#define _GNU_SOURCE
#define _DARWIN_C_SOURCE
#define _POSIX_C_SOURCE 200809L
#include "common/loop.h"
#include "platform/sync.h"

#include <errno.h>
#include <fcntl.h>
#include <poll.h>
#include <signal.h>
#include <stdatomic.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>

#define YASL_LOOP_INIT_CAP 32

typedef int (*yasl_loop_fd_fn)(int fd, short revents, void *ud);

struct yasl_fd_source {
    yasl_loop_fd_fn fn;
    void *ud;
    int id;
    int fd;
    short events;
};

struct yasl_timer_source {
    yasl_timer_cb fn;
    void *ud;
    uint64_t interval_ns;
    uint64_t deadline_ns;
    yasl_timer_id id;
    bool deleted;
};

struct yasl_post {
    void (*fn)(void *ud);
    void *ud;
    struct yasl_post *next;
};

/* Signals this loop watches; the previous actions are restored on unwatch. */
struct yasl_signal_watch {
    void (*cb)(void *ud, int signo);
    void *ud;
    int count;
    int signals[NSIG];
    struct sigaction old[NSIG];
};

struct yasl_loop {
    struct yasl_fd_source *fds;
    struct pollfd *pollfds;
    int count;
    int capacity;
    int next_id;
    bool running;
    int wake_rfd;
    int wake_wfd;
    int wake_watch;

    struct yasl_timer_source *timers;
    int timer_count;
    int timer_capacity;
    uint32_t next_timer_id;
    int timer_firing_depth;

    yasl_mutex post_mutex;
    struct yasl_post *post_head;
    struct yasl_post *post_tail;

    struct yasl_signal_watch *signals;
};

/* Handlers run on any thread, so they only touch lock-free atomics: they mark
 * the signal pending and write to the owning loop's wake pipe. The pipe is
 * stored as fd + 1 so that the zero initializer means no pipe. */
static atomic_int signal_wake_fd[NSIG];
static atomic_bool signal_pending[NSIG];
static struct yasl_loop *signal_owner[NSIG];

uint64_t yasl_loop_now_ns(void)
{
    struct timespec ts;
    clock_gettime(CLOCK_MONOTONIC, &ts);
    return (uint64_t)ts.tv_sec * 1000000000ull + (uint64_t)ts.tv_nsec;
}

static void run_posts(struct yasl_loop *l)
{
    yasl_mutex_lock(&l->post_mutex);
    struct yasl_post *p = l->post_head;
    l->post_head = l->post_tail = NULL;
    yasl_mutex_unlock(&l->post_mutex);
    while (p) {
        struct yasl_post *next = p->next;
        p->fn(p->ud);
        free(p);
        p = next;
    }
}

static void run_signals(struct yasl_loop *l)
{
    /* A callback may unwatch, so re-check the watch before each delivery. */
    for (int signo = 1; signo < NSIG; signo++) {
        struct yasl_signal_watch *w = l->signals;
        if (!w) return;
        if (signal_owner[signo] != l || !atomic_exchange(&signal_pending[signo], false))
            continue;
        w->cb(w->ud, signo);
    }
}

static int loop_wake_cb(int fd, short revents, void *ud)
{
    (void)revents;
    char buf[128];
    ssize_t n;
    while ((n = read(fd, buf, sizeof(buf))) > 0 || (n < 0 && errno == EINTR)) {}
    struct yasl_loop *l = ud;
    run_posts(l);
    run_signals(l);
    return 0;
}

struct yasl_loop *yasl_loop_create(void)
{
    struct yasl_loop *l = calloc(1, sizeof(*l));
    if (!l) return NULL;

    l->capacity = YASL_LOOP_INIT_CAP;
    l->fds = malloc((size_t)l->capacity * sizeof(*l->fds));
    l->pollfds = malloc((size_t)l->capacity * sizeof(*l->pollfds));
    if (!l->fds || !l->pollfds) {
        free(l->fds);
        free(l->pollfds);
        free(l);
        return NULL;
    }

    if (yasl_mutex_init(&l->post_mutex) != 0) {
        free(l->fds);
        free(l->pollfds);
        free(l);
        return NULL;
    }

