Matlab爬虫程序

% 本程序用于获取网站中的表格


% written by longwen36
% all rights reserved 


clc,clear;


warning off;


for year = 1991:1992 %年份
    for jidu = 1:4
        
        fprintf('%d年%d季度的数据...', year, jidu)
        [sourcefile, status] = urlread(sprintf('http://vip.stock.finance.sina.com.cn/corp/go.php/vMS_MarketHistory/stockid/000001/type/S.phtml?year=%d&jidu=%d', year,jidu));
        
        if ~status
            error('读取出错！\n')
        end
        
        expr1 = '\s+(\d\d\d\d-\d\d-\d\d)\s*'; %获取日期
        [datefile, date_tokens]= regexp(sourcefile, expr1, 'match', 'tokens');
        
        
        date = cell(size(date_tokens));
        
        for idx = 1:length(date_tokens)
            date{idx} = date_tokens{idx}{1};
        end
          
        expr2 = '<div align="center">(\d*\.?\d*)</div>'; %获取数据
        
        [datafile, data_tokens] = regexp(sourcefile, expr2, 'match', 'tokens');
        
        data = zeros(size(data_tokens));
        
        for idx = 1:length(data_tokens)
            data(idx) = str2double(data_tokens{idx}{1});
        end
        
        data = reshape(data, 6, length(data)/6 )'; %重排
        
        filename = sprintf('%d年',year);
        pathname = [pwd '\data'];
        
        if ~exist(pathname,'dir')
            mkdir(pathname);
        end
        
        fullfilepath = [pwd '\data\' filename];
        % 保存数据到Excel
        sheet = sprintf('第%d季度', jidu);
        xlswrite(fullfilepath, date' , sheet);
        range = sprintf('B1:%s%d',char(double('B')+size(data,2)-1), size(data,1));
        xlswrite(fullfilepath, data, sheet, range);
        fprintf('OK!\n')
        
    end
end


fprintf('全部完成！\n')

来源：http://blog.csdn.net/rs_network/article/details/8302020