Python实现简单HTML表格解析的方法

yipeiwu_com5年前Python基础

本文实例讲述了Python实现简单HTML表格解析的方法。分享给大家供大家参考。具体分析如下:

这里依赖libxml2dom,确保首先安装!导入到你的脚步并调用parse_tables() 函数。

1. source = a string containing the source code you can pass in just the table or the entire page code

2. headers = a list of ints OR a list of strings
If the headers are ints this is for tables with no header, just list the 0 based index of the rows in which you want to extract data.
If the headers are strings this is for tables with header columns (with the tags) it will pull the information from the specified columns

3. The 0 based index of the table in the source code. If there are multiple tables and the table you want to parse is the third table in the code then pass in the number 2 here

It will return a list of lists. each inner list will contain the parsed information.

具体代码如下:

#The goal of table parser is to get specific information from specific
#columns in a table.
#Input: source code from a typical website
#Arguments: a list of headers the user wants to return
#Output: A list of lists of the data in each row
import libxml2dom
def parse_tables(source, headers, table_index):
  """parse_tables(string source, list headers, table_index)
    headers may be a list of strings if the table has headers defined or
    headers may be a list of ints if no headers defined this will get data
    from the rows index.
    This method returns a list of lists
    """
  #Determine if the headers list is strings or ints and make sure they
  #are all the same type
  j = 0
  print 'Printing headers: ',headers
  #route to the correct function
  #if the header type is int
  if type(headers[0]) == type(1):
    #run no_header function
    return no_header(source, headers, table_index)
  #if the header type is string
  elif type(headers[0]) == type('a'):
    #run the header_given function
    return header_given(source, headers, table_index)
  else:
    #return none if the headers aren't correct
    return None
#This function takes in the source code of the whole page a string list of
#headers and the index number of the table on the page. It returns a list of
#lists with the scraped information
def header_given(source, headers, table_index):
  #initiate a list to hole the return list
  return_list = []
  #initiate a list to hold the index numbers of the data in the rows
  header_index = []
  #get a document object out of the source code
  doc = libxml2dom.parseString(source,html=1)
  #get the tables from the document
  tables = doc.getElementsByTagName('table')
  try:
    #try to get focue on the desired table
    main_table = tables[table_index]
  except:
    #if the table doesn't exits then return an error
    return ['The table index was not found']
  #get a list of headers in the table
  table_headers = main_table.getElementsByTagName('th')
  #need a sentry value for the header loop
  loop_sentry = 0
  #loop through each header looking for matches
  for header in table_headers:
    #if the header is in the desired headers list 
    if header.textContent in headers:
      #add it to the header_index
      header_index.append(loop_sentry)
    #add one to the loop_sentry
    loop_sentry+=1
  #get the rows from the table
  rows = main_table.getElementsByTagName('tr')
  #sentry value detecting if the first row is being viewed
  row_sentry = 0
  #loop through the rows in the table, skipping the first row
  for row in rows:
    #if row_sentry is 0 this is our first row
    if row_sentry == 0:
      #make the row_sentry not 0
      row_sentry = 1337
      continue
    #get all cells from the current row
    cells = row.getElementsByTagName('td')
    #initiate a list to append into the return_list
    cell_list = []
    #iterate through all of the header index's
    for i in header_index:
      #append the cells text content to the cell_list
      cell_list.append(cells[i].textContent)
    #append the cell_list to the return_list
    return_list.append(cell_list)
  #return the return_list
  return return_list
#This function takes in the source code of the whole page an int list of
#headers indicating the index number of the needed item and the index number
#of the table on the page. It returns a list of lists with the scraped info
def no_header(source, headers, table_index):
  #initiate a list to hold the return list
  return_list = []
  #get a document object out of the source code
  doc = libxml2dom.parseString(source, html=1)
  #get the tables from document
  tables = doc.getElementsByTagName('table')
  try:
    #Try to get focus on the desired table
    main_table = tables[table_index]
  except:
    #if the table doesn't exits then return an error
    return ['The table index was not found']
  #get all of the rows out of the main_table
  rows = main_table.getElementsByTagName('tr')
  #loop through each row
  for row in rows:
    #get all cells from the current row
    cells = row.getElementsByTagName('td')
    #initiate a list to append into the return_list
    cell_list = []
    #loop through the list of desired headers
    for i in headers:
      try:
        #try to add text from the cell into the cell_list
        cell_list.append(cells[i].textContent)
      except:
        #if there is an error usually an index error just continue
        continue
    #append the data scraped into the return_list    
    return_list.append(cell_list)
  #return the return list
  return return_list

希望本文所述对大家的Python程序设计有所帮助。

相关文章

基于python traceback实现异常的获取与处理

这篇文章主要介绍了基于python traceback实现异常的获取与处理,文中通过示例代码介绍的非常详细,对大家的学习或者工作具有一定的参考学习价值,需要的朋友可以参考下 1、trac...

Python中用psycopg2模块操作PostgreSQL方法

Python中用psycopg2模块操作PostgreSQL方法

其实在Python中可以用来连接PostgreSQL的模块很多,这里比较推荐psycopg2。psycopg2安装起来非常的简单(pip install psycopg2),这里主要重点...

Pytorch根据layers的name冻结训练方式

使用model.named_parameters()可以轻松搞定, model.cuda() # ######################################...

Python代码实现KNN算法

kNN算法是k-近邻算法的简称,主要用来进行分类实践,主要思路如下: 1.存在一个训练数据集,每个数据都有对应的标签,也就是说,我们知道样本集中每一数据和他对应的类别。 2.当输入一个...

Python Pandas 箱线图的实现

Python Pandas 箱线图的实现

各国家用户消费分布 import numpy as np import pandas as pd import matplotlib.pyplot as plt data...