude {
    fn id(&self) -> &'static str {
        "claude"
    }

    fn display_name(&self) -> &'static str {
        "Claude Code"
    }

    fn is_installed(&self) -> bool {
        claude_projects_dir().is_some_and(|p| p.is_dir())
    }

    fn parse_date_range(&self, start: &str, end: &str) -> HashMap<String, DailyActivity> {
        parse_date_range(start, end)
    }
}

// ── Private serde types — moved verbatim from src/jsonl_parser.rs ───────────

#[derive(Debug, Default, Deserialize)]
struct MessageUsage {
    #[serde(default)]
    input_tokens: Option<u64>,
    #[serde(default)]
    output_tokens: Option<u64>,
    #[serde(rename = "cache_read_input_tokens", default)]
    cache_read_tokens: Option<u64>,
    #[serde(rename = "cache_creation_input_tokens", default)]
    cache_creation_tokens: Option<u64>,
}

#[derive(Debug, Default, Deserialize)]
struct ContentBlock {
    #[serde(rename = "type", default)]
    block_type: Option<String>,
}

#[derive(Debug, Default, Deserialize)]
struct AssistantMessage {
    #[serde(default)]
    model: Option<String>,
    #[serde(default)]
    id: Option<String>,
    #[serde(default)]
    usage: Option<MessageUsage>,
    #[serde(default)]
    content: Option<Vec<ContentBlock>>,
}

#[derive(Debug, Default, Deserialize)]
struct ClaudeEntry {
    #[serde(rename = "type", default)]
    entry_type: Option<String>,
    #[serde(default)]
    subtype: Option<String>,
    #[serde(default)]
    timestamp: Option<String>,
    #[serde(rename = "durationMs", default)]
    duration_ms: Option<u64>,
    #[serde(default)]
    message: Option<AssistantMessage>,
    #[serde(rename = "messageCount", default)]
    message_count: Option<u32>,
    #[serde(rename = "requestId", default)]
    request_id: Option<String>,
}

// ── Private helpers — bodies moved verbatim from src/jsonl_parser.rs ────────

fn claude_projects_dir() -> Option<PathBuf> {
    std::env::var("HOME")
        .ok()
        .map(|h| PathBuf::from(h).join(".claude").join("projects"))
}

fn collect_jsonl_files(dir: &Path, acc: &mut Vec<PathBuf>) {
    let entries = match std::fs::read_dir(dir) {
        Ok(e) => e,
        Err(_) => return, // unreadable directory — skip silently
    };
    for entry in entries.flatten() {
        let path = entry.path();
        if path.is_dir() {
            collect_jsonl_files(&path, acc);
        } else if path.extension().and_then(|e| e.to_str()) == Some("jsonl") {
            acc.push(path);
        }
    }
}

fn earliest_timestamp(path: &Path) -> Option<String> {
    let file = std::fs::File::open(path).ok()?;
    let mut earliest: Option<String> = None;
    for line in BufReader::new(file).lines().map_while(Result::ok) {
        let entry: ClaudeEntry = match serde_json::from_str(&line) {
            Ok(e) => e,
            Err(_) => continue,
        };
        let Some(ts) = entry.timestamp else {
            continue;
        };
        if earliest.as_ref().is_none_or(|prev| ts < *prev) {
            earliest = Some(ts);
        }
    }
    earliest
}

fn parse_file(
    path: &Path,
    start: &str,
    end: &str,
    result: &mut HashMap<String, DailyActivity>,
    processed_hashes: &mut HashSet<String>,
) {
    let file = match std::fs::File::open(path) {
        Ok(f) => f,
        Err(_) => return, // unreadable file — skip silently
    };

    let mut session_date: Option<String> = None;
    let mut duration_ms: u64 = 0;
    let mut input_tokens: u64 = 0;
    let mut output_tokens: u64 = 0;
    let mut cache_read_tokens: u64 = 0;
    let mut cache_creation_tokens: u64 = 0;
    let mut session_models: BTreeMap<String, u64> = BTreeMap::new();
    let mut message_count: u32 = 0;
    let mut tool_uses: u32 = 0;

    for line in BufReader::new(file).lines().map_while(Result::ok) {
        let entry: ClaudeEntry = match serde_json::from_str(&line) {
            Ok(e) => e,
            Err(_) => continue, // malformed line — skip silently (NFR14)
        };

        // Capture session date from the first timestamp we encounter.
        if session_date.is_none() {
            if let Some(ts) = &entry.timestamp {
                if let Some(date) = ts.get(..10) {
                    session_date = Some(date.to_string());
                }
            }
        }

