EA2: Formatter issues

1) Formatter in the worksheet doesn't leave a blank line between sql statements.
If you have several statements in worksheet and format them, they are all formatted as one with no gap between them.
2) You can't just format the selection. If I select a statement and format, the whole worksheet is formatted.
3) It doesn't handle syntax errors.
I think the formatter should detect and report syntax errors (it is a parser after all).
In my worksheet, one of the statements had a syntax error (semi colon inside the parentheses of a sub-query instead of outside), and later statements in the worksheet are indented relative to the statement with the error. AFAICS, the subsequent statements are formatted correctly, but the left margin isn't reset after the error statement.

I agree with smitjb's statements. The formatter has other flaws as well like:
1) Columns and tables are simply indented based on the indentation setting rather than left aligned 1 space after the leading SELECT
2) [INNER, LEFT, RIGHT, FULL [OUTER]] JOIN statements are not right aligned with the select key word the same was as the FROM and WHERE key words
3) GROUP BY and ORDER BY are not aligned with the SELECT key word on the space between GROUP/ORDER and BY, alternatively, the SELECT, FROM and WHERE key words are not indented suffciently to right align with the BY key word.
Setting the formatter indent to 8 cause columns and table to line up 1 space to the right of the SELECT statement, but then nesting of PL/SQL statements shifts to the right very quickly. Perhaps these could be separate indents.
Selecting the option to have a line break follow select from and where statements is not preserved so is not honored.

Similar Messages

EA1 - SQL Formatter issues (JOINs and GROUPs and ORDER BY oh my ;)

Great job with improving the SQL Formatter, but it still has some bugs that need to be worked out.
The key words JOIN and it's modifiers INNER, LEFT, RIGHT and FULL OUTER are not recognized as master key words. As such they end up flush against the left margin Also when GROUP BY and/or ORDER BY key words are present in an outer most select statement the other key words are not indented far enough to be right aligned with the end of the word BY and are indented too far to be right aligned with the word GROUP or ORDER. In sub queries, GROUP and ORDER BY are correctly right aligned with their respective SELECT statements.

We're picking up and collating the Formatter issues. I'll add these.
Specific bug for these #7013462
Sue

Formatter issue

I have a strange issue at hand. I am trying to apply a formatter to format date to one of the fields but the formatter function never gets invoked.
I tried with XMl and JS views but non of them seems to work is there any trick I am missing.
I am also using lazy loading because I thought eager loding of views might be issue but no use.
Here is code in contention
var oTemplate = new sap.m.ColumnListItem({
type:"Navigation",
press:oController.handleProjectSelected,
cells:[
        new sap.m.ObjectIdentifier({title:"{displayName}"}),
        new sap.m.Text({
        path:"creationDate",
        formatter:"util.Formatter.date"
        new sap.m.Label({
        text:"{creatorUserName}"
        new sap.m.Text({
        path:"lastDataUpdateDate",
        formatter:"util.Formatter.date"
//Same for XML view
<ColumnListItem
type="Navigation"
press="handleProjectSelected">
<cells>
<ObjectIdentifier title="{displayName}"/>
<Text text="{
path:'creationDate',
formatter:'util.Formatter.date'
}"/>
<Text text="{creatorUserName}"/>
<Text text="{
path:'lastDataUpdateDate',
formatter:'util.Formatter.date'
}"/>
</cells>
<customData>
<core:CustomData
key="target"
value="Project"/>
</customData>
</ColumnListItem>
Here is Formatter code it never gets invoked
jQuery.sap.declare("util.Formatter");
jQuery.sap.require("sap.ui.core.format.DateFormat");
util.Formatter = {
date : function (value) {
if (value) {
var oDateFormat = sap.ui.core.format.DateFormat.getDateTimeInstance({pattern: "yyyy-MM-dd"});
return oDateFormat.format(new Date(value));
} else {
return value;

Hi,
Since you can specify a formatter for basically any attribute, you should also specify for which attribute you want tho use a formatter.
For instance, the following code uses a formatter for both 'text' and 'valueState' properties:
var oTV = new sap.ui.commons.TextView({
    text: {
        parts: ["mydatefield"],
        formatter: function(oValue) {
            if (oValue != undefined) {
                var yyyy = oValue.getFullYear().toString();
                var mm = (oValue.getMonth() + 1).toString(); // getMonth() is zero-based
                var dd = oValue.getDate().toString();
                return yyyy + '/' + (mm[1] ? mm : "0" + mm[0]) + '/' + (dd[1] ? dd : "0" + dd[0]);
            } else return "";
    valueState: {
        parts: ["mydatefield"],
        formatter: function(oValue) {
            return (oValue === undefined || oValue == null) ? sap.ui.core.ValueState.Error : sap.ui.core.ValueState.None;
Hope this explains!
R.

