On Sat, 2009-05-09 at 17:20 +1200, Keith Hopper wrote: > I have what might be a related problem - I don't know whether it has > something to do with encodings or messages files. I have a font family > called "Greek" with flavours Arial, Courier, Poly and Trinity. > > The firat two and last flavours seem to follow the ISO8859 Greek > variant. Poly, however, is a full Greek font including all accented > characters - including, in particular U+1f21 - which I noted when browsing > was merely displayed as the hex code square - but when the latest Netsurf > re-read all the font files it quite happily read these without complaint.
How do you know the font contains that glyph? Does it display correctly in other applications? > I suspect that there might need to be an encoding file - but haven't the > faintest idea how to build one which includes the Unicode code points (I do > have the Unicode standard book so could write one if someone were able to > tell me how - and where to put it and what to call it). > > Has anyone had this problem or a similar one? Is it possible to do > anything about setting up encoding and/or message files for this font (in > plain and italic variants)? Is the font a Symbol or Language font? (By which, I mean; do the IntMetrics and Outlines filenames end in a number?). If the files end in a number, then the font is a language font. The glyphs in the font are specified in Fonts:Encodings./BaseNN (where NN matches the number on the end of the font data files). If the files have no numbers at the end, then the font is a symbol font. If there is no Encoding file in the font's directory, then it will use the default encoding (Fonts:Encodings./Default on Unicode Font Manager systems, on non-UFM systems, this encoding is internal to the font manager). Otherwise, it will use the Encoding file provided. If the font in question is a language font, then there's not a great deal you can do -- the base encodings are shared between multiple fonts on the system and thus changing them is hazardous. If the font is a symbol font, then you can provide/modify the font's Encoding file to contain the correct glyphs. The format of the encoding file is very simple. Each glyph is specified on a separate line. Lines beginning with % are treated as comments. Glyph specifiers take the form "/name", where name is either the PostScript glyph name or, in the case of RISC OS 5 may be "uniXXXX" or "uniXXXXYYYY", where XXXX is the UCS-2 code for the glyph, and "XXXXYYYY" specify the surrogate pair for a codepoint that isn't in the base multilingual plane. If you like, you can send me the font and I'll take a look. John.
