HP 3000 Manuals HP 3000 Manuals
Data Description Entries (continued) SIGN Clause The SIGN clause is only used with a signed numeric data description item or description entry whose usage is DISPLAY, or a group item containing at least one such data description entry. It states the position of the sign, whether leading or trailing, as well as whether the sign was formed by overpunching in the first or last character of the data item (see the USAGE IS DISPLAY clause, below) or was formed separately. Syntax.Parameters. LEADING and indicates that the sign is at the beginning or end, TRAILING respectively, of the item. SEPARATE indicates that the sign is not overpunched; that is, the sign exclusively occupies the first or last character of this item. Description. Only one sign clause may be used per given numeric data description entry. Also, if the CODE-SET clause is specified in a file description, any signed numeric data description entry associated with that file must be described with the SIGN IS SEPARATE clause. Valid signs for data items, and their representations when overpunching is used, are shown in the table under the USAGE IS DISPLAY heading on the following pages. For a SIGN IS SEPARATE designation, the two valid operational signs (whether LEADING or TRAILING) are + or - for positive and negative quantities, respectively. For a signed numeric data description entry having no SIGN clause associated with it, the default is equivalent to SIGN IS TRAILING. That is, the sign is assumed to be overpunched in the last character of the item. In either the default case, or the case when the optional SEPARATE CHARACTER phrase is not used, the letter "S" in the PICTURE CLAUSE is not counted in determining the size of the item when represented in standard data format. To illustrate the SIGN clause: * The data to be entered is 123489F In this case, no SIGN clause is required because the default is SIGN IS TRAILING. However, if the PICTURE clause for this data item is PICTURE S9(7), the size of the data item is seven characters. * The data to be entered is +1409748 In this case, the SIGN clause should be: SIGN IS LEADING SEPARATE CHARACTER Also, since the sign is separate, the PICTURE clause for this data item, PICTURE S9(7) defines the data item to be eight characters long in order to hold the separate sign. SYNCHRONIZED Clause The SYNCHRONIZED clause is used to align items defined as USAGE IS COMPUTATIONAL or BINARY on word boundaries in order to facilitate arithmetic operations. A word size is defined by the operating[REV BEG] system environment (a word is 32 bits on MPE XL).[REV END] All other items are aligned on byte boundaries. Because the character (byte) is the smallest directly addressable unit within the COBOL language, the SYNCHRONIZED clause has no meaning when applied to an item with any usage other than COMPUTATIONAL. It is treated as a comment for items described as DISPLAY, INDEX, COMPUTATIONAL-3, or PACKED-DECIMAL. Syntax.Description. Because of the word structure used on HP computers, the LEFT and RIGHT options are irrelevant and are treated as comments by the compiler. The words SYNCHRONIZED and SYNC are equivalent. The compiler always aligns all level 01 and level 77 items on word boundaries[REV BEG] except in the LINKAGE section (see the section "OPTFEATURES" in Appendix G for details).[REV END] When the SYNCHRONIZED clause is specified for a data item whose description also contains an OCCURS clause, or in a data description entry of a data item subordinate to a description entry containing an OCCURS clause, each occurrence of the data item is synchronized. Any implicit filler (see "Slack Bytes", below) generated for other data items within that same table are generated for each occurrence of those data items. Slack Bytes. The SYNCHRONIZED clause specifies that the data described is to be aligned on word boundaries. If the SYNCHRONIZED item does not fall naturally on a word boundary, the compiler assigns the next highest boundary address to the item. The effect of adding a byte (or bytes) is equivalent to providing extra FILLER characters, known as slack bytes, just before the SYNCHRONIZED item. These slack bytes are not used for any other data item and are not counted in the size of the items. They are, however, included in the size of any group item or items to which the elementary item belongs, and are included in the character positions redefined when the SYNCHRONIZED item is the object of a REDEFINES clause. Therefore, when you use the REDEFINES clause in a data description that also contains a SYNCHRONIZED clause, you must ensure that the redefined item has the proper boundary alignment for the item that redefines