PCG Seed Data Invalid characters displayed in FRC

Hi, I'm not certain this is the correct forum or not.
Customer has PCG 7.3.2 on 11i, using FRC.
Many menus and concurrent program names & messages (all seed data) that use accents, are displaying garbled characters.
eg: GRC Controls : RÃ¨gles associÃ©es aux Ã©crans
when viewed through TOAD, Data looks the same, yet non PCG seed data with same words have the correct french accented characters.
NLS character set is WE8ISO8859P1, not UTF8. (which is standard in 11i FRC installs).
As per the PCG install guide, I have tried using both files (for UTF8 & WE8ISO8859P1) in $ORACLE_HOME/forms60/admin/resource/FRC, but the results do not change.
I think this may have had something to do with environment setting used during the installation process.
I would like to experiment with this, but cannot find the scripts/files that actually seed the FRC data for seeded menus and concurrent programs.
Can anybody guide to to these ? or provide an alternative or known (I can't find one) solutiuon ?

Hi Srini,
When customer is viewing output from the SRS window, then it contains invalid characters because it is in UTF-8 character set. Customer is on Oracle OnDemand so they cannot take the output generated on the server.Every time they have to raise a request to Oracle for the output file. So the concern here is, why don't the output from SRS window show output with valid characters ?
The reason could be conversion of ISO format to UTF-8. How could this be resolved ? Does SRS window output cannot generate in ISO format ?
A quick reply will be appreciated as customer is chasing for an update.
Thanks,
Saket
Edited by: 868054 on Aug 7, 2012 11:08 PM

Similar Messages

Invalid characters displayed in java-based tool

I'm having problems with a java applet that works fine on the vast majority of PCs, with 1 exception. The one in question runs Windows 7 32 bit professional, IE8. Java versions 27,29, and 31 have been tried, with the same results.
There are several fields, each of which has a title and a default value. On a working PC, the field names are:
Mortgage amount:
Term:
Interest rate:
Monthly payment (PI):
Annual property taxes:
Annual home insurance:
Monthly payment (PITI):
On a PC that its not working:
Kn` mHmendl `shmm9
Lnqsf` fd `l nt ms9
Hmsdqdrsq` sd9
Lnmaglx o` xl dms` OH(9
@mt ` koqnodqsx s wdr 9
@mt ` kgnl d hmtq` mbd9
L nmaglx o`xl dms ' OHSH(9#371-1/
There are more fields, but that's an example. it may not be exactly correct, since the spacing and character overlap makes it hard to differentiate some characters.
All the browser troubleshooting has been done (reset, cleared temps, deleted active X objects etc.)
Cleared all local cache, java cache and ran ccleaner
Checked Windows region, language, currency etc settings
Tried to reinstall java, tried 3 different versions (27, 29, 31)
Tried the local Admin profile and problem persists across multiple profiles.
If anyone has any ideas about which settings could possibly cause this character translation, I would be very appreciative. There appers to be a 1:1 correlation of characters, for example:
# = $
/ = 0
: = 9
R = S
Could it be some kind of java / unicode setting?

If I could post a picture it might help explain...
http://postimage.org/image/qtl15mdoz/
Uploaded, hopefully you can view it.
We've reset the browser, even tried IE9, with the same result. There's no toolbars installed and machine tested clean for adware. With no java installed, the webpage displays sans mortgage calculator. What seems strange is that some of the text displays properly, as you can see in the picture, so that's why I wonder if its just one particular font that's not being displayed properly? The entire rest of the web page, and everything else that uses java works fine, it seems like it's just this one particular calculator.
It's at the point where the OS may just need to be reinstalled, though it seems like a small, petty thing to reinstall the OS for... but without a solution i don't really have any other options.

Discoverer Report showing invalid characters in the report data

Hi,
Can you please help me. I have designed a discoverer report in discoverer desktop 4i, when it is running the data showing some invalid characters in the data like below.
kara¿ah L¿MAK HOLD¿NG A.¿
I have executed the same query in sql plus at that time the data shows correctly. Please let me know, is there any settings required.
Best Regards,
Umamaheshwar.

