pdf to html converter [SOLVED]

[ternor]

According to this page Zamzar is able to convert pdf to html.

For $7 dollars a month - I think I'll pass...  for now, anyway.

There is a free service and (if you want more than that provides) a paid service.  I have used it quite a few times without charge.

[horusfalcon]

+1 for Zamzar.com, (I've had good results with their free service in converting AutoCAD drawings to PDF) but a lot depends on the PDFs.  If you have PDFs that originated with scanned images, any conversion to HTML is going to be problematic at best (you'll wind up with HTML that embeds an image).  PDFs that are watermarked or have security features enabled may present other problems.

If the PDFs are mostly text with a few images, you'll see better (more usable) results, but any conversion process you use will likely leave some "warts" on the finished product which you will have to remove by judicious hand-editing.

Is the information you need not available in a more usable format somewhere else?  Might be a case of, "If I was going there... I wouldn't a started from here!"

Later On,
D

[The Chief]

+1 for Zamzar.com, (I've had good results with their free service in converting AutoCAD drawings to PDF) but a lot depends on the PDFs.  If you have PDFs that originated with scanned images, any conversion to HTML is going to be problematic at best (you'll wind up with HTML that embeds an image).  PDFs that are watermarked or have security features enabled may present other problems.

So I noticed.  The pdf's are tables of data.  Most everything I try extracts the data, but converts the table grid lines into background images.

Is the information you need not available in a more usable format somewhere else?  Might be a case of, "If I was going there... I wouldn't a started from here!"

I can't seem to locate it - no one will admit anything.  It is hand extracted date from will record books at the courthouse.  Five books worth.  I'm not about to start over, so I've resigned my self to a lot of concatenation coupled with extensive "find and replace."  It'll take awhile, but I'll just keep plugging away until I get there.  I would like to strangle whoever decided pdf's were the way to go. 

Of course, the guy running web site before me used images for EVERYTHING!  Even most of the text.  And absolute positions, so any change mucked things up pretty bad.  When I took it over,  a little over a year ago, most of the text appeared to be scanned images.  And the menu buttons were three slightly different graphic images used for initial, mouse over and visited. For a long time I was stuck adding more pages, as I couldn't manage to duplicate the buttons.  I finally got it all converted to pure html, using css and javascript to control the appearance. 

These pdf files are the last things I need to convert.

By the way - if you're interested, here's the url:

http://www.douglascountygensoc.org/

[kjpetrie]

As I said earlier, the table grid lines are not being replaced by background images by the HTML conversion process. They were replaced by the PDF program. That's how PDF works. Like an incompetent web designer, it only cares about how things look; not what they mean. If the same image is being repeated over and over again, kwrite's search and replace might help. It can do regular expression matches if the images are in tags with small differences.

Otherwise, it's the process I outlined before.

No program can do the impossible. The information has been corrupted by the PDF process and there is no automated way to get it back. Stop wasting your time looking for something which can't exist. Accept you will have to recreate it manually. The sooner you start, the sooner you'll finish.

Hope that doesn't sound harsh, but it's the harsh reality you're in.

[ternor]

If you are unfamiliar with html and css coding, there are applications for creating web pages.  I forget the names.

[horusfalcon]

{snip:  my previous}

So I noticed.  The pdf's are tables of data.  Most everything I try extracts the data, but converts the table grid lines into background images.

So, are you attempting to render this data in HTML as a set of tables (or table-like structures in CSS)?  This is beginning to sound like a task someone somewhere has already written a Perl script to accomplish, maybe?  The primary modus here would be to take your first pass conversion and run it through the Perl script to do the slice-n-dice and generate the output in a format required by your style sheet.

I wouldn't worry too much about that background image - that can be sliced out pretty easily once the first pass HTML is generated (it's really not needed, is it?) and the remaining data then processed.

I'm not a Perl monger, myself, (I've written a few very basic scripts) but know of a few places where scripts for useful purposes are available for download:

http://www.perlscriptsjavascripts.com/ - PerlScriptJavaScripts.com is a repository of a wide array of Perl scripts, some free and some for pay, and they also offer custom scripts for hire.

http://www.scripts.com/perl-scripts/ - Perl Scripts is another combination repository that's well organized by function.

http://savage.net.au/Perl.html - Ron Savage's Perl Scripts page - all of Ron's stuff is free, open source code.  Take particular note that he has already done some scripts for reading genealogy data.

http://www.bewley.net/perl/  - Dale Bewley's Perl Scripts page - Dale offers custom scripting services, too, and has several scripts up on his page for download (but nothing that looks like what you need)

These are just a few of the pages out there dedicated to scripting - you might conduct a search and see if it's possible someone has chewed some of the same ground you are chewing now...  a little searching on the front end might save you a lot of work on the back end.

Failing the discovery/creation of some kind a tool to help do this work, yeah, you're exactly where kjpetrie says... and the sooner out the sooner done.

{snip: more of my previous}

I can't seem to locate it - no one will admit anything.  It is hand extracted date from will record books at the courthouse.  Five books worth.  I'm not about to start over, so I've resigned my self to a lot of concatenation coupled with extensive "find and replace."  It'll take awhile, but I'll just keep plugging away until I get there.  I would like to strangle whoever decided pdf's were the way to go. 

Of course, the guy running web site before me used images for EVERYTHING!  Even most of the text.  And absolute positions, so any change mucked things up pretty bad.  When I took it over,  a little over a year ago, most of the text appeared to be scanned images.  And the menu buttons were three slightly different graphic images used for initial, mouse over and visited. For a long time I was stuck adding more pages, as I couldn't manage to duplicate the buttons.  I finally got it all converted to pure html, using css and javascript to control the appearance.

This is so typical.  Don't feel too harshly toward your predecessor - he  was likely just muddling through until he could do better.  So many folks don't keep up with web technologies any more 'cause they're such a crazy quilt.

He may have done everything as images at the advice of an ill-informed attorney who believed it's somehow harder to alter image data than HTML or text. 

These pdf files are the last things I need to convert.

By the way - if you're interested, here's the url:

http://www.douglascountygensoc.org/

I'll give it a look.  I hope you can find a script out there that gets where you want to go, otherwise it'll be a lot of work.

Later On,
D

[The Chief]

You can all consider the problem solved.  I found an on line converter (upload the pdf, download the html) that  works perfectly.  It is at:

http://www.pdfonline.com/convert-pdf-to-html/

Creates actual tables in the html, although a separate table for each page, but that's easy to fix.  A bit of global find and replace for some unneeded text formatting (trying to make it look exactly like the pdf, I suppose) and it's good to go.

Thanks to all...

[horusfalcon]

Good deal, man!  A little searching on the front end saved work on the back end, that's for sure.

Later On,
D

[kjpetrie]

Glad you found something. I wasn't expecting that, but good.

[The Chief]

Just to let everyone see the results, here is the first file (Book A - they go through Book E):

http://www.douglascountygensoc.org/will_index_book_a.html