I use reports quite intensively in my projects. It is important for me to use Polish special characters and it is not a problem when printing. If I want a PDF document, I simply capture the moment of printing and from the user interface I choose to save the printout to PDF and everything is fine.
Trying to print directly to PDF in Xojo works, but of course UTF8 - doesn’t work. Existing ‘rpt’ is a report while ‘rs’ is RecordSet.
var dpdf as new PDFDocument
var g as Graphics
g=dpdf.Graphics
if rpt.Run(rs,dpdf) then
rpt.Document.Print(g)
var f as FolderItem = SpecialFolder.Documents.Child("PDFpreview.pdf")
dpdf.Save(f)
end if
Unfortunately, I have many such reports in the project and I know that there are plugins that correctly generate PDFs with UTF8. But do any of these plugins allow me to capture such report printing directly to PDF? Otherwise, I will have a lot of work to recreate all these reports in a new way, which I definitely would not want and would not have to do if Xojo’s PDFDocument worked correctly with UTF8 at first place…
it is exactly good presentation what the problem is. Even text labels put on report page, when containing special characters - are ??? when PDF gets generated…
Var d As New PDFDocument
d.Language = "pl-PL"
Var g As Graphics = d.Graphics
g.FontSize = 20
g.FontName = "Helvetica"
// utf-8
var t As String = "ą, ć, ę, ł, ń, ó, ś, ź, ż"
// convert to 1 byte only extended charset
// trying to circumvent the 1 byte charset limit of Xojo PDFDocument
t = t.ConvertEncoding(Encodings.ISOLatin2)
g.DrawText(t, 20, 60, 400)
d.Save(SpecialFolder.Desktop.Child("test.pdf"))
// Still wrong
quit
Once Xojo could make it work using a Latin2 charset (Polish compatible), changing the report converting the string sources on the fly to latin2 could solve your problem.
Yes, into a bitmap yes, because it’s not writing texts, inserting “letters” as in a document as PDF, but drawing graphics, setting bits on/off on a bitmap.
PDF can handle UTF-8 if fully implemented, Xojo version only handles maps of 255 chars per charset, so many symbols can’t be translated to whatever map Xojo uses (Latin1?). If we could “define” that map to latin2 to your case, it could solve your presentation problem. “drawing text” in a PDF is just inserting a text object into some place of the document, not drawing bits. That’s how your report does it behind the scenes.
I set PDF.Language to “pl-PL”
I put a label on the report and set it’s text properties (in BeforePrint event) to text containing Polish character after ConvertingEncoding to ISOLatin2. no change, “???” instead of Polish characters..
The PDF output needs to be able to cope with anything UTF8 can throw at it in one single document at the same time. My application can do that, and Apple’s Save as PDF can do that. The PDF class needs to be able to cope with it.