        // Accumulate token usage and model stats from assistant entries.
        if entry.entry_type.as_deref() == Some("assistant") {
            if let Some(msg) = &entry.message {
                if let Some(usage) = &msg.usage {
                    if let (Some(input), Some(out)) = (usage.input_tokens, usage.output_tokens) {
                        let is_duplicate = if let (Some(message_id), Some(request_id)) =
                            (&msg.id, &entry.request_id)
                        {
                            let unique_hash = format!("{message_id}:{request_id}");
                            !processed_hashes.insert(unique_hash)
                        } else {
                            false
                        };

                        if !is_duplicate {
                            input_tokens += input;
                            output_tokens += out;
                            cache_read_tokens += usage.cache_read_tokens.unwrap_or(0);
                            cache_creation_tokens += usage.cache_creation_tokens.unwrap_or(0);

                            // Tally output tokens per model.
                            if let Some(model) = &msg.model {
                                *session_models.entry(model.clone()).or_insert(0) += out;
                            }
                        }
                    }
                }

                // Count tool_use content blocks.
                if let Some(blocks) = &msg.content {
                    for block in blocks {
                        if block.block_type.as_deref() == Some("tool_use") {
                            tool_uses += 1;
                        }
                    }
                }
            }
        }

        // Capture duration and message count from the turn_duration system entry.
        if entry.entry_type.as_deref() == Some("system")
            && entry.subtype.as_deref() == Some("turn_duration")
        {
            if let Some(ms) = entry.duration_ms {
                duration_ms = ms;
            }
            if let Some(mc) = entry.message_count {
                message_count = mc;
            }
        }
    }

    // Count this file as one session for its date (if in range).
    if let Some(date) = session_date {
        if date.as_str() >= start && date.as_str() <= end {
            let activity = result.entry(date).or_default();
            activity.sessions += 1;
            let session_minutes = (duration_ms / 60_000) as u32;
            activity.active_minutes += session_minutes;
            activity.input_tokens += input_tokens;
            activity.output_tokens += output_tokens;
            activity.cache_read_tokens += cache_read_tokens;
            activity.cache_creation_tokens += cache_creation_tokens;
            activity.message_count += message_count;
            activity.tool_uses += tool_uses;
            for (model, count) in session_models {
                *activity.models.entry(model).or_insert(0) += count;
            }
            // Track longest single session on this day.
            if session_minutes > activity.longest_session_minutes {
                activity.longest_session_minutes = session_minutes;
            }
        }
    }
}

fn parse_date_range(start: &str, end: &str) -> HashMap<String, DailyActivity> {
    let mut result: HashMap<String, DailyActivity> = HashMap::new();

    let projects_dir = match claude_projects_dir() {
        Some(p) => p,
        None => return result,
    };

    let mut jsonl_files: Vec<PathBuf> = Vec::new();
    collect_jsonl_files(&projects_dir, &mut jsonl_files);
    jsonl_files.sort_by(|a, b| {
        let a_ts = earliest_timestamp(a);
        let b_ts = earliest_timestamp(b);
        match (a_ts, b_ts) {
            (Some(a_ts), Some(b_ts)) => a_ts.cmp(&b_ts),
            (Some(_), None) => std::cmp::Ordering::Less,
            (None, Some(_)) => std::cmp::Ordering::Greater,
            (None, None) => std::cmp::Ordering::Equal,
        }
        .then_with(|| a.to_string_lossy().cmp(&b.to_string_lossy()))
    });

    let mut processed_hashes: HashSet<String> = HashSet::new();
    for path in &jsonl_files {
        parse_file(path, start, end, &mut result, &mut processed_hashes);
    }

    result
}

// ── Tests — moved verbatim from src/jsonl_parser.rs ─────────────────────────

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

    /// Write JSONL content to a unique temp file and return the path.
    /// Caller is responsible for cleanup.
    fn write_temp_jsonl(lines: &[&str]) -> std::path::PathBuf {
        use std::sync::atomic::{AtomicU64, Ordering};
        static COUNTER: AtomicU64 = AtomicU64::new(0);
        let n = COUNTER.fetch_add(1, Ordering::SeqCst);
        let name = format!(
            "vibestats_test_{}_{}_{}.jsonl",
            std::process::id(),
            std::time::SystemTime::now()
                .duration_since(std::time::UNIX_EPOCH)
                .unwrap_or_default()
                .as_nanos(),
            n
        );
        let path = std::env::temp_dir().join(name);
        let mut f = std::fs::File::create(&path).unwrap();
        for line in lines {
            writeln!(f, "{}", line).unwrap();
        }
        path
    }