EA1: A few formatter issues

1. Alignment of AND keywords in select statement is wrong.
function get_segment_size(ownname in varchar2,   segname in varchar2, segtype in varchar2) return number is indsize number(16);
begin
    begin
      select bytes
      into indsize
      from dba_segments
      where owner = ownname
       and segment_type = segtype
       and segment_name = segname;
    exception
    when others then
      raise;
    end;
    return indsize;
end;becomes
FUNCTION get_segment_size
    ownname IN VARCHAR2,
    segname IN VARCHAR2,
    segtype IN VARCHAR2 )
RETURN NUMBER
IS
indsize NUMBER ( 16 ) ;
BEGIN
BEGIN
     SELECT bytes
       INTO indsize
       FROM dba_segments
      WHERE owner = ownname
AND segment_type = segtype
AND segment_name = segname;
EXCEPTION
WHEN OTHERS THEN
    raise;
END;
RETURN indsize;
END;The ANDs are left aligned with the enclosing block rather than right aligned with the 'master keywords'.
2. NUMBER(?,?) and VARCHAR2(?) declarations have parentheses and sizes split over multiple lines.
index_list dba_indexes_tab;
rebuild_stmt varchar2(200);
end_time date;
start_size number(16,   0);
end_size number(16,   0);
batch app_support_batch_runs % rowtype;
batch_results resultset;becomes
index_list dba_indexes_tab;
rebuild_stmt VARCHAR2
    200
end_time DATE;
start_size NUMBER
    16, 0
end_size NUMBER
    16, 0
batch app_support_batch_runs % rowtype;
batch_results resultset;3. Some array subscripted expressions have their parentheses split across multiple lines.
    end_time := sysdate;
    batch_results := resultset();
    batch_results.extend;
    batch_results(batch_results.last).asjr_result_type := 'ELAPSED_TIME';
    batch_results(batch_results.last).asjr_result_key := 'ALL';
    batch_results(batch_results.last).asjr_result_value :=((end_time -batch.asbr_component_start_time) *(24 *60 *60));becomes
end_time      := sysdate;
batch_results := resultset
batch_results.extend;
batch_results
    batch_results.last
.asjr_result_type := 'ELAPSED_TIME';
batch_results
    batch_results.last
.asjr_result_key := 'ALL';
batch_results
    batch_results.last
.asjr_result_value :=
    ( end_time -batch.asbr_component_start_time ) * ( 24 *60 *60 )
;Another similar block later in the same package is left as single lines.
4. 3 seems to apply generally to expressions involving parentheses - array subscripts, procedure calls. Some are done properly, others are spread across multiple lines.
5. Various keywords are not recognised by the formatter, although they are by the syntax highlighter. (They are in bold, but not upper case)
raise
extend
last
immediate
6. There doesn't seem to be away to format the current selection rather than the whole buffer (in the PL/SQL editor)

