2

I got a set of shapes with some additional informations looking like extracted/exported metadata, such as sources, extents, summary, descriptions, contacts... All in digital pdf and without XML files I could try to play around with. Is there a way to get these information into ArcGis Metadata ? I was wondering how I could set up a metadata template with the differents types of information contained in the Pdf or if I could append the FGDC Metadata with specific field. I will welcome any suggestions or even links to tutorial.

I would like to create an xml formular (never done that, is there some kind of wizard/tutorial for it?) with different topics such as: Name, Fullname, sources, contacts, extents, scale of production, date of creation and date of update, related legislation, "legal rights", geometry... Then I'd like to import this formular as template for metadata and then filling the box then would be done by Hand.

I guess another way around would be to modify an existing ArcGis Metadata template. is there a tool to do that?

Here is a screenshot of the type of categories I’d like to get in a form/template. Well it's in german, but it says things like 1_General Information (sub: Editors, actualisation, contact...), 2_ sources, 3_related regulations, 4_scale, 5_ right of use, 6_technical Parameters, DCM...

Dokumentation

Improve this question
6
  • This is a bad idea. Is the information you need accessible in a machine-readable format elsewhere? Do you really need to do this?
    – blah238
    Commented Feb 5, 2014 at 21:21
  • Well, good question... I am new to metadata (shame on me) and wanted to start learning the steps and process with these data. But I guess, I'll Keep the learning easier and will do that another time with another dataset.
    – PMessager
    Commented Feb 6, 2014 at 7:59
  • What do you mean by formular? Do you just mean a form with blanks to fill in? Why not just use the ArcGIS metadata editor?
    – blah238
    Commented Feb 6, 2014 at 16:45
  • 1
    Let me see if I have this right: you have a bunch of geometries in shapefile(?) format, that came with PDFs describing their content. You would like to extract metadata from the PDFs that can be used by ArcGIS. Yes? If so, you're likely looking for some kind of pdf2xml conversion tool, among others. Really though, without a sample of the shapefiles and accompanying documentation, or a detailed description of them, no one will be able to give specific (e.g. useful) help.
    – matt wilkie
    Commented Feb 6, 2014 at 17:17
  • @Blah238: my objectives were to 1: create a template: I had a go to the metadata editor, wanting to Change the Name of the field (box's title?) in order to match the Pdf information's categories. 2: fill the template, either automatically or by hand. --> conversion of pdf into XML
    – PMessager
    Commented Feb 7, 2014 at 8:21

2 Answers 2

Reset to default
4

Thanks for posting the sample. I'd bet money that PDF was generated automatically from some other format or system that is structured.

Before heading down the road of trying to recreate the structure I would (attempt to) contact the data provider/author and ask for the structured metadata. Explain your goal, and they may be happy to give you a raw dump of the table/xml/whatever. (If the exchange is cordial, you could send them to PDF Sharing is Not Data Sharing–A Public Service Announcement and the links thereof to help encourage the process.) It could save a lot of time and headache.

If that doesn't work, Scraping for Journalism: A Guide for Collecting Data has some useful tips and advice (also from previous link), though it will take some study. A quick perusal of search:convert pdf to structured xml results shows some promise too.

And finally, don't overlook brute force and ignorance. As unpleasant as it is, sometimes just knuckling down and retyping and/or copy-pasting for a time will get the job done more quickly and with less overall frustration than a truly elegant automatic solution. Treat it as a typing tutor exercise. ;-)

Improve this answer
2
  • I am torn between affronting the joy of german bureaucracy and getting it done, banging inelegantly my head against the wall, which might be indeed less frustrating. Thanks for your supportive reply.
    – PMessager
    Commented Feb 10, 2014 at 7:48
  • @PMessager I hear and understand -- with empathy! We can't always change or scale the wall, but our attitude to it is entirely ours to adjust.
    – matt wilkie
    Commented Feb 11, 2014 at 21:26
2

NOTE: extract information from non-structured document is very difficult. Change your contract (not accepting PDF but a HTML with Microformat document or linking to online registration system).

If you REALLY NEED, let hands-on:

  1. extract content with something like pdftotext, producing from each file.pdf a file.txt
  2. use regular expressions with perl, PHP, python, java, javascript, C# ... any language where you build your parser... and transform file.txt into XML or another valid output.
  3. send the structured data (extracted text-patterns) to ArcGis Metadata (check manual of API for interaction).
Improve this answer
3
  • Hi Peter, thanks for your reply. I can do 1. and 3. but 2. looks like I have to go back to School, or teach myself or bribe someone or give up...thanks anyway. Having someone telling you that's a waste of time is helpful.
    – PMessager
    Commented Feb 6, 2014 at 8:02
  • If you can't do, or it is bored, delegate (!). There are a lot of "junior programmers" that can do it (extract data from TXT by regex), they are also at technical-schools, as trainees... Or contract "online India worker", it is very cheap. :-)
    – Peter Krauss
    Commented Feb 10, 2014 at 18:14
  • My GIS dream is to have my own personal developper...
    – PMessager
    Commented Feb 12, 2014 at 7:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.