Annoying feature with "autocorrect" or "autoinput" (?)

[wdolek]

uh, i'm getting mad - when i write to cell text "x1" (for naming row as data of variable x_1 !), ooo converts my "x1" to date format as "1.10.2001"... i have turned off AutoInput (Tools > Cell Contents > AutoInput) and in AutoCorrect there is no such setting to stop ooo converting my text to date X-(

did i miss something? is there any setting to stop this?

[huw]

Format the cell as text.

[wdolek]

Format the cell as text.

yep, i know this - but when i'm writing from scratch some values from chalkboard... uh first i write "x1", get date, then angry delete it, click for cell, change format, then write "x1" again... i know about this option, but i would prefer that when i write "x1" it will stay "x1" by default setting

[huw]

You can't turn off the auto-formatting for cells formatted as number. You could set up a template with all cells formatted as text, but I suspect that isn't much use, sorry.

Entering x1 is treated as text for me (British English locale) - what OS & version of OpenOffice are you using, and what are its language settings? This information may help others to help you.

Edit: typo & clarification

[wdolek]

run on Windows XP SP2 CZ, OpenOffice 2.3.1 with installed Czech and English user interface - i'm using English user interface with Czech locale... so.. it is matter of locale? :'(

[huw]

it is matter of locale?

That's my guess, but I don't know much about locale issues. Others here do, hopefully one of them can help.

[kingfisher]

Formats recognised as date formats depend on your locale. I don't know why 'x1' would be interpreted as a date but I'm not familiar with the Czech ways of doing things. I use an English locale and 'x1' does not convert to a date. It is not a number, however; it enters as text, as I'd expect.

[Villeroy]

Huh? Czech cells treat i, x and v as roman numerals for months, but only one of them.
Input Czech(ISO) blue means numeric date:
i8 2008-01-01
ii8 ii8
iii8 iii8
iv8 iv8
v8 2008-05-01
vi8 vi8
vii8 vii8
viii8 viii8
ix8 ix8
x8 2008-10-01
xi8 xi8
xii8 xii8
23i8 2008-01-23
23v8 2008-05-23
23x8 2008-10-23
23i 2008-01-23
23v 2008-05-23
23x 2008-10-23

Even if roman numerals are intended as shorthand for months, I would say it includes an implementation bug because it works with 3 single numerals I, V and X only.

You can enforce text values with a single quote prefix. '23 puts text "23" into a cell. You can set number format "Text" for any group of cells to prevent all number recognition. Formatting will not change existing values.

[wdolek]

well, im czech for 24 years and i have not heared about we use roman numbers.. lol.. maybe some laws and paragraphs but this?!.. but thanks for explanation - i'll try that "haxor" with using "" when inserting text.. but still it is not fine for me.. is it possible to edit locale rules?

[kingfisher]

You can change the OO locale with the menu Tools >Options >Language Settings >Languages : Locale setting.

[huw]

i'll try that "haxor" with using "" when inserting text.. but still it is not fine for me

You need to type a single quote as the first character of your text, like this 'x1 instead of just x1

[Villeroy]

The faulty Czech localization is made by the Czech project group http://cs.openoffice.org/. There you find some more promising links.

well, im czech for 24 years and i have not heared about we use roman numbers.. lol.. maybe some laws and paragraphs but this?

In Germany I've seen this quite frequently in hand-written letters

Berlin, 23. X '99 <signature>

[huw]

well, im czech for 24 years :D and i have not heared about we use roman numbers.. lol.. maybe some laws and paragraphs but this?

In Germany I've seen this quite frequently in hand-written letters

Berlin, 23. X '99 <signature>

I've seen it in the UK occasionally - I think it might sometimes be an attempt to avoid the confusion caused by US date format mm/dd/yyyy frequently being the default MS Office install setting, even in the UK (not to mention Australia). It might also be some peoples' "personal flourish" in dull administration rôles.

[Villeroy]

The Czech locale uses roman numerals as abbreviations of months ("MMM" in format string).

Code: Select all

D. MMM YYYY(Czech)	D. MMM YYYY(English)
23. II 1999	23. Feb. 1999
23. III 1999	23. Mar. 1999
23. IV 1999	23. Apr. 1999
23. V 1999	23. May. 1999
23. VI 1999	23. Jun. 1999
23. VII 1999	23. Jul. 1999
23. VIII 1999	23. Aug. 1999
23. IX 1999	23. Sep. 1999
23. X 1999	23. Oct. 1999
23. XI 1999	23. Nov. 1999
23. XII 1999	23. Dec. 1999
23. I 2000	23. Jan. 2000

Numeric cell input is a never ending story. Calc recognizes too many input strings as number by means of dates.

[acknak]

There seem to be lots of open issues with Czech localization, but this has not been reported as far as I can see. Someone should definitely report it: http://qa.openoffice.org

[Villeroy]

http://www.openoffice.org/issues/show_bug.cgi?id=87259

[oliv']

I've got the same kind of problem (french setting this time).

When I import a file (csv) with numerical values inside some of the values are transformed into date. Some numbers will be transformed into date some other not, for example : 1.4200 is magicaly (dark magic I presume) transformed into 01/01/4200
But the most annoying thing ( and I stay polite) is that this &#!*'feature wrecks the value into the cell: If at least I could get back the original value, but nothing can resurrect it, I ve got 840060 as an useless result. I changed the cell format to standard number, I changed all the language settings I could find to english, no good...

I'm getting used to add ' before writing any date with letters ( May will always be transformed into 01/05/08)... just writting this post I begin to shake with frustration... but this number to date mutation is a bit too much.

I beg the authors of OOO to take things in hand and get ride of this horrible feature !

[Villeroy]

Please open the csv in a text editor, copy an arbitrary line or the offending part of such a line, post it here. Tell us what you expect to get (text, number or date).
Finally tell us your locale setting in Tools>Options...LangSettings>Languages (the second setting).
Calc is able to import any flavour of valid csv perfectly.

[oliv']

Locale setting is "Default", probably french as I have a french version of Ubuntu.

Here is the line :
2;0.71;0;1.4200

Here is the result :
2 0.71 0 01/01/00
(in the formula bar 01/01/00 becomes 01/01/4200)

The expected result is :
2 0.71 0 1.42

The solution I've got is to replace dots by comas in the csv file. The result is then :
2 0,71 0 1,42

By the way I've got the same problem doing a cut and paste. If I copy a number with a dot it transforms into a date. With a coma it's OK.
I can live with Calc not recognising a number as a proper number (by marking it as text for example), but I would like it not to modify the value of the cell.

And what I'm puzzled about is why does Calc has to autoformat things this way. Or to be more precise why don't whe have the possibility to disable this feature : I want my "May" to stay a "May" and a 1.4200 staying a 1.4200.

[Villeroy]

In the text import dialogue, right-click the fields and set them to English(US). This assumes dot as decimal separator and dates like 12/31/99. For dates there are also "D/M/Y" and "Y/M/D" where "/" could be anything else between the digits.
Another option: Set the application locale to something English, do the import and reset it to French.
Another option that applies when you import unformatted text from the clipboard: Select the columns before import and set the number format locale of the cells to the foreign locale that fits the text data. Then "Standard" will fit the columns to be imported.