See the prerule group example in https://github.com/espeak-ng/espeak-ng/blob/master/docs/dictionary.md#rule-groups. For example:

.L01 လ // Consonants

.group မ်
    .L01) မ် am
          မ် m

[espeak-ng:master] New Comment on Pull Request #658 Changes for Shan language
By valdisvi:

In general, you should use some approaches of following:

...
.group လ
    လ        l   // default pronunciation (shortest string means lowest weight)
    လ   (မ်   lam // only la is consumed, analysis will go to ma as nex character
    လမ်       lam // all letters are consumed, rule analysis will go to character after asat
    လ   (ိမ်   lim // only la is consumed, analysis will go to i character
    လိမ်       lim // all characters are consumed, analysis will go character after asat
...

See details about pre and post groups.

Also, you could set your own letter groups for rules L01, L02 etc, and also you can redefine consonants (meaning of C in rules) and vowels (A in rules) in tr_languages.c file. For example, as it is set for letters A,B,C,F,G,H,Y in Arabic