Hello Qais, Thanks for taking some time to test out the complex script features being developed for FOP.
To help address these issues, could you: (1) provide input FO test files showing the specific problems you have found? it is best if each test file contains the minimum needed to demonstrate the problem; (2) indicate which complex script build you are using? fyi, the latest work is available at http://github.com/skynavga/fop; Regarding connectivity, shape selection, kerning, point placement, etc., keep in mind that FOP doesn't do anything special other than implement the substitutions and positioning operations defined in the OpenType GSUB/GPOS tables. I'm not certain if the Arabeyes or other fonts you have tried support these tables. If they do, then there may be a bug in the new TTF/OTF parsing of this table data. In any case, if you provide some additional input as requested above, I'm sure we can resolve this matter soon. Also, if you haven't already seen it, please take a look at the documentation and active tickets found at http://skynav.trac.cvsdude.com/fop. Regards, Glenn Adams On Wed, Jul 27, 2011 at 1:57 PM, Qais <[email protected]> wrote: > Hi, > > Since a week I am working closely with FOP, Cocoon and Co, especially with > the Complex Script (connectivity, direction and fonts of Arabic). Below is > the result. > > Testing the Complex Script of FOP (27.07.2011) > > 1) Supported Fonts by the Complex Script: > > 1.1) Automatically registered: > Arial Unicode MS, Lateef and Scheherazade in a fo:block: connectivity: ok, > direction: ok. > Arial Unicode MS, Lateef and Scheherazade in a fo:block-container, fo:table > or fo:list-block: connectivity: none, direction: ok. > > 1.2) Using font metrics: > Arial Unicode MS and Traditional Arabic in a fo:block, fo:block-container, > fo:table or fo:list-block: connectivity: none, direction: ok. > Simplified Arabic in a fo:block, fo:block-container, fo:table or > fo:list-block: connectivity: none, direction: ok, glyphs: partially. > Lateef and Scheherazade in a fo:block, fo:block-container, fo:table or > fo:list-block: glyphs: none. > > 2) Base-14 Fonts: > > 2.1) Automatically registered: > Arial, Times New Roman, Courier New and Tahoma in a fo:block: connectivity: > ok, direction: ok. > Arial, Times New Roman, Courier New and Tahoma in a fo:block-container, > fo:table or fo:list-block: connectivity: none, direction: ok. > > 3) Arabeyes Fonts: > www.arabeyes.org > > 3.1) Automatically registered: > KacstArt, KacstBook, KacstDecorative, KacstDigital, KacstFarsi, > KacstLetter, KacstNaskh, KacstOffice, KacstOne, KacstPoster, > KacstScreen, KacstTitle, KacstTitleL and Thabit in a fo:block, > fo:block-container, fo:table or fo:list-block: connectivity: none, > direction: ok. > > 3.2) Using font metrics: > KacstArt, KacstBook, KacstDecorative, KacstDigital, KacstFarsi, > KacstLetter, KacstNaskh, KacstOffice, KacstOne, KacstPoster, > KacstScreen, KacstTitle, KacstTitleL and Thabit in a fo:block, > fo:block-container, fo:table or fo:list-block: connectivity: none, > direction: ok. > > 4) Other Fonts: > > 4.1) Automatically registered: > Advertising Light, Advertising Bold, Advertising Extra Bold and Bassam > Ostorah in a fo:block, fo:block-container, fo:table or fo:list-block: > glyphs: none. > > 4.2) Using font metrics: > Advertising Light, Advertising Bold, Advertising Extra Bold and Bassam > Ostorah in a fo:block, fo:block-container, fo:table or fo:list-block: > connectivity: none, direction: ok. > > 5) Connectivity and Direction Problem: > > A small tool called Free Ressam can be used to pre adjust connectivity and > direction of the text parts. This > is of course not the accurate solution, especially dealing with long texts, > but the tool might give an idea about > the necesarry logic. Free Ressam was developed to adjust connectivity and > direction of arabic text by using of > some programs that don't properly support the display of arabic text, such > as older versions of Photoshop. > > Maher > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] >
