On Mon, 5 Apr 2021 16:30:53 +0000, Farley, Peter x23353 wrote: >True. There is m4 in *ix systems and going back a long time there was ML/1 (I >think there was an academic book published on that one, I think I have a copy >somewhere around here). Undoubtedly others I do not know of or remember. > >Probably also Wirth's "literate programming" suite, TeX I think it is called. > Knuth?
>But IMHO none easy to learn or use. > PL/I addresses ease of use by making the metalanguage similar to the base language. My favorite experience was with Mainsail which integrated its metalanguage with its compiler: not "IF DEFINED()" but "IF DECLARED()". And the metalanguage was aware of the language's block structure: the scope of a macro was only the block in which it was defined. There was little syntactic distinction between the SETC and MACRO equivalents. Evaluation of metasymbols could be either performed at the point of macro definition or deferred to the applied occurrence. I understand that BAL had some of that unification; lost in subsequent assemblers which distinguish EQU from SETA. >-----Original Message----- >From: Charles Mills >Sent: Monday, April 5, 2021 12:20 PM > >Aren't there "generic" meta languages: meta languages that may be used to >generate any sort of text, including source code for most any language? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
