These are defined in
EA-CatAndSkeletonToWords (cat skeleton): This is the main routine. It returns a list of words, each element corresponding to one of the possible inflectional forms for the feature-set
skeletonis a list of strings giving the root morphemes. In general, you will not construct a skeleton yourself, but will get it by calling
Most calls of this routine will only return a one-element list. However, in a few cases there will be several alternative forms. At present, these are (as far as I know) only hyphenation variants: see section 3.4.
treemust be a word-structure parse-tree, as returned by
'D-WORDSTRUCTUREformat. This function returns the corresponding word skeleton.
EA-LexicalEntryToSkeleton (entry): Returns a skeleton, given a lexical entry. This is so that if a function wants to look up a lexical entry directly, it doesn't have to know how to convert the entry into a skeleton.
EA-ParseTreeToCat (tree): Returns the top-level feature-set from
tree. This is just a utility so that if a function which needs the skeleton also needs the feature-set, it doesn't have to know the format of the parse tree.
EA-LoadIFI (filename): The IFI is a special index which is used to locate irregular forms such as the past tense of strong verbs. This function reads and returns the IFI stored in the file named
These functions may be changed in future versions. The comments at the
MORPHGEN.LSP will always give an up-to-date
specification. To use the functions, you need various other things:
GENINFL +feature added to inflectional suffixes in the lexicon, and a
GENINFL -feature for derivational suffixes. You also need a feature declaration for
The comments heading
MORPHGEN.LSP describe how to do all this,
and are where you should look if you are going to use the functions. I
also describe how to do these things below, but since this report is a
more general account of the problems encountered and solutions chosen,
it is less suitable as a programmer's reference.