5581383962
Add surflet-path to options (needed by admin-surflets.scm)
2003-04-13 20:24:56 +00:00
4b08eac262
use with-lock instead of obtain-lock/release-lock
2003-04-13 18:27:43 +00:00
956fd7bbf1
Add non-thunk version to with-lock.
...
Note: To follow naming conventions of scsh, the old with-lock is
renamed to with-lock*. The syntactic sugar version is with-lock now.
2003-04-13 17:36:39 +00:00
d0782f63a2
Factor out with-locks from surflet-handler
2003-04-13 17:23:11 +00:00
213090a51d
+ use search-trees for timeout-queue
...
+ export session-adjust-timeout! for admin-surflet.scm
2003-04-01 13:17:45 +00:00
61c3a4c216
Introduced field lifetime to session
2003-03-31 10:56:28 +00:00
33f1a2f063
Two small fixes related to options.
2003-03-17 19:29:45 +00:00
b74ca11656
+ Make text displayed on session timeout an option to the handler.
...
+ Fixed various options related bugs.
2003-03-17 12:31:49 +00:00
b93e275415
Replaced implementation of surveillance thread: We have now only one
...
thread which checks every second for sessions to be deleted.
2003-03-17 12:09:26 +00:00
3de0a9c480
Added options for surflet-handler.
2003-03-17 10:17:55 +00:00
b1e0c7df54
Add GET-SESSION
2003-03-13 11:32:59 +00:00
ec341accf9
Add SESSION-ALIVE?
2003-03-10 16:37:22 +00:00
87a4165f94
check existence of sessions in some locking areas
2003-03-10 13:10:29 +00:00
64371c9941
Use the time-to-live argument in really-session-adjust-timeout!
2003-03-10 13:01:45 +00:00
909ed39ce2
Add optional arguments to ADJUST-TIMEOUT! and SESSION-ADJUST-TIMEOUT!
...
This argument defaults to OPTIONS-SESSION-LIFETIME and represents
the life-time in seconds for the session.
2003-03-10 12:57:44 +00:00
4748a882c3
Add a bang to SET-OPTIONS-SESSION-LIFETIME, SET-OPTIONS-CACHE-SURFLETS? and ADJUST-TIMEOUT
2003-03-09 18:49:09 +00:00
3f1542466c
Rename misleading {SET,GET}-SURFLET-DATA to {SET,GET}-SESSION-DATA, as the data is specific to a session and not to a surflet.
2003-03-09 18:03:15 +00:00
f4baad5ab1
changed let&if to cond&=> in GET-SURFLET-RT-STRUCTURE
2003-03-03 10:29:18 +00:00
71650fbb41
catch every condition only for load-config et al.; otherwise catch only errors.
2003-02-19 09:43:29 +00:00
21f62d5d8e
+ REQUEST object extended for SUrflets by INPUT-PORT
...
+ SURFLET-REQUESTs used in SUrflets
+ Use MAKE-INPUT-RESPONSE to generate SUrflet responses
2003-02-17 10:09:24 +00:00
e32b78badb
Add comments about thread safety.
2003-01-25 16:33:50 +00:00
f2872877ad
Minor change in comments.
2003-01-25 16:30:09 +00:00
e3f216aee7
{S,s}urflet-->SUrflet in strings
2003-01-25 16:28:16 +00:00
afada4efb2
Look for RESUME-URL-IDS only in the file-name, not in the path.
2003-01-25 16:22:37 +00:00
e934340fb5
uncurry LOAD-SURFLET in GET-SURFLET-RT-STRUCTURE
2003-01-25 16:17:33 +00:00
26fa1ca033
correct bogus call to MAKE-DEFAULT-options
2003-01-25 16:13:42 +00:00
9aea76ee83
remove RETURN-CONTINUATION from INSTANCE; it's not used anyway
2003-01-25 16:11:30 +00:00
100ffcc010
remove SURFLET-PREFIX; it's not used anyway
2003-01-25 16:10:19 +00:00
3d3655960b
+ Outhouse some components of surflet-handler.
...
+ Textual restructuring of surflet-handler.
+ Comments added.
2003-01-25 16:09:03 +00:00
220ac99930
Add SURFLET-RESPONSE->STRING for debug purpose.
2003-01-25 13:30:45 +00:00
3f3bf045de
Add SEND-ERROR, that sends an error.
2003-01-24 16:02:29 +00:00
6358463b5a
+ Change interface of SEND/SUSPEND, SEND/FINISH, SEND:
...
They expect SURFLET-RESPONSE objects now (not HTTP-RESPONSE).
+ SEND-HTML/SUSPEND, SEND-HTML/FINISH, SEND-HTML respect this.
2003-01-24 15:23:51 +00:00
3c73d6a37f
Force SEND/... calls in SEND/FINISH and SEND, too.
...
Rename STRING-OUT-PORT to BUFFER.
2003-01-21 10:36:49 +00:00
61fc543af2
Rename SERVLET --> SURFLET
2003-01-19 16:57:27 +00:00
422a1db09f
Add note about danger of killing / deleting current continuation.
...
Add current session's id and current session's continuation's id in
this message.
2003-01-16 12:53:10 +00:00
62c597e7ec
Rename URI-PATH-LIST->PATH to URI-PATH->URI and SPLIT-URI-PATH to
...
SPLIT-URI.
2003-01-14 15:01:21 +00:00
da98e19193
Force the result of the servlet to be printed into a string-port.
...
The servlet programmer is (nearly) unaware of this.
This removes the problem that arises if send/suspend is called in HTTPD
while transmitting data to the browser.
2003-01-14 11:27:42 +00:00
86b0639a7f
Rename MAKE-HTTP-ERROR-RESPONSE -> MAKE-ERROR-RESPONSE to be
...
consistent with MAKE-RESPONSE.
2003-01-10 09:52:35 +00:00
0754b74963
Replace integer HTTP status codes by finite record type instances.
2003-01-09 15:05:30 +00:00
4606552c2b
rename INSTANCE to SESSION and SESSION to INSTANCE
2002-12-07 22:26:40 +00:00
ffa8ad13ae
remove unused SERVLET-PREFIX option.
2002-12-02 08:42:37 +00:00
641b8129ba
Renamed REQUEST:xxx to REQUEST-xxx.
2002-11-29 14:49:22 +00:00
8e22009ca1
Add another error handler in the reset from to catch errors occuring
...
while running the servlet (perhaps with restored continuations).
2002-11-09 18:25:20 +00:00
d4ea641e14
comment added
2002-11-05 22:20:47 +00:00
47a60a26d3
Redesigning error catching machinery.
...
Now no malicious servlet should be able to let the server crash.
2002-11-05 10:21:15 +00:00
2800b21fea
Browser caching may be a reason for wrong servlet request. Tell them
...
about it.
2002-11-03 14:37:53 +00:00
26683bd822
* Move error handling of the servlets to the right point, i.e. after
...
the reset.
* Remove redirect 'feature'. It is not well tought out.
2002-10-26 15:35:20 +00:00
d02d0e19d2
use DEFINE-RECORD-TYPE instead of DEFINE-RECORD
2002-10-21 08:25:58 +00:00
900f552ee1
Wrap error-handler around calls to LAUNCH-NEW-INSTANCE and RESUME-URL.
...
This will catch most errors, but will fail on some, e.g. procedure
called with wrong number of arguments. As far as I see, the problem is
the underlying error reporting system of Scsh/Scheme48. Maybe I'm
wrong.
2002-10-09 15:11:55 +00:00
12b2b96582
remove some superfluous comments/code
2002-10-04 15:56:58 +00:00
interp