- Characteristics of created data file
- Additional files created
- Managing passwords
- Managing fatal errors when creating a data file
- Managing non-Latin character sets
- Replication: Range of identifiers
- Priority for defining the location of log files
- Creating an existing table
- Constraints for referential integrity
- Managing indexes
- Creating the data file from its description
- DATE and TIME items with the Native SQL Server Connector
- HCreation and Native SQLite Connector
In french: HCreation
Not available with this kind of connection
Creates an empty data file (.FIC extension) with the index file and the memo file if necessary.
This function can also be used to delete all records from a data file. In this case, the record in memory is not canceled. To initialize it, use HReset
: The automatic creation of non-existing data files can be enabled in the project description ("Files" tab of the project description, "Automatically create the data files when they are opened"). HCreation
is automatically used if the data file does not exist when it is handled for the first time.
If the data file is logged, you can specify the location of the log files.
If integrity constraints have been defined and if they are supported, the links are updated on the server.
// Création du fichier de données Commande (décrit dans l'analyse)
<Result> = HCreation(<Data file> [, <Password> [, <Path of JNL file> [, <Path of JournalOperation file> [, <Option>]]]])
- True if the data file was created,
- False if a problem occurred. HError is used to identify the error. If a fatal error occurs, a specific process is required to continue the program and to check whether the data file was created (see the Notes).
<Data file>: Character string
Name of data file to create.
MySQL: the table name must contain no accented character (for example: "ÉlanFile"). This is a problem specific to MySQL (tested on versions 3.23.52-max-nt and 3.23.47-max-nt of the MySQL server).
<Password>: Optional character string
- Password associated with the data file.
The password of HFSQL files is not case-sensitive.
- Empty string ("") if no password is used.
<Path of JNL file>: Optional character string
Access path to the JNL file associated with the logged data file. This parameter is only taken into account if the data file is logged. In this case, <File name>JNL.fic will be created at the specified location.
This parameter must not be specified. In HFSQL Client/Server, the path of JNL file is defined and it cannot be modified.
<Path of JournalOperation file>: Optional character string
Access path to JournalOperation file and to JournalIdentification file associated with the logged file. The JournalOperation file contains all the operations performed on the file with logs. This parameter is only taken into account if the data file is logged.
This parameter must not be specified. In HFSQL Client/Server, the path of operation file is defined and it cannot be modified.
<Option>: Optional Integer constant
Options for creating the data file:
|hChangeCharset||The character set defined by ChangeCharset will be taken into account in the new data file. The sorts, the searches, ... on string keys (strings, characters, date and time) will be performed according to this character set.|
|hNoLink||Management of integrity: The links are not updated on the server (creation, modification or deletion).|
Reminder: by default, the links are updated on the server when creating a data file.
This constant is available in HFSQL Client/Server mode only.
This parameter is ignored.
Characteristics of created data file
The characteristics of created data file are as follows:
- Data file name:
Physical name defined during the file description in the data model editor. This name can be modified by HChangeName before creating the data file.
- Directory of data file:
Directory defined during the file description in the data model editor. This directory can be modified by HChangeDir and HSubstDir before creating the data file.
- Content of data file:
The created data file is empty, which does not mean that its size is null (the file contains a header describing its structure).
- The order of items created in the physical file is the physical order of items created in the analysis (and not the logical order).
If the data file already existed: If the data file already existed (same name, same directory), this data file is recreated: the data file contains no record. The index file and the memo file (if they exist) are also recreated. In this case, the integrity constraints are not checked.
: Re-creating the data file is possible only if the data file is not used by another application. If a file currently used must be cleared, use HDelete
during a browse. You can also use HDeleteAll
to delete all records from a data file.
- The password can be defined:
- when the data file is created by HCreation.
- before the data file is created by HPass.
- If the data file existed and if it was password protected, the data file is re-created, even if the password changes or does not exist anymore.
: The password is independent of data encryption. The data encryption must be configured during the file description in the data model editor ("Details" tab).
The password is not associated with a data file but with a database. This password must be specified during the connection (HConnect
Managing fatal errors when creating a data file
If a fatal error occurs when creating a data file, it may be interesting to continue the program while checking whether the data file was created or not. To do so, we recommend that you perform the following exception process:
WHEN EXCEPTION IN
Error("Erreur HFSQL: " + HError())
Managing non-Latin character sets
If the constant hChangeCharset
is specified, the alphabet used in the data file will be the one specified by the function ChangeCharset
. The sorts, the searches, ... on string keys (strings, characters, date and time) will be performed according to this character set.
The character set used can be modified by HIndex
Remark: The size of string items specified in the analysis is expressed in bytes and not in characters. Depending on the character set used, you may have to increase the item size (if the characters used are coded on 2 bytes for example).
Replication: Range of identifiers
The automatic identifiers for the files of subscriber replicas are using a range of identifiers that differs from the one used by the file of master replica and by the other files of subscriber replicas.
This range of automatic identifiers is written in the header of each data file (".FIC" extension).
This range is initialized:
Priority for defining the location of log files
The location of log files is defined according to the following priority order:
- Location specified when creating the logged data file (HCreation).
- Location specified by HChangeLogDir.
- Location specified in the analysis.
- Location of the logged data file.
If another application uses the HDeclareExternal
function on the logged a data file, the changes made by this application are recorded in the log files specified in the logged of the data file HFSQL header.
Creating an existing table
If the table existed, the HCreation function attempts to delete and then recreate the table.
For example: Using an Oracle database
- with the OLE DB "MSDAORA" provider: a new table cannot be created. If the table already exists, the records are erased.
- with the OLE DB "oraOLEDB.Oracle" provider: the creation of a new table is possible as well as the re-creation of an existing table. However, no index or memo item can be created.
When creating a table, WINDEV and WEBDEV try to create the items in the OLE DB type corresponding to the HFSQL type. If this type is not supported, WINDEV and WEBDEV successively try all the "compatible" OLE DB types until they find a type supported by the OLE DB database.
Constraints for referential integrity
When creating a table, the referential integrity constraints described in the analysis are not created on the database.
If adding an index fails, no error message is returned by the HCreationfunction. The table is created without the index that is causing the problem. In this case, duplicates may be found in an item defined as unique key by WINDEV and WEBDEV if the external database does not support the unique keys.
Creating the data file from its description
The data file can be directly created from its description in the data model editor. To do so, select "Create the data file" and select the directory where the data file must be created.
DATE and TIME items with the Native SQL Server Connector
When using HCreation, the Native SQLServer Connector creates DATE and TIME items if the server version and the client layer version correspond to SQL Server 2008 (or later).
HCreation and Native SQLite Connector
The HCreation function may fail ("database table is locked" error) if the following conditions are met:
- the file to create already exists.
- a browse is in progress on another file of the same database.
This is caused by a limitation of SQLite database.
This page is also available for…