July 1998 - XEmacs-Beta - XEmacs Mailing List Archives

b50 success on UnixWare 2.1.2 26 years, 2 months

Vin Shelton

Looks good so far. vin uname -a: UNIX_SV mithril 4.2MP 2.1.2 i386 x86at ../configure '--with-gcc' '--prefix=/usr/local/xemacs-21.0' '--site-includes=/usr/local/include' '--site-libraries=/usr/local/lib' '--with-gnu-make' '--with-mule=no' '--cflags=-O3 -march=pentiumpro -finline-functions -fomit-frame-pointer -ffast-math' '--x-includes=/usr/X11R6/include' '--x-libraries=/usr/X11R6/lib' '--with-dialogs=no' '--with-site-lisp' XEmacs 21.0 "Benadir" configured for `i386-pc-sysv4.2uw2.1.2'. Where should the build process find the source code? /home/acs/XEmacs/xemacs-21.0-b50 What installation prefix should install use? /usr/local/xemacs-21.0 What operating system and machine description files should XEmacs use? `s/usg5-4-2.h' and `m/intel386.h' What compiler should XEmacs be built with? gcc -O3 -march=pentiumpro -finline-functions -fomit-frame-pointer -ffast-math Should XEmacs use the GNU version of malloc? yes Should XEmacs use the relocating allocator for buffers? yes What window system should XEmacs use? x11 Where do we find X Windows header files? /usr/X11R6/include Where do we find X Windows libraries? /usr/X11R6/lib Additional header files: /usr/local/include Additional libraries: /usr/local/lib Runtime library search path: /usr/X11R6/lib Compiling in support for XAUTH. Compiling in support for XPM images. Compiling in support for JPEG image conversion. Compiling in support for PNG image conversion. Compiling in support for TIFF image conversion. Compiling in support for Berkeley DB. Compiling in support for proper session-management. Using Lucid menubars. Using Lucid scrollbars. Compiling in DLL support. movemail will use "dot-locking" for locking mail spool files.