it. The computation of boundary addresses is affected by the $CONTROL SYNC option. This option changes the alignment of SYNCHRONIZED data items, which affects the number of slack bytes generated in a record. Items with the SYNC clause are aligned along 16-bit (2 characters) boundaries, if SYNC16 is in effect. Items with the SYNC clause are aligned along 32-bit (4 character) boundaries, if SYNC32 is in effect. (Refer to the description of DATA DIVISION language dependencies and $CONTROL SYNC16/SYNC32 in "MPE XL System Dependencies" in Appendix H for synchronization alignment specifics.) This option may be coded more than once in a program to align one record along 16-bit boundaries, and another record along 32-bit boundaries. Alignment cannot be changed within a record, only between records. This option is especially useful to develop files to be used on other computer architectures, or to read files developed on other architectures. Whenever a SYNCHRONIZED item is referenced in your program, the original size of the item, as shown in the PICTURE clause, is used in determining any action that depends on size. Such actions include justification, truncation, or overflow. If the SYNCHRONIZED clause is not used, no space is reserved for slack bytes, and when a computation is performed on a data item described as COMPUTATIONAL, the compiler provides the code and space required to move the data item from its storage area to a work area. This work area has the alignment required to perform the computation. As an illustration of slack bytes (assuming 16-bit synchronization), consider the following data description entries: 01 ITEM-LIST. 02 ITEM-NUMBER PICTURE X(3). 02 ITEM-1 PICTURE X(4). 02 ITEM-2 REDEFINES ITEM-1 PICTURE S9(6) USAGE COMP SYNC. The above is an example of not taking into account the slack byte required because of the REDEFINES clause. To correct it, the description of ITEM-LIST should include an extra byte prior to ITEM-1: 01 ITEM-LIST. 02 ITEM-NUMBER PICTURE X(3). 02 SLACK-BYTE PICTURE X. 02 ITEM-1 PICTURE X(4). 02 ITEM-2 REDEFINES ITEM-1 PICTURE S9(6) USAGE COMP SYNC. This change is all that was needed, since all 01 level entries are aligned on word boundaries[REV BEG] except in the LINKAGE section (see the section "OPTFEATURES" in Appendix H for details).[REV END] Example. The following illustrates the use of $CONTROL SYNC: $CONTROL SYNC32 01 OUT-REC. 05 A PIC X. 05 B PIC S9999 BINARY SYNC. 05 C PIC X. $CONTROL SYNC16 01 IN-REC. 05 A PIC X. 05 B PIC S9999 BINARY SYNC. 05 C PIC X.In the above example, three slack bytes are inserted before B of OUT-REC. One slack byte is inserted before B of IN-REC.
NOTE [REV BEG] Due to the $CONTROL SYNC option,[REV END] boundary alignment within a record is constant. It only changes between records.
USAGE Clause The USAGE clause specifies how the data item being described is stored internally. When used with index, the USAGE clause specifies that the data item being described contains a value equal to the value of an index name associated with an occurrence number for a table element. Syntax.Parameters. PACKED-DECIMAL, equivalent; they specify packed-decimal format. COMPUTATIONAL-3, and COMP-3 The words COMPUTATIONAL-3 and COMP-3 are an HP extension to the 1985 ANSI COBOL standard. BINARY, equivalent; they specify two's complement binary COMPUTATIONAL, and integer format. COMP DISPLAY the default usage if no USAGE clause is specified. It specifies that data is to be stored internally as ASCII characters. INDEX specifies an index data item. Description. This clause is optional, with a USAGE IS DISPLAY being used by your program for any data item having no USAGE clause as part of its description, or as part of the description of a data item to which it is subordinate. You may use the USAGE clause at any level of organization. However, if written at a group level, the USAGE clause applies to each elementary item in the group, and any USAGE clause specified at the elementary level must be the same as at the group level. Although a USAGE clause does not affect the use of a data item, some of the statements in the PROCEDURE DIVISION may restrict the USAGE clause of the operands used. USAGE IS DISPLAY. When usage of a data item is defined (implicitly or explicitly) as DISPLAY, the data is stored internally as ASCII characters. This means that each character of data is stored as an 8-bit byte. If you are using the data item in a noncomputational manner (that is, printing or displaying it), this is the appropriate type of usage to be specified. However, for optimum use of your COBOL program, you should specify USAGE IS COMPUTATIONAL, or BINARY, COMPUTATIONAL-3, or PACKED-DECIMAL for data items intended for