...
| Code Block | ||
|---|---|---|
| ||
assert(file: 'TEST'); |
Which means use local assertion (expected) file named 'TEST', and a remote target file named 'TEST', using the assertion mode 'equals', and reference-file-type named 'TEST.fml'.
You can now say use a custom sql statement or sql script to process the source (actual) file and output in the target reference file format for comparison:
| Code Block | ||||
|---|---|---|---|---|
| ||||
@Test
goodSqlExtract()
{
assert(
file: 'delimited_target_sql',
sql: 'SELECT POS_NAME, INPUT, ISNUM FROM $file-table-name',
source-reference-file-type: 'delimited_target_sample.out',
target-reference-file-type: 'delimited_target_sql',
target-file-name: 'delimited_target_sample.out'
);
}
@Test
goodSqlScriptExtract()
{
assert(
file: 'delimited_target_sql',
sql-script: 'extract',
source-reference-file-type: 'delimited_target_sample.out',
target-reference-file-type: 'delimited_target_sql',
target-file-name: 'delimited_target_sample.out'
);
} |
In both of these situations, the source-reference-file-type (the actual file) is loaded using the delimited_target_sample.out.fml format, then the sql statement is applied to extract those contents into a local file which matches the target-reference-file-type (expected), then the two are compared.
Note the FROM clause on the SQL statement. The table name is the special identifier $file-table-name. This must be used because the table is virtual and does not have any guaranteed or predictable name.
If include / exclude lists are needed, they can be used like this:
| Code Block | ||
|---|---|---|
| ||
@Test
goodScriptExtractWithExcludeColumns()
{
assert(
file: 'delimited_target_sql_excl',
sql-script: 'extract',
column-list-mode: 'exclude',
column-list: ['INPUT'],
source-reference-file-type: 'delimited_target_sample.out',
target-reference-file-type: 'delimited_target_sql',
target-file-name: 'delimited_target_sample.out'
);
}
@Test
goodScriptExtractWithIncludeColumns()
{
assert(
file: 'delimited_target_sql_excl',
sql-script: 'extract',
column-list-mode: 'include',
column-list: ['POS_NAME', 'ISNUM'],
source-reference-file-type: 'delimited_target_sample.out',
target-reference-file-type: 'delimited_target_sql',
target-file-name: 'delimited_target_sample.out'
);
} |
In this case, the sql is used as a direct extract from the source (actual) reference file type, and the output is then transformed using the exclude / include lists, and that result is used to create a subset of the target reference file type and the files compared. If the target-reference-file-type had not been specified, the output of the sql script and the column list would have to exactly match the assertion file.