1 participants
0 comments

first cut at patch for ~username completion 26 years, 2 months

Greg Klanderman

OK, I've got user name completion working and it was pretty simple. Open issues are: - dired.c is probably not the best home for these functions. is minibuf.c better, or somewhere else? - #include <pwd.h> needs to be conditionalized on a preprocessor macro defined by configure. read-file-name-internal-1 use fboundp to see if we have the user name completion functions. - can someone suggest a better way than my ugly hack to return whether the completion was unique from user-name-completion? - fix other elisp, ie comint to do user name completion. what else am I forgetting? - cache user name lookup for speed. I welcome any comments, esp from Ben or Kyle on whether this is the right way to do this. As I said earlier, I really don't think file-name-completion is the right place for this stuff. enjoy, greg 1998-07-20 Greg Klanderman <greg(a)alphatech.com> * dired.c (Fuser_name_completion): new function. (Fuser_name_all_completions): new function. (user_name_completion): new function. (syms_of_dired): 2 new DEFSUBRs. 1998-07-20 Greg Klanderman <greg(a)alphatech.com> * minibuf.el (read-file-name-internal-1): do ~user completion. Index: src/dired.c =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs-20/src/dired.c,v retrieving revision 1.18 diff -u -r1.18 dired.c --- dired.c 1998/05/01 01:10:55 1.18 +++ dired.c 1998/07/21 07:39:10 ＠＠ -31,6 +31,9 ＠＠ #include "sysfile.h" #include "sysdir.h" +#include <pwd.h> + + Lisp_Object Vcompletion_ignored_extensions; Lisp_Object Qdirectory_files; Lisp_Object Qfile_name_completion; ＠＠ -528,6 +531,159 ＠＠ } +static Lisp_Object user_name_completion (Lisp_Object user, + int all_flag, + Lisp_Object uniq); + +DEFUN ("user-name-completion", Fuser_name_completion, 1, 2, 0, /* +Complete user name USER. +Returns the longest string common to all user names that start +with USER. If there is only one and USER matches it exactly, +returns t. Returns nil if there is no user name starting with USER. + +If second optional argument UNIQ is a cons cell, set its car to +t if the match is unique, and nil otherwise. +*/ + (user, uniq)) +{ + return user_name_completion (user, 0, uniq); +} + +DEFUN ("user-name-all-completions", Fuser_name_all_completions, 1, 1, 0, /* +Return a list of all completions of user name USER. +These are all user names which begin with USER. +*/ + (user)) +{ + return user_name_completion (user, 1, Qnil); +} + +static Lisp_Object +user_name_completion_unwind (Lisp_Object notused) +{ + endpwent(); + return Qnil; +} + +static Lisp_Object +user_name_completion (Lisp_Object user, int all_flag, Lisp_Object uniq) +{ + /* #### i'm not particularly proud of this UNIQ hack. */ + /* #### it would probably make sense to cache the usernames + * then maybe once a day rebuild it just in case */ + struct passwd *pw; + int matchcount = 0; + Lisp_Object bestmatch = Qnil; + Charcount bestmatchsize = 0; + int speccount = specpdl_depth (); + Charcount user_name_length; + struct gcpro gcpro1, gcpro2; + + GCPRO2 (user, bestmatch); + + CHECK_STRING (user); + + user_name_length = XSTRING_CHAR_LENGTH (user); + + setpwent(); + record_unwind_protect (user_name_completion_unwind, Qnil); + + while ((pw = getpwent())) + { + Bytecount len; + /* scmp() works in chars, not bytes, so we have to compute this: */ + Charcount cclen; + Bufbyte *d_name; + + /* #### This is a bad idea, because d_name can contain + control characters, which can make XEmacs crash. This + should be handled properly with FORMAT_FILENAME. */ + d_name = (Bufbyte *) pw->pw_name; + len = strlen (d_name); + cclen = bytecount_to_charcount (d_name, len); + + QUIT; + + if (cclen < user_name_length || + 0 <= scmp (d_name, XSTRING_DATA (user), user_name_length)) + continue; + + /* Update computation of how much all possible completions match */ + matchcount++; + + if (all_flag || NILP (bestmatch)) + { + Lisp_Object name = Qnil; + struct gcpro ngcpro1; + NGCPRO1 (name); + /* This is a possible completion */ + name = make_string (d_name, len); + if (all_flag) + { + bestmatch = Fcons (name, bestmatch); + } + else + { + bestmatch = name; + bestmatchsize = XSTRING_CHAR_LENGTH (name); + } + NUNGCPRO; + } + else + { + Charcount compare = min (bestmatchsize, cclen); + Bufbyte *p1 = XSTRING_DATA (bestmatch); + Bufbyte *p2 = d_name; + Charcount matchsize = scmp (p1, p2, compare); + + if (matchsize < 0) + matchsize = compare; + if (completion_ignore_case) + { + /* If this is an exact match except for case, + use it as the best match rather than one that is not + an exact match. This way, we get the case pattern + of the actual match. */ + if ((matchsize == cclen + && matchsize < XSTRING_CHAR_LENGTH (bestmatch)) + || + /* If there is no exact match ignoring case, + prefer a match that does not change the case + of the input. */ + (((matchsize == cclen) + == + (matchsize == XSTRING_CHAR_LENGTH (bestmatch))) + /* If there is more than one exact match aside from + case, and one of them is exact including case, + prefer that one. */ + && 0 > scmp_1 (p2, XSTRING_DATA (user), + user_name_length, 0) + && 0 <= scmp_1 (p1, XSTRING_DATA (user), + user_name_length, 0))) + { + bestmatch = make_string (d_name, len); + } + } + + bestmatchsize = matchsize; + } + } + + unbind_to (speccount, Qnil); /* does endpwent() */ + + UNGCPRO; + + if (CONSP (uniq)) + Fsetcar (uniq, matchcount == 1 ? Qt : Qnil); + + if (all_flag || NILP (bestmatch)) + return bestmatch; + if (matchcount == 1 && bestmatchsize == user_name_length) + return Qt; + return Fsubstring (bestmatch, Qzero, make_int (bestmatchsize)); +} + + Lisp_Object make_directory_hash_table (CONST char *path) { ＠＠ -689,6 +845,8 ＠＠ DEFSUBR (Fdirectory_files); DEFSUBR (Ffile_name_completion); DEFSUBR (Ffile_name_all_completions); + DEFSUBR (Fuser_name_completion); + DEFSUBR (Fuser_name_all_completions); DEFSUBR (Ffile_attributes); } Index: lisp/minibuf.el =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs-20/lisp/minibuf.el,v retrieving revision 1.14 diff -u -r1.14 minibuf.el --- minibuf.el 1998/06/20 00:58:23 1.14 +++ minibuf.el 1998/07/21 07:39:11 ＠＠ -1688,14 +1688,34 ＠＠ ;; Not doing environment-variable completion hack (let* ((orig (if (equal string "") nil string)) (sstring (if orig (substitute-in-file-name string) string)) - (specdir (if orig (file-name-directory sstring) nil))) - (funcall completer - action - orig - sstring - specdir - (if specdir (expand-file-name specdir dir) dir) - (if orig (file-name-nondirectory sstring) string))) + (specdir (if orig (file-name-directory sstring) nil)) + (name (if orig (file-name-nondirectory sstring) string)) + (direct (if specdir (expand-file-name specdir dir) dir))) + ;; ~user completion + (if (string-match "^[~]" name) + (let ((user (substring name 1)) + (uniq (cons nil nil))) + (cond ((eq action 'lambda) + nil) + ((eq action 't) + ;; all completions + (mapcar #'(lambda (p) (concat "~" p)) + (user-name-all-completions user))) + (t;; 'nil + ;; complete + (let* ((val (user-name-completion user uniq))) + (cond ((stringp val) + (concat "~" val (if (car uniq) "/" ""))) + ((eq val t) + (concat name "/")) + (t nil)))))) + (funcall completer + action + orig + sstring + specdir + direct + name))) ;; An odd number of trailing $'s (let* ((start (match-beginning 3)) (env (substring string

