<<

Contents

Search

>>

 

Index Files (Menu item File/Open/Index File)

 

Hyperlinks:

Copy Index File

 

 

The programs created with Ebus-J can access sequential and Random files as well as Ebus Index Files (ISAM files) and (via JDBC) any SQL databases.

 

Each Ebus Index File <name> consists of 3 physical files:

 

<name>
  or
<name>.dbf

contains the data with file type Ebus.

contains the data with file type Dbase.

 

 

<name>.i1 

contains the indices.

 

 

<name>.ebu

contains the index structure.

 

These three files are logically summarized into one Index File by the Ebus-J. Only the logical file format is important for the user, the physical data access is generated by Ebus-J. Therefore, the descriptions of the Index Files must be recorded in the Index File administration.

 

Sequential and random files are not recorded in the Index File administration. They can be called with the programming language of the Ebus-J by the instructions OPEN, INPUT, LINE INPUT, OUTPUT, PUT, GET and CLOSE.

 

The descriptions of Index Files are saved in the library ebusdlib. Each time a new working directory is used with the program generation, a new file library is created automatically.

 

Note: The Ebus-J needs the Index File description for the generation of the program; it contains the name and the format of the Index File. While the Ebus program is running, an Index File (which is completely independent of the file library ebusdlib) with the respective filename is opened. The data sets are saved in this Index File.

 

 

 

Filename: The file name of an Index File must begin with a letter and may contain letters and numbers only.

 

Button 'Save': Data are saved and program execution continues at Filename.

 

Button 'Cancel': Changed data are not saved and program execution continues at Filename.

 

Key <Esc>: Program execution returns to Filename. If data have been changed, the user will be asked first if the changes shall be saved.

 

The following function keys (Icons) are available:

F2: Select Index File

F3: Scroll forward: next file name

F4: Scroll backward: previous file name

F5: Delete: The entire Index File specification is deleted after confirmation.

      Note: With deleting a file description the description is deleted only, not the index file.

F7: Select Index File (History)

F8: Copy Index File

F9: Print Index File

 

 

Tab Fields

 

 

Every Index File cannot contain more than 2 to the power of 64 data sets.

 

Every data set consists of any number of fields. There are 3 different field types:

1) alphanumeric (Length maximum 99999 Byte)
2)
numeric (Length maximum 15 Byte)
3) Date (Length 6 or 8 Byte)

 

The sum of the field lengths of all fields defined in an Index File is the length of the data set.

 

The data are filed in the fields in ASCII notation and also saved like this on the hard disk.

 

Texts in alphanumeric fields are saved left justified and filled up with blanks.

 

Numbers in alphanumeric fields are saved as decimal numbers. Positive numbers are saved without sign; negative numbers are saved with sign. The number may be filled up with leading zeros in order to fill the field completely.

 

Example:

 

Field length 6

-00123
-12345
000000
000123
123456

 

With fields with post decimal positions the decimal point is not saved. Only the numbers are saved.

 

 Example:

 

Field length 6, 2 post decimal positions

Value

saved string of characters

- 1.23
- 123.45
0.00
1.23
1234.56

-00123
-12345
000000
000123
123456

 

Date fields are numeric fields and are always saved in the format YYYYMMDD (Length 8) or YYMMDD (Length 6) by Ebus-J.

 

Example:

 

Date

saved value

28.07.1991
28.07.91

19910728
910728

 

Each record of an Index File is composed of several fields.

 

Each field consists of the following parts:

 

Field Name: The field name is used for accessing the field of an Index File. In the macros fields of Index Files can be accessed with <Filename>.<Fieldname>. The field name may contain letters, numbers and dots only. Alphanumeric fields of Index Files must end with '$'.

 

Type: Specifies the field type of the Index File. There are 3 field types:

A = alphanumeric

N = numeric

D = date

 

Length: Specifies the whole length (decimals included) of the field. With numeric fields (N) the maximal length is 15. With date fields (D) only 2 different lengths are allowed: 6 (MM/DD/YY) and 8 (MM/DD/YYYY).

