PIBIO_STORAGE_QUERY_BY_CONTENT_FN回呼函式 (winbio_adapter.h)
由引擎配接器呼叫,以找出符合指定索引向量的範本。
語法
PIBIO_STORAGE_QUERY_BY_CONTENT_FN PibioStorageQueryByContentFn;
HRESULT PibioStorageQueryByContentFn(
[in, out] PWINBIO_PIPELINE Pipeline,
[in] WINBIO_BIOMETRIC_SUBTYPE SubFactor,
ULONG IndexVector[],
[in] SIZE_T IndexElementCount
)
{...}
參數
[in, out] Pipeline
與執行作業之生物特徵辨識單位相關聯的 WINBIO_PIPELINE 結構的指標。
[in] SubFactor
WINBIO_BIOMETRIC_SUBTYPE值,指定與範本相關聯的子因素。
IndexVector[]
[in] IndexElementCount
值,包含索引向量陣列中的專案數。 這必須符合建立資料庫時所指定的大小。 如果資料庫是以零長度索引建立的,此參數必須為零。
傳回值
如果函式成功,則會傳回S_OK。 如果函式失敗,它必須傳回下列其中一個 HRESULT 值,以指出錯誤。
傳回碼 | 描述 |
---|---|
|
SubFactor參數指定的引數無效。 |
|
強制指標引數為 Null。 |
|
無法為記錄標頭配置記憶體。 |
|
索引向量的大小與建立資料庫時所指定的索引大小不符。 |
|
查詢成功,但找不到相符的記錄。 |
|
資料庫已鎖定。 |
|
發生未指定的問題。 |
|
管線物件的 StorageCoNtext 成員為 Null 或 FileHandle 成員無效。 |
備註
如果這個方法傳回成功,即使查詢傳回空的集合,管線中的結果集也會由查詢的結果取代。
如果資料庫是以零長度索引向量建立,則結果集會包含範本子因數符合 SubFactor 參數的每個記錄。 在此情況下,如果呼叫端為 SubFactor 參數傳遞WINBIO_SUBTYPE_ANY,此函式會傳回資料庫中的所有記錄。
成功呼叫此函式之後,結果集資料指標應該位於集合中的第一筆記錄上。
重要
請勿嘗試驗證為 SubFactor 參數提供的值。 Windows 生物特徵辨識服務會先驗證提供的值,再將其傳遞至您的實作。 如果值 是WINBIO_SUBTYPE_NO_INFORMATION 或 WINBIO_SUBTYPE_ANY,請適當地驗證。
範例
下列虛擬程式碼顯示此函式的一個可能實作。 此範例不會編譯。 您必須調整它以符合您的用途。
/////////////////////////////////////////////////////////////////////////////////////////
//
// StorageAdapterQueryByContent
//
// Purpose:
// Locates templates that match a specified index vector.
//
// Parameters:
// Pipeline - Pointer to a WINBIO_PIPELINE structure associated with
// the biometric unit performing the operation.
// SubFactor - A WINBIO_BIOMETRIC_SUBTYPE value that specifies the sub-factor
// associated with the template.
// IndexVector - Pointer to an array of ULONG index values.
// IndexElementCount - A value that contains the number of elements in the index
// vector array.
//
static HRESULT
WINAPI
StorageAdapterQueryByContent(
__inout PWINBIO_PIPELINE Pipeline,
__in WINBIO_BIOMETRIC_SUBTYPE SubFactor,
__in ULONG IndexVector[],
__in SIZE_T IndexElementCount
)
{
HRESULT hr = S_OK;
BOOL lockAcquired = FALSE;
struct _MY_ADAPTER_FILE_HEADER fileHeader = {0};
SIZE_T remainingRecords = 0;
LARGE_INTEGER currentRecordOffset = {0};
struct _MY_ADAPTER_RECORD_HEADER *recordHeader = NULL;
SIZE_T recordHeaderSize = 0;
// Verify that the Pipeline parameter is not NULL.
if (!ARGUMENT_PRESENT(Pipeline) ||
!ARGUMENT_PRESENT(IndexVector))
{
hr = E_POINTER;
goto cleanup;
}
// Retrieve the context from the pipeline.
PWINBIO_STORAGE_CONTEXT storageContext = (PWINBIO_STORAGE_CONTEXT)Pipeline->StorageContext;
// Verify the pipeline state.
if (storageContext == NULL || storageContext->FileHandle == INVALID_HANDLE_VALUE)
{
hr = WINBIO_E_INVALID_DEVICE_STATE;
goto cleanup;
}
// WINBIO_SUBTYPE_ANY is a valid sub-factor.