    fn parse_test_file(
        path: &std::path::Path,
        start: &str,
        end: &str,
        result: &mut HashMap<String, DailyActivity>,
    ) {
        let mut processed_hashes = HashSet::new();
        parse_file(path, start, end, result, &mut processed_hashes);
    }

    fn parse_test_files(
        paths: &[std::path::PathBuf],
        start: &str,
        end: &str,
        result: &mut HashMap<String, DailyActivity>,
    ) {
        let mut processed_hashes = HashSet::new();
        for path in paths {
            parse_file(path, start, end, result, &mut processed_hashes);
        }
    }

    // Minimal valid JSONL session: one assistant entry with timestamp,
    // one system/turn_duration entry with durationMs.
    const SAMPLE_VALID: &[&str] = &[
        r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","uuid":"aaa","sessionId":"s1"}"#,
        r#"{"type":"system","subtype":"turn_duration","durationMs":3600000,"timestamp":"2026-04-10T15:00:00.000Z","uuid":"bbb","sessionId":"s1"}"#,
    ];

    #[test]
    fn valid_file_within_range_accumulates() {
        let path = write_temp_jsonl(SAMPLE_VALID);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date must be present");
        assert_eq!(day.sessions, 1);
        assert_eq!(day.active_minutes, 60); // 3_600_000 ms / 60_000 = 60
    }

    #[test]
    fn file_outside_range_not_included() {
        let path = write_temp_jsonl(SAMPLE_VALID);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-11", "2026-04-11", &mut result);
        let _ = std::fs::remove_file(&path);
        assert!(result.is_empty());
    }

    #[test]
    fn unknown_fields_silently_ignored() {
        let lines = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","unknownField":"xyz","uuid":"x"}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":60000,"timestamp":"2026-04-10T14:01:00.000Z","extraKey":42}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date must be present");
        assert_eq!(day.sessions, 1);
        assert_eq!(day.active_minutes, 1); // 60_000 ms / 60_000 = 1
    }

    #[test]
    fn malformed_lines_skipped_file_still_counted() {
        let lines = &[
            "NOT VALID JSON AT ALL {{{",
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","uuid":"x"}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result
            .get("2026-04-10")
            .expect("date found from valid line");
        assert_eq!(day.sessions, 1);
        assert_eq!(day.active_minutes, 0); // no turn_duration entry
    }

    #[test]
    fn empty_file_returns_no_entry() {
        let path = write_temp_jsonl(&[]);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        assert!(result.is_empty());
    }

    #[test]
    fn two_files_same_date_accumulate() {
        let path1 = write_temp_jsonl(SAMPLE_VALID);
        let path2 = write_temp_jsonl(SAMPLE_VALID);
        let mut result = HashMap::new();
        parse_test_files(
            &[path1.clone(), path2.clone()],
            "2026-04-10",
            "2026-04-10",
            &mut result,
        );
        let _ = std::fs::remove_file(&path1);
        let _ = std::fs::remove_file(&path2);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.sessions, 2);
        assert_eq!(day.active_minutes, 120); // 60 + 60
    }

    #[test]
    fn file_without_turn_duration_active_minutes_zero() {
        let lines = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","uuid":"x"}"#,
            r#"{"type":"user","timestamp":"2026-04-10T14:01:00.000Z","uuid":"y"}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.sessions, 1);
        assert_eq!(day.active_minutes, 0);
    }

    #[test]
    fn assistant_entry_with_usage_accumulates_tokens() {
        let lines = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","uuid":"a","message":{"model":"claude-sonnet-4-5","usage":{"input_tokens":100,"output_tokens":50,"cache_read_input_tokens":20,"cache_creation_input_tokens":10},"content":[{"type":"text"},{"type":"tool_use"}]}}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":120000,"messageCount":4,"timestamp":"2026-04-10T14:02:00.000Z","uuid":"b"}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.sessions, 1);
        assert_eq!(day.active_minutes, 2); // 120_000 ms / 60_000 = 2
        assert_eq!(day.input_tokens, 100);
        assert_eq!(day.output_tokens, 50);
        assert_eq!(day.cache_read_tokens, 20);
        assert_eq!(day.cache_creation_tokens, 10);
        assert_eq!(day.message_count, 4);
        assert_eq!(day.tool_uses, 1);
        assert_eq!(
            day.models.get("claude-sonnet-4-5").copied().unwrap_or(0),
            50
        );
    }