Hi Umamaheshwar
This could be a registry setting. Let me give you some background. In the Windows registy, every Oracle Home has a setting called NLS_LANG. This is the variable that controls, among other things, the numeric characters and the language used. The variable is made up of 3 parts. These are:
language_territory.characterset
Notice how there is an underscore character between the first two variables and a period between the last two. This is very important and must not be changed.
So, for example, most American settings look like this: AMERICAN_AMERICA.WE8MSWIN1252
The second variable, the territory, controls the default date, monetary, and numeric formats and must correspond to the name of a country. So if I wanted to use the Greek settings for numeric formatting, editing the NLS_LANG for Discoverer Desktop to this setting will do the trick:
AMERICAN_GREECE.WE8MSWIN1252
Can you please check your settings? Here's a workflow:
a) Open up your registry by running Regedit
b) Navigate to HKEY_LOCAL_MACHINE\SOFTWARE\ORACLE
c) Look for the Oracle Home corresponding to where Discoverer Desktop is installed. It's probably called KEY_ORANT
d) Clicking on the Oracle Home will display all of the variables
e) Take a look at the variable called NLS_LANG - if it is correct Exit the registry
f) If its not correct please right-click on it and from the pop-up select Modify
f) Change the variable to the right setting
g) Click the OK button to save your change
h) Exit the registry
Best wishes
Michael

Date / / contains invalid characters IN MODULE POOL

Hi,
I have a display field in infotype creation which will take always take current date.I have tried using the screen field as sy-datum and also tried using z-field with data type char or dats.but when ia am inserting values in a editable field on the screen and pressing enter ,it is giving the error: "Date / / contains invalid characters" .And afetr pressing cancel ,it goes back to the PBO and then pressing f8,it shows the infotype screen designed* but the other functionalities which have been coded in PAI of screen like activating checkboxes,radiobuttons etc. are not there .Even i have tried passing sy-datum to the date field in PBO of the screen.Pls help.

Hello Sir,
Were you able to sort out this issue.
Would appreciate if you can share the solution. One of our users have this issue in the ADOBE form.
Let me know please.
Thanks,
Akshay...

Loading master data : error invalid characters

Hello,
I load master data (example 0PROG_POS) into BW with InfoPackage and loading stops with error "invalid characters". In the logs, the error is due to a special character (in french Ä : a special A).
On BW side, is it possible to solve this problem (I don't understand the message in the help : "If this message appears during a data load, maintain the attribute in the PSA maintenance screens").
My BW is 3.5 patch 7
Thanks in advance for your help.
Daniel.

HI Daniel
all characterstic & special characterstic
goto TCODE: RSKC -> all_capital-> F8
again : rskc-> !"%&'()*+,-./:;<=>?_0123456789ABCDEFGHIJKLMNOPQRSTUVWXY ->F8
GOTO SE11: RSALLOWEDCHAR (maintain special characterstic)
also u can goto-> SE37->RSKC_ALLOWED_CHAR_GET
-> F8
THANKS & BEST REGARDS
SANGRAM KESHARI SUTAR

Invalid Characters during data load

Hi All,
We are in BI7 and its unicode compliant with code page 4102.
My setting in RSKC is ALL_CAPITAL. When i tried loading data from the source system, i got many error records in BW saying that it has invalid characters. On debugging, i found that hexadecimals 00-1F comes as part of the record. Why control characters like 00-1F are not allowed in BW by default? I changed the RSKC setting to ALL_CAPITAL_PLUS_HEX, but its of no use.
when i checked in i18n transaction, i could see that code 4102 can accept all control characters.
Is there any setting that has to be made in order to allow hexadecimals in BW [ without converting it in to space using a routine ]?
Thanks and Regards
Amith

Hi Amith Anand,
Please
check
Program
RSKC_ALLOWED_CHAR_MAINTAIN
Or
RSKCCHECK
Let me know.
Regards
Hari

XML data and "invalid characters"

Hello everyone,
I'm developing a custom XML report for TestStand and I'm a bit confused about how to handle certain characters. For example, lets say I have a test step titled "Check that current is < 50mA". Now, when this data is written into the XML report it can be found between the Step Name tags, like this
<Step Name>Check that current is <50mA</Step Name>
Now the character causing the problem is "<" as XML presumes it indicates a new tag. Looking through XML resources it seems one way to escape these characters is to use the CDATA tags like this:
<Step Name><![CDATA[Check that current is <50mA]]></Step Name>
or to specify in the .xsl file "disable-output-escaping="yes" in the for these affected fields, like step name. However I'm concerned that by using the "disable-output-escaping" method I will encounter problems when feeding the XML file into a parser.
Now I'm a bit curious about how different parsers act with the CDATA tag, if I parse the step name fields will I get "Check that current is <50mA" or will I get ><![CDATA[Check that.....]]>. Any ideas?
Currently I'm using a function from the modelsupport.dll to replace these invalid characters with entity references, for example "<" is replaced with < but this is not a good solution from the parser point of view.
Best regards and thanks in advance for any advice,
Aleksi

Hi,
>>>
Now I'm a bit curious about how different parsers act with the CDATA tag, if I parse the step name fields will I get "Check that current is <50mA" or will I get ><![CDATA[Check that.....]]>. Any ideas?
>>>
Is it easy just to try it and see what happens?
Regards
Ray Farmer
Regards
Ray Farmer

