Python calls crf interface + information extraction

Continued https://blog.csdn.net/m0_49621298/article/details/109896619

1. Python interface installation of crf++

Enter the command line under the CRF++-0.58\python path and execute the following command

python setup.py build
python setup.py install

2. Interface call

Error: ImportError: DLL load failed: %1 is not a valid Win32 application.

import CRFPP
  File "D:\Anaconda3\lib\site-packages\CRFPP.py", line 26, in <module>
    _CRFPP = swig_import_helper()
  File "D:\Anaconda3\lib\site-packages\CRFPP.py", line 22, in swig_import_helper
    _mod = imp.load_module('_CRFPP', fp, pathname, description)
  File "D:\Anaconda3\lib\imp.py", line 243, in load_module
    return load_dynamic(name, filename, file)
  File "D:\Anaconda3\lib\imp.py", line 343, in load_dynamic
    return _load(spec)
ImportError: DLL load failed: %1 不是有效的 Win32 应用程序。

After searching online, there is no solution. I thought of the file libcrfpp.dll, copied it and put it under \Anaconda3\Lib\site-packages, and no error was reported.

3. Model call and information extraction

Word segmentation and part-of-speech tagging use pyhanlp this time. For details, see the previous article https://blog.csdn.net/m0_49621298/article/details/109896619 . The code and results are as follows

def singlesentence(sentence,modelname):#单句执行
    print("句子：",sentence)
    CustomDictionary.add("%", "q 0")  # 自定义词性
    CustomDictionary.add("％", "q 0")  # 自定义词性
    sentence_seged = HanLP.segment(sentence)  # 分词、词性
    tagger = CRFPP.Tagger("-m"+modelname)
    tagger.clear()
    for x in sentence_seged:
        tagger.add(str(x.word)+ "\t"+str(x.nature))
    tagger.parse()
    size = tagger.size()
    xsize = tagger.xsize()
    prearr=[]#保存预测
    for i in range(0, size):
        temparr=[]
        for j in range(0, xsize):
            char = tagger.x(i, j)
            temparr.append(char)
        tag = tagger.y2(i)
        temparr.append(tag)
        prearr.append(temparr)
    # print(prearr)
    starr=[]#保存实体
    numarr=[]#保存数值
    dwarr=[]#保存单位
    for i in range(len(prearr)):
        # 搜索实体
        tempstr = ""
        if prearr[i][2] == "w":
            starr.append(prearr[i][0])
        if prearr[i][2] == "b":
            tempstr = tempstr + prearr[i][0]
            for j in range(i + 1, len(prearr)):
                if prearr[j][2] == "m" or prearr[j][2] == "e":
                    tempstr = tempstr + prearr[j][0]
                else:
                    break
            starr.append(tempstr)
        if prearr[i][1] == "m":
            if prearr[i-1][1]=='v':
                numarr.append([prearr[i-1][0],prearr[i][0]])
            else:
                numarr.append(prearr[i][0])
        if prearr[i][1] == "q" :
            dwarr.append(prearr[i][0])
    print("实体：",starr)
    print("数值：",numarr)
    print("单位：",dwarr)
    return starr,numarr,dwarr