Given this -
select prej.asbr_component_name,
prej.asbr_batch_start_time,
etime.asjr_result_value Elapsed,
prer.asjr_result_value Size_before,
postr.asjr_result_value Size_after,
prer.asjr_result_value -postr.asjr_result_value Shrinkage
from app_support_batch_runs prej
inner join app_support_job_result prer on prer.asjr_asbr_id = prej.asbr_id
inner join app_support_job_result postr on postr.asjr_asbr_id = prej.asbr_id
inner join app_support_job_result etime on etime.asjr_asbr_id = prej.asbr_id
where prer.asjr_result_type = 'START_SIZE'
and postr.asjr_result_type = 'END_SIZE'
and etime.asjr_result_type = 'ELAPSED_TIME'
order by prej.asbr_component_name;with [em]indent AND/OR[em] unchecked, you get this.
SELECT prej.asbr_component_name    ,
prej.asbr_batch_start_time        ,
etime.asjr_result_value Elapsed   ,
prer.asjr_result_value Size_before,
postr.asjr_result_value Size_after,
prer.asjr_result_value -postr.asjr_result_value Shrinkage
   FROM app_support_batch_runs prej
INNER JOIN app_support_job_result prer
     ON prer.asjr_asbr_id = prej.asbr_id
INNER JOIN app_support_job_result postr
     ON postr.asjr_asbr_id = prej.asbr_id
INNER JOIN app_support_job_result etime
     ON etime.asjr_asbr_id    = prej.asbr_id
WHERE prer.asjr_result_type = 'START_SIZE'
AND postr.asjr_result_type    = 'END_SIZE'
AND etime.asjr_result_type    = 'ELAPSED_TIME'
ORDER BY prej.asbr_component_name;The problems with this are.
1) The sql statement as a whole is indented (ie the select keyword)
2) Despite 1), INNER, AND and ORDER are NOT indented at all whereas the should be at least level with the select. I suspect the problem may be that it is trying to right align 'INNER JOIN' and 'ORDER BY' with the SELECT-FROM-WHERE.I thing the ON of the join clauses is right aligned.
3) The column list isn't right. The columns should be left aligned and positioned one space to the right of the select keyword ( or at a pinch on a new line, indented from the select keyword by 2 columns)
with [em]indent AND/OR[em] checked, you get this.
SELECT prej.asbr_component_name    ,
prej.asbr_batch_start_time        ,
etime.asjr_result_value Elapsed   ,
prer.asjr_result_value Size_before,
postr.asjr_result_value Size_after,
prer.asjr_result_value -postr.asjr_result_value Shrinkage
   FROM app_support_batch_runs prej
INNER JOIN app_support_job_result prer
     ON prer.asjr_asbr_id = prej.asbr_id
INNER JOIN app_support_job_result postr
     ON postr.asjr_asbr_id = prej.asbr_id
INNER JOIN app_support_job_result etime
     ON etime.asjr_asbr_id    = prej.asbr_id
WHERE prer.asjr_result_type = 'START_SIZE'
AND postr.asjr_result_type = 'END_SIZE'
AND etime.asjr_result_type = 'ELAPSED_TIME'
ORDER BY prej.asbr_component_name;This is better but still has the problems with right-alignment.
With [em]Right-align master keywords[em] unchecked...
SELECT prej.asbr_component_name     ,
prej.asbr_batch_start_time        ,
etime.asjr_result_value Elapsed   ,
prer.asjr_result_value Size_before,
postr.asjr_result_value Size_after,
prer.asjr_result_value -postr.asjr_result_value Shrinkage
FROM app_support_batch_runs prej
INNER JOIN app_support_job_result prer
ON prer.asjr_asbr_id = prej.asbr_id
INNER JOIN app_support_job_result postr
ON postr.asjr_asbr_id = prej.asbr_id
INNER JOIN app_support_job_result etime
ON etime.asjr_asbr_id        = prej.asbr_id
WHERE prer.asjr_result_type = 'START_SIZE'
AND postr.asjr_result_type = 'END_SIZE'
AND etime.asjr_result_type = 'ELAPSED_TIME'
ORDER BY prej.asbr_component_name;IMO This is best but I would like ON indented with respect to the INNER JOIN it belongs to.
So in terms of this sql statement, I think the only bug is the alignment of columns. The right alignment option doesn't work very well for longer 'keywords' (mainly the new-fangled sql-92 stuff). Perhaps a redefinition of what constitutes a master keyword would do better.

TOC formattting issue

I've been using framemaker for over 10 years, and I have a nagging problem in TOC's.
Give that you want standard toc's such as:
1.1 Introduction........................................1.3
Simply put, when you generate your TOC in a book, there may be times when the "paratext" is quite long and it causes the text to get to close to the page number or even wraps over to the next line.
For example:
1.1 Introduction to Framemaker 9.0 basics1.3
1.1 The small brown fox jumped over the b
      arrel1.1
How can we set up the heading1TOC parastyle to maintain a distance from the page number and if the paratext is to long, wrap to the next line and put the page number on the far right side.
my basic heading1toc style is set as follows:
Tab stops: 1.375 left
                 6.25 right (with leader ....)
on the reference pages, the heading1toc is
<$paranum> <$paratext>.................<$pagenum>
How can I set it to always "wrap" the text, say for example, when it reachs 6.00 inch so the above example would result in
1.1 Introduction to Framemaker 9.0 basics...1.3
1.1 The small brown fox jumped over the b
      arrel....................................................1.3
Anybody have a solution?

