Re: [fossil-users] 'fossil sqlite3' db init issues
On Sun, Sep 1, 2013 at 11:43 PM, David Given d...@cowlark.com wrote: What I'm actually trying to do is to add some functionality which is accessed via a couple of extra SQL functions. What's the preferred way to do this? Most of the custom funcs are registered in db.c:db_open(), but apparently that one isn't used by the sqlite command: [stephan@host:~/cvs/fossil/libfossil]$ f sqlite SQLite version 3.8.1 2013-08-30 06:20:23 Enter .help for instructions Enter SQL statements terminated with a ; sqlite select now(); Error: no such function: now It might help to split those func registrations into a separate function and call it from both db_open() and the sqlite command setup. -- - stephan beal http://wanderinghorse.net/home/stephan/ http://gplus.to/sgbeal ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] 'fossil sqlite3' db init issues
On 29/08/13 23:19, David Given wrote: [...] I enclose a patch which I think fixes it; please comment... A little bit more investigation shows that there's actually several ways in which the code initialises the database connection, and they all very slightly, and I'm now a bit confused. It looks like the extra functions described in sqlcmd_autoinit() only get registered when using 'fossil sqlite3' (subject to the bug mentioned earlier). Other db connections, such as the ones used when actually doing work, get re_add_sql_func() called on them to register 'regexp()' but don't get the other functions. What I'm actually trying to do is to add some functionality which is accessed via a couple of extra SQL functions. What's the preferred way to do this? -- ┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ─ http://www.cowlark.com ─ │ │ Ripley's Law: Never go further for the cat than the cat would go for │ you. --- Vexxarr Bleen (trans. Hunter Cressall) signature.asc Description: OpenPGP digital signature ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
[fossil-users] 'fossil sqlite3' db init issues
I've found a minor bug in the way the patched sqlite3 shell handles
opening the db. In particular, this code from shell.c:
if( data.zDbFilename==0 ){
#ifndef SQLITE_OMIT_MEMORYDB
data.zDbFilename = :memory:;
#else
fprintf(stderr,%s: Error: no database filename specified\n, Argv0);
return 1;
#endif
/* Begin Fossil Patch */
{
extern void fossil_open(const char **);
fossil_open(data.zDbFilename);
}
/* End Fossil Patch */
}
if( access(data.zDbFilename, 0)==0 ){
open_db(data);
}
This is manifesting in a particularly confusing way. Specifically:
(a) 'fossil sqlite3' calls sqlcmd_autoinit
(b) 'fossil sqlite3 foo.fossil' does not
What's happening with (a), I think, is that fossil_open is opening
:memory:, which registers the autoinit handler, and then further down
the database is reopened with open_db which calls the handler.
However, with (b), :memory: is not being opened at all, therefore the
handler is not being registered, so open_db doesn't call it.
I enclose a patch which I think fixes it; please comment...
--
┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ─ http://www.cowlark.com ─
│
│ Ripley's Law: Never go further for the cat than the cat would go for
│ you. --- Vexxarr Bleen (trans. Hunter Cressall)
--- src/shell.c
+++ src/shell.c
@@ -3230,27 +3230,25 @@
data.zDbFilename = :memory:;
#else
fprintf(stderr,%s: Error: no database filename specified\n, Argv0);
return 1;
#endif
-/* Begin Fossil Patch */
-{
- extern void fossil_open(const char **);
- fossil_open(data.zDbFilename);
-}
-/* End Fossil Patch */
}
+
data.out = stdout;
/* Go ahead and open the database file if it already exists. If the
** file does not exist, delay opening it. This prevents empty database
** files from being created if a user mistypes the database name argument
** to the sqlite command-line tool.
*/
- if( access(data.zDbFilename, 0)==0 ){
-open_db(data);
+ /* Begin Fossil Patch */
+ {
+extern void fossil_open(const char **);
+fossil_open(data.zDbFilename);