1 participants
0 comments

[Success] XEmacs 21.0-beta1 "Aeolus" on Linux 26 years, 2 months

Kazuyoshi Furutaka

[i686-GNU-linux] Source : fresh tarballs Make : GNU Make version 3.76.1 CC : pgcc-2.90.27 980315 (egcs-1.0.2 release) OS : Linux 2.0.34 (RedHat 5.0) X : X11R6.3 (Accelerated-X ver.4.1 w/ XFree86 3.3.1) gdb : GDB 4.16 [alpha-GNU-linux] Source : fresh tarballs Make : GNU Make version 3.75 CC : egcs-2.90.29 980515 (egcs-1.0.3 release) OS : Linux 2.1.108 (StataboWare 1.4beta) X : X11R6.3 (XFree86 3.3.2) gdb : GDB 4.16 -- Kazuyoshi Furutaka furutaka(a)tokai.pnc.go.jp furutaka(a)jb3.so-net.ne.jp

1 participants
0 comments

[Success] (sort of) XEmacs-21.0-b50/pre7 on Digital UNIX 4.0D with CDE 26 years, 2 months

Kevin Oberman

Built on DU 4.0d with DEC C, GNU make, xpm, CDE, JPEG, GIFreader, zlib. It built pretty well using "-O3 -std". There were a couple of warning during the compile. Serious problem: I am still unable to install 21.0. It insists that I need tooltalk and don't have it. In fact, I do have it, but XEmacs is not seeing that. I still have been unable to convince XEmacs that it is there. Any clues appreciated! R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman(a)es.net Phone: +1 510 486-8634

1 participants
0 comments

[Success] XEmacs 21.0-b50/pre7 "Benadir" on Linux 26 years, 2 months

Kazuyoshi Furutaka

[i686-GNU-linux] Source : patched up from 21.0 beta48 :-) Patch : patch 2.4 Make : GNU Make version 3.76.1 CC : pgcc-2.90.27 980315 (egcs-1.0.2 release) OS : Linux 2.0.34 (RedHat 5.0) X : X11R6.3 (Accelerated-X ver.4.1 w/ XFree86 3.3.1) gdb : GDB 4.16 [alpha-GNU-linux] Source : patched up from 21.0 beta48 :-) Patch : Patch version 2.1 Make : GNU Make version 3.75 CC : egcs-2.90.29 980515 (egcs-1.0.3 release) OS : Linux 2.1.108 (StataboWare 1.4beta) X : X11R6.3 (XFree86 3.3.2) gdb : GDB 4.16 -- Kazuyoshi Furutaka furutaka(a)tokai.pnc.go.jp furutaka(a)jb3.so-net.ne.jp