Archive guy wrote:
I've been using framemaker for over 10 years, and I have a nagging problem in TOC's.
Give that you want standard toc's such as:
1.1 Introduction........................................1.3
Simply put, when you generate your TOC in a book, there may be times when the "paratext" is quite long and it causes the text to get to close to the page number or even wraps over to the next line.
For example:
1.1 Introduction to Framemaker 9.0 basics1.3
1.1 The small brown fox jumped over the b
      arrel1.1
How can we set up the heading1TOC parastyle to maintain a distance from the page number and if the paratext is to long, wrap to the next line and put the page number on the far right side.
my basic heading1toc style is set as follows:
Tab stops: 1.375 left
                 6.25 right (with leader ....)
on the reference pages, the heading1toc is
<$paranum> <$paratext>.................<$pagenum>
How can I set it to always "wrap" the text, say for example, when it reachs 6.00 inch so the above example would result in
1.1 Introduction to Framemaker 9.0 basics...1.3
1.1 The small brown fox jumped over the b
      arrel....................................................1.3
Anybody have a solution?
This is an often-posted question. You can find past solutions with aGoogle search for "framemaker toc wrap long lines" and similar phrases, without quotes.
HTH
Regards,
Peter Gold
KnowHow ProServices

Decimal formatter issue

Hi All,
In a datagrid I use labelFunction method to format input numbers, this function will format decimal numbers with precision 2.
public static function formatDecimal(val:Number):String
            var decimalFormatter: NumberFormatter = new NumberFormatter();
            decimalFormatter.rounding = "up";
            decimalFormatter.precision = 2;
            decimalFormatter.useNegativeSign = false;
            return decimalFormatter.format(val);
Now, if the input is 2.55 only, it format the number to 2.56. It is a way to prevent this?
It should format 2.56 if the input is 2.555. correct?
Thanks
JFB

This is strange, it happens only with 8.55, it converts to 8.56 but my total calculations on the grid is reading 8.55 not 8.56.
Is not doing the same for 2.55, 3.55 or 4.55
Any ideas? Is this a bug?
Thanks!
JFB

Help needed with referencing single Excel cells and formatting resulting text

In InDesign CS5 I am putting together a 20pp catalogue of about 200 products. The plan is to have the product information, SKU code, quantity etc fixed, but have the prices (there are two i.e. pack price and individual price) being linked to an Excel spreadsheet. This is so that the prices can be updated in the Excel file and the InDesign file will pull the new prices through. In case you are wondering why I don't pull the whole set of information through, this is because there are a lot of copywriting changes done to the information once it's in InDesign - it's only going to be the prices that will be updated.
I am planning on having two single cell tables in their own text frame, duly formatted with cell style, table style and paragraph style for the two price variables. This I am then going to have to repeat 200 times making sure I link to the next row down in Excel. This is going to be a hideous task but I see know way of modifying the cell in InDesign to point it to the next row in Excel. That's my first problem.
My second problem is this. In the Excel sheet, the prices are formatted as UK currency and are therefore like this...
£2.00
£0.40
£1.43
£9.99
£0.99
£0.09
What I will require is once I import that data (and refresh the data via a newly saved Excel file) is that the prices end up like this...
£2.00
40p
£1.43
£9.99
99p
9p
So if the value is lower than £1.00 it needs a trailing 'p' added and the leading zero and '£' sign stripped off. If the value is lower than £0.10 it also needs the zero after the decimal point stripping off.
Then formatting wise, the '£' sign needs to be superscripted and the same for the 'p'. This I am assuming could be done via GREP?
In summary, can anyone help with the first task of referecing the Excel cells on a cell by cell basis, given that it is the same cell column each time, but the next row down, and also point me in the right direction of the price formattting issues.
Any help will be gratefully received.

