ragflow/rag/svr/task_executor.py

#
#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
#  Licensed under the Apache License, Version 2.0 (the "License");
#  you may not use this file except in compliance with the License.
#  You may obtain a copy of the License at
#
#      http://www.apache.org/licenses/LICENSE-2.0
#
#  Unless required by applicable law or agreed to in writing, software
#  distributed under the License is distributed on an "AS IS" BASIS,
#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
#  See the License for the specific language governing permissions and
#  limitations under the License.

# from beartype import BeartypeConf
# from beartype.claw import beartype_all  # <-- you didn't sign up for this
# beartype_all(conf=BeartypeConf(violation_type=UserWarning))    # <-- emit warnings from all code

import sys
from api.utils.log_utils import initRootLogger
CONSUMER_NO = "0" if len(sys.argv) < 2 else sys.argv[1]
CONSUMER_NAME = "task_executor_" + CONSUMER_NO
initRootLogger(CONSUMER_NAME)

import logging
import os
from datetime import datetime
import json
import hashlib
import copy
import re
import time
import threading
from functools import partial
from io import BytesIO
from multiprocessing.context import TimeoutError
from timeit import default_timer as timer
import tracemalloc

import numpy as np

from api.db import LLMType, ParserType
from api.db.services.dialog_service import keyword_extraction, question_proposal
from api.db.services.document_service import DocumentService
from api.db.services.llm_service import LLMBundle
from api.db.services.task_service import TaskService
from api.db.services.file2document_service import File2DocumentService
from api import settings
from api.versions import get_ragflow_version
from api.db.db_models import close_connection
from rag.app import laws, paper, presentation, manual, qa, table, book, resume, picture, naive, one, audio, \
    knowledge_graph, email
from rag.nlp import search, rag_tokenizer
from rag.raptor import RecursiveAbstractiveProcessing4TreeOrganizedRetrieval as Raptor
from rag.settings import DOC_MAXIMUM_SIZE, SVR_QUEUE_NAME, print_rag_settings
from rag.utils import rmSpace, num_tokens_from_string
from rag.utils.redis_conn import REDIS_CONN, Payload
from rag.utils.storage_factory import STORAGE_IMPL

BATCH_SIZE = 64

FACTORY = {
    "general": naive,
    ParserType.NAIVE.value: naive,
    ParserType.PAPER.value: paper,
    ParserType.BOOK.value: book,
    ParserType.PRESENTATION.value: presentation,
    ParserType.MANUAL.value: manual,
    ParserType.LAWS.value: laws,
    ParserType.QA.value: qa,
    ParserType.TABLE.value: table,
    ParserType.RESUME.value: resume,
    ParserType.PICTURE.value: picture,
    ParserType.ONE.value: one,
    ParserType.AUDIO.value: audio,
    ParserType.EMAIL.value: email,
    ParserType.KG.value: knowledge_graph
}

CONSUMER_NAME = "task_consumer_" + CONSUMER_NO
PAYLOAD: Payload | None = None
BOOT_AT = datetime.now().isoformat()
PENDING_TASKS = 0
LAG_TASKS = 0

mt_lock = threading.Lock()
DONE_TASKS = 0
FAILED_TASKS = 0
CURRENT_TASK = None


def set_progress(task_id, from_page=0, to_page=-1, prog=None, msg="Processing..."):
    global PAYLOAD
    if prog is not None and prog < 0:
        msg = "[ERROR]" + msg
    cancel = TaskService.do_cancel(task_id)
    if cancel:
        msg += " [Canceled]"
        prog = -1

    if to_page > 0:
        if msg:
            msg = f"Page({from_page + 1}~{to_page + 1}): " + msg
    d = {"progress_msg": msg}
    if prog is not None:
        d["progress"] = prog
    try:
        logging.info(f"set_progress({task_id}), progress: {prog}, progress_msg: {msg}")
        TaskService.update_progress(task_id, d)
    except Exception:
        logging.exception(f"set_progress({task_id}) got exception")

