Problem with unstructured data extractor

Posted January 23rd, 2008 by danpoon12

Hi

Firstly, great program!

I've been trying to extract data from Act6, using a modified report showing the data I need. Everything works OK and I get output from the data extractor, however some of the fields are occasionally blank (say Address 3 field), and that seems to mess up the outputted ascii file. What seems to happen is that it uses the last known data field entry for all the blank ones it finds until a non-blank entry is found, whereupon it then uses that until the next entry, and so on...

As I understand the output part of the data extractor, it buffers the first record until it finds the ACCEPT line, then writes the record out and starts again, overwriting the buffer as required. It seems that my problem occurs because the buffer isn't flushed at the start of each new record, so it's using data form previous records until that field is refreshed with data from the current record.

Is there a way round this? Either by setting something in the line styles, or during the output stage maybe?

Cheers

Dan

‹ "Data Extractor" capability to identify keyword or syntax patterns Goldmine Data Extract ›

RE: Data Extractor for Unstructured Text

On January 23rd, 2008 jterrell said:

Hi Dan,

Glad you like the Data Extractor!

Yes! There are two methods and places in the Data Extractor to select how you want data to be "flushed" or "propagated" from record to record for each of your defined data fields. Here's a brief description of each method/place:

Source Options

You can set a "global default" for this "flush or propagate" setting in the Source Options dialog box. However, you have to select which option you want for the default prior to defining your data fields, so this won't help you right now. But for future data extraction projects, you may want to set the default to "Flush data contents" before you start defining your data extraction rules.

Data Field Definition

You can also set the "flush or propagate" option for each data field. Here are the steps to follow to change the setting for each of your data fields:

  1. In the main menu, click the Field menu.
  2. Select Edit Data Field to open the data field definition dialog box.
  3. Select a Line Style Name in the Line Name picklist. A list of the data fields that are defined within that line style will display in the Field Name picklist.
  4. Select a Data Field Name from the Field Name picklist.
  5. Click on the "Data Collection/Output" tab.
  6. In the "Behavior on Accept" area, click the radio button for the "Flush Field Contents" option.
  7. Click the Update button to update the data field definition.
  8. Repeat steps 3 through 7 for each of your data fields.

    9. Please let us know if you need further assistance.

    - Jo

Thanks Jo, that worked

On January 24th, 2008 danpoon12 said:

Thanks Jo, that worked perfectly!

Great product with (unusually) good documentation and customer support... you should be proud!

Regards

Dan

RE: Data Extractor

On January 24th, 2008 jterrell said:

Glad that worked for you! And thanks for the compliments on the Data Extractor product, documentation, and support!

- Jo