A number of Khmer fonts have been designed and released by the Khmer Software Initiative (khmeros.info).
They're available at
The DejaVu Sans Basic Latin characters are more readable than those in the KhmerOS fonts. And DejaVu fonts contain IPA phonetic characters that KhmerOS fonts lack.
Thus it is currently impossible to put English, Khmer and pronunciations in the same text to create a dictionary, except in two word-processing programs that can handle Khmer ZWSP _and_ use more than one font in the same text.
Would it be possible to adapt one of the KhmerOS fonts (or more than one of them, for Sans, Serif and Mono) and provide Khmer characters for the 114 characters?
Until more programs are capable of displaying different Unicode sections with different fonts -- a la Firefox -- it is of critical importance to get all the characters into a single font.
The Khmer consonants, vowels and punctuation are U+1780 - U+17DD. The digits 0-9 are U+17E0 - U+17E9. And the numeric symbols for divination lore are U+17F0 - U+17F9.
PS: I'm not equipped to do this work myself, or I would volunteer to assist in it. From the other perspective, I have requested the KhmerOS group to add the missing IPA characters too.
PPS: I see bug 8420 concerns two Khmer characters, U+17D9 and U+17DA
Since this request was made, Khmer Software Initiative has released a new font KhmerOS Siemreap that is very much more consistent with a Sans Serif typeface than other Khmer fonts.
This font is released under the Lesser GPL license.
In addition to the characters identified earlier, there are roughly 40 ligatures and 33 alternate characters that should be substituted whenever the non-printing coeng (U+17D2) appears before consonants U+1780 - U+17A2.
What are the license considerations in copying the glyphs from KhmerOS Siemreap? These match DejaVu Sans fairly well, although some adjustment might need to be made in height (minor), width (insignificant) and stroke (minor).
Consonant clusters in Khmer put the second consonant under the first. Some vowels go below the consonant, and below that subscripted consonant in a cluster. Other vowels go above the first consonant and sometimes diacritics are used that go above that.
Consequently, the vertical space for Khmer needs to be somewhat taller than for Latin characters.
What adjustment would need to be made to keep line-spacing normal for LGC text but considerably increased for Khmer text when both appear in the same document?
It would be great to see Khmer added!