I would do this:
Create on line with the formatting.
Export as InDesign tagged text (TXT)
Read out these tags
In Excel exists a function to connect text from several cells and predfined text, there connect the content from cells with the paragraph styling tags. Do it in a seperate sheet. (Better would be to use a database like Access, there you can link your Excel sheet).
Export this sheet as txt file
Place this sheet as tagged text (there is an option in one of the sialog boxes).
In preferences < file handling you can specify that Tablecalculation Sheets and text is not embedded but linked, turn it on.

Manual formatted for C: and D: drive Via SCCM 2012 task sequence

Dear All
i have a problem with the formatting of the HDD by using SCCM 2012 , our Co. have laptops with C: &
D: drive and we started to implement our W7 migration by using SCCM 2012 and we face the a problem which is how to keep the data of the D drive either during the deploy the new W 7 image our after we deployed the image with C and D drive .
we create a task sequence to format the whole HDD and then make two partition .
but our target to keep the old users data in D: drive in case of system crash .
so can we do the format issue manually via task sequence to can control the size of the HDD and just
format the C: drive
i read in internet there is a Disk partition script can be apply during the task sequence activity to control
the formate issue.
or to stop the formatte issue for D: drive via SCCM.

Dear EminM
thanks
for your replay
you are right i mean the laptop has C and D partition
we make a TS which was configure to crate two partition
After we complete our Task sequence we face problem that TS format the two partition and that mean we will loss the data in D partition in case we need to apply this TS in the future to any laptop has system crash .
our target to have TS just format the C partition and leave the D partition without any action to be able to save the data in the D partition without any corrupted.
for that do you have an idea to do that TS .

Printing issue in 4.0 EA2

When printing my data model in 4.0 EA2, most field names do not print on the document.

Have the same issue with version 4.0.0.825...
Method : Print => Page A3 Landscap => Scale 75% (Preview shows all attributes) => Print to printer gives missing attributes.
Using print to PDF printer generates the same issue (different printer driver).
I can send you examples (not able to insert pictures , browser crashe) if you wish.
Current workaround: print diagram => to image file => print image file with jpg viewer.....
Wim

EA2--Oracle SQL Developer 1.5 issue with open,function, procedurres.

Hi Friends,
After install SQL Developer 1.5, I can connection to database and see table object, view. However, I am not able to view codes of package, procedure and function and got an error message :
an error was encountered performing the requested operation: closed Connection Vendor code 17008 when I try to explorer folder of package, function and procedure.
Any suggestion for this issue?
Thanks,
JImmy

I am on version 1.2.1 and I am getting this error message more and more when compiling packages. I am tired of restarting SQL developer.

Issue with viewing packages in SQL Developer EA2

I posted a message when EA1 came out asking why I couldn't see the package's body in EA1.
Now I couldn't see the any thing under '$schema/packages' in EA2.
Does anybody know why?
My OS: Windows XP Pro SP2
Oracle: 9i
SQL Developer version: EA2

In version 9 a very simple sql is supposed to be executed:
SELECT OBJECT_NAME, OBJECT_ID,
DECODE(STATUS, 'INVALID', 'TRUE', 'FALSE') INVALID,
'TRUE' runnable
FROM SYS.ALL_OBJECTS a
WHERE OWNER = :SCHEMA
AND OBJECT_TYPE = 'PACKAGE'
AND SUBOBJECT_NAME IS NULL
-- for packages
SELECT OBJECT_NAME, OBJECT_ID,
DECODE(STATUS, 'INVALID', 'TRUE', 'FALSE') INVALID,
'TRUE' runnable
FROM SYS.ALL_OBJECTS o
WHERE OWNER =:SCHEMA
AND OBJECT_NAME = :PARENT_NAME
AND OBJECT_TYPE = 'PACKAGE BODY'
-- for package bodies
If either of these doesn't return anything in worksheet, then the problem is right there. However, this part hasn't been changed since EA1 (The only change was adding SYS.DBA_PLSQL_OBJECT_SETTINGS requirement for version >=10). I can't reproduce the problem: in 9i scott sees any procedure/package I create.