// WINBIO_SUBTYPE_NO_INFORMATION is not a valid sub-factor.
if (SubFactor == WINBIO_SUBTYPE_NO_INFORMATION)
{
hr = E_INVALIDARG;
goto cleanup;
}
// Validate the IndexElementCount argument.
if (IndexElementCount != storageContext->IndexElementCount)
{
hr = WINBIO_E_DATABASE_BAD_INDEX_VECTOR;
goto cleanup;
}
if (storageContext->IndexElementCount > 0 &&
!ARGUMENT_PRESENT(IndexVector))
{
hr = E_POINTER;
goto cleanup;
}
// Clear the result set.
hr = StorageAdapterClearContext(Pipeline);
if (FAILED(hr))
{
goto cleanup;
}
// Lock the database for reading.
hr = _LockDatabase( Pipeline->StorageHandle, FALSE);
if (FAILED(hr))
{
goto cleanup;
}
lockAcquired = TRUE;
// Read the header block.
hr = _ReadFileHeader( Pipeline->StorageHandle, &fileHeader );
if (FAILED(hr))
{
goto cleanup;
}
// Scan through all records looking for index vector matches.
recordHeaderSize =
sizeof(struct _MY_ADAPTER_RECORD_HEADER) +
(SIZE_T)fileHeader.IndexElementCount * sizeof(ULONG);
currentRecordOffset = _MY_ADAPTER_FIRST_RECORD_OFFSET;
remainingRecords = fileHeader.TotalRecordCount;
while (remainingRecords > 0)
{
SIZE_T recordSize = 0;
BOOLEAN match = FALSE;
LARGE_INTEGER dataOffset = {0};
// If you did not give up the current header during the previous
// iteration of the loop, reuse it.
if (recordHeader == NULL)
{
recordHeader = (struct _MY_ADAPTER_RECORD_HEADER*)_AdapterAlloc( recordHeaderSize );
if (recordHeader == NULL)
{
hr = E_OUTOFMEMORY;
goto cleanup;
}
}
else
{
ZeroMemory(recordHeader, recordHeaderSize);
}
hr = _ReadRecordHeader(
Pipeline->StorageHandle,
currentRecordOffset,
recordHeader,
recordHeaderSize
);
if (FAILED(hr))
{
goto cleanup;
}
recordSize = recordHeader->RecordSize;
// Skip records marked for deletion.
if ((recordHeader->Flags & _MY_ADAPTER_FLAG_RECORD_DELETED) == 0)
{
// Call a custom function (_MatchIndexVector) that compares the index
// vector of the current record with the input index vector.
hr = _MatchIndexVector(
SubFactor,
IndexVector,
IndexElementCount,
recordHeader->SubFactor,
_GetIndexVector(recordHeader),
storageContext->IndexElementCount,
&match
);
if (FAILED(hr))
{
goto cleanup;
}
if (match == TRUE)
{
// Calculate the file offset of this record's data area.
dataOffset.QuadPart =
currentRecordOffset.QuadPart +
recordHeader->RecordHeaderSize;
// Add the matching record to the result set in the pipeline.
hr = _ResultSetAddElement(
&storageContext->ResultSet,
recordHeader,
dataOffset
);
if (FAILED(hr))
{
goto cleanup;
}
// The result set now owns the record header. Set the pointer
// to NULL.
recordHeader = NULL;
}
}
currentRecordOffset.QuadPart += recordSize;
--remainingRecords;
}
// If the search was successful, but the result set is empty, return
// WINBIO_E_DATABASE_NO_RESULTS
if (SUCCEEDED(hr))
{
SIZE_T elementCount = 0;
hr = _ResultSetGetCount(&storageContext->ResultSet, &elementCount);
}
cleanup:
if (recordHeader != NULL)
{
_AdapterRelease(recordHeader);
recordHeader = NULL;
}
if (lockAcquired == TRUE)
{
_UnlockDatabase( Pipeline->StorageHandle);
lockAcquired = FALSE;
}
if (FAILED(hr))
{
// Clear any partial result set from the pipeline.
StorageAdapterClearContext(Pipeline);
}
return hr;
}
規格需求
最低支援的用戶端 | Windows 7 [僅限傳統型應用程式] |
最低支援的伺服器 | Windows Server 2008 R2 [僅限桌面應用程式] |
目標平台 | Windows |
標頭 | winbio_adapter.h (包含 Winbio_adapter.h) |