    l->next_id = 1;
    l->next_timer_id = 1;
    l->wake_rfd = -1;
    l->wake_wfd = -1;

    int p[2];
#if defined(__linux__) && defined(O_CLOEXEC)
    if (pipe2(p, O_CLOEXEC | O_NONBLOCK) == 0) {
        l->wake_rfd = p[0];
        l->wake_wfd = p[1];
    }
#else
    if (pipe(p) == 0) {
        int f0 = fcntl(p[0], F_GETFL, 0);
        int f1 = fcntl(p[1], F_GETFL, 0);
        if (f0 >= 0) fcntl(p[0], F_SETFL, f0 | O_NONBLOCK);
        if (f1 >= 0) fcntl(p[1], F_SETFL, f1 | O_NONBLOCK);
        fcntl(p[0], F_SETFD, FD_CLOEXEC);
        fcntl(p[1], F_SETFD, FD_CLOEXEC);
        l->wake_rfd = p[0];
        l->wake_wfd = p[1];
    }
#endif

    if (l->wake_rfd >= 0) {
        l->wake_watch = yasl_loop_add_fd(l, l->wake_rfd, POLLIN, loop_wake_cb, l);
    }
    return l;
}

void yasl_loop_destroy(struct yasl_loop *l)
{
    if (!l) return;
    yasl_loop_unwatch_signals(l);
    for (struct yasl_post *p = l->post_head; p;) {
        struct yasl_post *next = p->next;
        free(p);
        p = next;
    }
    yasl_mutex_destroy(&l->post_mutex);
    if (l->wake_watch > 0)
        yasl_loop_del_fd(l, l->wake_watch);
    if (l->wake_rfd >= 0)
        close(l->wake_rfd);
    if (l->wake_wfd >= 0)
        close(l->wake_wfd);
    free(l->fds);
    free(l->pollfds);
    free(l->timers);
    free(l);
}

int yasl_loop_add_fd(struct yasl_loop *l, int fd, short events,
                     int (*fn)(int fd, short revents, void *ud), void *ud)
{
    if (!l || fd < 0)
        return -1;

    if (l->count >= l->capacity) {
        int new_cap = l->capacity ? l->capacity * 2 : YASL_LOOP_INIT_CAP;
        struct yasl_fd_source *new_fds = realloc(l->fds, (size_t)new_cap * sizeof(*new_fds));
        if (!new_fds) return -1;
        l->fds = new_fds;

        struct pollfd *new_pollfds = realloc(l->pollfds, (size_t)new_cap * sizeof(*new_pollfds));
        if (!new_pollfds) return -1;
        l->pollfds = new_pollfds;
        l->capacity = new_cap;
    }

    int id = l->next_id++;
    if (l->next_id <= 0) l->next_id = 1;

    l->fds[l->count] = (struct yasl_fd_source){
        .fn = fn, .ud = ud, .id = id, .fd = fd, .events = events,
    };
    l->pollfds[l->count] = (struct pollfd){
        .fd = fd, .events = events, .revents = 0,
    };
    l->count++;
    return id;
}

int yasl_loop_del_fd(struct yasl_loop *l, int id)
{
    if (!l || id <= 0) return -1;
    for (int i = 0; i < l->count; i++) {
        if (l->fds[i].id == id) {
            l->count--;
            l->fds[i] = l->fds[l->count];
            l->pollfds[i] = l->pollfds[l->count];
            return 0;
        }
    }
    return -1;
}

int yasl_loop_set_fd_events(struct yasl_loop *l, int id, short events)
{
    if (!l || id <= 0) return -1;
    for (int i = 0; i < l->count; i++) {
        if (l->fds[i].id == id) {
            l->fds[i].events = events;
            l->pollfds[i].events = events;
            return 0;
        }
    }
    return -1;
}

static void wake_write(int fd)
{
    /* A full pipe already wakes the loop, so EAGAIN needs no retry. */
    int saved = errno;
    char b = 1;
    ssize_t n;
    do {
        n = write(fd, &b, 1);
    } while (n < 0 && errno == EINTR);
    errno = saved;
}

