|Page 164 of 800||Search internet|
The dictionary of
"";;015B62F7DBABF8F0AEFE73374B36AE157FCF4F9390DFA1A68A89CCBB0806 ""P tiny ""R base ""D 0 f ( " ) g (",",") ""B f ( x ) + "abc"
has the following structure:
DICTIONARY g(",",") 002 003 f ( " ) 001 001 DictEnd 000
The tiny.lgs source file defines three constructs:
tiny index 0 arity 0 f ( " ) index 1 arity 1 g(",",") index 2 arity 3
lgc(1) assigns index 0 to the page name and assigns an accumulating index to all other constructs starting with 1. Furthermore, lgc determines the arity of each construct by counting the number of double quote characters in the construct. The page name always has arity zero.
The dictionary comprises the index and arity of each construct except the page name (which is known to have arity zero). All indexes and arities are expressed in septet representation and the constructs are sorted such that indexes appear in decreasing order.
lgc(1) merely allows the user to define constructs with a contiguous range of indexes, starting with 0, but the vector format allows other compilers to generate pages where the constructs do not have a contiguous range of indexes.
It is understood that every page has a construct whose index and arity is zero. The format has no features for changing that.
Note that after translation of tiny.lgs into a vector, the names of construct 0, 1, and 2 are lost. To avoid such loss one has to include name definitions in the source e.g. using the ""N escape sequence.
|Page 164 of 800||Search logiweb.eu|