Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.forth > #50485 > unrolled thread

gforth.el questions

Back to article view | Back to comp.lang.forth

Contents

  gforth.el questions anotherreepca@gmail.com - 2016-08-25 22:15 -0700
    Re: gforth.el questions anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-08-26 08:07 +0000
      Re: gforth.el questions anotherreepca@gmail.com - 2016-08-26 07:56 -0700
        Re: gforth.el questions anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-08-27 07:44 +0000
          Re: gforth.el questions albert@cherry.spenarnc.xs4all.nl (Albert van der Horst) - 2016-08-28 20:40 +0200
            Re: gforth.el questions anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-08-29 23:11 +0000
    Re: gforth.el questions anotherreepca@gmail.com - 2016-08-29 21:46 -0700
      Re: gforth.el questions anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-08-31 16:11 +0000
        Re: gforth.el questions anotherreepca@gmail.com - 2016-08-31 18:16 -0700
          Re: gforth.el questions anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-01 05:20 +0000
            Re: gforth.el questions anotherreepca@gmail.com - 2016-08-31 23:21 -0700
              Re: gforth.el questions anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-01 08:26 +0000
                Re: gforth.el questions anotherreepca@gmail.com - 2016-09-01 11:40 -0700
    Re: gforth.el questions rickman <gnuarm@gmail.com> - 2016-09-02 19:39 -0400
      Re: gforth.el questions anotherreepca@gmail.com - 2016-09-02 18:19 -0700

#50485 — gforth.el questions

Hi friends, I just realized after some reading of the gforth manual that I'm supposed to be able to complete words with tab-completion. When I attempt to do so in a normal terminal, it works fine, but not so in emacs. I *think* I have gforth.el properly installed, but whenever I type "du" and hit tab it just produces a bunch of spaces. Any idea how to make that work?

[toc] | [next] | [standalone]

#50490

anotherreepca@gmail.com writes:
>Hi friends, I just realized after some reading of the gforth manual that I'=
>m supposed to be able to complete words with tab-completion. When I attempt=
> to do so in a normal terminal, it works fine, but not so in emacs. I *thin=
>k* I have gforth.el properly installed, but whenever I type "du" and hit ta=
>b it just produces a bunch of spaces. Any idea how to make that work?

There is no tab-completion feature in gforth.el.  However, Emacs has
completion support
<https://www.gnu.org/software/emacs/manual/html_node/emacs/Symbol-Completion.html>,
in the case of Forth based on the TAGS file.  So you have to create a
TAGS file for your own words
<https://www.complang.tuwien.ac.at/forth/gforth/Docs-html/Emacs-Tags.html>
and/or use the Gforth TAGS file (e.g.,
/usr/local/share/gforth/<version>/TAGS) for the built-in words, with
M-x visit-tags-table, and then you can do the completion with M-<TAB>
(possibly caught by the Window manager, so also try the following),
C-M-i, or <ESC><TAB>.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#50497

On Friday, August 26, 2016 at 3:16:47 AM UTC-5, Anton Ertl wrote:
> anotherreepca@gmail.com writes:
> >Hi friends, I just realized after some reading of the gforth manual that I'=
> >m supposed to be able to complete words with tab-completion. When I attempt=
> > to do so in a normal terminal, it works fine, but not so in emacs. I *thin=
> >k* I have gforth.el properly installed, but whenever I type "du" and hit ta=
> >b it just produces a bunch of spaces. Any idea how to make that work?
> 
> There is no tab-completion feature in gforth.el.  However, Emacs has
> completion support
> <https://www.gnu.org/software/emacs/manual/html_node/emacs/Symbol-Completion.html>,
> in the case of Forth based on the TAGS file.  So you have to create a
> TAGS file for your own words
> <https://www.complang.tuwien.ac.at/forth/gforth/Docs-html/Emacs-Tags.html>
> and/or use the Gforth TAGS file (e.g.,
> /usr/local/share/gforth/<version>/TAGS) for the built-in words, with
> M-x visit-tags-table, and then you can do the completion with M-<TAB>
> (possibly caught by the Window manager, so also try the following),
> C-M-i, or <ESC><TAB>.
> 
> - anton
> -- 
> M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
> comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
>      New standard: http://www.forth200x.org/forth200x.html
>    EuroForth 2016: http://www.euroforth.org/ef16/