    close_connection()
    if cancel:
        if PAYLOAD:
            PAYLOAD.ack()
            PAYLOAD = None
        os._exit(0)


def collect():
    global CONSUMER_NAME, PAYLOAD, DONE_TASKS, FAILED_TASKS
    try:
        PAYLOAD = REDIS_CONN.get_unacked_for(CONSUMER_NAME, SVR_QUEUE_NAME, "rag_flow_svr_task_broker")
        if not PAYLOAD:
            PAYLOAD = REDIS_CONN.queue_consumer(SVR_QUEUE_NAME, "rag_flow_svr_task_broker", CONSUMER_NAME)
        if not PAYLOAD:
            time.sleep(1)
            return None
    except Exception:
        logging.exception("Get task event from queue exception")
        return None

    msg = PAYLOAD.get_message()
    if not msg:
        return None

    if TaskService.do_cancel(msg["id"]):
        with mt_lock:
            DONE_TASKS += 1
        logging.info("Task {} has been canceled.".format(msg["id"]))
        return None
    task = TaskService.get_task(msg["id"])
    if not task:
        with mt_lock:
            DONE_TASKS += 1
        logging.warning("{} empty task!".format(msg["id"]))
        return None

    if msg.get("type", "") == "raptor":
        task["task_type"] = "raptor"
    return task


def get_storage_binary(bucket, name):
    return STORAGE_IMPL.get(bucket, name)


def build_chunks(task, progress_callback):
    if task["size"] > DOC_MAXIMUM_SIZE:
        set_progress(task["id"], prog=-1, msg="File size exceeds( <= %dMb )" %
                                              (int(DOC_MAXIMUM_SIZE / 1024 / 1024)))
        return []

    chunker = FACTORY[task["parser_id"].lower()]
    try:
        st = timer()
        bucket, name = File2DocumentService.get_storage_address(doc_id=task["doc_id"])
        binary = get_storage_binary(bucket, name)
        logging.info("From minio({}) {}/{}".format(timer() - st, task["location"], task["name"]))
    except TimeoutError:
        progress_callback(-1, "Internal server error: Fetch file from minio timeout. Could you try it again.")
        logging.exception("Minio {}/{} got timeout: Fetch file from minio timeout.".format(task["location"], task["name"]))
        raise
    except Exception as e:
        if re.search("(No such file|not found)", str(e)):
            progress_callback(-1, "Can not find file <%s> from minio. Could you try it again?" % task["name"])
        else:
            progress_callback(-1, "Get file from minio: %s" % str(e).replace("'", ""))
        logging.exception("Chunking {}/{} got exception".format(task["location"], task["name"]))
        raise

    try:
        cks = chunker.chunk(task["name"], binary=binary, from_page=task["from_page"],
                            to_page=task["to_page"], lang=task["language"], callback=progress_callback,
                            kb_id=task["kb_id"], parser_config=task["parser_config"], tenant_id=task["tenant_id"])
        logging.info("Chunking({}) {}/{} done".format(timer() - st, task["location"], task["name"]))
    except Exception as e:
        progress_callback(-1, "Internal server error while chunking: %s" % str(e).replace("'", ""))
        logging.exception("Chunking {}/{} got exception".format(task["location"], task["name"]))
        raise

    docs = []
    doc = {
        "doc_id": task["doc_id"],
        "kb_id": str(task["kb_id"])
    }
    if task["pagerank"]:
        doc["pagerank_fea"] = int(task["pagerank"])
    el = 0
    for ck in cks:
        d = copy.deepcopy(doc)
        d.update(ck)
        md5 = hashlib.md5()
        md5.update((ck["content_with_weight"] +
                    str(d["doc_id"])).encode("utf-8"))
        d["id"] = md5.hexdigest()
        d["create_time"] = str(datetime.now()).replace("T", " ")[:19]
        d["create_timestamp_flt"] = datetime.now().timestamp()
        if not d.get("image"):
            _ = d.pop("image", None)
            d["img_id"] = ""
            d["page_num_int"] = []
            d["position_int"] = []
            d["top_int"] = []
            docs.append(d)
            continue

