On 8 March 2017 at 13:54, Marcel Kilgus <ql-us...@mail.kilgus.net> wrote:
> Jan Bredenbeek wrote: > > MOVE.W $12C,A0 > > JSR $4000(A0) call pa_graph to parse line > > Wow, the old and new names are mightily confusing. > > What once was called pa_graph is now sb_parse. > What once was called pa_table is now sb_graph. > I got the names from the Minerva source. Obviously they were 'invented' by Laurence since they are not documented in the original QDOS documentation :-) > > > Like I stated earlier, this code has been shamelessly copied from the TK2 > > ED command :-/. > > Now the interesting question is whether this will work under SMSQ/E. > > I don't think anything has changed. > > > The vectors are indeed present and point to several JMP instructions > > to the actual code. However the pa_table vector at $12E is not > > really the address of a subroutine but a pointer to a syntax table > > That is true for Minerva, too. > > > (as is pa_expr at $130 which is not used here) > > That, too, is as, which is needed for pa_graph. Interestingly, on > > SMSQ/E it still points to a JMP (but it might as well be ignored by > > pa_graph). > > According to the source there is no JMP? > The vector which is supposed to point to a syntax table points to a JMP in SMSQ/E (like all vectors, because the real code is in high memory where it can't be reached from a word-sized vector). I'm quite sure that there is no JMP in QDOS/Minerva. So is this vector going to work? Jan. -- *Jan Bredenbeek* | Hilversum, NL | j...@bredenbeek.net _______________________________________________ QL-Users Mailing List