use in computations. This is because data items described as USAGE IS DISPLAY must be converted to two's-complement binary or packed-decimal format before they can be used in computations, and this conversion takes time. Also, if you intend to use signed numeric data items for computational purposes, you must specify a sign (by using the S symbol) in the PICTURE clause for that item (see USAGE IS COMPUTATIONAL on the following page), whether its usage is specified as DISPLAY or otherwise. An unsigned numeric data item whose description specifies the USAGE IS DISPLAY clause is assumed to be positive. Numeric DISPLAY items without a clause that designates SIGN IS SEPARATE are represented in ASCII coded (8-bit) decimal digits (0 through 9) except for the units digit which carries the sign of the data item. The units digit, with the sign of its associated number being positive, negative, or no sign (absolute value) respectively, is represented in ASCII code as shown in Table 7-7 . Note that using signed numeric DISPLAY data items for computational purposes is more efficient than using unsigned numeric data items. Table 7-7. Overpunch Characters for Rightmost Digit in ASCII Coded Decimal Numbers --------------------------------------------------------- - Units Digit - Internal Representation (ASCII) - --------------------------------------------------------- - - Positive - Negative - No Sign - --------------------------------------------------------- - 0 - { - } - 0 - --------------------------------------------------------- - 1 - A - J - 1 - --------------------------------------------------------- - 2 - B - K - 2 - --------------------------------------------------------- - 3 - C - L - 3 - --------------------------------------------------------- - 4 - D - M - 4 - --------------------------------------------------------- - 5 - E - N - 5 - --------------------------------------------------------- - 6 - F - O - 6 - --------------------------------------------------------- - 7 - G - P - 7 - --------------------------------------------------------- - 8 - H - Q - 8 - --------------------------------------------------------- - 9 - I - R - 9 - --------------------------------------------------------- Signed decimal fields entered through punched cards are known as zone-signed fields. To represent a positive value, an overpunch is placed in the 12-zone above the rightmost digit of the field. To represent a negative value, an overpunch is placed in the 11-zone above the rightmost digit of the field. If no sign is desired, only the digits need be punched. Zone signs cause the signed digit to have the same punch configuration as certain other characters. This is the purpose of the S symbol in the PICTURE clause; it informs the compiler that the last digit in the field is to be interpreted as a number and a sign, and not as the character that it would otherwise represent. Table 7-7 shows the data character equivalents to each possible rightmost digit sharing a zone sign. USAGE IS BINARY or COMPUTATIONAL. When usage of a data item is defined as BINARY or COMPUTATIONAL, the data must be numeric. It is stored in two's-complement binary integer form, consisting of either two, four, or eight bytes each. The number of bytes used depends upon the size of the data item, as shown in Table 7-8 below. Table 7-8. Number of Bytes Used to Contain a BINARY Data Item --------------------------------------------------------------------------------------------- - PICTURE - Number of Bytes - --------------------------------------------------------------------------------------------- - S9 to S9(4) - 2 - - S9(5) to S9(9) - 4 - - S9(9) to S9(18) - 8 - --------------------------------------------------------------------------------------------- A data item whose usage is defined as BINARY or COMPUTATIONAL must have an unedited numeric PICTURE clause associated with it. It may contain up to 18 digits plus a sign. Also, if a group item is described as BINARY or COMPUTATIONAL, all of the elementary items in the group are computational and may be used in computations. However, the group item itself may not be used in computations because it is considered alphanumeric. A numeric data item that does not have a sign associated with it is assumed to be positive. As with numeric DISPLAY data items, a signed numeric data item whose USAGE IS BINARY is more efficient than an unsigned numeric data item with the same USAGE. USAGE IS PACKED-DECIMAL or COMPUTATIONAL-3. Data items described as PACKED-DECIMAL or COMPUTATIONAL-3 are subject to the same restrictions and are used in the same way as data items described as COMPUTATIONAL. Such items are, however, stored in packed-decimal format. In