    #[test]
    fn old_format_entry_no_message_field_still_counted() {
        // Old JSONL entries without a `message` field must still count sessions/minutes.
        let lines = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","uuid":"x","sessionId":"s1"}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":60000,"timestamp":"2026-04-10T14:01:00.000Z","uuid":"y"}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.sessions, 1);
        assert_eq!(day.active_minutes, 1);
        assert_eq!(day.input_tokens, 0);
        assert_eq!(day.output_tokens, 0);
        assert!(day.models.is_empty());
        assert_eq!(day.tool_uses, 0);
    }

    #[test]
    fn longest_session_minutes_tracks_max_across_sessions() {
        // Two files on the same day: 60 min and 120 min. longest_session_minutes = 120.
        let long_session = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T10:00:00.000Z","uuid":"a"}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":7200000,"timestamp":"2026-04-10T12:00:00.000Z","uuid":"b"}"#,
        ];
        let short_session = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T13:00:00.000Z","uuid":"c"}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":3600000,"timestamp":"2026-04-10T14:00:00.000Z","uuid":"d"}"#,
        ];
        let p1 = write_temp_jsonl(long_session);
        let p2 = write_temp_jsonl(short_session);
        let mut result = HashMap::new();
        parse_test_files(
            &[p1.clone(), p2.clone()],
            "2026-04-10",
            "2026-04-10",
            &mut result,
        );
        let _ = std::fs::remove_file(&p1);
        let _ = std::fs::remove_file(&p2);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.sessions, 2);
        assert_eq!(day.active_minutes, 180); // 120 + 60
        assert_eq!(day.longest_session_minutes, 120); // max, not sum
    }

    #[test]
    fn models_accumulated_across_multiple_assistant_entries() {
        // Two assistant entries with different models in one session.
        let lines = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","uuid":"a","message":{"model":"claude-sonnet-4-5","usage":{"input_tokens":10,"output_tokens":30},"content":[]}}"#,
            r#"{"type":"assistant","timestamp":"2026-04-10T14:01:00.000Z","uuid":"b","message":{"model":"claude-opus-4","usage":{"input_tokens":20,"output_tokens":70},"content":[]}}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":60000,"uuid":"c","timestamp":"2026-04-10T14:02:00.000Z"}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(
            day.models.get("claude-sonnet-4-5").copied().unwrap_or(0),
            30
        );
        assert_eq!(day.models.get("claude-opus-4").copied().unwrap_or(0), 70);
        assert_eq!(day.input_tokens, 30);
        assert_eq!(day.output_tokens, 100);
    }

    #[test]
    fn duplicate_message_id_request_id_counted_once() {
        let session_a = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","requestId":"req-1","message":{"id":"msg-1","model":"claude-sonnet-4-5","usage":{"input_tokens":100,"output_tokens":50}}}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":60000,"timestamp":"2026-04-10T14:01:00.000Z"}"#,
        ];
        let session_b = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T15:00:00.000Z","requestId":"req-1","message":{"id":"msg-1","model":"claude-sonnet-4-5","usage":{"input_tokens":100,"output_tokens":50}}}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":60000,"timestamp":"2026-04-10T15:01:00.000Z"}"#,
        ];
        let p1 = write_temp_jsonl(session_a);
        let p2 = write_temp_jsonl(session_b);
        let mut result = HashMap::new();
        parse_test_files(
            &[p1.clone(), p2.clone()],
            "2026-04-10",
            "2026-04-10",
            &mut result,
        );
        let _ = std::fs::remove_file(&p1);
        let _ = std::fs::remove_file(&p2);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.sessions, 2);
        assert_eq!(day.input_tokens, 100);
        assert_eq!(day.output_tokens, 50);
    }

    #[test]
    fn usage_without_input_or_output_tokens_is_ignored() {
        let lines = &[
            r#"{"type":"assistant","timestamp":"2026-04-10T14:00:00.000Z","message":{"model":"claude-sonnet-4-5","usage":{"output_tokens":50},"content":[{"type":"tool_use"}]}}"#,
            r#"{"type":"system","subtype":"turn_duration","durationMs":60000,"timestamp":"2026-04-10T14:01:00.000Z"}"#,
        ];
        let path = write_temp_jsonl(lines);
        let mut result = HashMap::new();
        parse_test_file(&path, "2026-04-10", "2026-04-10", &mut result);
        let _ = std::fs::remove_file(&path);
        let day = result.get("2026-04-10").expect("date present");
        assert_eq!(day.input_tokens, 0);
        assert_eq!(day.output_tokens, 0);
        assert_eq!(day.tool_uses, 1);
    }
}