2.1 Final - Another Minor Formatter Indentation Issue

Looks like when you're declaring functions in a package, that it indents them when it shoudln't be, and causes a shift in following lines.
Example:
CREATE OR REPLACE
PACKAGE pkg
AS
PROCEDURE p1;
PROCEDURE p2;
FUNCTION f1
RETURN NUMBER;
FUNCTION f2
RETURN NUMBER;
PROCEDURE p3;
END pkg; Should be:
CREATE OR REPLACE
PACKAGE pkg
AS
PROCEDURE p1;
PROCEDURE p2;
FUNCTION f1
RETURN NUMBER;
FUNCTION f2
RETURN NUMBER;
PROCEDURE p3;
END pkg;

Hi Marwim,
Thanks for this feedback.
I have replicated the issue and logged a bug.
Bug 9231534 - FORUM:USER DEFINED REPORT FOLDER MISSING ACTIONS IN NON ENGLISH LOCALES
Regards,
Dermot
SQL Developer Team

2.1 EA1/EA2/RC1: Case change quirk from Formatter

Hi,
I love how the editors react to the automatic case change (as defined in the Formatter, not the button in the worksheet), but unfortunately I get its benefits less than half the time.
The problem is the formatter only kicks in after pressing enter/return, and only for the current line.
E.g. I go in an existing statement and want to add a new where condition. If I type on an existing line, it doesn't get formatted. If I insert a new line, the previous line gets formatted.
So this usually results in having 1 line different to the rest and/or not getting formatting where I wanted.
So my suggestions are to change the case whenever moving to another line (regardless if it was with enter/return, cursors, or mouse), and/or after typing words (on any whitespace), and/or change the case on the whole (current) statement.
Thanks for considering,
K.

Hi K,
Enhancement Request Logged:
9201516 - RC1:FORUM:CAPITALISE ON WHITESPACE AND MOVE AWAY FROM LINE
-Turloch

EA2 - Issues with Reconnection after DB restart

The support for lost connections in SQL Developer (ie if the DB is shutdown and then restarted) is improving, but there are still some problems.
EA2 JDK 1.6.0_04
Basic connection types
Don't save passwords on connections
In the SQL Worksheet, most things work well - the first attempt at running a query as a statement after the DB restart gives a "No more data to read from socket" error, but after that (or maybe another repeat of the error), queries work fine when run as statements. Note that this does not prompt for the password, even though I have not saved passwords on my connection.
However, it appears that you can no longer run as script. Also, the DBMS Output tab no works - even enabling/disabling dbms output doesn't produce the "set serveroutput on/off" lines in the output.
Worse for the navigator pane - it's connection never seems to be reconnected. Each attempt at accessing data through the navigator pane returns a "Connection Closed Vendor Code 17008" error.
Obviously the workaround is to disconnect and then reconnect (which could close certain windows based on preferences), but if we can detect the lost connection and reconnect for the SQL Worksheet connection, can we do the same for the navigator connection? Also, if the fix is separate from reconnecting the navigator connection, can we improve the handling of the reconnect in the worksheet so that running as scripts and DBMS Output work as well?
theFurryOne

Yes, lost connections have been one of the most annoying problems in sqldev: a cascade of exceptions (like when entering preferences) and having to close the connection manually (although it's obviously closed already), which makes you lose opened windows.
I understand not all situations can be accounted for (at least it's obvious it'll take a long time to enumerate them), which justifies the need for a manual reconnect option. As we tried asking this on various occasions (Exchange, forum and Kris' blog), but judging from the states on the requests this won't get addressed, I've added a specific new one: http://htmldb.oracle.com/pls/otn/f?p=42626:39:2998242762554303::NO::P39_ID:10421. To all who ever had problems with this, please vote this feature in!
Thanks,
K.

EA2 TNSNAMES and LDAP issues

great to see the long awaited LDAP option available. Some remarks after first try's:
1. there is no field to set the ldap context.
We have several contexts for different environment. All are using the same dbname, but different ports and target ip adresses. Now, the loaded list shows all databases several times. I can't see, which is the right entry. If i can set the context, this list should be unique for the instance name.
Hope, there is not the same problem as explained later in my point 3.
2. LDAP Admin and LDAP passwords are not needed to retrieve the database list, it's only misleading our security staff which is afraid to use the admin passwords for normal login
3. in case of normal tnsnames connection entries, the nasty feature(bug?) to show a instance entry several times if you have some ifiles into your tnsnames.ora. This was reported also for the past 1.2 release last year and should be fixed soon. It can't be a problem to unique these list entries.
Maybe these small issues can be fixed for the release version
Thomas

Thomas, We're taking a look.
Barry

EA2: Formatter issues

Similar Messages

Maybe you are looking for