Website collector PHP

I need a script that will connect to website (EXAMPLE: [login to view URL] or [login to view URL]) Grab all internal/same domain links. Then go to each url using multicurl and put the page into a string. Then process the string through a keyword function I already have and it will give you all the most used keywords. for each page. Then I need to match the most popular keyword(s) with a category with the least amount of words, which will be in a database table that is already created. Then grab sentences from the same page that has the category words in them. As each page is processed it should echo the data found to screen as the script continues to run.

If only 20 keywords are found on a page I want the script to re-process the page using the keyword function I will give you with different paramaters until the keywords are at least 50 or more. If I can not get enough keywords then the script will skip the page and not post data for it. If no category is found for the top 10 keywords the script should skip the page. The options mentioned above should be configurable in a config file.

When this is all complete I should have data for every page found, neatly on the screen. Then I can click a button and it should insert all the information into a database which again is already created.

Any configurable settings should be neatly organized in a file called [login to view URL] and that [login to view URL] file can be included in the main script. This way I can easily make configuration changes.

I will give the winning bider an FTP site, url to FTP data and databse access.

I know this job may become longer than anticipated so I do not have a problem sending out bonuses over and beyond what the actual bid price is as work gets completed for extra stuff not mentioned in here. I also would like someone who has a lot of time on there hands so I can send them future jobs.

Ingenieurwesen MySQL PHP Projektmanagement Softwarearchitektur Testen von Software Web Hosting Website-Management Testen von Webseiten

Projekt-ID: #2797308

Über das Projekt