Build Multi-part Search Expressions

You can define your searches to meet a specific set of criteria by connecting several search terms together. Search expressions can include several parts, for example:

[author] CO john doe AND [cc] NC jane smith or gene summer

([author] CO doe OR smith) NC summer

Review the information below to learn more about building multi-part search expressions. For information about basic search operators and search syntax, see Full-text Search Options.

How Multi-part Search Expressions are Processed

The following rules are used for multi-part search expressions:

Search operations are performed in sequence from left to right.
Search expressions can be enclosed in parentheses to alter the sequence of operations.
The use of nested parentheses is supported, but if your case includes page-level OCR, it may be necessary to specify the field(s) involved to eliminate confusion or inconsistencies when page-level OCR is included in the search results. See examples 6 and 7 in the following table.

Multi-part Search Examples

The following examples will help you understand how you can construct multi-part search expressions.

No.	Search Expression/Results
1	annual, profit + deficit
1	This search first creates a set of documents containing either “annual” OR “profit” and then reduces that subset to only those documents that ALSO contain the word “deficit.”
2	annual, (profit + deficit)
2	This search first creates a set of documents containing “annual” and then adds to that set any documents containing both “profit” AND “deficit.”
3	[author] CO jane smith AND [recipient] NC john doe
3	This search first creates a set of documents in which the AUTHOR field contains both “jane” and “smith,” then reduces that set to only those documents in which the RECIPIENT field does not contain both “john” and “doe.”
4	[author] smith:PH, jones, carpenter:FL AND[ocrtext] stock convertible ST
4	This search first creates a set of all documents in which the AUTHOR field contains words that sound like “smith” or “jones” or words similar to “carpenter.” This set is then reduced to a set containing only those documents in which the OCRTEXT field contains words that begin with “stock” and “convert.”
5	[date] ~ 01/01/2001 - 12/31/2005 OR [date] = NULL
5	This search creates a set of documents in which the DATE field contains any date between January 1, 2001 and December 31, 2005, then adds to it a set of documents in which the DATE field is blank.
6	(jones, (smith + hall)) ! (carpenter, blakely)
6	This search first creates a set of documents containing “jones” and then adds to that set any documents containing both “smith” AND “hall.” It then removes from the set any records containing either “carpenter” OR “blakely.”
7	[author] (jones, (smith + hall)) ! (carpenter, blakely)
7	This search is identical to example 6, except that it limits the search to the AUTHOR field.

Incorrect Search Expressions

Multiple join operators or proximity operators should not be used between search words/terms. Users should apply only one operator at a time between search terms, and can further define their search by stringing together many such terms. These operators include:

and, or, not
, (comma), + (plus), ! (exclamation point)

The following table shows examples of the correct and incorrect approach:

Wrong	Right
car and not boat	car NOT boat
car w/5 house and n/5 boat	car w/5 house n/5 boat

Optimize Searches

The following is a valid search expression:

[Author] CO smith AND [Title] NC automobile

The following search expression, however, will provide the same results in a more optimized manner:

[Author] CO smith NOT [Title] CO automobile

It is more efficient for the search engine to join two searches together with a NOT operation than it is for the engine to find every record that does not contain or does not equal a certain value.

The following is also a valid search expression:

[DOC_TAG] Intermediate Review Status

Using quotation marks for is more efficient, however:

[DOC_TAG] “Intermediate Review Status”

The use of quotation marks is recommended for [DOC_TAG], [PAGE_TAG], and [REDACTION] searches when the tag name or category contains spaces.