        try:
            output_buffer = BytesIO()
            if isinstance(d["image"], bytes):
                output_buffer = BytesIO(d["image"])
            else:
                d["image"].save(output_buffer, format='JPEG')

            st = timer()
            STORAGE_IMPL.put(task["kb_id"], d["id"], output_buffer.getvalue())
            el += timer() - st
        except Exception:
            logging.exception("Saving image of chunk {}/{}/{} got exception".format(task["location"], task["name"], d["_id"]))
            raise

        d["img_id"] = "{}-{}".format(task["kb_id"], d["id"])
        del d["image"]
        docs.append(d)
    logging.info("MINIO PUT({}):{}".format(task["name"], el))

    if task["parser_config"].get("auto_keywords", 0):
        st = timer()
        progress_callback(msg="Start to generate keywords for every chunk ...")
        chat_mdl = LLMBundle(task["tenant_id"], LLMType.CHAT, llm_name=task["llm_id"], lang=task["language"])
        for d in docs:
            d["important_kwd"] = keyword_extraction(chat_mdl, d["content_with_weight"],
                                                    task["parser_config"]["auto_keywords"]).split(",")
            d["important_tks"] = rag_tokenizer.tokenize(" ".join(d["important_kwd"]))
        progress_callback(msg="Keywords generation completed in {:.2f}s".format(timer() - st))

    if task["parser_config"].get("auto_questions", 0):
        st = timer()
        progress_callback(msg="Start to generate questions for every chunk ...")
        chat_mdl = LLMBundle(task["tenant_id"], LLMType.CHAT, llm_name=task["llm_id"], lang=task["language"])
        for d in docs:
            d["question_kwd"] = question_proposal(chat_mdl, d["content_with_weight"], task["parser_config"]["auto_questions"]).split("\n")
            d["question_tks"] = rag_tokenizer.tokenize("\n".join(d["question_kwd"]))
        progress_callback(msg="Question generation completed in {:.2f}s".format(timer() - st))

    return docs


def init_kb(row, vector_size: int):
    idxnm = search.index_name(row["tenant_id"])
    return settings.docStoreConn.createIdx(idxnm, row.get("kb_id",""), vector_size)


def embedding(docs, mdl, parser_config=None, callback=None):
    if parser_config is None:
        parser_config = {}
    batch_size = 16
    tts, cnts = [], []
    for d in docs:
        tts.append(rmSpace(d.get("docnm_kwd", "Title")))
        c = "\n".join(d.get("question_kwd", []))
        if not c:
            c = d["content_with_weight"]
        c = re.sub(r"</?(table|td|caption|tr|th)( [^<>]{0,12})?>", " ", c)
        cnts.append(c)

    tk_count = 0
    if len(tts) == len(cnts):
        tts_ = np.array([])
        for i in range(0, len(tts), batch_size):
            vts, c = mdl.encode(tts[i: i + batch_size])
            if len(tts_) == 0:
                tts_ = vts
            else:
                tts_ = np.concatenate((tts_, vts), axis=0)
            tk_count += c
            callback(prog=0.6 + 0.1 * (i + 1) / len(tts), msg="")
        tts = tts_

    cnts_ = np.array([])
    for i in range(0, len(cnts), batch_size):
        vts, c = mdl.encode(cnts[i: i + batch_size])
        if len(cnts_) == 0:
            cnts_ = vts
        else:
            cnts_ = np.concatenate((cnts_, vts), axis=0)
        tk_count += c
        callback(prog=0.7 + 0.2 * (i + 1) / len(cnts), msg="")
    cnts = cnts_