For some reason my gforth directory was in /usr/share/ instead of /usr/local/share/. I visited it with M-x visit-tags-table, but typing "du" and then C-M-i just causes the minibuffer to say "no match" (same for <ESC><TAB>). Also, I tried to generate a TAGS file as per https://bernd-paysan.de/gforth-manual/Emacs-Tags.html but "require etags.fs" causes an error where the comment word \ is undefined (specifically the comment is "our dictionary search order becomes the law ( -- )", and subsequently makes it impossible to use any words. Version is 0.7.2, installed from ubuntu 16.04 repositories. Should I try a manually-compiled more recent version?

[toc] | [prev] | [next] | [standalone]

#50504

anotherreepca@gmail.com writes:
>For some reason my gforth directory was in /usr/share/ instead of /usr/loca=
>l/share/.

That's normal if you install gforth through the distribution.

> I visited it with M-x visit-tags-table, but typing "du" and then =
>C-M-i just causes the minibuffer to say "no match" (same for <ESC><TAB>).

I just tried this on Debian Jessie, and it showed me a selection of
all the words starting with "du".  When I tried to complete "duxx", it
did not do anything (or, if this is at the end of buffer, complains
about that).  On Ubuntu 14.04 (with gforth 0.7.0), pretty much the
same.

> A=
>lso, I tried to generate a TAGS file as per https://bernd-paysan.de/gforth-=
>manual/Emacs-Tags.html but "require etags.fs" causes an error where the com=
>ment word \ is undefined (specifically the comment is "our dictionary searc=
>h order becomes the law ( -- )", and subsequently makes it impossible to us=
>e any words.

You tried to insert "require etags.fs" and then load Gforth's source
files for building an image?  Yes, that can be tricky.

>Version is 0.7.2, installed from ubuntu 16.04 repositories. Sh=
>ould I try a manually-compiled more recent version?

Actually, this should work fine in 0.7.2 (and earlier), but you never
know what butchering the Debian/Ubuntu maintainers applied to the
version you have.  You can look if the TAGS file contains anything,
and if so, try if various other tags functions work, such as

M-. du<tab> 

(should show you the completions of the words starting with "du", but
in the Emacs mini-buffer), or

M-x tags-apropos du .*

or 

M-x list-tags /usr/share/gforth/0.7.2/prim

Independently of this, it's a good idea to build and install gforth
from the sources (either a tarball, or from the git (installed gforth
needed for git)) on Debian/Ubuntu; advantages are that you get the
manual installed, and that Gforth is considerably faster
(Debian/Ubuntu disables one of the performance features).