void yasl_loop_kick(struct yasl_loop *l)
{
    if (!l || l->wake_wfd < 0) return;
    wake_write(l->wake_wfd);
}

bool yasl_loop_post(struct yasl_loop *l, void (*fn)(void *ud), void *ud)
{
    if (!l || !fn || l->wake_wfd < 0) return false;
    struct yasl_post *p = malloc(sizeof(*p));
    if (!p) return false;
    *p = (struct yasl_post){.fn = fn, .ud = ud};
    yasl_mutex_lock(&l->post_mutex);
    if (l->post_tail)
        l->post_tail->next = p;
    else
        l->post_head = p;
    l->post_tail = p;
    yasl_mutex_unlock(&l->post_mutex);
    wake_write(l->wake_wfd);
    return true;
}

static void signal_handler(int signo)
{
    atomic_store(&signal_pending[signo], true);
    int fd = atomic_load(&signal_wake_fd[signo]) - 1;
    if (fd >= 0) wake_write(fd);
}

static bool signal_valid(int signo)
{
    return signo > 0 && signo < NSIG && signo != SIGKILL && signo != SIGSTOP;
}

bool yasl_loop_watch_signals(struct yasl_loop *l, const int *signals, int n,
                             void (*cb)(void *ud, int signo), void *ud)
{
    if (!l || !signals || n <= 0 || !cb || l->wake_wfd < 0) {
        errno = EINVAL;
        return false;
    }
    bool wanted[NSIG] = {0};
    for (int i = 0; i < n; i++) {
        if (!signal_valid(signals[i])) {
            errno = EINVAL;
            return false;
        }
        if (signal_owner[signals[i]] && signal_owner[signals[i]] != l) {
            errno = EBUSY;
            return false;
        }
        wanted[signals[i]] = true;
    }
    struct yasl_signal_watch *w = calloc(1, sizeof(*w));
    if (!w) return false;
    w->cb = cb;
    w->ud = ud;

    /* Keep the old watch's saved actions for signals that stay watched, so
     * that unwatch restores what the application had before any watch. */
    struct yasl_signal_watch *old = l->signals;
    struct sigaction sa = {.sa_handler = signal_handler, .sa_flags = SA_RESTART};
    sigemptyset(&sa.sa_mask);
    for (int signo = 1; signo < NSIG; signo++) {
        if (!wanted[signo]) continue;
        int prev = -1;
        for (int i = 0; old && i < old->count; i++)
            if (old->signals[i] == signo) prev = i;
        if (prev >= 0) {
            w->old[w->count] = old->old[prev];
        } else {
            atomic_store(&signal_pending[signo], false);
            atomic_store(&signal_wake_fd[signo], l->wake_wfd + 1);
            if (sigaction(signo, &sa, &w->old[w->count]) != 0) {
                int err = errno;
                atomic_store(&signal_wake_fd[signo], 0);
                for (int i = 0; i < w->count; i++) {
                    int done = w->signals[i];
                    bool kept = false;
                    for (int j = 0; old && j < old->count; j++)
                        if (old->signals[j] == done) kept = true;
                    if (kept) continue;
                    sigaction(done, &w->old[i], NULL);
                    atomic_store(&signal_wake_fd[done], 0);
                }
                free(w);
                errno = err;
                return false;
            }
        }
        w->signals[w->count++] = signo;
    }

    /* Restore signals the old watch had that the new one drops. */
    for (int i = 0; old && i < old->count; i++) {
        int signo = old->signals[i];
        if (wanted[signo]) continue;
        sigaction(signo, &old->old[i], NULL);
        atomic_store(&signal_wake_fd[signo], 0);
        atomic_store(&signal_pending[signo], false);
        signal_owner[signo] = NULL;
    }
    for (int i = 0; i < w->count; i++)
        signal_owner[w->signals[i]] = l;
    free(old);
    l->signals = w;
    return true;
}

void yasl_loop_unwatch_signals(struct yasl_loop *l)
{
    if (!l || !l->signals) return;
    struct yasl_signal_watch *w = l->signals;
    l->signals = NULL;
    for (int i = 0; i < w->count; i++) {
        int signo = w->signals[i];
        sigaction(signo, &w->old[i], NULL);
        atomic_store(&signal_wake_fd[signo], 0);
        atomic_store(&signal_pending[signo], false);
        signal_owner[signo] = NULL;
    }
    free(w);
}