}
+ /* End Fossil Patch */
/* Process the initialization file if there is one. If no -init option
** is given on the command line, look for a file named ~/.sqliterc and
** try to process it.
*/
signature.asc
Description: OpenPGP digital signature
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3
Od: Edward Berner e...@bernerfam.com Komu: fossil-users@lists.fossil-scm.org Datum: 31.12.2012 08:53 Předmět: Re: [fossil-users] fossil sqlite3 On 12/30/2012 10:43 PM, Michael Richter wrote: Is there any way to execute SQL statements from the command line using fossil sqlite3? The docs for this http://www.fossil-scm.org/xfer/help?cmd=sqlite3 are a bit skimpy (to say the least). Like what are the /?OPTIONS?/ mentioned, precisely? What I'm specifically trying to accomplish is to extract the project ID from the repository in a script file. If there's another way to do this I'm happy to use that instead, of course. I think fossil info will get what you're looking for. eg: $ ../fossil info -R ~/fossil/test.fossil project-name: unnamed project-code: 07660919180b651835b3a51776a58007ca9fb6f9 fossil info calls it project-code but it seems to be the same thing that fossil new and fossil clone call project-id. using fossil sqlite3: echo select value as 'project ID' from config where name = 'project-code'; statement.sql fossil sqlite3 -R .\myfossilclone.fossil .\statement.sql (myfossilclone.fossil is my clone of fossil repository) Peter ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3
On 31 December 2012 15:53, Edward Berner e...@bernerfam.com wrote: fossil info calls it project-code but it seems to be the same thing that fossil new and fossil clone call project-id. Waitwhat? My version of Fossil (This is fossil version 1.25 [558a17a686] 2012-12-22 13:48:31 UTC) doesn't show anything about project-id for fossil new/clone. -- Perhaps people don't believe this, but throughout all of the discussions of entering China our focus has really been what's best for the Chinese people. It's not been about our revenue or profit or whatnot. --Sergey Brin, demonstrating the emptiness of the don't be evil mantra. ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3
On 12/31/2012 12:24 AM, Michael Richter wrote: On 31 December 2012 15:53, Edward Berner e...@bernerfam.com mailto:e...@bernerfam.com wrote: fossil info calls it project-code but it seems to be the same thing that fossil new and fossil clone call project-id. Waitwhat? My version of Fossil (This is fossil version 1.25 [558a17a686] 2012-12-22 13:48:31 UTC) doesn't show anything about project-id for fossil new/clone. What do you get when you create a test repository? It should, I think, display the project-id, etc., after creating the repository. eg: $ ./fossil new test.fossil project-id: 2d7cade36dce2af94df648e178d588e5a3b00a14 server-id: af94c0a2462bc18103a19c20fd0b3918c3f9a2a3 admin-user: erb (initial password is 811b05) -- Edward ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3
On 31 December 2012 17:27, Edward Berner e...@bernerfam.com wrote: Waitwhat? My version of Fossil (This is fossil version 1.25 [558a17a686] 2012-12-22 13:48:31 UTC) doesn't show anything about project-id for fossil new/clone. What do you get when you create a test repository? It should, I think, display the project-id, etc., after creating the repository. eg: $ ./fossil new test.fossil project-id: 2d7cade36dce2af94df648e178d588**e5a3b00a14 server-id: af94c0a2462bc18103a19c20fd0b39**18c3f9a2a3 admin-user: erb (initial password is 811b05) Ah. I misunderstood. I thought you meant there was a command line option related to the project ID. -- Perhaps people don't believe this, but throughout all of the discussions of entering China our focus has really been what's best for the Chinese people. It's not been about our revenue or profit or whatnot. --Sergey Brin, demonstrating the emptiness of the don't be evil mantra. ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3
there might have been email overlap. fossil info also spits out the project-id. That's what Petr was saying I believe. Date: Mon, 31 Dec 2012 17:51:31 +0800 From: ttmrich...@gmail.com To: fossil-users@lists.fossil-scm.org Subject: Re: [fossil-users] fossil sqlite3 On 31 December 2012 17:27, Edward Berner e...@bernerfam.com wrote: Waitwhat? My version of Fossil (This is fossil version 1.25 [558a17a686] 2012-12-22 13:48:31 UTC) doesn't show anything about project-id for fossil new/clone. What do you get when you create a test repository? It should, I think, display the project-id, etc., after creating the repository. eg: $ ./fossil new test.fossil project-id: 2d7cade36dce2af94df648e178d588e5a3b00a14 server-id: af94c0a2462bc18103a19c20fd0b3918c3f9a2a3 admin-user: erb (initial password is 811b05) Ah. I misunderstood. I thought you meant there was a command line option related to the project ID. -- Perhaps people don't believe this, but throughout all of the discussions of entering China our focus has really been what's best for the Chinese people. It's not been about our revenue or profit or whatnot. --Sergey Brin, demonstrating the emptiness of the don't be evil mantra. ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
