Obituary database

[rcraig]

We have about 5000 obituary records that need to be scanned and put on a CD that would include a linked name index and could be read by any browser. What program could be used to facilitate this project? It is anticipated that the data would also be used on a Web site.

[Deadeye]

if you want something thats portable and be able to be read right off the CD, you would want to use MS Access. mysql, oracle, and sql server would need to be installed somewhere and put a backup of the database on the CD.

my vote is access.

Wesley

[Allen]

somehow I guess this is spam but if it isn't then the first step is getting the obits scanned using OCR - the issue is going to be depending on how old the papers are, whether the characters will be picked up correctly.

After that I agree with Deadeye that Access would work well.

[rcraig]

Thanks for the replies. Access seems like an overkill for what we need to do. The obituaries are pasted on 5 X 8 cards, often with other information written on the cards. It is too easy for these cards to be pilfered from the library. I want to scan the cards to obtain image files (.jpeg, etc.) and save them using the name of each subject in alphabetical folders. All that is needed is to be able to easily search the file names using CD(s) in the library computer and link to the image files using HTML.

[Deadeye]

well if you want to do it right you might as well go with access. I'm fairly certain you can imbed images inside a record in the database. You can always expand the database while you get more info, cards, or what ever the case may be. Its a lot easier to parse though a few hundred records in a database than it is to do a windows/mac search though a bunch of images.

I would setup the access database as such
FirstName
LastName
DOB
OtherRelaventInformation
ImbeddedImage

Just my 2 cents though

Wesley

[zoul]

If you only have 5,000 or so records ... and you have a specific way of searching for them "by name" (for example) ... then, you can easily build a folder structure, like you proposed ... and use javascript in a simple HTML page to find the results ...

Here's how you do it ...

Create a Javascript file that contains an object that has the name, and the path:

var database = { {name:'Zoul',path:'relative/path/to/image'},{name:'Deadeye','relative/path/to/image'}};

Then, simply use generic javascript functionality to search through this ... as this search will most likely be time consuming (depending on the computer ... 5000 iterations could take 2 seconds, could take 45 seconds ... ) ... you'll want to handle it properly, and only search through a few at a time, using a setTimeout ... just keep track of the last record you search (using a numeric indexing) and every 100 or so records ... stop searching, remember the last record, set a timeout for like 100ms ... and keep searching ... and just keep updating the UI to let the user know 'the search is currently in progress' ...

This does NOT require a database, a database would be the preferred way to pull this off ... but if the requirement is beyond the scope of your target audience ... then you don't NEED one.

A good example of a javascript indexed system is Garage Game's documentation for the Torque Game Builder, which you can download a free 30-day trial of on there website (Game Development Tools And Software :. GarageGames) -- the documentation is extremely thorough and contains a ton of stuff, and it even includes a simple, yet elegant 'find' feature ... every word on every web page is indexed and stored in one large massive javascript file ...

Yes, this requires you to load a large javascript file into memory ... but, seriously ... loading MDAC or Access into memory isnt any better (memory wise) ...


So ... if you have any more questions related to this, just let me know ... but, don't feel like you NEED a database, it's easy enough to implement a simple Hash style db client-side with javascript ...

Heck, if you use something like jQuery or whatever ... you can even store the database as a hidden TABLE in the HTML and just look through it with jQuery ... or, use XmlHttpRequest and the built-in DOM features to sift through some XML format with X-Path or XQuery ...

Lots of options, all can be done with plain HTML and Javascript ...

-- David

[zoul]

Oh, I might add ... that if you also implement this on a web site with a server-side technology like PHP or .NET ... that you do use a Database engine on the web site ... but for a stand-alone 'shippable' thing ...

Just use an Index.html on the CD, and use Autorun.inf to launch the index.html file ... viola, done ...

I'll even do it for you, if you want to contract me

[rcraig]

I finally discovered I was asking the wrong question because my problem is one of indexing. The website "Software for HTML indexing" seems to be a good start for me.
Thanks for your replies.