Creating Dataviews for Sequential Files › Defining Modeled Sequential Files › Rules for Sequential File Dataviews

Rules for Sequential File Dataviews

CA Ideal imposes certain rules on dataviews used for sequential files. The CA Ideal CATALOG command enforces the rules. For more information, see the Command Reference Guide.

Dataview Status and Related Occurrences

A dataview can be successfully cataloged in TEST or PROD status. A dataview that is cataloged in TEST status has all elements in TEST status and all version numbers of the elements are the same as the dataview version. A dataview that is cataloged in PROD status has all elements in PROD status. The elements in PROD status do not have to have the same version number as the dataview. The dataview must be related to one element. The element must be related to a single record, and the record to a file.

Note: Exceptions exist for dataviews created before Datadictionary r2.4. Refer to Datadictionary documentation.

Dataview Attributes

For a dataview that CA Ideal accesses, the only DATAVIEW attribute whose specified value is used is UPDATE‑INTENT. You can specify Y for updateable or N for non‑updateable. (U for updateable is also supported for compatibility with earlier releases.) CA Ideal interprets any other value as N, non‑updateable. If the DATABASE entity‑occurrence name is DATA‑DICT or UNIVERSAL, CA Ideal assumes a value of N, non‑updateable.

The attributes DBMS‑USED and ACCESS‑METHOD for the dataview attribute are updated with values CA Ideal supplies when the dataview is cataloged.

FILE Attributes

For CA Ideal to access a sequential dataview, FILE attributes must be defined as follows:

• DBMS‑USED must be NONE
• ACCESS‑MTHD must be QSAM or SAM
• RECORD‑FORMAT must be FBLK or FUNB
• MONITOR‑NAME must be supplied (see following)
• Under VSE, the LABEL, DEVICE, MAX‑RECORD‑SIZE, and MAX‑BLOCK‑SIZE attributes must be specified for SAM files

This information is summarized in the following table.

For more information, see Run‑Time Considerations for Sequential File Dataviews in this chapter.

Sequential File MONITORNAME

For a sequential file dataview, the MONITOR‑NAME attribute of the FILE entity‑occurrence must contain a logical data set identifier. (The term MONITOR‑NAME indicates that this is the logical name by which the file is known to the host monitor.) In batch, the host monitor is actually the host operating system. In a given run, each dataview can have a unique MONITOR‑NAME relating each dataview to a separate physical data set. Or two (or more) dataviews can have the same MONITOR‑NAME value, which causes the two dataviews to share the same physical dataset (for output files, these results in interleaved output).

Under z/OS or VSE

All runs in a CICS partition that use the same sequential dataview for output write interleaved records to the same physical file. In addition, MONITOR‑NAME connects the dataview to the host environment in the following ways:

• CICS The first four characters of the MONITOR‑NAME must appear as the DESTID in a DFHDCT entry. This DESTID relates the dataview to the JCL as with any CICS DFHDCT.
• Batch (z/OS) The MONITOR‑NAME becomes the ddname for the file and must conform to the rules for z/OS ddnames.
• Batch (VSE) The MONITOR‑NAME becomes the DLBL name for DISK files or the TLBL name for standard label tape (SLTAPE) files and, in these cases, must conform to the rules for VSE DLBL and TLBL names. For all other device types, MONITOR‑NAME is not used in the JCL.

Variable Length Records

Not supported.

Element Consisting of Contiguous Fields

For a sequential dataview, there can be only one element. It must explicitly include all the fields in the record. It is possible to accidentally violate this rule by inserting a field through DEFINE FILE, but forgetting to include the field in the appropriate DEFINE or MODIFY element.

Levels of Data Structure Nesting

The maximum number of nested levels of groups and fields in a dataview is 15. Since the dataview itself is the outermost level (level 1), the highest level number allowed is 16. Since the dictionary assigns level 1 to all columns with a PARENT of START, the level numbers that CA Ideal assigns is always at least one greater than the level numbers assigned by the dictionary.

Number of Levels of Occurrences

CA Ideal supports a maximum of three nested levels of occurring fields (that is, tables of up to three dimensions are supported).

Compound Fields

Compound fields (group fields) should not have initial values. If they do, the value is ignored and a warning message issued.

Data Types of Fields

CA Ideal supports the following data types:

• Alphanumeric (display) (C)
• Zoned decimal (N)
• Packed decimal (D), and
• Binary (B)

CA Ideal converts all other unknown data types to alphanumeric and issues a warning message.

CA Ideal supports numeric fields up to the equivalent of 31 decimal (base 10) digits (for zoned or packed decimal) and up to the equivalent of 9 decimal (base 10) digits (full word or half word binary) for binary fields. CA Ideal supports only half word (2 bytes) and full word (4 bytes) binary fields. You can define binary fields of other than 2 or 4 bytes (including data types 2, 4, and 8), but CA Ideal treats them as alphanumeric fields (with a warning message) and the user must call a non‑CA Ideal subprogram to handle them. CA Ideal does not support floating point items of any precision.

Other Field Attributes

CA Ideal ignores FIELD occurrences with CLASS=I (INDEXED BY). The BLANK‑WHEN‑ZERO attribute is also ignored. If either the LOW‑RANGE or HIGH‑RANGE attribute is used with a CLASS=V FIELD, a warning message occurs.

