Let me make a distinction between values known a fortiori, like those
of the keywords or reserved words of a statement-level language and
classes of values defined by initial letter, length in characters, or
the like.

It is correct that IBM REXX treats single-letter identifiers
differently from longer ones.  MC has said so in two other fora.

Organizing a symbol table/dictionary by identifier length is an old
idea.  For example, the Lowry-Medlock FORTRAN H  compiler, to which I
have referred here before, kept one-, two-, three-, four-, five- and
six-character identifiers is separate subdictionaries.   (The
feasibility of this scheme rerflects a FORTRAN constraint thasty in
turn reflects the architecture of the IBM 704 and its sequelæ.  An
identifier of at most six six-bit characters could always be kept in a
36-bit register.)

There is strong evidence, derived from studies of large samples of
source programs, that single-character identifiers are both fewer and
referenced much more frequently than longer identifiers, at least in
languages in the FORTRAN-ALGOL tradition.   This probably reflects the
contaminating influence of mathematical notation.  (I have more than
once been dismayed to see that COBOL programmers, who are largely free
of this malign influence and who do not use subscripts/indices a lot,
often give them long identifiers like
'in-batch-subgroup-index-position' when they do use them.)

John Gilmore, Ashland, MA 01721 - USA

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to