> Furthermore: the more modern languages like Pascal, C and Java etc. > forbid the use of reserved symbols as variable names. This may be > restrictive, but makes the compilers much much simpler.
The cardinal sin in language design is to make the compiler simpler at the expense of the user. An enhancement to a language with reserved word can render a previously valid program invalid. Contrast this with PL/I, where several times keywords have been added without affecting existing code. -- Shmuel (Seymour J.) Metz http://mason.gmu.edu/~smetz3 ________________________________________ From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of Bernd Oppolzer <bernd.oppol...@t-online.de> Sent: Tuesday, July 16, 2019 3:06 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: Where put the notional constant in a condition (Was RE: JCL COND Parameter) Space is permitted almost everywhere in PL/1, not in the middle of identifiers (in contrast to FORTRAN). It is permitted before the left paranthesis of subscripts, so there is no chance for the compiler to decide before the semicolon or THEN in this case: IF (possible index expr1) = (complicated and long expr2); /* valid assignment */ IF (possible index expr1) = (complicated and long expr2) THEN ... /* IF statement */ This decision has to be done by the Parser, before if knows, if there is a definition for a IF array. The problem for this ambiguity is a problem of the PL/1 grammar. Pascal and other languages avoid this by defining different symbols for comparison and assignment. Furthermore: the more modern languages like Pascal, C and Java etc. forbid the use of reserved symbols as variable names. This may be restrictive, but makes the compilers much much simpler. Kind regards Bernd Am 16.07.2019 um 20:37 schrieb Seymour J Metz: > AFAIK a space is not permitted in PL/I between an identifier and the left > parenthesis for the subscripts. If that is correct then > > IF (1) = (2); > > can easily be recognized as an IF statement. > > Id do, however, agree that it is very bad form to knowingly use a keyword as > an identifier. > > > -- > Shmuel (Seymour J.) Metz > http://mason.gmu.edu/~smetz3 > > ________________________________________ > From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of > Bernd Oppolzer <bernd.oppol...@t-online.de> > Sent: Tuesday, July 16, 2019 2:22 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: Where put the notional constant in a condition (Was RE: JCL COND > Parameter) > > This double meaning of =, together with the absence of reserved words > makes PL/1 parsing extremely hard. Consider for example > > IF (1) = (2); > > now what does that mean? > > Given a declaration > > DCL IF (25) BIN FIXED (31); > > that is, if IF is an array of integers, the "IF" statement above is a > valid assignment > of 2 to the array element 1. The PL/1 compiler does not know that the > statement > above is no IF statement before it encounters the semicolon; before that > it has > to consider both possibilities ... this is completely different from > most other > languages (only Fortran has similar issues). > > I always thought that nobody would be as silly as to name his or her > identifiers IF, > but then I observed a program containing the following: > > DCL IA BIN FIXED (15) /* index for TABA */ > DCL IB BIN FIXED (15) /* index for TABB */ > DCL IC BIN FIXED (15) /* index for TABC */ > DCL ID BIN FIXED (15) /* index for TABD */ > DCL IE BIN FIXED (15) /* index for TABE */ > DCL IF BIN FIXED (15) /* index for TABF */ > > the coder probably didn't even notice the use of IF as variable name :-)) > > Long live PL/1 :-))) > > Kind regards > > Bernd > > > Am 16.07.2019 um 20:04 schrieb Seymour J Metz: >> One of the things that I liked about Algol was that "=" was always a >> comparison operator; assignment was ":=". But even in PL/I that particular >> error is impossible: the "=" in "IF foo=bar" can never be interpreted as >> assignment. >> >> >> -- >> Shmuel (Seymour J.) Metz >> http://mason.gmu.edu/~smetz3 >> > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
