|
SPLIT stage v1.2 |
Pipelines v2.1 |
Part of examples section does not wrap!
|
Syntax |
┌─AT───────────────────────────────────────────┐>>──SPLIT──┼──────────────────────────────────────────────┼──────────────────────────><
│ ┌─AT─────┐ │ └─┼────────┼──┬─────────┬──┬─────────────────┬─┘ ├─BEFORE─┤ └─ANYCase─┘ ├─charrange───────┤ └─AFTER──┘ ├─STRing─┬─string─┘ ├─REGexp─┤└─ANYof──┘
|
Purpose |
Use the SPLIT stage to divide input records into multiple output records. SPLIT reads records from its primary input stream; splits the records and writes the resulting records to its primary output stream. If no operands are specified; SPLIT divides records at whitespace characters (both space (X'20) and tab (X'09') characters are considered to be whitespace) and discards them. If you only specify the AT, BEFORE or AFTER operands; records are split at whitespace characters, before whitespace characters or after whitespace characters, respectively. With additional operands, records are split relative to occurrences of a specified target.
|
Operands |
|
● |
AT causes input records to be split at the specified target. The target characters are discarded. AT is the default. |
||
|
● |
BEFORE causes input records to be split before the specified target. The target characters are retained. |
||
|
● |
AFTER causes input records to be split after the specified target. The target characters are retained. If you specify
AFTER with STRING, the records are split after any columns that contain the
last character of the string. |
||
|
● |
ANYCase specifies that
charrange or string is compared with the input record in uppercase. In effect
this means that a non-case-sensitive comparison is made when selecting
characters that cause the records to be split. |
||
|
● |
is a character range. A split occurs when
any one of the characters in the range is matched. |
||
|
● |
STRing specifies that the string
operand is a literal string of characters to locate. A split occurs only when
the entire string is matched. |
||
|
● |
REGexp specifies that the string operand is a regular expression of characters to locate. A split occurs when the expression is matched. |
||
|
● |
ANYof specifies that the string
operand is a list of characters to locate. A split occurs when any one of the characters in the list is
matched. |
||
|
|
|
|
Streams |
The following streams are used by the SPLIT stage:
|
|
Usage |
|
1. |
SPLIT
does not delay the records. |
|
2. |
If
the SPLIT stage discovers that its primary input or output streams are not
connected, the SPLIT stage ends. |
|
3. |
SPLIT copies null input records to its primary output stream. It does not
generate null output records. |
|
4. |
SPLIT
verifies that its secondary
input and output streams are not connected and then begins execution. |
|
Examples |
|
Given the input file: input.txt, below; the following four examples demonstrate the before and after operands of the SPLIT stage. input.txt (input) ...|...+....1....+....2....+....3.... **** Top of file ****1 1234512345 2 5432154321 **** End of file **** |
|
1. |
To
split input records before the column that is 1 column to the left of each
occurrence of the character 4, use
the following: 'pipe
< input.txt | split 1 before string /4/ | console' output: 12 34512 345 54321 54321 |
|
2. |
To
split input records after the column that is 2 columns to the right of each
occurrence of the last character in string 12, use the following: 'pipe
< INPUT DATA | split 2 after string /12/ | console' output: 1234 51234 5 5432154321 |
|
3. |
To
split input records between the characters 5 and 4 in occurrences
of the string 54, use the
following: 'pipe
< input data | split -1 before string /54/ | console' output: 1234512345 5 43215 4321 |
|
4. |
To
split input records between the characters 2 and 3 in occurrences
of the string 123, use the
following: 'pipe
< input data | split -1 after string /123/ | console' output: 12 34512 345 5432154321 |
|
Other
miscellaneous examples: |
|
|
1. |
The following example utilises the
SPLIT stage command in a pipeline which determines the number of bytes that
could be saved by removing trailing whitespace. **** Top of file ****1 Address Rxpipe 23 'pipe < myfile.txt ', 4 '| locate', /* Discard blank-lines. */ 5 '| xlate w-1;* x20 @ x09 @', /* Change spaces/tabs to at(@) chars. */ 6 '| split before str /@/', /* Split at each at(@), start new record. */ 7 '| strip trailing anyof /@/', /* Reduce records to length zero. */ 8 '| nlocate 1', /* Select only null/empty records. */ 9 '| count', /* Count the records. */ 10 '| specs /The number of bytes which could be saved is:/ 1 1-* nw', 11 '| cons' /* Display the result. */ 12 13 Exit 0 **** End of file **** |
|
Related |
CHOP, JOIN, PAD, STRIP