- Handling a deleted record
- Number of deleted records
- Deleting a record from a query
- Version of data file
- Size of the data file
- Deletion and lock
In french: HSupprime
Not available with this kind of connection
Deletes a record from a data file (query or view). The record is deleted logically and physically. It cannot be restored (unlike the records crossed by HCross
The following operations are performed:
- the record is deleted: the record is inactive and it is not referenced in the index anymore. The data is not kept.
- the indexes (corresponding to the record keys) are deleted from the index file.
- the memos associated with the record are deleted from the memo file.
After running HDelete
, the record is inserted in the list of deleted records of the data file and will be re-used by HAdd
The deleted record is no longer visible by the browse on the current item. This record will no longer be visible for the browses on the other items once these browse operations have been re-initialized.
WINDEV, WINDEV Mobile and WEBDEV propose an automatic management of errors for the following types of errors: Duplicate error, integrity error, password error, modification conflict error and status error during modification conflict, lock error. For more details, see Assisted management of HFSQL errors
: To delete all records from a data file, use HDeleteAll
// Suppression des commandes d'un client
HReadSeekFirst(Commande, NumCli, ValNumCli)
WHILE HFound() = True
// Suppression des commandes d'un client
FOR EACH Commande where NumCli = ValNumCli
<Result> = HDelete([<Data file> [, <Record number> [, <Options>]]])
- True if the record was deleted,
- False if a problem occurred (locked record, etc. (see the special cases)). HErrorInfo is used to identify the error.
If the record to delete is already deleted, HDelete returns True.
<Data file>: Optional character string (with or without quotes)
Name of HFSQL data file or view used. If this parameter corresponds to an empty string (""), HDelete handles the last data file used by the last HFSQL management function (functions that start with H).
<Record number>: Optional integer
Number of the record to delete. If this parameter is not specified (if it is equal to 0 or to the hCurrentRecNum constant), the current record will be deleted. However, the current record will remain the deleted record.
CAUTION: Do not confuse the record number with the automatic identifier associated with the record.
In most cases, these two numbers are not identical. The record number is returned by HRecNum
This parameter must correspond to the current record number (hCurrentRecNum
<Options>: Optional constant
Configures the integrity management performed on the deleted record.
|hIgnoreIntegrity||Ignores the integrity check for this operation, even if the automatic integrity check is enabled (HSetIntegrity)|
Handling a deleted record
- A deleted record will not be read:
- by the browse functions: HFirst, HLast, HNext, HPrevious, HReadFirst, HReadLast, HReadNext, HReadPrevious, HForward, HBackward.
- by the search functions: HSeekFirst, HReadSeekFirst, HSeekLast, HReadSeekLast.
- A deleted record can be read by HRead. Its content will be incomplete or inconsistent.
- The physical location of a deleted record can be immediately re-used by HAdd for example.
Number of deleted records
is used to find out the number of records deleted from the data file.
Remark: For backward compatibility with Hyper File 5.5, the H.NbDel variable can also be used to find out the number of deleted records.
Deleting a record from a query
Deleting a record from a query can be performed:
- in the query result
- in the data files handled by the query (hModifyFile constant used in HExecuteQuery or HExecuteSQLQuery). In this case, the hWithFilter constant is automatically selected.
Regardless of the query execution mode (with or without the hWithFilter constant), HDelete can only be used on single-file queries.
- On data files not in HFSQL format, there is no need to specify the hModifyFile constant in HExecuteQuery or HExecuteSQLQuery: the data files taking part in the query are automatically modified when modifying the query result.
- No record can be deleted from a query that is using the groups or aggregates.
- The management of integrity and duplicates is not performed during the operations on the multi-file queries: we recommend that you use the transactions to avoid any problem.
Version of data file
is successful, the data file version number is modified. This version number is returned by HVersion
Size of the data file
The deletion of one or more records has no incidence on the total file size on disk. The locations left empty in the data files by the deletions will be re-used during the next additions of records (HAdd
). This feature is used to get better performance, the reduction of file size being costly in time.
However, if after many deletions it is necessary to reduce the size of a data file, it can be reindexed (via HIndex
, WDMap or the HFSQL Control Center).
Deletion and lock
If a lock problem occurs when "deleting" a record (attempt to delete a record locked in write mode for example), the record is not deleted and HErrorLock
have the same effect: the record is crossed but it is not deleted. Therefore, the record can be re-enabled by HModify
. In this case, HState
returns the hStateCrossed
constant for a deleted record (instead of hStateDel
To physically delete the crossed records, you must use HIndex
is not available
This page is also available for…