PHP MYSQL and DW Question: Invalid characters and new lines entry/viewing

Hello all,
I have followed David Powers' two php/dw/mysql books and I've
set up PHP
pages that insert, edit, delete, etc. listings. Everything
works fine
except for invalid characters being inserted into the
database, and then
displaying with a question mark in firefox.
Is there a trick to cleaning up the data that a user may
paste into a
textbox on a web page for insertion in the database before
the data is
inserted or as it is inserted? Em dashes, en dashes, 'curly
quotes' etc.
are giving me problems.
Also, when I enter new lines into the php page while I'm
typing out
information for a record, these new lines are not being
displayed either
in the database (using phpmyadmin to view) or on the
resulting details
pages, even though I'm using nl2br, etc. Is there a way to
make this
work reliably?
Thank you for any help you may give me!
-John

Hello all,
I have followed David Powers' two php/dw/mysql books and I've
set up PHP
pages that insert, edit, delete, etc. listings. Everything
works fine
except for invalid characters being inserted into the
database, and then
displaying with a question mark in firefox.
Is there a trick to cleaning up the data that a user may
paste into a
textbox on a web page for insertion in the database before
the data is
inserted or as it is inserted? Em dashes, en dashes, 'curly
quotes' etc.
are giving me problems.
Also, when I enter new lines into the php page while I'm
typing out
information for a record, these new lines are not being
displayed either
in the database (using phpmyadmin to view) or on the
resulting details
pages, even though I'm using nl2br, etc. Is there a way to
make this
work reliably?
Thank you for any help you may give me!
-John

Invalid Characters shown in UTF-8 character set

There is an XMLP report whose template output character set is ISO-8859-1. The character set ISO-8859-1 is required for this report as per Spanish Authorities. When the report is run, output gets generated in the output directory file of application server. This output file doesn't contain any invalid characters.
But when the output is opened from SRS window, which opens it in a browser, the invalid characters are shown for characters like Ñ , É etc.
Investigation done:
Found that the output generated on the server is having ISO encoding and hence doesn't contain any invalid characters. Whereas the output generated from SRS window, it is in UTF encoding, so it seems the invalid characters are displayed when conversion takes place from ISO to UTF-8 format.
Created the eText output using the data xml and template using BI publisher tool, the output is in ISO encoding. So if i go and change the encoding to UTF-8 by opening it in explorer or Notepad++, invalid charcters are shown for Ñ, É etc.
Is there any limitation, that output from SRS window will show only in UTF-8 encoding? If not then please suggest.
Thanks,
Saket
Edited by: 868054 on Aug 2, 2012 3:05 AM
Edited by: 868054 on Aug 2, 2012 3:05 AM

Hi Srini,
When customer is viewing output from the SRS window, then it contains invalid characters because it is in UTF-8 character set. Customer is on Oracle OnDemand so they cannot take the output generated on the server.Every time they have to raise a request to Oracle for the output file. So the concern here is, why don't the output from SRS window show output with valid characters ?
The reason could be conversion of ISO format to UTF-8. How could this be resolved ? Does SRS window output cannot generate in ISO format ?
A quick reply will be appreciated as customer is chasing for an update.
Thanks,
Saket
Edited by: 868054 on Aug 7, 2012 11:08 PM

Error message if invalid characters entered in PM order oper.short text fld

Hi ALL,
I need to display a error message, when user enters the invalid characters in work order operation short text field. The error message should be displayed during work order u201CSaveu201D (during creation and change mode).
Kindly let me know is there is any user exit / BADI can be used for this.
Note: I have checked some user exit, such as IWO10009 but I am not getting the value of Operation short text (AFVGD-LTXA1) to compare and throw the error message
Regards
Ramu

hi
the exit is correct but don't use structure field,
follow this to get text
get AUFPL(Routing number of operations in the order)from CAUFV and paa it in AFVC -AUFPLyou get the short text
it can be usefull
-ashok

Value 'KIRBYu00A0' of characteristic 0SORTL contains invalid characters

hello experts!
i am facing this error while activating 0vendor. data not available either in info object master neither in psa.
Error Message: Value 'KIRBY ' (hex. '004B004900520042005900A0') of characteristic 0SORTL contains invalid characters
Message no. BRAIN060
Diagnosis
Only the following standard characters are valid in characteristic values by default:
!"%&''()*+,-./:;<=>?_0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ.
Furthermore, characteristic values that only consist of the character # or that begin with ! are not valid.
You are trying to load the invalid characteristic value 1. (hexidecimal representation 004B004900520042005900A0).
Procedure for System Administration
Try to change the invalid characters to valid characters.
If you want values that contain invalid characters to be admitted into the system, make the appropriate setting in BW Customizing. Refer to the documentation describing the requirements for special characters and the possible consequences.
For more information on the problems involved with valid and invalid characters, click here.
regards
vijay