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#50517

In article <2016Aug27.094438@mips.complang.tuwien.ac.at>,
Anton Ertl <anton@mips.complang.tuwien.ac.at> wrote:
>anotherreepca@gmail.com writes:
>>For some reason my gforth directory was in /usr/share/ instead of /usr/loca=
>>l/share/.
>
>That's normal if you install gforth through the distribution.
>
>> I visited it with M-x visit-tags-table, but typing "du" and then =
>>C-M-i just causes the minibuffer to say "no match" (same for <ESC><TAB>).
>
>I just tried this on Debian Jessie, and it showed me a selection of
>all the words starting with "du".  When I tried to complete "duxx", it
>did not do anything (or, if this is at the end of buffer, complains
>about that).  On Ubuntu 14.04 (with gforth 0.7.0), pretty much the
>same.
>
>> A=
>>lso, I tried to generate a TAGS file as per https://bernd-paysan.de/gforth-=
>>manual/Emacs-Tags.html but "require etags.fs" causes an error where the com=
>>ment word \ is undefined (specifically the comment is "our dictionary searc=
>>h order becomes the law ( -- )", and subsequently makes it impossible to us=
>>e any words.
>
>You tried to insert "require etags.fs" and then load Gforth's source
>files for building an image?  Yes, that can be tricky.
>
>>Version is 0.7.2, installed from ubuntu 16.04 repositories. Sh=
>>ould I try a manually-compiled more recent version?
>
>Actually, this should work fine in 0.7.2 (and earlier), but you never
>know what butchering the Debian/Ubuntu maintainers applied to the
>version you have.  You can look if the TAGS file contains anything,
>and if so, try if various other tags functions work, such as
>
>M-. du<tab>
>
>(should show you the completions of the words starting with "du", but
>in the Emacs mini-buffer), or
>
>M-x tags-apropos du .*
>
>or
>
>M-x list-tags /usr/share/gforth/0.7.2/prim
>
>Independently of this, it's a good idea to build and install gforth
>from the sources (either a tarball, or from the git (installed gforth
>needed for git)) on Debian/Ubuntu; advantages are that you get the
>manual installed, and that Gforth is considerably faster
>(Debian/Ubuntu disables one of the performance features).

Do you mean the man pages, or the info files?

>
>- anton
>--
>M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
>comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
>     New standard: http://www.forth200x.org/forth200x.html
>   EuroForth 2016: http://www.euroforth.org/ef16/
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#50561

albert@cherry.spenarnc.xs4all.nl (Albert van der Horst) writes:
>In article <2016Aug27.094438@mips.complang.tuwien.ac.at>,
>>Independently of this, it's a good idea to build and install gforth
>>from the sources (either a tarball, or from the git (installed gforth
>>needed for git)) on Debian/Ubuntu; advantages are that you get the
>>manual installed, and that Gforth is considerably faster
>>(Debian/Ubuntu disables one of the performance features).
>
>Do you mean the man pages, or the info files?

info files etc.  The man page is not the Gforth manual, despite its
name, just a reminder on how to invoke Gforth.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#50565

Unsure where else to put this, it isn't strictly speaking related to the topic title, but I tried compiling gforth git from scratch. As I finished up with "sudo make install" I got the error message "cannot stat 'gforth.elc': No such file or directory", causing make (target 'install-elc') to fail. Naturally, googling that didn't explain much of anything, and I'm not very familiar with the internal workings of makefiles. What should I do to get it working?

> You tried to insert "require etags.fs" and then load Gforth's source
files for building an image?  Yes, that can be tricky. 

All I did was type "require etags.fs" like it said I was supposed to to generate a TAGS file, I didn't directly do anything with loading source files for building an image as far as I can tell. 

[toc] | [prev] | [next] | [standalone]

#50580

anotherreepca@gmail.com writes:
>Unsure where else to put this, it isn't strictly speaking related to the to=
>pic title, but I tried compiling gforth git from scratch. As I finished up =
>with "sudo make install" I got the error message "cannot stat 'gforth.elc':=
> No such file or directory", causing make (target 'install-elc') to fail. N=
>aturally, googling that didn't explain much of anything, and I'm not very f=
>amiliar with the internal workings of makefiles. What should I do to get it=
> working?

Apparently building the .elc file failed for you.  It's hard to say
more without seeing a log of the build.

>> You tried to insert "require etags.fs" and then load Gforth's source
>files for building an image?  Yes, that can be tricky.=20
>
>All I did was type "require etags.fs" like it said I was supposed to to gen=
>erate a TAGS file, I didn't directly do anything with loading source files =
>for building an image as far as I can tell.=20

Ok, my guess is that it tries to load search.fs, although it is
already loaded.  If you start gforth and do

.included

you should see the search.fs somewhere in the output, e.g.,
"/usr/share/gforth/0.7.2/search.fs".  If you now do

require search.fs

nothing should happen, but I guess you will see the same problem as
above.  If you do

s" search.fs" open-fpath-file throw type

you should ideally see the filename also shown by .included.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#50588