static void timer_cleanup_deleted(struct yasl_loop *l)
{
    if (!l || l->timer_firing_depth > 0) return;
    int dest = 0;
    for (int src = 0; src < l->timer_count; src++) {
        struct yasl_timer_source t = l->timers[src];
        l->timers[dest] = t;
        dest += !t.deleted;
    }
    l->timer_count = dest;
}

bool yasl_loop_timer_armed(const struct yasl_loop *l, yasl_timer_id id)
{
    if (!l || id == 0) return false;
    for (int i = 0; i < l->timer_count; i++)
        if (l->timers[i].id == id)
            return !l->timers[i].deleted;
    return false;
}

int yasl_loop_poll_set(const struct yasl_loop *l, const struct pollfd **fds)
{
    if (!l) {
        *fds = NULL;
        return 0;
    }
    *fds = l->pollfds;
    return l->count;
}

int yasl_loop_next_timeout_ms(const struct yasl_loop *l)
{
    if (!l || l->timer_count <= 0) return -1;
    uint64_t now = yasl_loop_now_ns();
    uint64_t min_deadline = UINT64_MAX;
    bool has_active = false;
    for (int i = 0; i < l->timer_count; i++) {
        if (!l->timers[i].deleted) {
            has_active = true;
            if (l->timers[i].deadline_ns < min_deadline)
                min_deadline = l->timers[i].deadline_ns;
        }
    }
    if (!has_active) return -1;
    if (min_deadline <= now) return 0;
    uint64_t diff = min_deadline - now;
    uint64_t ms = (diff + 999999ull) / 1000000ull;
    return (ms > (uint64_t)INT_MAX) ? INT_MAX : (int)ms;
}

yasl_timer_id yasl_loop_add_timer(struct yasl_loop *l, uint64_t interval_ns,
                                  yasl_timer_cb cb, void *ud)
{
    if (!l || !cb || interval_ns == 0) return 0;

    if (l->timer_count >= l->timer_capacity) {
        int new_cap = l->timer_capacity ? l->timer_capacity * 2 : 16;
        struct yasl_timer_source *new_timers = realloc(l->timers, (size_t)new_cap * sizeof(*new_timers));
        if (!new_timers) return 0;
        l->timers = new_timers;
        l->timer_capacity = new_cap;
    }

    uint32_t id = l->next_timer_id++;
    if (l->next_timer_id == 0) l->next_timer_id = 1;

    uint64_t now = yasl_loop_now_ns();
    l->timers[l->timer_count++] = (struct yasl_timer_source){
        .fn = cb,
        .ud = ud,
        .id = id,
        .interval_ns = interval_ns,
        .deadline_ns = now + interval_ns,
        .deleted = false,
    };

    yasl_loop_kick(l);
    return id;
}

bool yasl_loop_del_timer(struct yasl_loop *l, yasl_timer_id id)
{
    if (!l || id == 0) return false;
    for (int i = l->timer_count - 1; i >= 0; i--) {
        if (l->timers[i].id == id && !l->timers[i].deleted) {
            l->timers[i].deleted = true;
            if (l->timer_firing_depth == 0) {
                if (i == l->timer_count - 1) {
                    while (l->timer_count > 0 && l->timers[l->timer_count - 1].deleted)
                        l->timer_count--;
                } else {
                    timer_cleanup_deleted(l);
                }
            }
            return true;
        }
    }
    return false;
}

int yasl_loop_step(struct yasl_loop *l, int timeout_ms)
{
    if (!l) return -1;

    int timer_to = yasl_loop_next_timeout_ms(l);
    int effective_timeout = timeout_ms;
    if (timer_to >= 0) {
        if (effective_timeout < 0 || timer_to < effective_timeout)
            effective_timeout = timer_to;
    }