    title_w = float(parser_config.get("filename_embd_weight", 0.1))
    vects = (title_w * tts + (1 - title_w) *
             cnts) if len(tts) == len(cnts) else cnts

    assert len(vects) == len(docs)
    vector_size = 0
    for i, d in enumerate(docs):
        v = vects[i].tolist()
        vector_size = len(v)
        d["q_%d_vec" % len(v)] = v
    return tk_count, vector_size


def run_raptor(row, chat_mdl, embd_mdl, callback=None):
    vts, _ = embd_mdl.encode(["ok"])
    vector_size = len(vts[0])
    vctr_nm = "q_%d_vec" % vector_size
    chunks = []
    for d in settings.retrievaler.chunk_list(row["doc_id"], row["tenant_id"], [str(row["kb_id"])],
                                             fields=["content_with_weight", vctr_nm]):
        chunks.append((d["content_with_weight"], np.array(d[vctr_nm])))

    raptor = Raptor(
        row["parser_config"]["raptor"].get("max_cluster", 64),
        chat_mdl,
        embd_mdl,
        row["parser_config"]["raptor"]["prompt"],
        row["parser_config"]["raptor"]["max_token"],
        row["parser_config"]["raptor"]["threshold"]
    )
    original_length = len(chunks)
    chunks = raptor(chunks, row["parser_config"]["raptor"]["random_seed"], callback)
    doc = {
        "doc_id": row["doc_id"],
        "kb_id": [str(row["kb_id"])],
        "docnm_kwd": row["name"],
        "title_tks": rag_tokenizer.tokenize(row["name"])
    }
    if row["pagerank"]:
        doc["pagerank_fea"] = int(row["pagerank"])
    res = []
    tk_count = 0
    for content, vctr in chunks[original_length:]:
        d = copy.deepcopy(doc)
        md5 = hashlib.md5()
        md5.update((content + str(d["doc_id"])).encode("utf-8"))
        d["id"] = md5.hexdigest()
        d["create_time"] = str(datetime.now()).replace("T", " ")[:19]
        d["create_timestamp_flt"] = datetime.now().timestamp()
        d[vctr_nm] = vctr.tolist()
        d["content_with_weight"] = content
        d["content_ltks"] = rag_tokenizer.tokenize(content)
        d["content_sm_ltks"] = rag_tokenizer.fine_grained_tokenize(d["content_ltks"])
        res.append(d)
        tk_count += num_tokens_from_string(content)
    return res, tk_count, vector_size


def do_handle_task(task):
    task_id = task["id"]
    task_from_page = task["from_page"]
    task_to_page = task["to_page"]
    task_tenant_id = task["tenant_id"]
    task_embedding_id = task["embd_id"]
    task_language = task["language"]
    task_llm_id = task["llm_id"]
    task_dataset_id = task["kb_id"]
    task_doc_id = task["doc_id"]
    task_document_name = task["name"]
    task_parser_config = task["parser_config"]

    # prepare the progress callback function
    progress_callback = partial(set_progress, task_id, task_from_page, task_to_page)
    try:
        # bind embedding model
        embedding_model = LLMBundle(task_tenant_id, LLMType.EMBEDDING, llm_name=task_embedding_id, lang=task_language)
    except Exception as e:
        error_message = f'Fail to bind embedding model: {str(e)}'
        progress_callback(-1, msg=error_message)
        logging.exception(error_message)
        raise

    # Either using RAPTOR or Standard chunking methods
    if task.get("task_type", "") == "raptor":
        try:
            # bind LLM for raptor
            chat_model = LLMBundle(task_tenant_id, LLMType.CHAT, llm_name=task_llm_id, lang=task_language)