[fossil-users] fossil sqlite3
Is there any way to execute SQL statements from the command line using fossil sqlite3? The docs for thishttp://www.fossil-scm.org/xfer/help?cmd=sqlite3are a bit skimpy (to say the least). Like what are the *?OPTIONS?* mentioned, precisely? What I'm specifically trying to accomplish is to extract the project ID from the repository in a script file. If there's another way to do this I'm happy to use that instead, of course. -- Perhaps people don't believe this, but throughout all of the discussions of entering China our focus has really been what's best for the Chinese people. It's not been about our revenue or profit or whatnot. --Sergey Brin, demonstrating the emptiness of the don't be evil mantra. ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3
On 12/30/2012 10:43 PM, Michael Richter wrote: Is there any way to execute SQL statements from the command line using fossil sqlite3? The docs for this http://www.fossil-scm.org/xfer/help?cmd=sqlite3 are a bit skimpy (to say the least). Like what are the /?OPTIONS?/ mentioned, precisely? What I'm specifically trying to accomplish is to extract the project ID from the repository in a script file. If there's another way to do this I'm happy to use that instead, of course. I think fossil info will get what you're looking for. eg: $ ../fossil info -R ~/fossil/test.fossil project-name: unnamed project-code: 07660919180b651835b3a51776a58007ca9fb6f9 fossil info calls it project-code but it seems to be the same thing that fossil new and fossil clone call project-id. -- Edward ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3 shell vs regular sqlite3 shell
On Sat, Jun 11, 2011 at 7:01 AM, Joerg Sonnenberger jo...@britannica.bec.de wrote: On Fri, Jun 10, 2011 at 10:09:52PM -0400, Martin Gagnon wrote: When use the fossil sqlite3 shell, the arrow key doesn't work and I get something like ^[[A caracter appearing on screen when I try the up arrow to access command history. (similar for left and right to move cursor) It doesn't provide the readline/libedit functionality. It can if you recompile with the right command-line options. I think: -DHAVE_READLINE ... -lreadline -lncurses The precompiled binaries do not do this because the necessarily libraries are not available on every system. Joerg ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users -- D. Richard Hipp d...@sqlite.org ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
Re: [fossil-users] fossil sqlite3 shell vs regular sqlite3 shell
Le 2011-06-11 à 08:01, Richard Hipp d...@sqlite.org a écrit : On Sat, Jun 11, 2011 at 7:01 AM, Joerg Sonnenberger jo...@britannica.bec.de wrote: On Fri, Jun 10, 2011 at 10:09:52PM -0400, Martin Gagnon wrote: When use the fossil sqlite3 shell, the arrow key doesn't work and I get something like ^[[A caracter appearing on screen when I try the up arrow to access command history. (similar for left and right to move cursor) It doesn't provide the readline/libedit functionality. It can if you recompile with the right command-line options. I think: -DHAVE_READLINE ... -lreadline -lncurses Thanks, I'll try this, I already compile it myself anyway. The precompiled binaries do not do this because the necessarily libraries are not available on every system. That's logical... -- Martin___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
[fossil-users] fossil sqlite3 shell vs regular sqlite3 shell
When use the fossil sqlite3 shell, the arrow key doesn't work and I get something like ^[[A caracter appearing on screen when I try the up arrow to access command history. (similar for left and right to move cursor) I'm get used to the regular sqlite3 shell because our product use sqlite3 database. With the regular sqlite3 shell, I can use the up and down arrow key to get through command history, and left and right arrow to move cursor on command line without any problem. It seems there's something different on the fossil sqlite3 shell... I notice this problem on Mac OSX (regular Terminal.App) and Ubuntu Linux. For some reason, I tried on windows and it work (I don't really use windows; but it might be useful to know..) Someone else encounter the same problem ? Thanks -- Martin ___ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