    if (l->count <= 0) {
        if (effective_timeout > 0) {
            struct timespec req = {
                .tv_sec = effective_timeout / 1000,
                .tv_nsec = (long)(effective_timeout % 1000) * 1000000L,
            };
            nanosleep(&req, NULL);
        }
    } else {
        int n;
        do {
            n = poll(l->pollfds, (nfds_t)l->count, effective_timeout);
        } while (n < 0 && errno == EINTR);

        if (n < 0) return -1;
    }

    int handled = 0;

    if (l->count > 0) {
        int count = l->count;
        struct yasl_fd_source sstack[32];
        struct pollfd pstack[32];
        struct yasl_fd_source *scopy = (count <= 32) ? sstack : malloc((size_t)count * sizeof(*scopy));
        struct pollfd *pcopy = (count <= 32) ? pstack : malloc((size_t)count * sizeof(*pcopy));
        if (!scopy || !pcopy) {
            if (scopy && scopy != sstack) free(scopy);
            if (pcopy && pcopy != pstack) free(pcopy);
            return -1;
        }

        memcpy(scopy, l->fds, (size_t)count * sizeof(*scopy));
        memcpy(pcopy, l->pollfds, (size_t)count * sizeof(*pcopy));

        for (int i = 0; i < count; i++) {
            if (pcopy[i].revents) {
                short revents = pcopy[i].revents;
                // Fast-path: check if source is still at index i before linear scan
                bool valid = false;
                if (i < l->count && l->fds[i].id == scopy[i].id) {
                    l->pollfds[i].revents = 0;
                    valid = true;
                } else {
                    for (int j = 0; j < l->count; j++) {
                        if (l->fds[j].id == scopy[i].id) {
                            l->pollfds[j].revents = 0;
                            valid = true;
                            break;
                        }
                    }
                }
                if (valid && scopy[i].fn) {
                    scopy[i].fn(scopy[i].fd, revents, scopy[i].ud);
                    handled++;
                }
            }
        }

        if (scopy != sstack) free(scopy);
        if (pcopy != pstack) free(pcopy);
    }

    if (l->timer_count > 0) {
        uint64_t now = yasl_loop_now_ns();
        l->timer_firing_depth++;
        int initial_count = l->timer_count;
        for (int i = 0; i < initial_count; i++) {
            if (i >= l->timer_count) break;
            if (l->timers[i].deleted || l->timers[i].deadline_ns > now)
                continue;

            yasl_timer_id tid = l->timers[i].id;
            uint64_t interval = l->timers[i].interval_ns;
            void *ud = l->timers[i].ud;
            struct yasl_timer_source timer_src = l->timers[i];

            uint64_t next_interval = timer_src.fn(ud, tid, interval);
            handled++;

            // Fast-path: timer is usually still at index i unless an earlier timer was deleted
            if (i < l->timer_count && l->timers[i].id == tid) {
                if (next_interval == 0 || l->timers[i].deleted) {
                    l->timers[i].deleted = true;
                } else {
                    l->timers[i].interval_ns = next_interval;
                    if (l->timers[i].deadline_ns + next_interval > now)
                        l->timers[i].deadline_ns += next_interval;
                    else
                        l->timers[i].deadline_ns = now + next_interval;
                }
            } else {
                for (int j = 0; j < l->timer_count; j++) {
                    if (l->timers[j].id == tid) {
                        if (next_interval == 0 || l->timers[j].deleted) {
                            l->timers[j].deleted = true;
                        } else {
                            l->timers[j].interval_ns = next_interval;
                            if (l->timers[j].deadline_ns + next_interval > now)
                                l->timers[j].deadline_ns += next_interval;
                            else
                                l->timers[j].deadline_ns = now + next_interval;
                        }
                        break;
                    }
                }
            }
        }
        l->timer_firing_depth--;
        if (l->timer_firing_depth == 0) {
            timer_cleanup_deleted(l);
        }
    }

    return handled;
}

int yasl_loop_run(struct yasl_loop *l)
{
    if (!l) return -1;
    l->running = true;
    while (l->running) {
        if (yasl_loop_step(l, -1) < 0)
            return -1;
    }
    return 0;
}

void yasl_loop_quit(struct yasl_loop *l)
{
    if (l) l->running = false;
}