Carlos Villegas commented on FOP-2355:

A thai consonant (base character) can have one or more vowels or tonemarks 
around it. Some of these diacritics can be in the same area like above the 
character. If there are two diacritics they are stacked one on top of the 
other. Thus the position of the diacritics depends on the shape of the consonat 
(some consonants are taller) and whether there are other diacritics together. 
Sometimes a piece of the character is removed altogether to make room for the 
vowel. I don't know much about TrueType but my understanding is that in fonts 
like AriaUnicodeMS this information is in GSUB and GPOS tables. Diacritics have 
there own Unicode mappings, but these fonts also provide alternate forms to 
cope with the different positions of the diacritics. For example, The MAI THO 
tonemark in ArialUnicodeMS has unicode code U0E49 and corresponds to the "high 
right" position, UF714 is the same character at the "high left" position, UF70B 
the "low right" position and UF706 the "low left" position. The right glyph 
need to be chosen depending on the consonant and other diacritics associated 
with the same consonant. For an introduction to Thai standards see:

> Issues with stacking of diacritics while trying to render PDFs in Thai 
> language
> -------------------------------------------------------------------------------
>                 Key: FOP-2355
>                 URL: https://issues.apache.org/jira/browse/FOP-2355
>             Project: Fop
>          Issue Type: Bug
>          Components: fonts, pdf
>    Affects Versions: 1.1, trunk
>         Environment: Win 7, JDK1.6
>            Reporter: Sougata Bhattacharya
>         Attachments: ConsoleLog.txt, PDFGenerator.java, example pdf.jpg, 
> fopConfig.xml, input.xml, op.pdf, transformer.xsl
> Let us take 3 examples, where the first two are rendering ok in the PDF, 
> whereas the 3rd is not.
> 1. ยี = ย + ี (consonant + vowel)
> 2. ยี่ = ย + ี + ่ (consonant + vowel + diacritic)
> 3. ย่ = ย + ่ (consonant + diacritic)
> The third complex alphabet is showing ok when you see it in 
> notepad/MSWord/webpage, etc. but in the PDF, it renders with a vertical space 
> in between the diacritic and the consonant, as if a vowel is missing.
> Though this does not alter the meaning of the script, it looks 
> unprofessional. 
> Details on how i want the characters to be rendered in the PDF, can be found 
> in the 2nd example of the link: 
> http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=CmplxRndExamples
> Fonts used: PSL096pro.ttf, PSL095pro.ttf, PSL094pro.ttf, PSL053pro.ttf, 
> PSL052pro.ttf, PSL051pro.ttf, PSL050pro.ttf, etc..

This message was sent by Atlassian JIRA

Reply via email to