Hi Vijay,
after KIRBY is a characteristic value which cannot be interpreted by BW properply / is not in the valid char list.
KRIBY is hex value 004B0049005200420059
the invalid char is hex value 00A0 -> online hex converter didn't found a ASCI match.
To solve the problem you will need to fix the record in the source. Or you go to PSA -> delete all in teh field and re-enter KRIBY.
Best regards,
Axel

The File name property is not valid. The file name is a device or contains invalid characters

I have an SSIS task that has run successfully for years. They just moved the dtsx to a new server and now it is failing.
Issue:
The task has a data flow that writes a raw file destination. It then has a subsequent data flow that reads the raw file destination.
The path and name of the raw file is passed to each of the data flows via the same variable.
The raw file resides in a folder on the server in the same path that the dtsx resides in.
The task fails with the "The File name property is not valid. The file name is a device or contains invalid characters" error when trying to access the raw file.
What we have done to troubleshoot:
1. I ran the task successfully on my local machine so it is definitely a server issue.
2. The first thing we were seeing was that it failed trying to write to the raw file. The task is set to "Create Always".
3. We saw that the dba's copied over the entire directory structure from the old server which included a previously created raw file, so we deleted the raw file.
4. We reran the task and it successfully wrote a new raw file but then failed on reading the raw file with the same error as above.
5. We reran the task AGAIN and this time it failed trying to write the raw file throwing the same error as above.
6. The dba looked at the directory and the account that performs the task has full control.
7. The dba looked at the raw file that was created and verified that the account that performs the task has full control of the file and that it is the owner of the file.
Summary:
The task fails when trying to access the file EXCEPT when the file does not exist. In that situation it can write the file but subsequently fails in accessing it again.

Hi Whalensdad,
Based on my research, the issue can be caused by the following reasons:
There are some invalid characters in the File name at package runtime. In this scenario, just as Russ said, could you please post the value that is in the File name variable? Also use a Script Task with Messagebox.show to show the variable values at runtime.
May be it changes to others at run time.
The account runs the package not having access to all the folders in the path. Since you are moving the dtsx file from one server to the current server, do you also move the raw file to the same folder in the current server? Does the user runs the package
have access to all the folders in the path? To solve this, please ensure that the user has access to all folders and the raw file in the path on the current server.
Thanks,
Katherine Xiong
If you have any feedback on our support, please click
here.
Katherine Xiong
TechNet Community Support
Please see the response above. I can't use a message box on a server so I logged the values to the database. The resource account has full control to the folder. I haven't been able to get the DBA to check the SQL Agents permissions, but I was always under
the impression that if you use "Run As" on the task, it will not use the SQL Agent but the account you identify it to "Run As". The "Run As" account has full control on the folder and file.

11.5.10.2 to R12.1.1 upgrade: Error loading seed data for GL_DEFAS_ACCESS_SETS: DEFINITION_ACCESS_SET = SUPER_USER_DEFAS, ORA-06508: PL/SQL: could not find program unit being called

