Support Forums > Report Indexing

Character removal question

(1/2) > >>

DDP021:
Hi there...Stupid question...We see on the Application setup under the LOAD INFORMATION tab options for Character Removal (Leading, imbedded, trailing) for App Grp names.  We also see these options under the APP GRP setup under the FIELD INFORMATION tab.   If you wanted to incorporate these, where would set them, the APP setup, AppGrp setup or both?

Darrell Bryant:
Character removal on the Application Group > Field Information tab is used to 'cleanup' search values entered by the end user before querying the database.
From the Administrator Client help text:

String Embedded. You can define from one to eight characters that Content Manager OnDemand removes from search values typed by the user before querying the database. Content Manager OnDemand can remove embedded, leading, and trailing characters. For example, suppose you define a field to store index values that contain phone numbers. The values in the database contain only numbers. That is, the values do not include the dash character. However, when users query the database with a phone number, they typically type search strings such as 303-555-1212. Since the database values do not include the dash character, Content Manager OnDemand would not find a document to match the query. However, you can use the Embedded field to enable users to type values that include the dash character. Simply type a dash character in the Embedded field. That way, the user can type 303-555-1212 (or 3035551212) in the search field. Content Manager OnDemand removes all occurrences of the character specified in the Embedded field from the search string typed by the user before querying the database.

Justin Derrick:
You may also find this interesting if you're loading CMOD Generic Index files, and doing character removal at load time to normalize data:  https://cmod.wiki/index.php?title=Default_Values_and_the_Generic_Indexer

-JD.

DDP021:
Thanks everyone for the response.  I guess my question is, with the option to set these parameters on both the Application and Application Group setups,  which option is the better place to set turn them on (ie Application or Appliction group definition).

Justin Derrick:
Hrm.  Interesting question.

It depends on how you want to 'normalize' the document metadata, so that you can ensure the maximum level of success in FINDING a document you're searching for. 

In the vanilla thick client & ICN, since there's not a lot of options for modifying queries entered by end users to match specific formats (like a US Social Security Number or phone number) it's probably best to just store them as a series of digits in a string.  It doesn't make sense to treat them as numbers -- because you won't be doing comparisons like greater-than or less-than.

If you ensure that SSNs are loaded with spaces, dashes, or dots removed, then ensure that spaces, dashes, or dots are removed at query time, you're dramatically increasing the chance that someone can find what they're looking for.

Hope that helps.

-JD.

Navigation

[0] Message Index

[#] Next page