Visual C++ ERRATA.WRI Part 1-2: Class Library Reference

• Microsoft Visual C++ for Windows, 16-bit edition, version 1.5

SUMMARY

This article is a reproduction of parts 1 and 2 of the ERRATA.WRI file distributed with Microsoft Visual C++, version 1.5. The Visual C++ setup installs ERRATA.WRI in the MSVC\HELP directory.

MORE INFORMATION

Part 1: Notes on Class Library Reference

• Page 32, File Manager Drag and Drop:
  
  Replace the last paragraph with the following:

  To enable file drag-and-drop in your application, AppWizard writes a call to the CWnd member function DragAcceptFiles (if you chose to assign a file extension to your document class) for your main frame window in your InitInstance. You can remove this call if you do not want to implement file drag-and-drop.




• Page 577, CObject::GetRuntimeClass:
  
  Replace the second paragraph with the following:

  const char* m_lpszClassName A null-terminated string containing the ASCII class name.




• Page 737, CScrollView::SetScaleToFitSize(SIZE szTotal):
  
  Replace the last sentence in the first paragraph with the following:

  Usually cx and cy are greater than zero. However, if you wish to invert either the x-axis or the y-axis (as is common when emulating standard mapping modes such as MM_LOENGLISH), specify a negative cx or cy.

Part 2: Notes on Database Classes

• Page 10:
  
  In Table 2.1, column 3 of the Enrollment row, place an asterisk after StudentID. The full primary key for the Enrollment table is CourseID + SectionID + StudentID.



• Page 14:
  
  In the second full paragraph, which begins "A record view is a specialized view class that...", replace the sentence

  CRecordView supports end-user navigation through records, one at a time, using Move First, Move Next, Move Previous, and Move Last commands.

  with this sentence:

  CRecordView supports end-user navigation through records, one at a time, using the Move First, Move Next, Move Previous, and Move Last commands of the associated CRecordset object.




• Page 34:
  
  Change the last line on the page from

  
        pDoc->m_courseSet.m_strSort = "CourseID:; 

  to read as follows:

  
        pDoc->m_courseSet.m_strSort = "CourseID"; 

  (That is, change the colon to a double quotation mark.)



• Page 53:
  
  In the second paragraph of the "AppWizard" article, complete the sentence

  
        To run AppWizard, choose 

  to read as follows:

  To run AppWizard, choose the AppWizard command on the Project menu in Visual Workbench.




• Page 86:
  
  In the third full paragraph, replace "CRecordset::Open" with "CRecordset::dynaset".



• Page 114:
  
  The second line of text at the top of the page reads:

  
        you get a recordset that is dynamic but doesn't allow updates or
        backward scrolling 

  Delete the words "is dynamic but". (A forward-only recordset allows forward scrolling only, does not allow updates, and should not be considered "dynamic.")



• Page 122:
  
  Step 2 of the procedure "To specify a filter for a recordset object" contains two lines of code. In both code lines, add single quotation marks around the literal string "MATH101". For example:

  
        m_pSet->m_strFilter = "CourseID = 'MATH101'"; 

  For an explanation, see the "Caution" on page 31.



• Page 172:
  
  Under the section "SQL Statement Construction," after the following paragraph:

  To override the default SELECT statement, pass a string containing a complete SELECT statement when you call Open. Instead of constructing its own default string, the recordset uses the string you supply. If your replacement statement contains a WHERE clause, don't specify a filter in m_strFilter because you would then have two filter statements. If your replacement statement contains an ORDER BY clause, don't specify a sort in m_strSort so that you will not have two sort statements.

  add the following "Caution" note:

  *Caution*

  In the Enroll tutorial application, filter strings typically use a parameter placeholder, "?", rather than assigning a specific literal value, such as "MATH101", at compile time. If you do use literal strings in your filters (or other parts of the SQL statement), you may have to "quote" such strings with a DBMS-specific "literal prefix" and "literal suffix" character (or characters). For example, the code in this section uses a single quote character to bracket the value assigned as the filter, "MATH101". You may also encounter special syntactic requirements for operations such as outer joins, depending on your DBMS. Use ODBC functions to obtain this information from your driver for the DBMS. For example, call ::SQLGetTypeInfo for a particular data type, such as SQL_VARCHAR, to request the LITERAL_PREFIX and LITERAL_SUFFIX characters. If you are writing database-independent code, see Appendix C in the ODBC Programmer's Reference for detailed syntax information.



• Page 201:
  
  In the list of values that can be returned in the CDBException::m_nRetCode data member, add the following values (alphabetically):

  
        * AFX_SQL_ERROR_DATA_TRUNCATED  You requested more data than you have
        provided storage for. For information on increasing the provided data
        storage for CString or CByteArray data types, see the nMaxLength
        argument for RFX_Text and RFX_Binary under "Macros and Globals."
  
        * AFX_SQL_ERROR_LOCK_MODE_NOT_SUPPORTED  Your request to lock records
        for update could not be fulfilled because your ODBC driver does not
        support locking.
  
        * AFX_SQL_ERROR_NO_POSITIONED_UPDATES  Your request for a dynaset
        could not be fulfilled because your ODBC driver does not support
        positioned updates.
  
        * AFX_SQL_ERROR_ODBC_V2_REQUIRED  Your request for a dynaset could
        not be fulfilled because a Level 2-compliant ODBC driver is required. 




• Page 221:
  
  In the description of the CRecordset::GetStatus member function, change the description of m_lCurrentRecord to the following:

  Contains the zero-based index of the current record in the recordset, if known. If the index cannot be determined, this member contains AFX_CURRENT_RECORD_UNDEFINED (-2). If IsBOF is TRUE (empty recordset or attempt to scroll before first record) then m_lCurrentRecord is set to AFX_CURRENT_RECORD_BOF (-1). If on the first record, then it is set to 0, second record 1, and so on.




• Pages 224, 225, 234, and 235:
  
  Add the indicated information below to the Remarks for the CRecordset functions listed below:

  
        IsFieldDirty            p. 224
        IsFieldNull             p. 225
        IsFieldNullable         p. 225
        SetFieldDirty           p. 234
        SetFieldNull            p. 235 

  Add the following:
  
  

  Using NULL for the first argument of the function will apply the function only to outputColumns, not params. For example, the call

  
           SetFieldNull(NULL); 

  sets only outputColumns to NULL. Params will be unaffected.
  
  To work on params, you must supply the actual address of the individual param you want to work on (i.e. SetFieldNull(&m_strParam)). This means you cannot set all params NULL, as you can with outputColumns.




• Page 230:
  
  

  In the description of the forwardOnly option for the nOpenType parameter to the CRecordset::Open member function, delete the word "dynaset". (A forward-only recordset allows forward scrolling only, does not allow updates, and should not be considered a dynaset.)

Additional query words: 1.50

Keywords          : kb16bitonly 
Version           : 
Platform          : 
Issue type        : 

Last Reviewed: July 29, 1999