Hello,
EBS version : 11.5.10.2
DB version : 11.2.0.3
OS version : AIX 6.1
As a part of 11.5.10.2 to R12.1.1 upgrade, while applying merged 12.1.1 upgrade driver(u6678700.drv), we got below error :
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
ATTENTION: All workers either have failed or are waiting:
           FAILED: file glsupdas.ldt on worker 3.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
drix10:/fmstop/r12apps/apps/apps_st/appl/admin/FMSTEST/log>tail -20 adwork003.log
Restarting job that failed and was fixed.
Time when worker restarted job: Wed Aug 07 2013 10:36:14
Loading data using FNDLOAD function.
FNDLOAD APPS/***** 0 Y UPLOAD @SQLGL:patch/115/import/glnlsdas.lct @SQLGL:patch/115/import/US/glsupdas.ldt -
Connecting to APPS......Connected successfully.
Calling FNDLOAD function.
Returned from FNDLOAD function.
Log file: /fmstop/r12apps/apps/apps_st/appl/admin/FMSTEST/log/US_glsupdas_ldt.log
Error calling FNDLOAD function.
Time when worker failed: Wed Aug 07 2013 10:36:14
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
drix10:/fmstop/r12apps/apps/apps_st/appl/admin/FMSTEST/log>tail -20 US_glsupdas_ldt.log
Current system time is Wed Aug 7 10:36:14 2013
Uploading from the data file /fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/import/US/glsupdas.ldt
Altering database NLS_LANGUAGE environment to AMERICAN
Dumping from LCT/LDT files (/fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/import/glnlsdas.lct(120.0), /fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/import/US/glsupdas.ldt) to staging tables
Dumping LCT file /fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/import/glnlsdas.lct(120.0) into FND_SEED_STAGE_CONFIG
Dumping LDT file /fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/import/US/glsupdas.ldt into FND_SEED_STAGE_ENTITY
Dumped the batch (GL_DEFAS_ACCESS_SETS SUPER_USER_DEFAS , GL_DEFAS_ACCESS_SETS SUPER_USER_DEFAS ) into FND_SEED_STAGE_ENTITY
Uploading from staging tables
Error loading seed data for GL_DEFAS_ACCESS_SETS: DEFINITION_ACCESS_SET = SUPER_USER_DEFAS, ORA-06508: PL/SQL: could not find program unit being called
Concurrent request completed
Current system time is Wed Aug 7 10:36:14 2013
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Below is info about file versions and INVALID packages related to GL.
PACKAGE BODY GL_DEFAS_ACCESS_SETS_PKG is invalid with error component 'GL_DEFAS_DBNAME_S' must be declared.
I can see GL_DEFAS_DBNAME_S is a VALID sequence accessible by apps user with or without specifying GL as owner.
SQL> select text from dba_source where name in ('GL_DEFAS_ACCESS_DETAILS_PKG','GL_DEFAS_ACCESS_SETS_PKG') and line=2;
TEXT
/* $Header: glistdds.pls 120.4 2005/05/05 01:23:16 kvora ship $ */
/* $Header: glistddb.pls 120.16 2006/04/10 21:28:48 cma ship $ */
/* $Header: glistdas.pls 120.4 2005/05/05 01:23:02 kvora ship $ */
/* $Header: glistdab.pls 120.5 2006/03/13 19:56:21 cma ship $ */
SQL> select * from all_objects where object_name in ('GL_DEFAS_ACCESS_DETAILS_PKG','GL_DEFAS_ACCESS_SETS_PKG')
2 ; OWNER OBJECT_NAME SUBOBJECT_NAM OBJECT_ID DATA_OBJECT_ID OBJECT_TYPE CREATED LAST_DDL_TIME TIMESTAMP STATUS T G S NAMESPACE
EDITION_NAME
APPS GL_DEFAS_ACCESS_DETAILS_PKG 1118545 PACKAGE 05-AUG-13 05-AUG-13 2013-08-05:18:54:51 VALID N N N 1
APPS GL_DEFAS_ACCESS_SETS_PKG 1118548 PACKAGE 05-AUG-13 06-AUG-13 2013-08-05:18:54:51 VALID N N N 1
APPS GL_DEFAS_ACCESS_SETS_PKG 1128507 PACKAGE BODY 05-AUG-13 06-AUG-13 2013-08-06:12:56:50 INVALID N N N 2
APPS GL_DEFAS_ACCESS_DETAILS_PKG 1128508 PACKAGE BODY 05-AUG-13 05-AUG-13 2013-08-05:19:43:51 VALID N N N 2
SQL> select * from all_objects where object_name='GL_DEFAS_DBNAME_S'; OWNER OBJECT_NAME SUBOBJECT_NAM OBJECT_ID DATA_OBJECT_ID OBJECT_TYPE CREATED LAST_DDL_TIME TIMESTAMP STATUS T G S NAMESPACE
EDITION_NAME
GL GL_DEFAS_DBNAME_S 1087285 SEQUENCE 05-AUG-13 05-AUG-13 2013-08-05:17:34:43 VALIDN N N 1
APPS GL_DEFAS_DBNAME_S 1087299 SYNONYM 05-AUG-13 05-AUG-13 2013-08-05:17:34:43 VALIDN N N 1
SQL> conn apps/apps
Connected.
SQL> SELECT OWNER, OBJECT_NAME, OBJECT_TYPE, STATUS
FROM DBA_OBJECTS
WHERE OBJECT_NAME = 'GL_DEFAS_ACCESS_SETS_PKG'; 2 3 OWNER OBJECT_NAME OBJECT_TYPE STATUS
APPS GL_DEFAS_ACCESS_SETS_PKG PACKAGE VALID
APPS GL_DEFAS_ACCESS_SETS_PKG PACKAGE BODY INVALID SQL> ALTER PACKAGE GL_DEFAS_ACCESS_SETS_PKG COMPILE; Warning: Package altered with compilation errors. SQL> show error
No errors.
SQL> ALTER PACKAGE GL_DEFAS_ACCESS_SETS_PKG COMPILE BODY; Warning: Package Body altered with compilation errors. SQL> show error
Errors for PACKAGE BODY GL_DEFAS_ACCESS_SETS_PKG: LINE/COL ERROR
39/17 PLS-00302: component 'GL_DEFAS_DBNAME_S' must be declared
drix10:/fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/odf>cat $GL_TOP/patch/115/sql/glistdab.pls|grep -n GL_DEFAS_DBNAME_S
68: SELECT GL.GL_DEFAS_DBNAME_S.NEXTVAL
81: fnd_message.set_token('SEQUENCE', 'GL_DEFAS_DBNAME_S');
SQL> show user
USER is "APPS"
SQL> SELECT GL.GL_DEFAS_DBNAME_S.NEXTVAL
FROM dual; 2                         -- with GL.
NEXTVAL
1002
SQL> SELECT GL_DEFAS_DBNAME_S.NEXTVAL from dual;               --without GL. or using synonym.
NEXTVAL
1003
drix10:/fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/odf>strings -a $GL_TOP/patch/115/sql/glistdab.pls|grep '$Header'
REM | $Header: glistdab.pls 120.5 2006/03/13 19:56:21 cma ship $ |
/* $Header: glistdab.pls 120.5 2006/03/13 19:56:21 cma ship $ */
drix10:/fmstop/r12apps/apps/apps_st/appl/gl/12.0.0/patch/115/odf>strings -a $GL_TOP/patch/115/sql/glistdas.pls |grep '$Header'
REM | $Header: glistdas.pls 120.4 2005/05/05 01:23:02 kvora ship $ |
/* $Header: glistdas.pls 120.4 2005/05/05 01:23:02 kvora ship $ */

