home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!sdd.hp.com!think.com!enterpoop.mit.edu!bloom-picayune.mit.edu!mintaka.lcs.mit.edu!ai-lab!life.ai.mit.edu!mly
- From: mly@ai.mit.edu (Richard Mlynarik)
- Newsgroups: alt.lucid-emacs.bug
- Subject: Environment variable completion
- Date: 21 Dec 92 23:45:22
- Organization: ITS Preservation Society
- Lines: 40
- Distribution: world
- Message-ID: <MLY.92Dec21234522@august.ai.mit.edu>
- References: <cbwc02D52fvw01@JUTS.ccc.amdahl.com>
- NNTP-Posting-Host: august.ai.mit.edu
- In-Reply-To: <cbwc02D52fvw01@JUTS.ccc.amdahl.com>
-
- From: tlp00@ras.amdahl.com (Tibor Polgar)
- Newsgroups: alt.lucid-emacs.help
- Subject: Environment variable completion
- Date: 22 Dec 92 00:06:56 GMT
-
- Anyone out there written an environment variable expansion to be
- used as part of file name completion. Basicly its something that
- calls "substitute-in-file-name" before it calls
- "minibuffer-complete-word". Yeh, i know i could write something
- but if someone already has.... For instance:
-
- Find file: $SRC<space>
-
- expands to:
-
- Find file: /usr/project/src/
-
- and the rest is same as normal...
-
- [...]
-
- WOW, i get to follow up my own posting... in looking at it further,
- seems that completion DOES "try" to substitute in an environment
- variable but there is a bug somewhere, all i ever get is a "/". If
- i enter an invalid environment variable then I get "bad format
- environment variable substitution" so something is trying.
-
-
- I've rewritten all the minibuffer code in Lisp and in the process
- fixed the longstanding bug in minibuffer-complete-word versus pathname
- environment variables (amongst a slew of other little problems.) I
- believe this code will be in the next Lucid Emacs release.
-
- At first I misunderstood the question you were asking and thought you
- were repeating the longstanding feature-request for completion of
- environment-variable names (eg if environment variables "el"
- and "es" are defined, then completing "$e" will offer two
- possibilities instead of complaining that "$e" isn't defined.)
- It wasn't very hard to implement this and so I did it. It may or may
- not be in Lucid 19.4 depending on Jamie's judgement/whims.
-