            # run RAPTOR
            chunks, token_count, vector_size = run_raptor(task, chat_model, embedding_model, progress_callback)
        except Exception as e:
            error_message = f'Fail to bind LLM used by RAPTOR: {str(e)}'
            progress_callback(-1, msg=error_message)
            logging.exception(error_message)
            raise
    else:
        # Standard chunking methods
        start_ts = timer()
        chunks = build_chunks(task, progress_callback)
        logging.info("Build document {}: {:.2f}s".format(task_document_name, timer() - start_ts))
        if chunks is None:
            return
        if not chunks:
            progress_callback(1., msg=f"No chunk built from {task_document_name}")
            return
        # TODO: exception handler
        ## set_progress(task["did"], -1, "ERROR: ")
        progress_callback(msg="Generate {} chunks".format(len(chunks)))
        start_ts = timer()
        try:
            token_count, vector_size = embedding(chunks, embedding_model, task_parser_config, progress_callback)
        except Exception as e:
            error_message = "Generate embedding error:{}".format(str(e))
            progress_callback(-1, error_message)
            logging.exception(error_message)
            token_count = 0
            raise
        progress_message = "Embedding chunks ({:.2f}s)".format(timer() - start_ts)
        logging.info(progress_message)
        progress_callback(msg=progress_message)
    # logging.info(f"task_executor init_kb index {search.index_name(task_tenant_id)} embedding_model {embedding_model.llm_name} vector length {vector_size}")
    init_kb(task, vector_size)
    chunk_count = len(set([chunk["id"] for chunk in chunks]))
    start_ts = timer()
    doc_store_result = ""
    es_bulk_size = 4
    for b in range(0, len(chunks), es_bulk_size):
        doc_store_result = settings.docStoreConn.insert(chunks[b:b + es_bulk_size], search.index_name(task_tenant_id), task_dataset_id)
        if b % 128 == 0:
            progress_callback(prog=0.8 + 0.1 * (b + 1) / len(chunks), msg="")
    logging.info("Indexing {} elapsed: {:.2f}".format(task_document_name, timer() - start_ts))
    if doc_store_result:
        error_message = f"Insert chunk error: {doc_store_result}, please check log file and Elasticsearch/Infinity status!"
        progress_callback(-1, msg=error_message)
        settings.docStoreConn.delete({"doc_id": task_doc_id}, search.index_name(task_tenant_id), task_dataset_id)
        logging.error(error_message)
        raise Exception(error_message)

    if TaskService.do_cancel(task_id):
        settings.docStoreConn.delete({"doc_id": task_doc_id}, search.index_name(task_tenant_id), task_dataset_id)
        return

    DocumentService.increment_chunk_num(task_doc_id, task_dataset_id, token_count, chunk_count, 0)

    time_cost = timer() - start_ts
    progress_callback(prog=1.0, msg="Done ({:.2f}s)".format(time_cost))
    logging.info("Chunk doc({}), token({}), chunks({}), elapsed:{:.2f}".format(task_id, token_count, len(chunks), time_cost))


def handle_task():
    global PAYLOAD, mt_lock, DONE_TASKS, FAILED_TASKS, CURRENT_TASK
    task = collect()
    if task:
        try:
            logging.info(f"handle_task begin for task {json.dumps(task)}")
            with mt_lock:
                CURRENT_TASK = copy.deepcopy(task)
            do_handle_task(task)
            with mt_lock:
                DONE_TASKS += 1
                CURRENT_TASK = None
            logging.info(f"handle_task done for task {json.dumps(task)}")
        except Exception:
            with mt_lock:
                FAILED_TASKS += 1
                CURRENT_TASK = None
            logging.exception(f"handle_task got exception for task {json.dumps(task)}")
    if PAYLOAD:
        PAYLOAD.ack()
        PAYLOAD = None


def report_status():
    global CONSUMER_NAME, BOOT_AT, PENDING_TASKS, LAG_TASKS, mt_lock, DONE_TASKS, FAILED_TASKS, CURRENT_TASK
    REDIS_CONN.sadd("TASKEXE", CONSUMER_NAME)
    while True:
        try:
            now = datetime.now()
            group_info = REDIS_CONN.queue_info(SVR_QUEUE_NAME, "rag_flow_svr_task_broker")
            if group_info is not None:
                PENDING_TASKS = int(group_info["pending"])
                LAG_TASKS = int(group_info["lag"])