this format, there are two digits per byte, with a sign in the low order 4-bits of the rightmost byte. Each PACKED-DECIMAL or COMPUTATIONAL-3 item may contain up to 18 digits plus a sign. If the picture for the item does not contain a sign, the sign position in the data field is occupied by a bit configuration that is interpreted as positive. Table 7-9 illustrates the bit configurations used to represent signs in packed-decimal fields. Notice that the bit configuration 1100 specifies a positive value and that the 4-bit configuration 1111 represents the unsigned (assumed positive) value when an unsigned picture is specified. For negative values, the 4-bit configuration is 1101. Table 7-9. COMPUTATIONAL-3 or PACKED-DECIMAL Sign Configuration ---------------------------------------------------------------------------------------------- | | Bit | Hexadecimal | | Sign | Configuration | Value | ---------------------------------------------------------------------------------------------- - + - 1100 - C - ---------------------------------------------------------------------------------------------- - - - 1101 - D - ---------------------------------------------------------------------------------------------- - Unsigned - 1111 - F - ---------------------------------------------------------------------------------------------- Table 7-10 gives a graphic illustration of packed-decimal fields as they might appear in memory or in a file. Notice that these items follow the normal rules for truncation, even though the field may include an unused half-byte position. The contents of this half-byte are unpredictable when data is interchanged with other computer systems. In the table below, each box in the result column represents a byte. Table 7-10. PACKED-DECIMAL Fields in Memory or in a File ------------------------------------------------------------------------------------------------------- | Value to be | PICTURE of | Result | | Stored | Result | | ------------------------------------------------------------------------------------------------------- - +1234. - S9999 - - 01 - 23 - 4C - ------------------------------------------------------------------------------------------------------- - +12345. - S99999 - - 12 - 34 - 5C - ------------------------------------------------------------------------------------------------------- - 12345 - S9999 - - 12 - 34 - 5F - ------------------------------------------------------------------------------------------------------- - -1.2 - S999V999 - 00 - 01 - 20 - 0D - ------------------------------------------------------------------------------------------------------- - -.5 - S999V999 - 00 - 00 - 50 - 0D - ------------------------------------------------------------------------------------------------------- - +1.22172 - S999V999 - 00 - 01 - 22 - 1C - ------------------------------------------------------------------------------------------------------- - -12345. - 99999 - - 12 - 34 - 5F - ------------------------------------------------------------------------------------------------------- Note that the third and last number in the table were stored as unsigned (assumed positive) numbers because the receiving field is unsigned according to its PICTURE. USAGE IS INDEX. An elementary data item whose usage is defined as INDEX is called an index data item. Its purpose is to hold the contents of a table index while the table is being processed. Therefore, any value within the index data item must correspond to an occurrence number of an element in a table. An index data item is stored as a synchronized unsigned computational integer, both internally and externally. (Refer to "System Dependencies" in Appendix H for the correct size of an index data item.) An index data item cannot be a conditional variable, and can only be referenced explicitly in a SEARCH or SET statement, a relation condition, the USING phrase of the PROCEDURE DIVISION header, or USING phrase of a CALL statement. Do not use the SYNCHRONIZED, JUSTIFIED, PICTURE, VALUE, and BLANK WHEN ZERO clauses to describe a group of elementary items whose usage is defined as INDEX. Index data items are automatically SYNCHRONIZED. In ANSI COBOL'85, if a group item is described with a USAGE IS INDEX clause, all of its elementary items are index data items, but the group itself is not an index data item and cannot be used in the SEARCH or SET statements, or in an alphanumeric comparison in the PROCEDURE DIVISION. As an HP extension to the ANSI COBOL standard, HP COBOL II allows a group item described with USAGE IS INDEX to be used in an alphanumeric comparison. VALUE Clause The VALUE clause is used to define the values of constants and to initialize the values of