Re: [fpc-pascal] postgres3 async notify mem leak
Brad Campbell wrote: G'day all, I'm using Postgres 8.1 and have async notifies working nicely based on select(). The problem is each time I get a notify I leak memory. I've commented out the PQfreemem call and find exactly the same size leak (tested over 10,000 notifies). I assume from that I'm doing something very wrong but I just can't see where. I've been unable to find much in the way of examples out there so I assume I'm doing something wrong. And I'm not, the leak is in the LCL TLabel component. Sorry for the noise. Brad -- "Human beings, who are almost unique in having the ability to learn from the experience of others, are also remarkable for their apparent disinclination to do so." -- Douglas Adams ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] html link extractor
Andrew Haines wrote: > Is there a unit somewhere that can extract links from html pages? I want > to be able to recursively add pages to a chm archive. > I think this from l505 will work: http://lists.freepascal.org/lists/fpc-announce/2007-February/000536.html but are there any units that come default with fpc that can do this? Andrew ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] html link extractor
Is there a unit somewhere that can extract links from html pages? I want to be able to recursively add pages to a chm archive. Thanks, Andrew Haines ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] fpCanvas and FPColor property
On Sat, 30 Jun 2007, Graeme Geldenhuys wrote: > Hi, > > TFPCustomFont has a property as follows: > property FPColor : TFPColor ... > > TFPColor is a record compared to Lazarus and Delphi's Color property > being TColor with is a LongWord. > If I persist a Form containing a Label to a .dfm or .lfm or whatever > extension I want to use. Will the internal TWriter of TComponent be > able to write a TFPColor property to file? Or will I have to do some > magic in TWriter or implement a Color: TColor property in my > components? > > If the later is needed, why doesn't TFPCustomCanvas implement the > FPColor property like Lazarus and Delphi using TColor (LongWord). Because TFPColor is a 64-bit structure. When it was implemented, streaming was not a consideration. Michael. ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Documentation for sqldb - further adventures
On Fri, 2007-06-29 at 18:35 +1000, John & Marg Sunderland wrote:
> Following earlier advice, I constructed a simple query ("select * from
> organiser.durn_type"), set readonly to false and tried to modify some
> data and commit it.
>
> (At this point I should note that I am doing this from Lazarus 0.9.22
> with the supplied fpc, 2.0.4 I think, and I am connecting to a
> postgresql database v 8.1.4 on the same PC, all on Window$ XP SP2.)
>
> I have no trouble opening the query to display the data in a grid
> (readonly or not) but when I made some changes, ApplyUpdates gave a
> database error: "syntax error at end of input at character 127"
>
> I first tried to trace the execution (in Lazarus IDE), but was unable to
> trace into the db or sqldb modules - "step into" ran straight to the
> error, and breakpoints reverted to disabled once the program started to
> run. Is there any reason why this should be so ? Or should I ask in
> the Lazarus discussion areas ?
>
> I then turned on lots of logging in Postgresql, and found the error to
> be in a statement:
>
> 2007-06-29 14:50:16 LOG: statement: EXECUTE [PREPARE:
> prepare prepst0 as select * from organiser.durn_type]
> 2007-06-29 14:50:16 LOG: statement: prepare prepst1 as
> select ic.relname as indexname, tc.relname as tablename, ia.attname,
> i.indisprimary, i.indisunique
> from pg_attribute ta, pg_attribute ia, pg_class tc, pg_class ic,
> pg_index i
> where( (i.indrelid = tc.oid)
> and (ta.attrelid = tc.oid)
> and (ia.attrelid = i.indexrelid)
> and (ic.oid = i.indexrelid)
> and (ta.attnum = i.indkey[ia.attnum-1])
> and (upper(tc.relname)='ORGANISER.DURN_TYPE') )
> order by ic.relname;
This is what is done: to create the update/insert/delete it has to
detect what the primary-key is. For that purpose it executes the second
query.
> I then ran this in an SQL window from PGAdmin, and it ran, but returned
> no rows. I then removed the "ORGANISER." prefix on the table, and it
> ran and returned some rows. So, I removed the "organiser." schema
> prefix from the query in Lazarus, and it worked ! YAY! (The database
> connection is logged in under that schema) However, is it definite that
> I can't specify a schema prefix for a table, and should I document it as
> a restriction, or does this rate as a bug ? I can see from looking at
> the pg_class table that the above query is never going to work - relname
> is a simple object name, and the schema owner is an oid in another column.
To fix this the table-name has to be parsed, to get the schema-prefix.
That's not an easy fix.
You could solve this problem by setting the update/delete/insert queries
yourself. (and set parsesql to false) Or you could set
UsePrimaryKeyAsKey to false, and set updatemode to UpWhereAll, or make
one of the fields a key-field.
> Finally, as this table is a set of fairly static codes, and therefore
> would not expect a lot of activity, I tried constructing the editing
> form as a dialog box, only applying updates and committing when the OK
> button is used to close the form, and canceling the updates otherwise.
> Again, does this make sense ? It seems to work, except that: When I
> only change one line and close (committing) it is fine, but if I change
> several lines, I get a string of errors "EVariantTypeCastError - could
> not convert type (String) to type (OleStr). " It still works - I
> presume the errors must be "handled", as I do not see them if I run from
> outside the IDE, and the changes are saved.
This is normal. The exception is handled, but the debugger used by
lazarus, can't detect this. But check that your changes are really
saved.
Joost
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] How to analyze a core dump?
