CodePattern/13 多计算单元/数据分包/多进程.py

# -*- coding: utf-8 -*-
import multiprocessing
from collections import Counter
from cppy.cp_util import *


#
# 多进程: 因为创建进程相比计算过程开销太大，结果最慢
#
stop_words = get_stopwords()

def process_chunk(chunk):
    # 过滤停用词    
    words = [ w for w in chunk if ( not w in stop_words ) and len(w) >= 3 ]
    return Counter(words)

def merge_counts(counts_list):
    # 合并多个Counter对象
    total_counts = Counter()
    for counts in counts_list:
        total_counts += counts
    return total_counts


@timing_decorator
def main():
    # 读取文件内容，分割文件内容为多个块，每个块由一个进程处理    
    chunks = get_chunks(testfilepath,1000)

    # 使用多进程处理每个块
    pool = multiprocessing.Pool(processes=multiprocessing.cpu_count())
    counts_list = pool.map(process_chunk, chunks)
    pool.close()
    pool.join()

    # 合并计数
    total_counts = merge_counts(counts_list)

    # 输出最高频的n个词
    print_word_freqs(total_counts.most_common(10))


if __name__ == '__main__':
    main()