Hello,
I've committed changes to commac.lisp and displa.lisp to make
backslash a line continuation character, as was discussed briefly
a few weeks ago.
(%i1) 100!;
(%o1) 9332621544394415268169923885626670049071596826438162146859\
2963895217599993229915608941463976156518286253697920827223758251\
185210916864000000000000000000000000
(%i2) 9332621544394415268169923885626670049071596826438162146859\
2963895217599993229915608941463976156518286253697920827223758251\
185210916864000000000000000000000000 - 100!;
(%o2) 0
I've appended the cvs commit message.
Hope this helps,
Robert Dodier
--------------- begin commit message ---------------
Changes to make backslash a line continuation character.
Output: print backslash (instead of hash mark) at the end of
a line when a long atom is broken across lines. (e.g., 100!)
Input: recognize the sequences ,
, and ,
and eat them. Otherwise backslash is not processed by the
new code; other functions farther up the food chain process
backslash as before.
- src/commac.lisp: revise TYI and TYI-PEEK, new functions TYI-RAW
and EAT-CONTINUATIONS
- src/displa.lisp: in DIMENSION-STRING, print a backslash at the end of a line
(instead of hash mark)
- tests/rtestparser_continuations_{cr,nl,crnl}.mac: new test scripts
cr = end-of-line is marked by return only
nl = end-of-line is marked by newline only
crnl = end-of-line is marked by return, then newline
When tested on Linux with GCL 2.6.7, SBCL 0.9.4, and Clisp 2.34,
batch ("rtestparser_continuations_cr.mac", test); succeeds,
same with rtestparser_continuations_nl.mac and
rtestparser_continuations_crnl.mac.
batch (); (i.e., without the test argument) succeeds for all three.
load (); succeeds for all three.
When tested on Windows with GCL 2.6.7, tests with
rtestparser_continuations_nl.mac and rtestparser_continuations_crnl.mac
succeed, but not rtestparser_continuations_cr.mac;
Maxima gets all the way through the test without problems,
then barfs at the very end. batch (, test); tickles
a bug in GCL's READ-CHAR-NO-HANG when the last line is terminated
by . However, batch ("rtestparser_continuations_cr.mac"); succeeds,
as batch () (without test argument) does for the other two.
load (); fails on rtestparser_continuations_cr.mac
(same READ-CHAR-NO-HANG bug) and succeeds with the other two.