most powerful web automation

[scripter]

hello!

I still havent find an easy way for web automation

learning JS, Greasemonkey and such is kinda cumbersome and has severe limitations (eg. cannot write data to a simple text file)

I have so far, found these tools:
Mechanize
UBot
Selenium
Watir

are there any more?

which is the best? the most powerful? even if I have to learn more

I want to (hopefully easily) be able to automate web browsing actions, for example:

1) be able to set triggers, such as:
- a specific url (that matches a regex pattern) has loaded
- a specific website control (eg. button, form, drop down menu, etc) has loaded
- a specific drop down menu has been selected
- a specific text appears in the loaded webpage
- a specific link has been loaded
- a specific form contains specific text
etc

2) which above triggers will trigger specific actions, such as:
- click a specific button in a webpage
- open a specific webpage
- close a specific tab
- save a specific website file (that matches a regex pattern)
- bookmark a specific url
- select a specific entry in a specific drop down menu
- copy text from specific area of a webpage, and use that text to save it in a text file or to do other actions, like fill forms with specific text, etc
- reload a specific webpage
- read a txt, xls or whatever file and extract specific text from it to fill a specific website form

3) also triggers could be specific mouse or keyboard shortcuts that trigger specific actions

- wait until a specific control (eg. a radio button with specific text, a specific link with specific label, etc) in a specific webpage is loaded
- wait until a specific control gets a specific value (eg a specific tickbox gets ticked, a specific form gets specific text inside it, etc)
- be able to specify controls, by their size, color, position in the webpage, name, type (eg button), etc
- be able to select specific area in a webpage or specific control(s) and extract the text from it, or save a screenshot of it, or save that portion of the webpage as a html/mht, or change its background color, etc
- be able to match text inside forms that I type, or that already are present in, by regex
- be able to perform all kind of clicks (right, left, middle, etc) and keypresses in specific coordinates, control, etc of the webpage

being able to interact with Windows API would be best too!!

in general, I want to be able to do what a human can do in the web, browse, see, check, save, interact, fill, extract, everything in any website, automatically

thanks!

[VictorN]

And how and where is such an application supposed to be used?

[Igor Vartanov]

..and what this all has to do with Visual C++ Programming?

being able to interact with Windows API would be best too!!

in general, I want to be able to do what a human can do in the web, browse, see, check, save, interact, fill, extract, everything in any website, automatically

In general this severely conflicts with browser security concepts. Oops.

[Igor Vartanov]

And how and where is such an application supposed to be used?

My best guess is: QA department.

[scripter]

In general this severely conflicts with browser security concepts. Oops.

Lily programming language, doesnt do that exactly?

[OReubens]

you can do all of that if you write your own application that hosts a IWebBrowser2. Though it requires quite a bit of work to tie it all together.

The bad side of it is that changes to the needs will require changes to the C++ code (and recompiles/redeployments), which is why tools such as greasemonkey exist where you just change script.

You can of course implement your own scripting as well, but ... that's a whole lot of extra work too.

[scripter]

you can do all of that if you write your own application that hosts a IWebBrowser2. Though it requires quite a bit of work to tie it all together.

I would prefer it to work with Firefox API or Opera API, since I never use Internet Explorer