 

PDP: PDP means "post decimal positions"; they need to be specified with numeric fields (N) only. Length is the whole length of the field, PDP must therefore be <= Length.

 

Remark: The remark is used for information only. With memo fields (Dbase) the first entry must be MEMO in capital letters; after that any kind of text can be entered. With logical field (Dbase) the first entry must be LOGIC in capital letters; after that any kind of text can be entered.

 

 

Tab File Type

 

 

The file type specifies which format is used for the Index File when the program is running.

 

The Dbase-III format is used for the file type "Dbase". However, only the data file <FILENAME>.dbf is created in Dbase-III format, the index (<FILENAME>.i1) is created in Ebus format.

 

With the file type JDBC the Index File is opened in a SQL database. The required parameters can be set by the following commands:

JDBC_DRIVER specifies the JDBC driver for SQL database access.

JDBC_DATABASE specifies the name of the database in which the Index File is opened.

JDBC_TABLE specifies the table name in the database.

JDBC_USERID specifies the user name for database access.

JDBC_PASSWORD specifies the password for database access.

 

The following restrictions apply to Dbase:

The field name may not exceed 10 characters.

6-digit date fields are not allowed.

Memo fields contain the type <A> (alphanumeric) and are exactly 10 units  long. The first entry under "Remark" must be MEMO in capital letters; after  that any kind of text can be entered. Memo fields can only be written with  the instruction PUT_MEMO and can only be read with the function GET_MEMO$

Logic fields contain the type <A> (alphanumeric) and are exactly 1 unit  long. The first entry under "Remark" must be LOGIC in capital letters; after  that any kind of text can be entered.

 

Import Dbase File Description from file: Specifies the dbase file name from which the dbase file description is imported.

 

Button ' Import Dbase File Description ': The dbase file description is imported from the specified file.

 

 

Tab Keys

 

 

Every data set of an Index Files is clearly identified by a primary key. In addition, up to 32 secondary keys exist.

 

A key consists of up to 10 fields of the data set. The key length is the sum of the field lengths. It may not exceed 100 bytes.

 

The contents of the fields of the key are the value of the key, by which the sorting order of the data sets is specified. The values of the primary key must be unique for every data set. Secondary keys can be ambiguous, which means that several data sets may have the same secondary key value. Within similar secondary key values the sorting order is random.

 

Since key values of numeric fields are calculated according to their ASCII notation, the higher number corresponds to the higher key value with positive numbers. Since the ASCII value of the minus sign is lower than that of zero, all negative numbers have a lower key value than positive numbers. However, negative numbers with a higher absolute value have a higher key value than negative numbers with a lower absolute value, because the ASCII values of the decimal numbers rise with their value.


Example:

 

Number

ASCII display
Field length 6

- 1
- 123
- 99999
0
1
123
99999

-00001   lowest key value
-00123                  .
-99999                  .
000000                  .
000001                  .
000123                  .
999999  highest key value

 

By saving date fields in the way YYMMDD or YYYYMMDD the higher date has the higher key value.

 

For each Index File a primary key has to be specified.

 

The primary key must be unique for every record; it identifies the respective record.

 

Additionally, up to 32 secondary keys can be defined. Secondary keys do not have to be unique, which means that several records can have the same secondary key.

 

Every key can consist of up to 10 fields of the Index File. The size of a key may not exceed 100 bytes.

 

With F2 a selection window is opened in the mask fields Field Name.

 

Button 'New Secondary Key': A new secondary key is created.

 

Button 'Delete Key': The current secondary key is deleted. The primary key cannot be deleted.

 

Distinguish between capital and small letters: If this checkbox is set, small and capital letters are differentiated in alphanumeric fields.

 

Example: The name "John" is saved in a record. If you try to read the record with a respective key and the respective field is defined as "JOHN", the record cannot be found, since "John" and "JOHN" are read as different strings.

 

 

Tab Description

 

 

 

Here you can add a short and a detailed description of the Index File.

 

The short description is displayed with the file selection.

 

The detailed description gives information only, it has no influence on the run of the program.