On Wednesday, August 31, 2016 at 11:32:18 AM UTC-5, Anton Ertl wrote:
> anotherreepca@gmail.com writes:
> >Unsure where else to put this, it isn't strictly speaking related to the to=
> >pic title, but I tried compiling gforth git from scratch. As I finished up =
> >with "sudo make install" I got the error message "cannot stat 'gforth.elc':=
> > No such file or directory", causing make (target 'install-elc') to fail. N=
> >aturally, googling that didn't explain much of anything, and I'm not very f=
> >amiliar with the internal workings of makefiles. What should I do to get it=
> > working?
> 
> Apparently building the .elc file failed for you.  It's hard to say
> more without seeing a log of the build.
> 
> >> You tried to insert "require etags.fs" and then load Gforth's source
> >files for building an image?  Yes, that can be tricky.=20
> >
> >All I did was type "require etags.fs" like it said I was supposed to to gen=
> >erate a TAGS file, I didn't directly do anything with loading source files =
> >for building an image as far as I can tell.=20
> 
> Ok, my guess is that it tries to load search.fs, although it is
> already loaded.  If you start gforth and do
> 
> .included
> 
> you should see the search.fs somewhere in the output, e.g.,
> "/usr/share/gforth/0.7.2/search.fs".  If you now do
> 
> require search.fs
> 
> nothing should happen, but I guess you will see the same problem as
> above.  If you do
> 
> s" search.fs" open-fpath-file throw type
> 
> you should ideally see the filename also shown by .included.
> 
> - anton
> -- 
> M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
> comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
>      New standard: http://www.forth200x.org/forth200x.html
>    EuroForth 2016: http://www.euroforth.org/ef16/

require search.fs causes all words to become undefined (can't even run "bye"), and .included does show search.fs. Here's my session log (let me know if c.l.f has a preferred paste dump site): http://pastebin.com/UR6tAYCr

That should be a bug, shouldn't it? If it's already been loaded and then is REQUIREd, it shouldn't be loaded again...

And here's the build log: http://pastebin.com/prBcHztV
I looked through it and the relevant part seems to be on line 567. Apparently something is trying to run a command named "t"...? (I tried "apt-cache search t" and hit ctrl-C after about 40000 results).

[toc] | [prev] | [next] | [standalone]

#50591

anotherreepca@gmail.com writes:
>require search.fs causes all words to become undefined (can't even run "bye"), and .included does show search.fs. Here's my session log (let me know if c.l.f has a preferred paste dump site): http://pastebin.com/UR6tAYCr
>
>That should be a bug, shouldn't it? If it's already been loaded and then is REQUIREd, it shouldn't be loaded again...

Gforth uses the absolute file name for determining if a file is
already loaded, but for some reason the included files in your system
only contain relative names, and therefore
/usr/share/gforth/0.7.2/search.fs will be loaded again.

>And here's the build log: http://pastebin.com/prBcHztV
>I looked through it and the relevant part seems to be on line 567. Apparently something is trying to run a command named "t"...? (I tried "apt-cache search t" and hit ctrl-C after about 40000 results).

That's due to the emacs test producing "t" in line 142.  I have no
idea why that is; it should find "emacs" or say "no".  What autoconf
version are you using?  As a workaround, you can change the EMACS
definition in Makefile (after running "configure"), and then run
"make".

