Tips/Traps - Tweaks for CSV Output

DBohner-(db)

Legendary Poster
Don't we all love CSV Output? I've been doing a few Flat-File Exports (CSV), lately - and thought I post a few comments (tips), so I can find them here (instead of trying to remember). Here Goes:

CSV Tips / Traps:
1. Make sure your Unicode / Flat File settings are set correctly. Starting in one of the 8.94 tools releases, Oracle flipped a switch that requires us to use CP1252 Encoding. P93081->Create a new entry, defaulting everything to CP1252 -> Then remember to ACTIVATE it.
a. If you do not turn on the Unicode Switch – you will get a one column entry, when you open in CSV.
b. The cause is, actually, that the file data is Unicode (6 6 6 . 0 0 instead of 666.00 – if you look at it through DOS Edit)
2.In Layout, Set your Horizontal (Columns) Grid Alignment to 52 (yes, that is a magical number – and it usually works)
3.In Layout, set your Vertical (Rows) Grid Alignment to 11(Not a Magical Number – just something Experience has taught me). Setting to 11 places a small gap between the rows – allowing a bit more visual definition between rows. BUT – it does not garuntee that the row will be aligned (<grrrrr>)
4. If you have a row that does not want to align, one or more objects keep dropping to the next row; Select the ENTIRE ROW->Hold the CTRL Key and select the object that keeps dropping to the next line. This will place a darkened box around that specific Object ->then Bottom Align the entire row.
a. The Forces within the bOrgacle have not figured out that, for some reason; Text, Text Variable, Numeric Variable, Date Variable and Bolded objects are of different heights (WHY???). Doing this ‘Bottom Align’ with the ‘tallest’ object – seems to fix this issue

I know there are a ton more tricks – Share yours, please!

ps, Terry D....
[ QUOTE ]
Just posting, because I don't really know (and, I'm trying get to my 2000th JDEList Post):

[/ QUOTE ]

I finally made it!

(db)
 
[ QUOTE ]
ps, Terry D....
Just posting, because I don't really know (and, I'm trying get to my 2000th JDEList Post): I finally made it!


[/ QUOTE ]

These are awesome tips Daniel and certainly the very best way to hit that milestone. I plan on sharing these CSV Tips/Traps with my fellow OYO Geospace IT staff.

You know...I would have been perfectly content with hearing more about chasing springer through a field of pheasants or swapping photography tips between two avid photography enthusiasts.
grin.gif
 
Hi Daniel,

1.)
In many region the text file created with .txt extension and the list separator also can be different. The following jde.ini settings had been published here more times:

[UBE] section
prtCSVExtension=.csv
prtCSVSeparator=,

2.)
Not JDE specific, but useful:
The Comma Separated Value (CSV) File Format

These were my 2 cents.

Regards,

Zoltán
 
Great tips Daniel!

I have a few cells insist on dropping to the nex row, so I tried your #4 tip and started to see good result. However, when fixing the last cell on the far right, the cells on the far left were messed up.

I think it's because I was not able to select the entire row as you suggested. My report width is too wide and you can't see (and select) the entire row. Is there a trick to select the entire row while scrolling?

Thanks!
 
Wooi,

You don't have to select the entire row - if you have an 'extended' row.

For instance, select the first half of the row and align. Then, select the second half of the row and press the CTL key while selecting the last column of the first half. That last column on the first half will now be the Anchor for the alignment of the second half.

The column / Filed that has the BOLD block around it is the Anchor Column for alignment.

Hope this helps!

(db)
 
Oracle acually has a decent article on formatting a report to csv format. It saved me a lot of time. Log on oracle support and search for article ID 626452.1
 
Let us keep in mind that BI Publisher is now another option to export to Excel without changing the original UBE.
 
Something that I started doing, on any report that might be used for CSV and PDF - Use the DL01 Data Dictionary Item for the Column Headers.

Group Section: For each column header, use DL01 instead of the boring typical Text. Inside the ER, you set the value of each column header. If you set the length of the DL01 to 10, it will wrap in a PDF and will be full-width, one-cell in a CSV. Sometimes you may have to play with the width, to get it to wrap just right, but the Version Translation between PDF and CSV is much easier.

I'll post more about it on the JDE Research site, next week.

(db)

ps - it also gives you the ability to have dynamic headers =D
 
Back
Top