            with mt_lock:
                heartbeat = json.dumps({
                    "name": CONSUMER_NAME,
                    "now": now.isoformat(),
                    "boot_at": BOOT_AT,
                    "pending": PENDING_TASKS,
                    "lag": LAG_TASKS,
                    "done": DONE_TASKS,
                    "failed": FAILED_TASKS,
                    "current": CURRENT_TASK,
                })
            REDIS_CONN.zadd(CONSUMER_NAME, heartbeat, now.timestamp())
            logging.info(f"{CONSUMER_NAME} reported heartbeat: {heartbeat}")

            expired = REDIS_CONN.zcount(CONSUMER_NAME, 0, now.timestamp() - 60 * 30)
            if expired > 0:
                REDIS_CONN.zpopmin(CONSUMER_NAME, expired)
        except Exception:
            logging.exception("report_status got exception")
        time.sleep(30)


def analyze_heap(snapshot1: tracemalloc.Snapshot, snapshot2: tracemalloc.Snapshot, snapshot_id: int, dump_full: bool):
    msg = ""
    if dump_full:
        stats2 = snapshot2.statistics('lineno')
        msg += f"{CONSUMER_NAME} memory usage of snapshot {snapshot_id}:\n"
        for stat in stats2[:10]:
            msg += f"{stat}\n"
    stats1_vs_2 = snapshot2.compare_to(snapshot1, 'lineno')
    msg += f"{CONSUMER_NAME} memory usage increase from snapshot {snapshot_id - 1} to snapshot {snapshot_id}:\n"
    for stat in stats1_vs_2[:10]:
        msg += f"{stat}\n"
    msg += f"{CONSUMER_NAME} detailed traceback for the top memory consumers:\n"
    for stat in stats1_vs_2[:3]:
        msg += '\n'.join(stat.traceback.format())
    logging.info(msg)


def main():
    logging.info(r"""
  ______           __      ______                     __            
 /_  __/___ ______/ /__   / ____/  _____  _______  __/ /_____  _____
  / / / __ `/ ___/ //_/  / __/ | |/_/ _ \/ ___/ / / / __/ __ \/ ___/
 / / / /_/ (__  ) ,<    / /____>  </  __/ /__/ /_/ / /_/ /_/ / /    
/_/  \__,_/____/_/|_|  /_____/_/|_|\___/\___/\__,_/\__/\____/_/                               
    """)
    logging.info(f'TaskExecutor: RAGFlow version: {get_ragflow_version()}')
    settings.init_settings()
    print_rag_settings()
    background_thread = threading.Thread(target=report_status)
    background_thread.daemon = True
    background_thread.start()

    TRACE_MALLOC_DELTA = int(os.environ.get('TRACE_MALLOC_DELTA', "0"))
    TRACE_MALLOC_FULL = int(os.environ.get('TRACE_MALLOC_FULL', "0"))
    if TRACE_MALLOC_DELTA > 0:
        if TRACE_MALLOC_FULL < TRACE_MALLOC_DELTA:
            TRACE_MALLOC_FULL = TRACE_MALLOC_DELTA
        tracemalloc.start()
        snapshot1 = tracemalloc.take_snapshot()
    while True:
        handle_task()
        num_tasks = DONE_TASKS + FAILED_TASKS
        if TRACE_MALLOC_DELTA > 0 and num_tasks > 0 and num_tasks % TRACE_MALLOC_DELTA == 0:
            snapshot2 = tracemalloc.take_snapshot()
            analyze_heap(snapshot1, snapshot2, int(num_tasks / TRACE_MALLOC_DELTA), num_tasks % TRACE_MALLOC_FULL == 0)
            snapshot1 = snapshot2
            snapshot2 = None


if __name__ == "__main__":
    main()