How can unicode fonts be added to adobe reader on an iPad to support Unicode documents? Appending a 38mb font to every document doesn't seem very efficient if it is possible at all.
You shouldn't need any Unicode font for the Adobe Reader.
Most PDF files that use characters other than Western Latin ones always embed a subset of the font or fonts covering the glyph usage in the PDF file. Thus, unless you had a PDF file that had text with at least one of every Unicode character, you would never see that 38MB font file embedded.
Embedded font subsets in your PDF files is the answer!
That is an unfortunate, though expected, response. In this day in age with global support a standard in so many facets of business many businesses have standardized documentation by simply using a unicode font standard across many communications. In this case PC's / Mac's will work every time because fonts can be added to them as needed.
However, to read the same font on an iPad or other mobile device means that either the font needs to be embedded or available on the device. iOS doesn't have a way to add fonts unless you hack the OS, - not recommended -, and this is the same way with most Android devices, root access is needed. This means that documentation created using unicode font sets cannot be read on mobile devices.
QuickOffice can read these documents and is fast becoming the mobile office suite standard. From an end user perspective it simply looks like Adobe Reader mobile addition has a bug. All PCs can read this, the only device that looks to have a problem from an end user perspective are mobile devices using Adobe Reader.
Technically, the reality of the situation is that some apps on iOS and Android embed certain unicode fonts in their app directly, but all the end user sees is an aparent bug in Reader mobile. It would seem simple to resolve this by allowing users to add a font to the app somehow. For certain platforms there are language packs that can be downloaded, it would seem that something like this would help on the mobile platform.
I'll try to be a bit more explicit. Virtually every professional PDF creation program or library follows the practice of embedding the glyphs for non-Western Latin characters referenced in whatever the source document for the PDF file was. For example, the Standard joboptions of Adobe Acrobat do not call for embedding of the Arial typeface. However if Acrobat via either Distiller or PDFMaker encounters references to characters outside the Western Latin range of characters in the Arial font (let's say characters representing Arabic, Hebrew, or Russian), those particular characters' glyphs of Arial are embedded into the PDF file (even though the Western Latin characters used in English and Western European languages are not embedded). Again, those font subsets are typically very compact unless you have some form of document that is attempting do display/print every Unicode character or major portions of the various Unicode ranges.
If you have a PDF file that you think is triggering what you believe to be a bug, we'd be more than happy to look at it and see what the situation is.
I'd like to elaborate further on the reason why we don't ship fonts with the Reader. Fonts are very large and we're sensitive to the size of the application (in fact, it's possible that adding just 1-2 fonts would more than double the size of the existing app). Using subsets in documents means that users have only incur the overhead of having the data they need rather than many many fonts they don't need. We do attempt to replace non-subsetted and non-embedded fonts with fonts available on the system in the case of CJK text. You are welcome to share your document and your use case with us at firstname.lastname@example.org so we can determine whether there is a significant case that we've missed.