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#50592

On Thursday, September 1, 2016 at 12:37:59 AM UTC-5, Anton Ertl wrote:
> anotherreepca@gmail.com writes:
> >require search.fs causes all words to become undefined (can't even run "bye"), and .included does show search.fs. Here's my session log (let me know if c.l.f has a preferred paste dump site): http://pastebin.com/UR6tAYCr
> >
> >That should be a bug, shouldn't it? If it's already been loaded and then is REQUIREd, it shouldn't be loaded again...
> 
> Gforth uses the absolute file name for determining if a file is
> already loaded, but for some reason the included files in your system
> only contain relative names, and therefore
> /usr/share/gforth/0.7.2/search.fs will be loaded again.
> 
> >And here's the build log: http://pastebin.com/prBcHztV
> >I looked through it and the relevant part seems to be on line 567. Apparently something is trying to run a command named "t"...? (I tried "apt-cache search t" and hit ctrl-C after about 40000 results).
> 
> That's due to the emacs test producing "t" in line 142.  I have no
> idea why that is; it should find "emacs" or say "no".  What autoconf
> version are you using?  As a workaround, you can change the EMACS
> definition in Makefile (after running "configure"), and then run
> "make".
> 
> - anton
> -- 
> M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
> comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
>      New standard: http://www.forth200x.org/forth200x.html
>    EuroForth 2016: http://www.euroforth.org/ef16/

autoconf 2.69 - the workaround worked fine, thanks. Emacs version is 24, if that helps. I'm very pleased to see the speedup, great work. Running the git version, "require search.fs" and "require etags.fs" both work just fine. Weird how git is *more* stable in my case than the distribution release...

[toc] | [prev] | [next] | [standalone]

#50594

anotherreepca@gmail.com writes:
>On Thursday, September 1, 2016 at 12:37:59 AM UTC-5, Anton Ertl wrote:
>> That's due to the emacs test producing "t" in line 142.  I have no
>> idea why that is; it should find "emacs" or say "no".
...
>autoconf 2.69

The same version I am using, so that should be ok.  In the meantime I
have an idea what's wrong: If you have an environment variable EMACS
with value "t", you will get that result.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#50601

On Thursday, September 1, 2016 at 3:33:52 AM UTC-5, Anton Ertl wrote:
> anotherreepca@gmail.com writes:
> >On Thursday, September 1, 2016 at 12:37:59 AM UTC-5, Anton Ertl wrote:
> >> That's due to the emacs test producing "t" in line 142.  I have no
> >> idea why that is; it should find "emacs" or say "no".
> ...
> >autoconf 2.69
> 
> The same version I am using, so that should be ok.  In the meantime I
> have an idea what's wrong: If you have an environment variable EMACS
> with value "t", you will get that result.
> 
> - anton
> -- 
> M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
> comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
>      New standard: http://www.forth200x.org/forth200x.html
>    EuroForth 2016: http://www.euroforth.org/ef16/

echo $EMACS produces nothing in a shell for my user, or for root.

[toc] | [prev] | [next] | [standalone]

#50630

On 8/26/2016 1:15 AM, anotherreepca@gmail.com wrote:
> Hi friends, I just realized after some reading of the gforth manual that I'm supposed to be able to complete words with tab-completion. When I attempt to do so in a normal terminal, it works fine, but not so in emacs. I *think* I have gforth.el properly installed, but whenever I type "du" and hit tab it just produces a bunch of spaces. Any idea how to make that work?

Is gforth.el a forth written by Jor-El?

-- 

Rick C

[toc] | [prev] | [next] | [standalone]

#50631

On Friday, September 2, 2016 at 6:39:53 PM UTC-5, rickman wrote:
> On 8/26/2016 1:15 AM, anotherreepca@gmail.com wrote:
> > Hi friends, I just realized after some reading of the gforth manual that I'm supposed to be able to complete words with tab-completion. When I attempt to do so in a normal terminal, it works fine, but not so in emacs. I *think* I have gforth.el properly installed, but whenever I type "du" and hit tab it just produces a bunch of spaces. Any idea how to make that work?
> 
> Is gforth.el a forth written by Jor-El?
> 
> -- 
> 
> Rick C

gforth.el is the elisp code for interactively developing with gforth in emacs. I don't know who wrote it. The one I have in my .emacs.d says it is originally written by Goran Rydqvist, and updated by "anton" and "David". Although I suppose that it's not the one that came with my git install, that one is one I think I just downloaded by itself from somewhere since I couldn't find where the distribution install put it. Let me check that quickly... the one in /usr/local/share/emacs/site-lisp/ is the same.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.forth

csiph-web