SEARCH itab FOR pattern [IN {BYTE|CHARACTER} MODE]
[STARTING AT idx1] [ENDING AT idx2]
[ABBREVIATED]
[AND MARK].
1. ... IN { BYTE |CHARACTER } MODE
2. ... [STARTING AT idx1] [ENDING AT idx2]
3. ... ABBREVIATED
4. ... AND MARK
This statement searches the rows of the index table itab for a pattern specified in pattern. For hashed-tables, SEARCH cannot be used. The statement always searches the internal table and not a potentially existing header line.
For pattern, a byte- resp. character-type data object can be specified depending on the processing type. The patterns in pattern have the same forms as the statement SEARCH has for Byte- resp. character chain processing.
The search ends at the first hit and sy-tabix is set to
the index of the found table row. sy-fdpos is set to the
offset of the found byte- or character sequence resp. to the found word in the table row. If the pattern
is not found, then sy-fdpos and sy-tabix are set to 0.
Return Value
| sy-subrc | Relevance |
| 0 | Pattern found in itab. |
| 4 | Pattern not found in itab. |
Since release 6.10, you should - if possible - use the new statement
FIND within a loop defined with LOOP ... ASSIGNING instead of SEARCH.
... IN { BYTE |CHARACTER } MODE
With the addition IN BYTE MODE resp. IN CHARACTER MODE, you determine whether
byte- or character chain processing is carried out. The row type
of the internal table must be suitable for the chosen processing type. Without specifying the addition and before release 6.10, the search is carried out character by character.
... [STARTING AT idx1] [ENDING AT idx2]
With the additions STARTING AT and ENDING AT, you can restrict the search to a part of the the table rows of table itab. For idx1 and idx2, data objects of the type i are expected. The value in idx1 specifies from which row and to which row the value idx2 is searched for. If only one of the additions is specified, the search is carried out from the first to the last row.
The search is not carried out and sy-subrc is set to 4, if:
... ABBREVIATED
When you search character by character, you can specify a shortened pattern in
pattern for character chain processing with the addition ABBREVIATED< (just as with statement SEARCH)
... AND MARK
When searching character by character, you can convert a character sequence resp. a word found in
itab to upper case with the statement AND MARK
(just as with the statement SEARCH for character chain processing).
The search character by character is successful and sets sy-tabix to the index (2) of the respective row and sy-fdpos to the offset (7) of the word "see" in the row. After execution of the statement, the second table row contains the content "you'll SEE the line" via the addition AND MARK.
DATA text_table TYPE TABLE OF string.
APPEND: 'Sweet child in time' TO text_table,
'you''ll see the line' TO text_table,
'the line between' TO text_table,
'good and bad.' TO text_table.
SEARCH text_table FOR '.see.' AND MARK.