The next few steps apply only to files that Analyzer identifies that do not store or contain any internal information about their field names or file structure.
Analyzer can read the contents of the data file, but needs you to provide additional information available from the data owner. In the following panels you need to define the file and field attributes. These include confirming the type of data in each field and providing field names and descriptions. This information is usually contained in a document called a data definition.
Analyzer displays a partial View of the data in the file and asks you to confirm whether the data is aligned vertically in columns. If it is, Analyzer has correctly defined the file to this point. Click [Next] to continue. If not, see Skewed or Snaking Records.
When the data in records do not line up, you will need to make some adjustments. If records appear staggered, the record length is incorrect.
Record length refers to the length of each record in a file that has fixed record lengths or the length of the longest record in a file that has records of variable length. To correct the alignment, increase or decrease the record length.
If the records appear to snake, it may indicate the presence of unnecessary bytes or a header record that has a different length than the data records that follow. Report files sometimes cause the records to snake as well. To correct this you may need to skip a number of bytes that correspond to this record’s length at the beginning of the file. Specify the number of bytes to skip to exclude data at the beginning of a file from being analyzed.
The following table will help you to correct misaligned records:
|
Symptom |
Problem |
Solution |
|
Fields in the records are skewed down to the right |
The record length is likely too short |
Increase record length |
|
Fields in the records are skewed down to the left |
The record length is likely too long |
Decrease record length |
|
The records area consists mostly of blank spaces and punctuation characters |
You have incorrectly changed the character set identified by Analyzer (ASCII or EBCDIC) |
Go Back in the Wizard and choose the other character set |
|
The record length appears too large |
The record length is probably a multiple of (2, 3 or 4 times) the actual record length. |
Try entering a value in the Record Length text box equal to a fraction (such as 1/2) of the displayed record length |
|
The records appear to “snake” |
The skip length is too small. |
Increase the skip length |
If you are working with a multiple-record type file, such as a print image file, you can skip the header records that appear only at the top of the file by entering a value in the Bytes to Skip box until these records are no longer visible.
For example, if the first 317 bytes contain only header information, you would specify Bytes to Skip as 317.
Analyzer adjusts the data display to show only the records that follow the skipped bytes.
Once you have finished, click [Next] to accept the Wizard’s analysis and to let the Wizard continue defining the data.
In rare cases, complex data structures cannot be adequately defined by the Wizard and must be defined manually. To define a file manually, select the Skip Field Identification checkbox in the File Properties panel and click [Next]. You are taken directly to Final panel and are prompted to enter a name for the table layout. When you click [Finish], Analyzer displays the Edit Table Layout tab. For more information about defining files manually, see Defining Data Manually.
To View the data in hexadecimal format, check the Hex checkbox. This option is useful when working with unprintable characters or compressed data, such as packed numeric data. When the Hex checkbox is checked, two lines appear that are read vertically from left to right displaying the two digit hex value of each character.