The SEARCH statement is used to search a table for a table element that satisfies the specified condition and to adjust the
associated index-name to indicate that table element.
General Format for Format 1
General Format for Format 2
Syntax Rules for Format 1
- Condition-1, condition-2, and so on, can be any condition as described in the section
Conditional Expressions in the chapter
Procedure Division.
- Identifier-1, the table element being searched, can be an internal or external floating-point item.
Syntax Rules for Format 2
- All referenced condition-names must be defined as having only a single value. The data-name associated with a condition-name
must appear in the KEY clause of identifier-1. Each data-name-1, data-name-2 can be qualified. Each data-name-1, data-name-2
must be indexed by the first index-name associated with identifier-1 along with other indices or literals as required, and
must be referenced in the KEY clause of identifier-1. Identifier-3, identifier-4, or identifiers specified in arithmetic-expression-1,
arithmetic-expression-2 must not be referenced in the KEY clause of identifier-1 or be indexed by the first index-name associated
with identifier-1.
When a data-name in the KEY clause of identifier-1 is referenced, or when a condition-name associated with a data-name in
the KEY clause of identifier-1 is referenced, all preceding data-names in the KEY clause of identifier-1 or their associated
condition-names must also be referenced.
- Identifier-1, the table element being searched, cannot be a floating-point item.
- Neither data-name-1 nor data-name-2, the key data items, can be floating-point items. However, identifier-3, identifier-4,
literal-1, or literal-2, the items against which the key is compared, can be floating-point items.
General Rules for All Formats
- The scope of a SEARCH statement can be terminated by any of the following:
- After execution of imperative-statement-1 or imperative-statement-2 that does not terminate with a GO TO statement, control
passes to the next executable sentence.
- If identifier-1 is a data item subordinate to a data item that contains an OCCURS clause (providing for a two- or three-dimensional
table), an index-name must be associated with each dimension of the table through the INDEXED BY phrase of the OCCURS clause.
Only the setting of the index-name associated with identifier-1 (and the data item identifier-2 or index-name-1, if present)
is modified by the execution of the SEARCH statement. To search an entire two- or three-dimensional table it is necessary
to execute a SEARCH statement several times. Prior to each execution of a SEARCH statement, SET statements must be executed
whenever index-names must be adjusted to appropriate settings.
General Rules for Serial search Format 1
- If Format 1 of the SEARCH statement is used, a serial type of search operation takes place, starting with the current index
setting.
- If, at the start of execution of the SEARCH statement, the index-name associated with identifier-1 contains a value that corresponds
to an occurrence number that is greater than the highest permissible occurrence number for identifier-1, the SEARCH is terminated
immediately. The number of occurrences of identifier-1, the last of which is the highest permissible, is discussed in the
topic
The OCCURS Clause in the chapter
Data Division - File and Data Description.) Then, if the AT END phrase is specified, imperative-statement-1 is executed; if the AT END phrase is not specified, control
passes to the next executable sentence.
- If, at the start of execution of the SEARCH statement, the index-name associated with identifier-1 contains a value that corresponds
to an occurrence number that is not greater than the highest permissible occurrence number for identifier-1 (the number of
occurrences of identifier-1, the last of which is the highest permissible is discussed in the topic
The OCCURS Clause), the SEARCH statement operates by evaluating the conditions in the order that they are written, making use of the index
settings, wherever specified, to determine the occurrence of those items to be tested. If none of the conditions is satisfied,
the index name for identifier-1 is incremented to obtain reference to the next occurrence. The process is then repeated using
the new index-name settings unless the new value of the index-name settings for identifier-1 corresponds to a table element
outside the permissible range of occurrence values, in which case the search terminates as indicated in General Rule 4a above.
If one of the conditions is satisfied upon its evaluation, the search terminates immediately and the imperative statement
associated with that condition is executed; the index-name remains set at the occurrence which caused the condition to be
satisfied.
- If the VARYING phrase is not used, the index-name that is used for the search operation is the first (or only) index-name
that appears in the INDEXED BY phrase of identifier-1. Any other index-names for identifier-1 remain unchanged.
- If the VARYING index-name-1 phrase is specified, and if index-name-1 appears in the INDEXED BY phrase of identifier-1, that
index-name is used for this search. If this is not the case, or if the VARYING identifier-2 phrase is specified, the first
(or only) index-name given in the INDEXED BY phrase of identifier-1 is used for the search. In addition, the following operations
occur:
- If the VARYING index-name-1 phrase is used, and if index-name-1 appears in the INDEXED BY phrase of another table entry, the
occurrence number represented by index-name-1 is incremented by the same amount as the index-name associated with identifier-1,
and at the same time.
- If the VARYING identifier-2 phrase is specified, and identifier-2 is an index data item, then the data item referenced by
identifier-2 is incremented by the same amount as the index-name associated with identifier-1, and at the same time. If identifier-2
is not an index data item, the data item referenced by identifier-2 is incremented by the value (1) at the same time as the
index referenced by the index-name associated with identifier-1 is incremented.
General Rules for Non-serial search Format 2
- The results of the SEARCH ALL operation are predictable only when:
- The data in the table is ordered in the same manner as described in the ASCENDING/DESCENDING KEY clause associated with the
description of identifier-1, and:
- The contents of the key(s) referenced in the WHEN clause are sufficient to identify a unique table element.
- If Format 2 of the SEARCH statement is used, a non-serial type of search operation can take place; the initial setting of
the index-name for identifier-1 is ignored and its setting is varied during the search operation with the restriction that
at no time is it set to a value that exceeds the value which corresponds to the last element of the table, or that is less
than the value that corresponds to the first element of the table. The length of the table is discussed in the topic
The OCCURS Clause). If any of the conditions specified in the WHEN clause cannot be satisfied for any setting of the index within the permitted
range, control is passed to imperative-statement-1 of the AT END phrase, when specified, or to the next executable sentence
when this phrase is not specified; in either case the final setting of the index is not predictable. If all conditions can
be satisfied, the index indicates an occurrence that allows the conditions to be satisfied, and control passes to imperative-statement-2.
- The index-name that is used for the search operation is the first (or only) index-name that appears in the INDEXED BY phrase
of identifier-1. Any other index-names for identifier-1 remain unchanged.
- Neither imperative-statement-2 nor NEXT SENTENCE is required. Without them, the SEARCH statement sets the index to the value
in the table that matched the condition.
Comments:
Figure 1 shows a flowchart of a Format 1 SEARCH operation containing two WHEN phrases.