|
Apache OpenOffice (AOO) Bugzilla – Full Text Issue Listing |
| Summary: | Exporting a base table into CSV text file: Unicode code replaced by ? sign | ||
|---|---|---|---|
| Product: | Base | Reporter: | nbrouard <nbrouard> |
| Component: | code | Assignee: | dbaneedsconfirm <needsconfirm> |
| Status: | CLOSED IRREPRODUCIBLE | QA Contact: | issues@dba <issues> |
| Severity: | Trivial | ||
| Priority: | P3 | CC: | drewjensen.inbox, issues, r4zoli, thierry.munoz |
| Version: | OOO300m7 | Keywords: | needmoreinfo |
| Target Milestone: | --- | ||
| Hardware: | Unknown | ||
| OS: | Windows XP | ||
| URL: | http://user.services.openoffice.org/en/forum/viewtopic.php?f=13&t=5009&st=0&sk=t&sd=a&start=10 | ||
| Issue Type: | DEFECT | Latest Confirmation in: | --- |
| Developer Difficulty: | --- | ||
| Attachments: | |||
clu->nbrouard: 1. may you add a small sample database file that shows the problem 2. have you tried that manually (without sql code)? thx p.s. how fast an issue is confirmed, mostly depends on the quality of the description (fast and easy is the best), the complexity of the issue, the reproducability and how many user it may concern (rather not the priorisation) .. this issue is not a prio 2 issue, so please do not send it as an p2 (wrong prio do not help the issue but bugs the confirming guys) .. please read: http://qa.openoffice.org/scdocs/ddIssues_EnterModify.html#priority thx Same for me results with OOo 3.0.0 and DEV300m37 under winXP. See demo database, attached. Created attachment 59143 [details]
database with sample table created in 3.0.0
Created attachment 59144 [details]
database with sample table created in 3.0.0
Created attachment 59145 [details]
database with sample table created in 3.0.0
Created attachment 59146 [details]
database with sample table created in 3.0.0
Created attachment 59147 [details]
database with sample table created in 3.0.0
Created attachment 59148 [details]
exported text from 3.0.0
Created attachment 59149 [details]
export from m37
I uploaded sample file, site not responded as usual give an empty page. I tried again and again. Sorry for this. If you used db as data source in Calc, F4, select table and click into left upper corner(select all data), drag and drop into sheet, save as csv, you can export with correct coding. If you select database table in Base and copy with CTRL+C, CTRL+V into sheet, you can have some trouble. see my attached Calc spreadsheet. Created attachment 59153 [details]
ods file with copied data from db table
Thank you for confirming the bug with your nice files examples. The work around which is also a simpler way was also mentioned in the quoted forum. Using OOdev-300m7 build 9371 I can confirm that opening a new sheet, typing F4 to register the database given as an example, clicking on the top-left grey empty square of the table of the database in order to select all data and draging into the empty sheet fills the sheet with correct hungarian accentuation. I haven't been able to reproduce the bug of CTRL_C from Base to CTRL-V in a new spread sheet: I get only the top left element (0) to be copied. But if I drag directly from Base and drop into the empty sheet it works: no need to use F4. And all characters are OK: U umlaut is still a U umlaut. The original first reported bug comes probably from the fact that on Windows as well as on Linux, while writing or reading a text file like a CSV, calc asks for the underlying code, UTF-8 or another old one. Anything against UTF-8 by default? CTRL+C and CTRL+V works as you decribed in open database table, but if you select whole table (in main table window)and copy into the ods file, will show my results. Őő, and Űű are oldest problems with Hungarian keyboard drivers and setting, it allways caused problems until UTF was introduced, but haunting now. Under winXP when saved from Calc to csv needs to select win1250 coding for correct characters. If I save as UTF-8 total mess. >CTRL+C and CTRL+V works as you decribed in open database table, but if you >select whole table (in main table window)and copy into the ods file, will show >my results. Ok I understood the différence (copying the whole table) but on my French XP there is no errors and I get: ÁRVÍZTŰRŐTÜKÖRFÚRÓGÉP >Őő, and Űű are oldest problems with Hungarian keyboard drivers and setting, it >allways caused problems until UTF was introduced, but haunting now. On Linux there is no error too. >Under winXP when saved from Calc to csv needs to select win1250 coding for >correct characters. >If I save as UTF-8 total mess. Did you use a simple notepad to read it? (or a recent emacs). Because I verified that the output in utf-8 was correct too. At least csv files in utf-8 are read correctly by oocalc on Windows (and Linux). Concerning Excel it can be read as is, I agree. ÁRVÍZTŰRŐTÜKÖRFÚRÓGÉP Hi, I confirm the problem with the "translated" file with the command : SELECT * INTO TEXT "output_csv_file_name" FROM "your_table_name" All the characters with an accent (acute/grave/circumflex) are replaced by "?". It's an encoding problem (I use the french language under Windows Vista). I found that it was possible to specify the type of encoding to read a file text in OOo Base with an instruction like that : SET TABLE "MaTable" SOURCE "monFichier.csv;encoding=UTF-8" Maybe, there is a solution by this way ? Regards You can also change the HSQLdb GLOBAL varialbe textdb.encoding To UTF-8, by default it is ASCII So I close this one. Steps to go are: SET TABLE "MaTable" SOURCE "MaTable.csv;encoding=UTF-8" SELECT * INTO TEXT "MaTable" FROM "your_table_name" and the chars should be o.k. Closing |
From the forum discussion link it appears that a way to export a table from a base database into a CSV consists in the following steps. It works but ...: -------------------------Solution quoted----- To start this ( and, this can also be accomplished by a macro if you so desire ) . . . the steps are: 1. Open your database 2. From the menu: Tools -> SQL... 3. under Command to execute, enter your SQL 4. an example might be ( change output_csv_file_name to your desired OUTPUT name, and, your_table_name to the HSQL table you are using ) Code: Select all Expand viewCollapse view SELECT * INTO TEXT "output_csv_file_name" FROM "your_table_name" 5. Press the Execute button 6. Press the Close button The bottom line . . . this will create a TEXT CSV file for you . . . located in the same directory where your database ODB file is located. The file will have a file type of CSV . Additionally, your database will have a NEW TEXT 'table' as defined by the name you gave it in the SELECT above. This 'database' file may be deleted. Also, you can include a WHERE clause above if you wish to only output a part of the table to your CSV file. And, an ORDER BY clause will further define the CSV output. ------------------------- ... the problem encountered, even on latest 3.00m7 version is that my accented characters disappeared. I know that Windows has difficulties is writing text files in Unicode but I am not sure that it is a Windows problem. I haven't found an option for the SELECT command or for the INTO TEXT command to add something like UTF-8 encoding or similar. In both, the .CSV file as well as the new table created into the database, all accented characters have been replaced by ? signs. PS: Priority is set to P2 because last time I used P3 it last more than a year. I won't be frustrated if you confirm the bug and that if there is no easy work around to be downgraded to a lower priority.