En/na Luca Olivetti ha escrit: Maybe it's not a good idea to mix c multithreaded libraries and pascal code? Any special unit I should use? (I already tried cmem and it made no difference). If I cannot solve it I think I'll have to write a small backend program in c that communicates with pascal either through stdin/stdout redirection or with a socket. Well, I actually did that and it works pretty well: I hacked the existing linphonec console program to output status codes on stdout, I spawn it using a TProcess to read its input and sending it commands. Everything else in my program is pretty much the same (where before I called a linphone routine, now I write to the TProcess input, where before I gathered status information from a callback[*] now I read from the TProcess output), so I think there's no glaring error in the logic of my program, just something "strange" mixing pascal and c code (note that I'm still using libusb with no apparent issues). I'd really like to debug the original problem, since the current workaround is too hackish to be published, but, alas, every effort to debug it so far has failed. [*]note that the callbacks were not called at random, but only during a specific call to the linphone library, so there are no synchronization differences between the library calls and the TProcess polling. Bye -- Luca ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] postgres3 async notify mem leak
G'day all, I'm using Postgres 8.1 and have async notifies working nicely based on select(). The problem is each time I get a notify I leak memory. I've commented out the PQfreemem call and find exactly the same size leak (tested over 10,000 notifies). I assume from that I'm doing something very wrong but I just can't see where. I've been unable to find much in the way of examples out there so I assume I'm doing something wrong. Aside from the memory leak it all works very, very well. (I'm using this in Lazarus but not using any of the lcl or fcl database components but the postgres3 unit directly as none of those seem to support PQsocket) procedure chk_notify; var Notify : PpgNotify; begin Notify := PQnotifies(conn); while Notify <> Nil do Begin PQfreemem(Notify); Inc(queue) Notify := PQnotifies(conn); End; end; Anybody have any ideas? Regards, Brad -- "Human beings, who are almost unique in having the ability to learn from the experience of others, are also remarkable for their apparent disinclination to do so." -- Douglas Adams ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] fpCanvas and FPColor property
Hi, TFPCustomFont has a property as follows: property FPColor : TFPColor ... TFPColor is a record compared to Lazarus and Delphi's Color property being TColor with is a LongWord. If I persist a Form containing a Label to a .dfm or .lfm or whatever extension I want to use. Will the internal TWriter of TComponent be able to write a TFPColor property to file? Or will I have to do some magic in TWriter or implement a Color: TColor property in my components? If the later is needed, why doesn't TFPCustomCanvas implement the FPColor property like Lazarus and Delphi using TColor (LongWord). Regards, - Graeme - ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] fpcanvas and double buffering support
Hi, Does TFPCustomCanvas have any mechanism to help with double buffering support? What does the ManageResources method do in TFPCustomCanvas? Graeme. ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Documentation for sqldb - further adventures
Following earlier advice, I constructed a simple query ("select * from
organiser.durn_type"), set readonly to false and tried to modify some
data and commit it.
(At this point I should note that I am doing this from Lazarus 0.9.22
with the supplied fpc, 2.0.4 I think, and I am connecting to a
postgresql database v 8.1.4 on the same PC, all on Window$ XP SP2.)
I have no trouble opening the query to display the data in a grid
(readonly or not) but when I made some changes, ApplyUpdates gave a
database error: "syntax error at end of input at character 127"
I first tried to trace the execution (in Lazarus IDE), but was unable to
trace into the db or sqldb modules - "step into" ran straight to the
error, and breakpoints reverted to disabled once the program started to
run. Is there any reason why this should be so ? Or should I ask in
the Lazarus discussion areas ?
I then turned on lots of logging in Postgresql, and found the error to
be in a statement:
2007-06-29 14:50:16 LOG: statement: EXECUTE [PREPARE:
prepare prepst0 as select * from organiser.durn_type]
2007-06-29 14:50:16 LOG: statement: prepare prepst1 as
select ic.relname as indexname, tc.relname as tablename, ia.attname,
i.indisprimary, i.indisunique
from pg_attribute ta, pg_attribute ia, pg_class tc, pg_class ic,
pg_index i
where( (i.indrelid = tc.oid)
and (ta.attrelid = tc.oid)
and (ia.attrelid = i.indexrelid)
and (ic.oid = i.indexrelid)
and (ta.attnum = i.indkey[ia.attnum-1])
and (upper(tc.relname)='ORGANISER.DURN_TYPE') )
order by ic.relname;
I then ran this in an SQL window from PGAdmin, and it ran, but returned
no rows. I then removed the "ORGANISER." prefix on the table, and it
ran and returned some rows. So, I removed the "organiser." schema
prefix from the query in Lazarus, and it worked ! YAY! (The database
connection is logged in under that schema) However, is it definite that
I can't specify a schema prefix for a table, and should I document it as
a restriction, or does this rate as a bug ? I can see from looking at
the pg_class table that the above query is never going to work - relname
is a simple object name, and the schema owner is an oid in another column.
Finally, as this table is a set of fairly static codes, and therefore
would not expect a lot of activity, I tried constructing the editing
form as a dialog box, only applying updates and committing when the OK
button is used to close the form, and canceling the updates otherwise.
Again, does this make sense ? It seems to work, except that: When I
only change one line and close (committing) it is fine, but if I change
several lines, I get a string of errors "EVariantTypeCastError - could
not convert type (String) to type (OleStr). " It still works - I
presume the errors must be "handled", as I do not see them if I run from
outside the IDE, and the changes are saved.
thanks for any help,
John
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