Renaming files with invalid characters in their names on NTFS partitions, introduced by operating systems other than Windows

Essentially, Linux created some files with colons (:) in the name on a NTFS partition where I have Windows installed. I have since uninstalled Linux, but now I can only view these files in Windows Explorer. I can't open them, I can't even rename them to
correct the problem. It's as if they don't exist, because of the invalid search paths.
If I try to rename them in Windows Explorer I get following message.
The file name you specified is no valid or too long.
Specify a different file name.
Well isnt' that something?... isn't that nice? Windows is able to display these files, but it doesn't allow me to open them and it certainly doesn't like me to rename them. So why is it whining about it then, when I'm trying to help? It says "try a different
file name". Yeah, right! Like I haven't tried that one already! It doesn't matter what file name I input it will never accept it.
So what am I supposed to do now? Ditch Windows and go back to Linux? Surely, Microsoft doesn't like the sound of that. Sure, I could reinstall Linux or run a Linux live system to correct the problem. But what good is Windows then? I might as well switch to
Linux altogether.
After doing some research I now know by fact that it's (kind of) possible to rename files from UNIX and UNIX-like operating systems to those compliant with Windows by using something called file name character translation. To some level this is essential and
necessary for Windows interoperability with other operating systems (Windows is not the only operating system in the world). But this seems to be very complicated and I can't get my head around it. My brain is in overload. I don't know where to start.
Once there was a...
There's the Windows Services for UNIX (SFU) 1.0, 2.0, 3.0, 3.5. The first two versions were based on MKS Toolkit, a package licensed by Microsoft from MKS Inc. The later versions were based on the similar Interix product, after Microsoft purchased the company
that made it.
Then there's the new Subsystem for UNIX-based Applications (SUA). These are services for UNIX components. They are supposed to have Client for NFS v3 included as well. But the server components from the SFU line is missing (e.g. Server for NFS). These are included
in Server editions of Windows.
Then there's the Microsoft Knowledge Base article
289627: "How to Enable File Name Character Translation". This article seems to describe exactly my situation.
Windows and UNIX operating systems have restrictions on valid characters that can be used in a file name. The list of illegal characters for each operating system, however, is different. For example, a UNIX file name can use a colon (:), but a Windows
file name cannot use a colon (:). If a UNIX user attempts to create a file with a Windows illegal character on a Windows Services for UNIX network file system (NFS) share, the attempt is unsuccessful and the UNIX client computer receives an input or output
error.
It goes further than that. At first glance, this KB article also seems to offer a solution to this exact problem, with examples as shown below.
For example, the following maps the UNIX colon (:) to a Windows dash (-):
0x3a : 0x2d ; replace client : with - on server
I checked these values in charmap.exe and they are correct. Except for 2D not being a "dash", it's rather a hyphen ("hyphen minus" to be exact), but these two have pretty much the same appearance and they get interchanged a lot, I'm sure
they are used to it by now. (Yes, the characters! They don't mind.)
Then there's this registry key.
HKEY_LOCAL_MACHINE\Software\Microsoft\Server For NFS\CurrentVersion\Mapping
Well, of course, I don't have Server for NFS. So this is a dead end. Well, actually, it was a dead end from the beginning...
1. First of all, I'm not working with a network share on a NAS or SAN storage. The files are on the local disk drive where Windows is installed, so that's a DAS for you.
2. I don't have SFU! Well obviously, I'm on Windows Vista! So that means SUA!
3. SUA are service components only. No server components. Can you guess what that means? Yeah... no "Server for NFS" since it's a server component.
4. Windows Vista is a client side operating system! Server for NFS is only offered for use with Windows Server systems.
5. Back to square one!
So there you have it. They all lived happy for the rest of their lives...
I'm stuck here. Can someone tell me what to do? I mean beyond the obvious option to use Linux to fixa a Windows problem? The NTFS file system itself supports colons in file names. It's Windows that doesn't, and so by default it proclaims it invalid character.
Surely, even a Windows client operating system like Windows Vista should be able to allow the user to at least rename files with invalid characters to something more sensible (from the system point of view) and valid, if not being able to open them as they
are. Just add some crazy voodoo code to it and it will work. If you can make it possible on Windows Server with UNIX user-mode subsystem on NT kernel, then what's stopping you from giving the Windows client system the same benefit?
So what now? Purchase a Windows Server 2012 R2 license, copy my invalid files to a NAS share with NFS on a UNIX or Linux system, and have a go at the Windows registry and Server for NFS? Yeah... you're right, it's probably a bit over the top...
On a second thought... I might as well install Linux again. There are countless situations where Linux has helped me solve problems related to, and more often than not caused by Windows.