WORKING-STORAGE data items. For information on VALUE clauses in condition names, refer to "Condition Names" later in this chapter. Syntax. [VALUE IS literal-1] Parameters. literal-1 the value assigned to the data item being described. Description. The above format for the VALUE clause can only be used in the WORKING-STORAGE SECTION. If used, the VALUE clause causes the item to which it is associated to assume the specified value at the start of the object program, irrespective of any BLANK WHEN ZERO or JUSTIFIED clause. If the VALUE clause is not used in an item's description, the initial value of the item is undefined. A VALUE clause may be used with data items (or descendants of data items) containing an OCCURS clause to initialize tables. A VALUE clause (VALUE is literal-1) specified in a data description entry that contains an OCCURS clause, or in an entry that is subordinate to an OCCURS clause, causes every occurrence of the associated data item to be assigned the specific value. If a VALUE clause is specified in a data description entry of a data item that is associated with a variable occurrence data item, {{the initialization of the data item behaves as if the value of the DEPENDING ON phrase in the OCCURS clause is set to the maximum number of occurrences specified by that OCCURS clause. A data item is associated with a variable occurrence data item in any of the following cases: * When it is a group data item that contains a variable occurrence data item. * When it is a variable occurrence data item. * When it is a data item that is subordinate to a variable occurrence data item. If a VALUE clause is associated with the data item referenced by a DEPENDING ON phrase, that value is considered to be placed in the data item after the variable occurrence data item is initialized. Restrictions on the Use of the VALUE Clause. The following restrictions apply to the use of the VALUE clause. They also apply to the VALUE clause in condition names. * The VALUE clause cannot be used in a data description entry containing a REDEFINES clause (except when used with a condition name), or in an entry subordinate to an entry containing a REDEFINES clause. Nor can it be used for a group containing items with descriptions including JUSTIFIED, SYNCHRONIZED, or USAGE (except USAGE IS DISPLAY). * The VALUE clause must not conflict with any other clauses in the data description of the item, or in the data description within the hierarchy of the item. * The VALUE clause must not be used with any external record, except for condition-name entries associated with external records. Literals in the VALUE Clause. The literals used in the VALUE clause are subject to the following rules: * Figurative constants may be substituted for literals. * A signed numeric literal must have a signed numeric PICTURE and character string associated with it. * All numeric literals must have a value within the range of values indicated by the PICTURE clause, and must not have a value which would require truncation of nonzero digits. Nonnumeric literals must not exceed the size indicated by the PICTURE clause. * If the category of the item being described is numeric, all literals in the VALUE clause must be numeric. If the literal defines the value of a working-storage item, the literal is aligned in the data item according to the standard alignment rules. * If the category of the item being described is any other than numeric, all literals in the VALUE clause must be nonnumeric. The literal is aligned in the data item as if the data item had been described as alphanumeric. Editing characters in the PICTURE clause are included in determining the size of the data item, but have no effect on initialization. Therefore, the VALUE for an edited item must be presented in an edited form. * If the VALUE clause is used in an entry at the group level, the literal must be a figurative constant or a nonnumeric literal, and the group area is initialized without consideration for the individual elementary or group items contained within the group. A VALUE clause cannot be used for elements of a group that has a VALUE clause assigned to it at the group level. RENAMES Clause The RENAMES clause permits alternative, possibly overlapping groupings of elementary items. This clause is always associated with a 66 level entry. Syntax.The level number (66) and data-name-1 are not part of the RENAMES clause, but are used to clarify the purpose of the clause. Parameters. THROUGH and equivalent. THRU data-name-1 the name used to rename the item or items referenced by data-name-2 and data-name-3. It cannot be a qualifier, and can only be qualified by the names of the associated 01, FD, or SD level