Drupal:
http://www.google.com/search?hl=en&q=drupal
Videos: Drupal introduction:
http://www.google.com/search?hl=en&q=%22Drupal%20introduction%22&um=1&ie=UTF-8&tbo=u&tbs=vid:1&source=og&sa=N&tab=wv
The web site is a database that is created by the admin user by clicking and filling out forms. Can be for Linux or Windows. Supposedly uses MySQL, Apache, and modules as needed.
Web site themes have been developed and can be applied quickly. If it has been done before, bingo, you are done (if you are not too fussy, or can mess with changing it).
CMS - Content Management System - Open Source
Alternatively, ...
Keywords >"PHP code generator" catalog OR retail OR merchandise OR inventory search application<
http://www.google.com/search?hl=en&q=%22php%20code%20generator%22%20catalog%20OR%20retail%20OR%20merchandise%20OR%20inventory%20search%20application
Some users or development groups have packaged together generic web sites for use with PHP and MySQL, etc. A friend of mine used such a thing, and had a big web site over night. Troubles came only after tweaking caused bugs that he could not easily figure out how to undo / fix in the PHP code / image links, etc.
You can use Web-based PHP or ASP or ASPX or ASP.NET or JSP with browser cookies to showcase a listing of cars the web user has not seen, and be able to accept web form filtering / searches to narrow down what cars to show the visitor in closer detail. This is server side web database application programming and HTML coding work.
Learn about Drag 'N Drop programming to permit the used car lot owner guy to take 3 photos of each car, then drag 'n drop them into a newly created used car database record. Once the dropped item(s) are done, you can rename the files, crop / thumbnail scale / smart resize via pixel interpolation, publish / copy them to the web server folder where they can be accessed by the public.
The car lot owner still has to provide typing input to complete the addition of a used car entry by supplying the asking price, the make, the model, the general condition, and more detailed description. This could be done by the same application software that does the drag 'n drop of the photo files. ODBC, JDBC, MS SQL server, MS Access, MySQL, Oracle, can all be accessed by this application by appropriate sequences of data source name connection strings, port numbers, authentications, etc, and SQL transactions. It is possible to have 2 databases--one local to the car place, and one local to the web server. When records from the local system are "ready," "approved," etc, by management, they can be sent to the remote web server.
You will have to provide a means of saying that the vehicle on file was sold up to one month ago, and have a means of adding, updating/modifying (marking as sold or changing descriptions / pricing), archiving older data, then deleting it after a while. This is fundamental database functionality.
You could use C#, VB, C++, VBA, Java, or any number of programming languages to do the Drag 'n Drop / database input programming--whatever you can learn quickest that has the greatest flexibility to do the required job. If you have already done drag 'n drop programming before, then just a few changes to add the database functionality would be needed.
The search function could be done web server-side using a natural language / free-format regular expression compiler in C/C++, then using the compiled expression, scan all of the text fields of all of the used car database (a little slow if thousands of cars). Search the following keywords >Henry Spencer regex.h regcomp example qnx OR gcc OR g++<
For each provided web visitor keyword spaced from the next, you would assume a "logical and" between each such keyword listed.
The other way for this is to pre-index all of these database text fields to build up a searchable keyword index.
Keywords >"full text index" OR "inverted index"<
If each of these car records were themselves separate text or html or doc or pdf files, then Microsoft Index server (or modern day equivalent) could be used in ASP, ASPx, or ASP.NET code on the server side to query the content index of these documents within the specific web server folders containing these files. The filenames returned by the hit results are within a content index record set. These filenames could be used to index / query find the corresponding photos / thumbnails, description text, etc so that a small picture of the vehicle could be presented with the query result doc links and brief description text. A filter could also be established / applied on price range so that the record set would not even contain vehicles out of the price range consideration. Similarly with vehicle manufacturing date or other considerations.
There was a Microsoft internet access API that could be used to connect to a web server, then send / update the files there so that web updates could be completely automated.
Keywords >WinInet API c OR c++<
FrontPage Server extensions, Web Publishing Wizard, MS Posting Acceptor, etc, have similar functionality if the end web server has write enabled the folders in question for the given username and password.
Most of this depends on the software running on the platforms that you will be using such as MS Windows Server or gnu / Linux.