Essentially, Linux created some files with colons (:) in the name on a NTFS partition where I have Windows installed. I have since uninstalled Linux, but now I can only view these files in Windows Explorer. I can't open them, I can't even rename them to
correct the problem. It's as if they don't exist, because of the invalid search paths.
If I try to rename them in Windows Explorer I get following message.
The file name you specified is no valid or too long.
Specify a different file name.
Well isnt' that something?... isn't that nice? Windows is able to display these files, but it doesn't allow me to open them and it certainly doesn't like me to rename them. So why is it whining about it then, when I'm trying to help? It says "try a different
file name". Yeah, right! Like I haven't tried that one already! It doesn't matter what file name I input it will never accept it.
So what am I supposed to do now? Ditch Windows and go back to Linux? Surely, Microsoft doesn't like the sound of that. Sure, I could reinstall Linux or run a Linux live system to correct the problem. But what good is Windows then? I might as well switch to
Linux altogether.
After doing some research I now know by fact that it's (kind of) possible to rename files from UNIX and UNIX-like operating systems to those compliant with Windows by using something called file name character translation. To some level this is essential and
necessary for Windows interoperability with other operating systems (Windows is not the only operating system in the world). But this seems to be very complicated and I can't get my head around it. My brain is in overload. I don't know where to start.
Once there was a...
There's the Windows Services for UNIX (SFU) 1.0, 2.0, 3.0, 3.5. The first two versions were based on MKS Toolkit, a package licensed by Microsoft from MKS Inc. The later versions were based on the similar Interix product, after Microsoft purchased the company
that made it.
Then there's the new Subsystem for UNIX-based Applications (SUA). These are services for UNIX components. They are supposed to have Client for NFS v3 included as well. But the server components from the SFU line is missing (e.g. Server for NFS). These are included
in Server editions of Windows.
Then there's the Microsoft Knowledge Base article
289627: "How to Enable File Name Character Translation". This article seems to describe exactly my situation.
Windows and UNIX operating systems have restrictions on valid characters that can be used in a file name. The list of illegal characters for each operating system, however, is different. For example, a UNIX file name can use a colon (:), but a Windows
file name cannot use a colon (:). If a UNIX user attempts to create a file with a Windows illegal character on a Windows Services for UNIX network file system (NFS) share, the attempt is unsuccessful and the UNIX client computer receives an input or output
error.
It goes further than that. At first glance, this KB article also seems to offer a solution to this exact problem, with examples as shown below.
For example, the following maps the UNIX colon (:) to a Windows dash (-):
0x3a : 0x2d ; replace client : with - on server
I checked these values in charmap.exe and they are correct. Except for 2D not being a "dash", it's rather a hyphen ("hyphen minus" to be exact), but these two have pretty much the same appearance and they get interchanged a lot, I'm sure
they are used to it by now. (Yes, the characters! They don't mind.)
Then there's this registry key.
HKEY_LOCAL_MACHINE\Software\Microsoft\Server For NFS\CurrentVersion\Mapping
Well, of course, I don't have Server for NFS. So this is a dead end. Well, actually, it was a dead end from the beginning...
1. First of all, I'm not working with a network share on a NAS or SAN storage. The files are on the local disk drive where Windows is installed, so that's a DAS for you.
2. I don't have SFU! Well obviously, I'm on Windows Vista! So that means SUA!
3. SUA are service components only. No server components. Can you guess what that means? Yeah... no "Server for NFS" since it's a server component.
4. Windows Vista is a client side operating system! Server for NFS is only offered for use with Windows Server systems.
5. Back to square one!
So there you have it. They all lived happy for the rest of their lives...
I'm stuck here. Can someone tell me what to do? I mean beyond the obvious option to use Linux to fixa a Windows problem? The NTFS file system itself supports colons in file names. It's Windows that doesn't, and so by default it proclaims it invalid character.
Surely, even a Windows client operating system like Windows Vista should be able to allow the user to at least rename files with invalid characters to something more sensible (from the system point of view) and valid, if not being able to open them as they
are. Just add some crazy voodoo code to it and it will work. If you can make it possible on Windows Server with UNIX user-mode subsystem on NT kernel, then what's stopping you from giving the Windows client system the same benefit?
So what now? Purchase a Windows Server 2012 R2 license, copy my invalid files to a NAS share with NFS on a UNIX or Linux system, and have a go at the Windows registry and Server for NFS? Yeah... you're right, it's probably a bit over the top...
On a second thought... I might as well install Linux again. There are countless situations where Linux has helped me solve problems related to, and more often than not caused by Windows.