entries. data-name-2 and must be names of elementary items or groups of data-name-3 elementary items in the same logical record. They must not be the same name, and neither may have an OCCURS clause in its data description, or be subordinate to an item that has an OCCURS clause in its description. Furthermore, no item within the range of the portion of the logical record being renamed can be variable in size, or can contain such an item. Data-name-2 and data-name-3 may be qualified. If data-name-2 is used alone (that is, the optional THROUGH phrase is unused), data-name-2 can be either a group or an elementary item. Description. When data-name-2 is a group item, data-name-1 is treated as a group item; when data-name-2 is an elementary item, data-name-1 is treated as an elementary item. If the THROUGH phrase is used, data-name-1 is a group item that includes all elementary items starting from data-name-2 and concluding with elementary item data-name-3. Or, if data-name-2 and data-name-3 are also group names, data-name-1 is a group name that begins with the first elementary item in data-name-2 and concludes with the last elementary item in data-name-3. Because of the way in which data-name-1 is defined, there are restrictions on the area described by data-name-2 and data-name-3. That is, the area described by data-name-3 must not begin to the left of the first character in the area described by data-name-2, and it must end to the right of the last character of the area of data-name-2. Example. 01 STUDENT-REC. 03 NAME PIC X(20). 03 ID-NO PIC X(9). 03 MAJOR PIC X(3). 03 CLASSES OCCURS 5 TIMES. 05 CLASS-ID PIC X(3). 05 DEPT PIC X(3). 66 STUD-INFO RENAMES NAME THRU MAJOR. 66 MAJOR-DEPT RENAMES MAJOR. In the above example, CLASSES, CLASS-ID, and DEPT may not be named in a RENAMES clause because of the OCCURS clause.
NOTE The above paragraph implies that data-name-3 cannot be subordinate to data-name-2.
You can use more than one RENAMES entry for a single logical record. However, all RENAMES entries referring to data items within a given logical record must immediately follow the last data description entry of the associated record description entry. You cannot use a level 66 entry to rename another level 66 entry or a 77, 88, or 01 level entry. Condition Names A condition name is always subordinate to another data item called the conditional variable. The level number 88 is used to describe it. The characteristics of a condition name are implicitly those of its conditional variable. This must be reflected in the value or values assigned to the condition name. A condition name is assigned one or more values. The condition name can later be used to specify comparison with the conditional variable (see Chapter 8 , "PROCEDURE DIVISION", for information on condition name conditions). Syntax. The VALUE clause in a condition name has the following format:Parameters. condition-name- any valid user-defined COBOL word. 1 literal-1 and the values assigned to the condition name. literal-2 THROUGH and equivalent and can be used interchangeably. THRU Description. The VALUE clause and the condition name itself are the only two clauses permitted in the data description entry. The VALUE clause can be used in any of the sections of the DATA DIVISION and must be used for condition names. Wherever the THROUGH phrase is used, literal-1 must be less than literal-2. When a VALUE clause is used in a level 88 entry, you can specify no more than 127 ranges of values for the related condition name. A range of values is either a single literal or two literals related by the THROUGH (or THRU) keyword. Additional rules applying to the VALUE clause in condition names are described under the headings "Restrictions on the Use of the VALUE Clause" and "Literals in the VALUE Clause", earlier in this chapter. The condition name entries for a particular conditional variable must follow the entry describing the item with which the condition name is associated (that is, the conditional variable). Each condition name in your program must have a separate level 88 entry associated with it. A condition name cannot be associated with any data description entry containing a level number 66, another condition name, or a group item with descriptions including JUSTIFIED, SYNCHRONIZED, or USAGE (other than DISPLAY). Example. 01 CONDVAR PIC 9(5) USAGE DISPLAY. 88 COND1 VALUE 10 THRU 25, 100 THRU 250. 01 ALPHAVAR PIC A. 88 ALPHACOND VALUE "A" , "M" THROUGH "Z".
![[]](../../pic3k/M011820/FFN533c50.gif)
![[]](../../pic3k/M011820/FFN543c50.gif)
![[]](../../pic3k/M011820/FFN553c50.gif)
![[]](../../pic3k/M011820/FFN563c50.gif)
![[]](../../pic3k/M011820/FFN573c50.gif)
![[]](../../pic3k/M011820/FFN583c50.gif)