The REDEFINES attribute has the following values:

• N (Does not redefine) is the default
• P or S Treated as P (redefines the previous FIELD occurrence at the same level)

Edit Patterns

You can specify any valid COBOL edit pattern (except Sterling currency symbols) for a field in the dictionary (attribute EDIT‑PATTERN). It is used in CA Ideal reports unless overridden by RDF (see the Generating Reports Guide). PDF also uses the edit pattern specified in the dictionary for dynamically generated panel fields.

Field Column Headings

You can specify field column headings for fields in the dictionary (attributes HEADING‑1 and HEADING‑2). However, headings longer than 20 characters are truncated to 20 characters in the CA Ideal cataloged dataview. PDF uses the field column headings specified in the dictionary for dynamically generated panel fields. They are also used in CA Ideal reports unless overridden by RDF.

Note: CA Ideal uses a not‑sign () as the begin‑line character in report headings. For more information, see the Generating Reports Guide.

PDL Reserved Words

You cannot use a PDL reserved word as the name of a dataview (see the reserved word list in the Command Reference Guide). You can use reserved words for field names; however, qualification of such field names with the dataview name is required. For example, assume there is a dataview called SHIPMENT with elements containing fields called TO and FROM. It is valid for a PDL procedure to have statements:

MOVE SHIPMENT.TO TO LOCATION
SUBTRACT SHIPMENT.FROM FROM AMOUNT

However, the following command is ambiguous and causes a program compilation error.

MOVE TO TO LOCATION

Use of Field Entity Names

For names of fields, CA Ideal uses the 32‑character field entity‑occurrence name, not the 30‑character compiler name. Field names in CA Ideal must be unique in the dataview. This occurs automatically since all elements must belong to one record and the dictionary requires field names to be unique in the record name.

MultiValue Condition Names

CA Ideal supports one unique value for each condition name (class V field in the dictionary). Lists or ranges of values are not supported.

Figurative Constants

CA Ideal supports the following figurative constants as values in the dictionary:

SPACE
SPACES
ZERO
ZEROS
ZEROES

The VALUE attribute for a condition‑name (class V field in the dictionary) cannot be blank. A value of all spaces can only be represented by the figurative constants SPACE or SPACES.

Optional FIELD Attributes

CA Ideal recognizes the following considerations for optional FIELD attributes:

• SEMANTIC‑TYPE Entered for date fields only. Enter a SEMANTIC‑TYPE of CA‑DATE for a numeric field (TYPE=N, D, or B) that contains an internal date stored as the number of days since (positive number) or preceding (negative number) the date January 1, 1901. A five‑digit number can contain values from ‑99,999 through 99,999 or the equivalent of almost 274 years before and after the base date. This is from approximately the year 1627 to approximately the year 2175. A six‑digit number can represent dates approximately 2740 years before and after the base date, and so on. SQL semantic types may not be used for sequential dataviews.
• DECIMALS Total number of decimal (fractional) places in the number. CA Ideal determines the number of integer places by calculating the total number of decimal (base 10) digits that fit in the field (based on the TYPE) and subtracting the DECIMALS value.
• DEPENDING ON Name of a non occurring simple numeric field with DECIMALS=0 that precedes this field in the record. The Advantage
  CA-Ideal restriction that a field with a DEP ON clause can be a simple or compound field, but must not be followed in the dataview by any other fields (unless they are subordinate fields to a compound field).
• EDIT‑PATTERN COBOL‑like edit pattern used when printing the field. The CA Ideal Report Definition Facility and the CA Ideal Panel Definition Facility use this pattern as the default for the field. If you omit this edit pattern, the following default edit pattern is used:
  • X(n) Alphanumeric fields (where n is the length)
  • 9(i).(9d) Numeric fields (where i = integers, d = decimals)
• HEADING‑1 and HEADING‑2 One or two literals (without quotes) used as the default column headings for the CA Ideal Report Definition Facility and for dynamic field definition using the CA Ideal Panel Definition Facility. For example, if ACCOUNT is specified as the value for HEADING‑1 and NUMBER is specified as the value for HEADING‑2, the column header appears as follows:

  ACCOUNT
  NUMBER
  

• SIGN N to force “unsigned” (the default for numeric fields is Y).
• VALUE For numeric fields, a valid numeric literal or one of: ZERO, ZEROS, ZEROES.

  For alphanumeric fields, enter the characters to appear as the initial value without quotes (or any other delimiter) or enter one of: SPACE, SPACES.

  For a field with CLASS=V (value, or condition‑name in CA Ideal), you must type at least one non‑blank character for VALUE. If the condition‑name corresponds to a value of spaces, type the word SPACE or SPACES.

  If a value contains embedded blanks, type in the value with no delimiters. For example, enter a three‑character value “A B” as A in position 1, blank in position 2, B in position 3.

  CA Ideal does not support any of the following:

  • Multiple values for a single condition‑name.
  • A range of values for a field, either in the VALUE attribute or the LOW‑RANGE and HIGH‑RANGE attributes.
  • COBOL figurative constants:
    • HIGH‑VALUE(S)
    • LOW‑VALUE(S)
    • QUOTE
    • ALL literal
  • NULL-INDICATOR is not supported for sequential dataviews.