//
// Copyright 2012 Josh Blum
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with io_sig program.  If not, see <http://www.gnu.org/licenses/>.

#include "element_impl.hpp"
#include <gras_impl/messages.hpp>
#include <boost/foreach.hpp>
#include <algorithm>

#define REALLY_BIG size_t(1 << 30)

using namespace gnuradio;

void ElementImpl::mark_done(const tsbe::TaskInterface &task_iface)
{
    if (this->block_state == BLOCK_STATE_DONE) return; //can re-enter checking done first

    //flush partial output buffers to the downstream
    for (size_t i = 0; i < task_iface.get_num_outputs(); i++)
    {
        if (not this->output_queues.ready(i)) continue;
        SBuffer &buff = this->output_queues.front(i);
        if (buff.length == 0) continue;
        task_iface.post_downstream(i, buff);
        this->output_queues.pop(i);
    }

    this->interruptible_thread.reset();

    //mark down the new state
    this->block_state = BLOCK_STATE_DONE;

    //release upstream, downstream, and executor tokens
    this->token_pool.clear();

    //release allocator tokens, buffers can now call deleters
    this->output_buffer_tokens.clear();

    //release all buffers in queues
    this->input_queues.flush_all();
    this->output_queues.flush_all();

    //tell the upstream and downstram to re-check their tokens
    //this is how the other blocks know who is interested,
    //and can decide based on interest to set done or not
    for (size_t i = 0; i < task_iface.get_num_inputs(); i++)
    {
        task_iface.post_upstream(i, CheckTokensMessage());
    }
    for (size_t i = 0; i < task_iface.get_num_outputs(); i++)
    {
        task_iface.post_downstream(i, CheckTokensMessage());
    }

    if (ARMAGEDDON) std::cerr
        << "==================================================\n"
        << "== The " << name << " is done...\n"
        << "==================================================\n"
        << std::flush;
}

void ElementImpl::handle_task(const tsbe::TaskInterface &task_iface)
{
    #ifdef WORK_DEBUG
    WorkDebugPrinter(this->name);
    #endif

    //------------------------------------------------------------------
    //-- Decide if its possible to continue any processing:
    //-- Handle task may get called for incoming buffers,
    //-- however, not all ports may have available buffers.
    //------------------------------------------------------------------
    if (not(
        this->block_state == BLOCK_STATE_LIVE and
        this->input_queues.all_ready() and
        this->output_queues.all_ready()
    )) return;

    const size_t num_inputs = task_iface.get_num_inputs();
    const size_t num_outputs = task_iface.get_num_outputs();
    //const bool is_source = (num_inputs == 0);
    //const bool is_sink = (num_outputs == 0);
    this->work_io_ptr_mask = 0; //reset

    //------------------------------------------------------------------
    //-- sort the input tags before working
    //------------------------------------------------------------------
    for (size_t i = 0; i < num_inputs; i++)
    {
        if (not this->input_tags_changed[i]) continue;
        std::vector<Tag> &tags_i = this->input_tags[i];
        std::sort(tags_i.begin(), tags_i.end(), Tag::offset_compare);
        this->input_tags_changed[i] = false;
    }

    //------------------------------------------------------------------
    //-- initialize input buffers before work
    //------------------------------------------------------------------
    size_t num_input_items = REALLY_BIG; //so big that it must std::min
    bool inputs_done = false;
    size_t output_inline_index = 0;
    for (size_t i = 0; i < num_inputs; i++)
    {
        inputs_done = inputs_done or this->input_tokens[i].unique();

        ASSERT(this->input_queues.ready(i));
        bool potential_inline;
        const SBuffer buff = this->input_queues.front(i, potential_inline);
        void *mem = buff.get();
        const size_t items = buff.length/this->input_items_sizes[i];

        this->work_io_ptr_mask |= ptrdiff_t(mem);
        this->input_items[i]._mem = mem;
        this->input_items[i]._len = items;
        this->work_input_items[i] = mem;
        this->work_ninput_items[i] = items;
        num_input_items = std::min(num_input_items, items);
        this->consume_called[i] = false;

        //inline dealings, how and when input buffers can be inlined into output buffers
        //continue;
        if (
            potential_inline and
            input_inline_enables[i] and
            output_inline_index < num_outputs and
            buff.get_affinity() == this->buffer_affinity
        ){
            //copy buffer reference but push with zero length, same offset
            SBuffer new_obuff = buff;
            new_obuff.length = 0;
            this->output_queues.push_front(i, new_obuff); //you got inlined!
            output_inline_index++; //done do this output port again
        }
    }

    //------------------------------------------------------------------
    //-- initialize output buffers before work
    //------------------------------------------------------------------
    size_t num_output_items = REALLY_BIG; //so big that it must std::min
    bool outputs_done = false;
    for (size_t i = 0; i < num_outputs; i++)
    {
        outputs_done = outputs_done or this->output_tokens[i].unique();

        ASSERT(this->output_queues.ready(i));
        const SBuffer &buff = this->output_queues.front(i);
        void *mem = buff.get(buff.length);
        const size_t bytes = buff.get_actual_length() - buff.length - buff.offset;
        const size_t items = bytes/this->output_items_sizes[i];

        this->work_io_ptr_mask |= ptrdiff_t(mem);
        this->output_items[i]._mem = mem;
        this->output_items[i]._len = items;
        this->work_output_items[i] = mem;
        num_output_items = std::min(num_output_items, items);
    }

