Automatically filling out web forms and returning

问题:

This is my first time posting here. I greatly appreciate any and all guidance on this subject.

I'm trying to make a program that automatically fills in web forms and submits the data, returning the resulting page to the program so it can continue to 'browse' the page, allowing it to recursively submit even more data.

The main problems I'm having are:

• The 'submit' button is coded in Javascript, so I don't know where the form data goes when making the page request.
• I want to fill in the forms using data from an Excel table, so I need to be able to access data from outside the page.
• I need to be able to navigate the resulting page to continue to submit more data.

More specifically, I'm trying to first login to the Practice Mate website, navigate to 'Manage Patients', hit 'Add Patients', and fill in the proper forms and submit. I'm filling in the forms from an Excel table thousands of rows long.
Sorry I can't be more clear on this without providing a username and password.

What I've been trying to do is use Javascript to make page requests from a page that retrieves information from the Excel document using PHP. I still can't seem to get anything to work with this method though.

I apologize for being a relative novice at this. Thanks in advance.

回答1:

You can use PHP cURL to browse & submit forms to websites, but it does depend on how the website is setup. Most have security checks in place to prevent bots and can be tricky to get everything to work right.

I spent a little bit of time and came up with this login script. Without a valid username and password I can't verify that it is successful, but should do what you need. This short example first browses to the page to set any cookies and scrape a __VIEWSTATE value needed to submit the form. It then submits the form using the username/password you provide.

<?php

// Login information
$username = 'test';
$password = 'mypass';
$utcoffset = '-6';
$cookiefile = '/writable/directory/for/cookies.txt';

$client = new Client($cookiefile);

// Retrieve page first to store cookies 
$page = $client -> get("https://pm.officeally.com/pm/login.aspx");
// scrape __VIEWSTATE value
$start = strpos($page, '__VIEWSTATE" value="') + 20;
$end = strpos($page, '"', $start);
$viewstate = substr($page, $start, $end - $start);

// Do our actual login
$form_data = array(
    '__LASTFOCUS' => '', 
    '__EVENTTARGET' => '',
    '__EVENTARGUMENT' => '',
    '__VIEWSTATE' => $viewstate,
    'hdnUtcOffset' => $utcoffset,
    'Login1$UserName' => $username,
    'Login1$Password' => $password,
    'Login1$LoginButton' => 'Log In'
);
$page = $client -> get("https://pm.officeally.com/pm/login.aspx", $form_data);

// cURL wrapper class    
class Login {
    private $_cookiefile;

    public function __construct($cookiefile) {
        if (!is_writable($cookiefile)) {
            throw new Exception('Cannot write cookiefile: ' . $cookiefile);
        }
        $this -> _cookiefile = $cookiefile;
    }

    public function get($url, $referer = 'http://www.google.com', $data = false) {
        // Setup cURL
        $ch = curl_init();
        curl_setopt($ch, CURLOPT_URL, $url);
        curl_setopt($ch, CURLOPT_REFERER, $referer);
        curl_setopt($ch, CURLOPT_COOKIEFILE, $this -> _cookiefile);
        curl_setopt($ch, CURLOPT_COOKIEJAR, $this -> _cookiefile);
        curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);
        curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
        curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
        curl_setopt($ch, CURLOPT_AUTOREFERER, true);
        curl_setopt($ch, CURLOPT_MAXREDIRS, 10);

        // Is there data to post
        if (!empty($data)) {
            curl_setopt($ch, CURLOPT_POST, 1);
            curl_setopt($ch, CURLOPT_POSTFIELDS, http_build_query($data));
        }

        return curl_exec($ch);
    }

}

回答2:

Also check out the question I asked that is similar. That is how to do it in C# atleast for a website I encountered.

Submitting a web form using C#