Unicode Test Plan

This document outlines a plan for testing the Unicode enhancements in APPX 6.0.

Overview

Release 6.0.0 of the APPX Utility includes support for Unicode data types.

Major Features:

DD Changes:

Maximum Alpha Field Size increased to 4GB

Maximum Record Size increased to 4GB

Encoding Type Attribute added for Alpha fields

FMS Changes:

APPXIO Support for Unicode fields

Oracle Support for Unicode fields

ILF Support:

Literals (Tested/Ok)

SET (Tested/Ok)

SET TEMP **** (Tested/Ok)

IF / AND / OR**** (Tested/Bugs)
(KAD - Should be fixed)

APPEND (Tested/Ok)

STORE / RESTORE (Tested/Ok)

PUSH / POP (Tested/Ok)

ERROR / WARNING / MESSAGE / CANCEL / REFRESH**** (Tested/Bugs)
(KAD - should be fixed)

BEG AT/ END AT (Tested/Ok)

RUN**** (Tested/Bugs)
(KAD - should be fixed)

CALL (No Changes Required)

PASS**** (Tested/Bugs)
(KAD - should be fixed)

RECEIVE / LOCAL*** (Tested/Bugs)
(KAD - should be fixed)

CNV BIN**** (Tested/Ok)

CNV PORT**** (Tested/Ok)

PACK / UNPACK**** (Tested/Bugs)
(KAD - should be fixed)

CNV TEXT**** (Tested/Bugs)
(KAD - should be fixed)

SET DAY / SET MNTH**** (Tested/Bugs)
(KAD - should be fixed)

DEFINE / UNDEFINE**** (Tested/Bugs)

VERIFY**** (Tested/Bugs)
(KAD - should be fixed)

IF DIFF (Tested/Ok)

COMPUTE / TOTAL**** (Tested/Bugs)
(KAD - should be fixed)

SCAN**** (Tested/Bugs)
(KAD - 0-app bug)

PRINT**** (Tested/Bugs)

BROWSE (No Changes Required)

...

Misc Bugs

1. New debugger does not work, crashes with error when trying to read --- WIDGET. * FIXED * April 24 engine
2. Old debugger does not display entire contents of national/unicode fields, even though the data is there. * Not a bug * The old debugger is not intended for use with Unicode.
3. Process Generation toolbox. I used the toolbox to generate a scrolling input on the file TST BEGUNI, which contains 2 unicode fields, a 10 character key and a 30 character description. The heading size over the description field is incorrect. It should have 30 ='s underneath it, but it only has enough for the word 'Description'. * FIXED* March 28 engine
4. Invalid (blank) occurrence number in Image Editor after hitting F11 then attempting to edit Item. ** FIXED ** March 21 engine
5. File Management functions (create, restruct, etc) crash with a 'Value Does Not Fit in Mask' error when setting MESSAGE2 MSG TXT into STAT2 MSG. * FIXED * April 24 engine
6. Engine should be transcoding when fields are moved, and throwing a runtime error (like numeric overflow) if the encoding of the receiving field does not contain the necessary characters. ** No, will not be implemented this way **
7. The 'Global EPs for Images' is not highlighted when I add code to the global start of frame. None of the EP's in the Global section turn on highlighting when code is entered. * FIXED * tested with April 24 engine, seems to be ok now.
8. Cross Reference display is messed up. Do an x-ref on APPEND NORMAL in App TST, then click 'References' to see what I mean. Retested with April 24 engine, better but still different from 4.2.a. In 4.2.a, in the 'Context' column the word RECORD is immediately followed by the frame & image information. In the Unicode version, the word RECORD is followed by many spaces, then the rest of the context.
9. Put the field --- X on an input image, run the process & try to enter data. If you enter '12345', you will get 'Invalid Numeric Value' error. Add a decimal point to the end and it's accepted. * FIXED * tested with May 23 engine.
10. The ILF editor does not accept valid characters in the 8859-15 character set, ie, the £ symbol exists in 8859-15, but I cannot enter it on an ILF command.
11. Changing the value of a --- TEMP field in the new debugger during execution seems to corrupt the field.
12. Appx doesn't complain when a Unicode character is entered into a Raw field. The character is silently discarded, but either the engine or the client should give the user some kind of error, probably the same error when entering invalid characters into a formatted field.
13. If I have a National or Unicode field on an image with no GUI attributes, I can enter Unicode/National data into it. if I call the record up in Inquire mode, the Unicode/National data does not show properly (reverse image diamond with embedded question mark). However, if I call it up in change mode, the data shows properly. Why the difference? If I make the field RAW TEXT, then it shows properly in all modes.

Language Translation Support:

Translation Dictionary

Query Support:

Sort OrderSort Order based on Unicode Collating Sequence

Predefine Fields and Files:

TEMP xxxxx

COMPARE File

STD Report Parameters File

Restructure Support:

Restructure APPXIO Files

Oracle Tables

Printing Support: **** (Not Fully Tested/Bugs)

Text Print Files

PDF Print Files

APPX Client Support **** (Not Fully Tested/Bugs):

UTF-8 data stream

Field validation based on Encoding Type

Callable Functions:

Transcoding fields

-- SteveFrizzell - 2011-03-02