    //if we have outputs and at least one port has no downstream subscibers, mark done
    if (outputs_done)
    {
        this->mark_done(task_iface);
        return;
    }

    //------------------------------------------------------------------
    //-- forecast
    //------------------------------------------------------------------
    if (this->forecast_enable)
    {
        forecast_again_you_jerk:
        fcast_ninput_items = work_ninput_items;
        block_ptr->forecast(num_output_items, fcast_ninput_items);
        for (size_t i = 0; i < num_inputs; i++)
        {
            if (fcast_ninput_items[i] <= work_ninput_items[i]) continue;

            num_output_items = num_output_items/2; //backoff regime
            if (num_output_items) goto forecast_again_you_jerk;

            this->forecast_fail = true;
            this->conclusion(task_iface, inputs_done);
            return;
        }
    }
    this->forecast_fail = false;

    //------------------------------------------------------------------
    //-- the work
    //------------------------------------------------------------------
    work_noutput_items = num_output_items;
    if (this->enable_fixed_rate) work_noutput_items = std::min(
        work_noutput_items, myulround((num_input_items)*this->relative_rate));
    this->work_task_iface = task_iface;
    this->work_ret = -1;
    if (this->interruptible_thread)
    {
        this->interruptible_thread->call();
    }
    else
    {
        this->task_work();
    }
    this->work_task_iface.reset();
    const size_t noutput_items = size_t(work_ret);

    if (work_ret == Block::WORK_DONE)
    {
        this->mark_done(task_iface);
        return;
    }

    //------------------------------------------------------------------
    //-- process input consumption
    //------------------------------------------------------------------
    bool input_allows_flush = true;
    for (size_t i = 0; i < num_inputs; i++)
    {
        ASSERT(enable_fixed_rate or work_ret != Block::WORK_CALLED_PRODUCE);
        const size_t items = (this->consume_called[i])? this->consume_items[i] : (myulround((noutput_items/this->relative_rate)));
        this->consume_items[i] = 0;

        this->items_consumed[i] += items;
        const size_t bytes = items*this->input_items_sizes[i];
        input_allows_flush = input_allows_flush and this->input_queues.consume(i, bytes);
    }

    //------------------------------------------------------------------
    //-- process output production
    //------------------------------------------------------------------
    for (size_t i = 0; i < num_outputs; i++)
    {
        const size_t items = (work_ret == Block::WORK_CALLED_PRODUCE)? this->produce_items[i] : noutput_items;
        this->produce_items[i] = 0;
        if (items == 0) continue;

        SBuffer &buff = this->output_queues.front(i);
        this->items_produced[i] += items;
        const size_t bytes = items*this->output_items_sizes[i];
        buff.length += bytes;

        //only pass output buffer downstream when the input is fully consumed...
        //Reasoning: For the sake of dealling with history, we can process the mini history input buffer,
        //and then call work again on the real input buffer, but still yield one output buffer per input buffer.
        if (input_allows_flush)
        {
            task_iface.post_downstream(i, buff);
            this->output_queues.pop(i);
        }
    }

    //------------------------------------------------------------------
    //-- trim the input tags that are past the consumption zone
    //-- and post trimmed tags to the downstream based on policy
    //------------------------------------------------------------------
    for (size_t i = 0; i < num_inputs; i++)
    {
        std::vector<Tag> &tags_i = this->input_tags[i];
        const size_t items_consumed_i = this->items_consumed[i];
        size_t last = 0;
        while (last < tags_i.size() and tags_i[last].offset < items_consumed_i)
        {
            last++;
        }

        //follow the tag propagation policy before erasure
        switch (this->tag_prop_policy)
        {
        case Block::TPP_DONT: break; //well that was ez
        case Block::TPP_ALL_TO_ALL:
            for (size_t out_i = 0; out_i < num_outputs; out_i++)
            {
                for (size_t tag_i = 0; tag_i < last; tag_i++)
                {
                    Tag t = tags_i[tag_i];
                    t.offset = myullround(t.offset * this->relative_rate);
                    task_iface.post_downstream(out_i, t);
                }
            }
            break;
        case Block::TPP_ONE_TO_ONE:
            if (i < num_outputs)
            {
                for (size_t tag_i = 0; tag_i < last; tag_i++)
                {
                    Tag t = tags_i[tag_i];
                    t.offset = myullround(t.offset * this->relative_rate);
                    task_iface.post_downstream(i, t);
                }
            }
            break;
        };

        //now its safe to perform the erasure
        if (last != 0) tags_i.erase(tags_i.begin(), tags_i.begin()+last);
    }

    //------------------------------------------------------------------
    //-- Message self based on post-work conditions
    //------------------------------------------------------------------
    this->conclusion(task_iface, inputs_done);
}

inline void ElementImpl::conclusion(const tsbe::TaskInterface &task_iface, const bool inputs_done)
{
    //if there are inputs, and not all are provided for,
    //tell the block to check input queues and handle done
    if (inputs_done)
    {
        this->block.post_msg(CheckTokensMessage());
        return;
    }

    //still have IO ready? kick off another task
    if (this->input_queues.all_ready() and this->output_queues.all_ready())
    {
        this->block.post_msg(SelfKickMessage());
        return;
    }
}