1 participants
0 comments

asp-mode? 26 years, 2 months

Jin Choi

Does anyone know of or has anyone given any thought to an asp-mode, for Active Server Page scripting? Something that would perhaps be useful not only for Microsoft's ASP syntax, but also for other similar HTML templating languages. The problem is that in a lot of these systems, HTML is interspersed with code blocks delimited by some magic characters. Actually writing code in these systems is frustrating because emacs won't indent the language properly for you. I was thinking a system that would do some clever mode switching would be very useful for this sort of thing. It would have to hide the HTML when editing code somehow in order not to confuse the syntax parser for the language mode, and vice versa. What do y'all think, is this something that is feasible? What might be the best way to go about it, actually temporarily remove the non-code (or non-HTML) bits into another buffer, or can you just make them invisible somehow from the language mode parsers? -Jin

1 participants
0 comments

32 process limit on NT 26 years, 2 months

Gleb Arshinov

I use p4.el (Perforce VC mode), which executes call-process every time it opens a file to check whether the file is under source control. After a while, I get: Signaling: (file-error "Spawning child process" "Resource temporarily unavailable") call-process-internal("\\Perforce\\p4.exe" nil "*P4 Output*" nil "files" "c:\\src\\session\\src\\ssexplain.cpp") apply(call-process-internal "\\Perforce\\p4.exe" nil "*P4 Output*" nil ("files" "c:\\src\\session\\src\\ssexplain.cpp")) call-process("\\Perforce\\p4.exe" nil "*P4 Output*" nil "files" "c:\\src\\session\\src\\ssexplain.cpp") p4-is-vc() p4-check-mode() p4-find-file-hook() run-hooks(p4-find-file-hook) after-find-file(nil t) find-file-noselect("c:\\src\\session\\src\\ssexplain.cpp") find-file("c:\\src\\session\\src\\ssexplain.cpp" nil) call-interactively(find-file) I presume this is the 32 process maximum discusses in "ediff eats handles" thread. Are there any workarounds? Interestingly, the limit does not seem to apply to processes called from compile. Thanks, Gleb PS This is in 21.0 "Zhong Wei" Fabrice's binaries.

1 participants
0 comments

[Benadir] redisplay glitches 26 years, 2 months

Mark Borges

I'm not sure if this is a recent problem, or if cosmetic problems need be fixed before the final release. But, I'm able to produce redisplay "turds" in b50 by doing the following: (1) xrdb < /dev/null (2) xemacs -vanilla (3) \C-h a horizontal (4) \C-x \C-f /usr/local/bin # or any directory that gives > page of output (5) \C-x \C-b *Hyper # or any buffer with a horizontal scrollbar (6) \C-x \C-b *scratch* => the last dired line is still visible in the *scratch* buffer. -- -mb-

1 participants
0 comments

Re: [Failure] XEmacs-21.0-b50/pre7 on Digital UNIX 4.0D 26 years, 2 months

SL Baur

Kevin Oberman <oberman(a)es.net> writes in xemacs-beta(a)xemacs.org: >> ./configure > checking host system type... alphaev56-dec-osf4.0c > checking whether ln -s works... yes > checking the configuration name > Invalid configuration `alphaev56-dec-osf4.0c': machine `alphaev56-dec' not recognized > Unfortunately, while config.guess has been modified for this, > configure has not been modified to accept this. Did a patch get lost > here? Uh oh. Martin sent me an updated config.guess and I added it. I think other things being equal, it better be backed out of 21.0.

3 participants
2 comments

how to implement ~user completion [was: Re: Any (brave :-) mew users?] 26 years, 2 months

Greg Klanderman

>>>>> "Hrvoje" == Hrvoje Niksic <hniksic(a)srce.hr> writes: Hrvoje> >> >> Given the recent changes to expand-file-name I assumed we were >> trying to keep this kind of stuff out of the low level file >> primitives. Hrvoje> Hrvoje> Maybe, and maybe not. I still think file_name_completion is the right Hrvoje> place to change, but if you present arguments otherwise, I can stand Hrvoje> corrected. I've looked into this a bit more. The problems I see with changing file_name_completion are: - As things corrently stand, Ffile_name_completion is getting called with file="~userna" and dirname=default-directory. So you'd need to have file-name-completion return username completions any time the file name begins with "~", no matter what the directory. It seems highly bogus to ignore the directory and do username completion in this function. It might be OK to do it if the dirname were empty or nil. - This approach does not work for read-directory-name, as it does not call file-name-completion. So it seems to make more sense to hack read-file-name-internal-1 in minibuf.el, in the same place that environment variables are handled. This will work for read-directory-name too. So I'd add two primitives, user-name-completion and user-name-all-completions, and fix the code in minibuf.el to use them. greg

1 participants
0 comments

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

XEmacs-Beta July 1998