I see. You are executing a sentence of a character FF like this ". FF
I think only space and TAB are white spaces in J. If you really want other white spaces, you may change them to CS in the ctype array of t.c in j engine source code. > chances of getting the change incorporated into J804? I am not sure but probably no chances. Вс, 14 дек 2014, Michal Wallace написал(а): > Yep, I am doing the same thing, but on windows 7. > I just now put the file in my J install directory so I wouldn't have to > specify a directory path, double clicked on 'jconsole.exe', > and typed the command: > > load'cheq.ijs' > > |spelling error > > | ♀ > > | ^ > > |[-31] C:\Users\michal\j64-803\bin\cheq.ijs > > JVERSION > > Engine: j803/2014-10-19-11:11:11 > > Library: 8.03.10 > > Platform: Win 64 > > Installer: J803 install > > InstallPath: c:/users/michal/j64-803 > > I tried saving it with both LF and CRLF line endings, but this didn't make > a difference. > > I get the same behavior with jqt.exe. > > > Also, I tried opening the file in the jqt editor, copying the character > (which renders as > > the "female" symbol) over to the jqt terminal, and paste it into the > expression: > > > a.i.' ' NB. (It's between the quotes) > > 12 > > > If i paste it into jconsole.exe, the character renders in ASCII notation, > but produces the same result: > > > a.i.'^L' > > 12 > > > This indicates to me that J is correctly recognizing it as the form feed > character (rather than > > converting it to (u:16b2640), the unicode "FEMALE SIGN" code point (♀). > Interestingly, > > the phrase (u:12 16b2640) produces two distinct glyphs in jqt. > > > Out of curiosity, I installed J602 for windows and tried loading the same > file (clicking on > > the blue/cyan j): > > > 1!:44'c:\users\michal\j64-803' > > > load'cheq.ijs' > > |spelling error > > | > > | ^ > > |[-31] c:\users\michal\j64-803\cheq.ijs > > JVERSION > > Installer: j602a_win64.exe > > Engine: j602/2008-03-03/16:45 > > Library: 6.02.023 > > > I also tried going back to J504, before unicode support was added. The > script > uses the 'u:' primitive (which seems to have had a different meaning back > then) > so it gives a syntax error before it reaches the character, but a simple > script > that just includes a comment and form feed gives a spelling error: > > 1!:44'c:\users\michal\j64-803' > > loadd'simple.ijs' > NB. form feed on next line: > > |spelling error > | > | ^ > |[-1] c:\users\michal\j64-803\simple.ijs > JVERSION > Binary: j504binb_win > Library: j504libb > Help: j504hlpb > Engine: j504/2005-03-30/13:35 > > > All of this suggests to me that support for form feeds in J source code was > never > > intentional, and I'm probably the only person ever to try it. > > > This isn't the end of the world for me, but it seems like a small change. > If I go through > > the trouble of patching J to treat this character as whitespace on windows, > what are my > > chances of getting the change incorporated into J804? > > > > On Sat, Dec 13, 2014 at 8:28 AM, bill lam <[email protected]> wrote: > > > > 12{a. is valid utf8 character. I tested load'cheq.ijs' is successful on > > debian64 jconsole. I am not sure how you load the script. > > > > Сб, 13 дек 2014, Michal Wallace написал(а): > > > The version that works with the form feed character is J701. > > > It's my custom compile for FreeBSD (but I didn't change anything in the > > > parser). > > > https://github.com/tangentstorm/j/tree/freebsd64 > > > > > > JVERSION > > > > > > Engine: j701/2011-02-23/15:25 > > > > > > Library: 7.01.088 > > > > > > Platform: FreeBSD 64 > > > > > > Installer: this file is only here because pacman.ijs tests for it > > > > > > InstallPath: /usr/home/michal/ver/j/j > > > > > > The version that gives the spelling error is the latest(?) version of > > > > > > J803 / JQt on windows: > > > > > > > > > JVERSION > > > > > > Engine: j803/2014-10-19-11:11:11 > > > > > > Library: 8.03.10 > > > > > > Qt IDE: 1.3.1/5.3.2 > > > > > > Platform: Win 64 > > > > > > Installer: J803 install > > > > > > InstallPath: c:/users/michal/j64-803 > > > > > > > > > A simple, standalone file that uses the character is here: > > > > > > > > > https://github.com/tangentstorm/syndir/blob/master/cheq.ijs > > > > > > > > > Github doesn't render the character, but it's on line 32. > > > > > > In JQt it appears as: > > > > > > > > > ♀ > > > > > > > > > This is the old MS-DOS code page 437 rendering of (12{a.), > > > > > > so maybe J803 and/or JQT are assuming the file is an ancient > > > > > > DOS text file and converting it to a unicode symbol? > > > > > > > > > > > > On Sat, Dec 13, 2014 at 7:25 AM, bill lam <[email protected]> wrote: > > > > > > > > Michal, please tell us more detail of your testing environmen, eg > > JVERSION > > > > or other info that may help. I assume you are using jconsole. > > > > On Dec 13, 2014 9:04 PM, "Michal Wallace" <[email protected]> > > > > wrote: > > > > ---------------------------------------------------------------------- > > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > -- > > regards, > > ==================================================== > > GPG key 1024D/4434BAB3 2008-08-24 > > gpg --keyserver subkeys.pgp.net --recv-keys 4434BAB3 > > gpg --keyserver subkeys.pgp.net --armor --export 4434BAB3 > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm -- regards, ==================================================== GPG key 1024D/4434BAB3 2008-08-24 gpg --keyserver subkeys.pgp.net --recv-keys 4434BAB3 gpg --keyserver subkeys.pgp.net --armor --export 4434BAB3 ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
