pymupdf提取pdf表格

pymupdf提取pdf表格非常快速，相比其他库是个更好的选择．

一个行列多的表格打印成pdf后会由于页宽分页原因变成多个表格，提取的多个表格需要合并为一个表格，再来处理数据．

下面代码中merge函数用于合并表格．addOneAxis0用于合并仅分页导致的多子表．

def addOneAxis0(data,one):if len(data)==0:#firstr=oneelse:r=data+one[1:]#remove first title rowreturn r    def merge(data):output=data[0]titles=data[0][0][1:]for i in range(1,len(data)):newtitle=data[i][0][1:]# print(titles)# print(newtitle)# input("here")if newtitle[0] in titles:#repeatfor one in data[i][1:]:#add rowoutput.append(one)else:#if newtitle[0] in output[0]:#add column without titlethedata=data[i][1:]#remove titlen=len(thedata)out=output[-n:]for j in range(len(out)):out[j]+=thedata[j][1:]passelse:#add column with titlen=len(data[i])out=output[-n:]for j in range(len(out)):#add columnout[j]+=data[i][j][1:]passreturn output
def getDataMass(file_name):global doc,curvePagedoc=fitz.open(file_name) # open documentdata=[]mass=[]tables=[]i=Nonefor i in range(doc.page_count).__reversed__():page = doc[i]tabs=page.find_tables()if len(tabs.tables)==0:breakelse:tables.append(tabs[0].extract())tables.reverse()for one in tables:if one[0][0]=="溶液标签":data.append(one)#data.append(one)else:mass=addOneAxis0(mass,one)curvePage=iprint(data)data=merge(data)print(data)return (data,mass)