Invalid characters found in the generated out put file

hi ,
I had written a report to generate a file using open dataset , And In the generated file there are invalid characters generated for few fields. can any body help me with this.
regards
nelakonda

Hi,
Try below code.
DATA : l_w_file LIKE rlgrap-filename .
FIELD-SYMBOLS : <l_w_record> TYPE string,
                  <l_table>    TYPE STANDARD TABLE,
                  <l_line>     TYPE ANY,
                  <l_field>    TYPE ANY.
DATA         : new_table TYPE REF TO data   ,
                 new_line TYPE REF TO data   .
TYPE-POOLS : slis.
DATA      : wa_fieldcat             TYPE lvc_s_fcat         ,
              wa_fieldcat1            TYPE slis_fieldcat_alv ,
              fieldcat                TYPE slis_t_fieldcat_alv,
              gt_fieldcat             TYPE lvc_t_fcat         .
DATA : l_w_length TYPE i.
DATA : wa TYPE string.
CLEAR : p_subrc .
l_w_file = p_file .
OPEN DATASET l_w_file FOR INPUT IN BINARY MODE.
IF sy-subrc NE 0 .
    p_subrc = 1 .
    EXIT.
ENDIF .
ASSIGN wa TO <l_w_record> CASTING.
DO.
    READ DATASET l_w_file INTO <l_w_record>.
    CONDENSE <l_w_record>.
    IF sy-subrc EQ 0.
    ELSE.
      EXIT.
    ENDIF.
    APPEND <l_w_record> TO p_it_itab.
*This method will create a structure required for the
*runtime table generation
    l_w_length = STRLEN( <l_w_record> ).
      CLEAR wa_fieldcat.
   wa_fieldcat-fieldname    = wa.
   wa_fieldcat-seltext      = wa.
      wa_fieldcat-outputlen    = l_w_length          .
      APPEND wa_fieldcat       TO gt_fieldcat         .
      CALL METHOD cl_alv_table_create=>create_dynamic_table
        EXPORTING
                    I_STYLE_TABLE             =
          it_fieldcatalog           = gt_fieldcat
                    I_LENGTH_IN_BYTE          =
        IMPORTING
          ep_table                  = new_table
        EXCEPTIONS
          generate_subpool_dir_full = 1
          OTHERS                    = 2
      IF sy-subrc <> 0.
        MESSAGE ID sy-msgid TYPE sy-msgty NUMBER sy-msgno
                   WITH sy-msgv1 sy-msgv2 sy-msgv3 sy-msgv4.
      ENDIF.
*The structure is assigned to the table <l_table>
      ASSIGN new_table->* TO <l_table>.
*Structure of Work Area of the table created.
      CREATE DATA new_line LIKE LINE OF <l_table>.
*The structure of work area is assigned to <l_line>
      ASSIGN new_line->* TO <l_line>.
      MOVE wa TO <l_line>.
      ASSIGN COMPONENT wa OF STRUCTURE <l_line> TO <l_field>.
      INSERT <l_line> INTO TABLE <l_table>.
      CLEAR : <l_w_record>.
ENDDO.
CLOSE DATASET l_w_file .
IF sy-subrc NE 0 .
    p_subrc = 1 .
    EXIT .
ENDIF .
Refere above code as per your requirement.

PCG Seed Data Invalid characters displayed in FRC

Similar Messages

Maybe you are looking for