Web Query Application with Cross Reference Query (Update)

I need two php applications that will extract data from 2 county records websites. The first application will need to query case numbers sequentially on a county court records website to find cases of a particular type. This data then needs to be used as input for another application which queries a county real estate records website to extract address information of the plaintiff and defendant and output the results to a MySQL database and a downloadable .csv file. ## Deliverables I had a RentACoder programmer create this application for me a couple of years ago. My deliverables for that original project are included below. The county websites have changed so it doesn't work any more. You are welcome to change/fix/enhance his code or start over from scratch. I would like to add a couple of features that the original scripts didn't include. I would like the output to include all the property data fields that is available from the county instead of just selected fields. I would also like to add other surrounding counties once it is working for this county, though that doesn't need to be part of this project and should probably be a separate bid. I would also like to add another script (that will probably use much of the same code) to take a .csv input of property addresses and give me the same output as this application gives. The county input for court cases has changed. The url has changed too. It now uses a 4 digit year. They also added an additional court (the 429th) that needs to be searched too. The searches need to be conducted on the Collin County "Civil & Family Case Records" search page. To help you get an understanding of how the case numbers are now done, the first divorce case in the county in 2010 was case number 296-50001-2010. If you get on the county website and search for that case I think you will better understand what I'm trying to get. Here's the original deliverable request... The purpose of this application is to allow me to direct market to people who are in divorce and own real estate in my county. I want the first application to start with the user inputting the 5 digit case number of the first court case where the queries are to start, the 2 digit year, and optionally an end date. The application should query the following website: [login to view URL] under “Search By: District/County Court Case Number??. The cases should be searched one after another until you get to the end of the cases filed to date or to the end of the (optionally) specified date range. The case numbers are formatted as follows: CCC-ccccc-YY CCC ??" The three digit Court number. There are only 8 possible numbers. They are 199, 219, 296, 366, 380, 401, 416, and 417. ccccc ??" The case number which is assigned sequentially. The first case of the year filed is 50001, the second is 50002, and so on. YY ??" The two digit year. So for example, one case may be assigned a case number of 199-50904-07 which means that it is the 904th case filed in 2007 and it was assigned to the 199th District Court. The next case filed may be 296-50905-07. The case numbers are set sequentially but the 3 digit Court number is assigned at the time the suite is filed based on each court’s case load. Because of the way the website is set up, to find each case you will have to start with the case number and year set and sequence through up to all 8 of the possible Court numbers until it returns a case. Then capture the data for that case if it is the right case type, increment the case number by 1, and do it all again. Occasionally there is no case assigned a particular 5 digit case number (maybe case was filed and then deleted or something). So the queries will need to go 15 case numbers beyond the last case found before it should assume that it got to the end of the cases filed to date and stop querying. The case data that this application should capture is the data for is DIVORCE cases (Type of Case = Civil ??"DIVORCE). It will need to capture the following information: Case Number Plaintiff Name Defendant Name Petition Date The output from this set of queries then needs to be to a .csv file with a file name like [login to view URL] where the 20071226 part of the file name is the date of the query. See the sample output data file. The second application should use a selected file (which will be the output of the first application) as the input for queries to the county appraisal district website. Its purpose will be to extract property addresses, mailing addresses, and other data for the people who have filed divorce if they own real estate in this county. The url is [login to view URL] For each input divorce case the application will need to do two queries, one for the plaintiff’s name and another for the defendant’s name. The names have to be input in the form Last Name, space, First Name. If no results are returned for either name then the application should not capture any data and go on to the next divorce case. If there is a result for one of the names (or if both names produce the same result) then the application should output the data for that record. If each name produces a different result then the application should output each result in a separate record. If either name produces multiple results then the record that also includes the other party’s (plaintiff or defendant) first name should be the one selected. If either query returns multiple records which can not be narrowed down by using the other party’s name then the application should output the name and case data for this query and flag it so that the address information can be looked up manually. The following will need to be captured and output to a .csv file. Case Number First Name Middle Name (or initial) Last Name Property Street Address Property City Mailing Street Address Mailing City Mailing State Mailing Zip Code Subdivision (first part of the legal description before the “,??) Year Built (“Actual Year Built?? for Main Area) Square Feet (“Total Living Area??) Tax Value (“Total Market Value??) The output should be a file named [login to view URL] where the 20071226 part is the date of the query. See the sample output data file. * * *This broadcast message was sent to all bidders on Monday Jun 7, 2010 8:40:21 PM: I have added a file to this job request that contains the existing code that I have now. This code can be modified or completely replaced, whichever is easier for you. Tres