The other points where CONTINUE records are used are
- cells with very long text (> about 8000 bytes)
I export this file to the csv, ant the longest line about 220 char.
- cells with rich-text formatting, i.e. individual formatting of characters (such as sub/superscript of individual characters, bold of individual words etc).Does your file contain such features? If yes, again create a new file with this feature and see if the error is there.
Individual formatting of characters, yes you are right, this file contains this. I create sample (with Excel 2003), and yes, I get error messages from Lazarus environment:
first: "system read error" (this is different!)
second: "cannot read file: wrong, unknown or defective file format?"
An idea how you could locate the problematic cell:
- Make a backup copy of the file, and modify this backup copy only
- In Excel, delete the second half of all rows. Save under a temp name. Try to open with fpspreadsheet and check for the error
- If the error occurs again delete the second half of this temp file now and repeat.
- If there is no error go back to the previous version of the file, and now remove the first half of all rows. Do the test with this version.
- Repeat the procedure until you have only one (or a few rows) left.
- Repeat with the columns in the same way
- Store again under a new name (to remove the undo buffer). If required modify the cell text and remove any confidential data. Test this file again. It it still shows the error upload it here.
Very thank you for this detailed instructions, but in home I have only excel 2003 and 2007.
If I open the problematic (office XP) file with office 2003, and I press save, the problem is gone, I can open this file from fpspreadsheet. But
this is not solution for me! Because my job is important, and my app must open all (simple) files without problem
But interesting, If I open the problematic (office XP) file with office 2007, and I press save, the problem is stay, I can't open this file from fpspreadsheet. This is great, because is good for your instructions!
If I delete "D" column, the problem is gone. If I delete all column, except: "D", the problem is stay.
This is mean, the problem is the "D" column. This column contains individual formatting chars
only.
If I delete lines with individual formatting chars (19 lines), the file is opened by fpspreadsheet, without error.
Or, if I open original problematic file, and I delete individual formatting chars (select lines and press bold icon twice) then problem is gone too.
Okay, possible the errors come from the " individual formatting chars", but
these individual formatting must be present in the source file, and I can open this file without any magic and save. This is possible anyhow? If individual formatting lost after open from FPSpreadsheet, that isn't a big problem, because I only need the real datas from the source file.
I attached a simple sample with individual formatting chars.
virustotalThanks for the many help and instructions![/list]