RE: Too many network errors for this server, skipped

[Briggs, Gary]

I get this sometimes, too
It's probably because a while ago there would have been a load of errors
from, say, a particular page. It the ignores everything else that comes
under the server url.

Gary (-;

> -Original Message-
> From: Caffeinate The World [SMTP:[EMAIL PROTECTED]]
> Sent: Tuesday, March 13, 2001 5:31 AM
> To:   [EMAIL PROTECTED]
> Subject:  Too many network errors for this server, skipped
> 
> after indexing for about 24 hours, i start seeing this error below. but
> when i check the url on a browser on the same network, i can load the
> page fine. i've seen this in the previous versions but i've not been
> able to narrowed it down. upon restarting indexer, things are back to
> normal again and the same url get indexed just fine.
> 
> some relevant settings:
> ...
> Period 180d
> #Tag 
> #Category FFAABBCCDD
> MaxHops 56
> MaxNetErrors 32
> ReadTimeOut 120s
> DocTimeOut 120s
> NetErrorDelayTime 1d
> Robots yes
> Clones yes
> BodyWeight 2
> ...
> 
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0662.html
> [1] Too many network errors for this server, skipped
> [1] http://www.marshallschool.org/school/cours/course/handerson/top.htm
> [1] http://www.ci.east-grand-forks.mn.us/water/gsoc.htm
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0663.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0664.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0665.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0666.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0667.html
> [1] Too many network errors for this server, skipped
> 
> __
> Do You Yahoo!?
> Yahoo! Auctions - Buy the things you want at great prices.
> http://auctions.yahoo.com/
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
> 
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Php and phrase search

[Sergey Kartashoff]

Hi!

Monday, March 12, 2001, 8:54:10 PM, you wrote:


F> Maybe it's a stupid question, but is the PHP frontend supporting the phrase
F> search mode ?
F> I've read a message in the archive about it, but it was not clear for me
F> (was about the mnogo php module )

F> If it's possible, how does it works ? A simple "Phrase yes" in the template ?

New PHP frontend with all search.cgi features will be written after
release of php-4.0.5.

-- 
Regards, Sergey aka gluke.


___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Too many network errors for this server, skipped

[Smileyq]

I've been having the same exact problem when indexing a ftp server that
I own. If you fine out anything please let me know. Thanks.

Caffeinate The World wrote:
> 
> after indexing for about 24 hours, i start seeing this error below. but
> when i check the url on a browser on the same network, i can load the
> page fine. i've seen this in the previous versions but i've not been
> able to narrowed it down. upon restarting indexer, things are back to
> normal again and the same url get indexed just fine.
> 
> some relevant settings:
> ...
> Period 180d
> #Tag 
> #Category FFAABBCCDD
> MaxHops 56
> MaxNetErrors 32
> ReadTimeOut 120s
> DocTimeOut 120s
> NetErrorDelayTime 1d
> Robots yes
> Clones yes
> BodyWeight 2
> ...
> 
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0662.html
> [1] Too many network errors for this server, skipped
> [1] http://www.marshallschool.org/school/cours/course/handerson/top.htm
> [1] http://www.ci.east-grand-forks.mn.us/water/gsoc.htm
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0663.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0664.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0665.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0666.html
> [1] Too many network errors for this server, skipped
> [1] http://www.e-democracy.org/mn-politics-archive/9802/0667.html
> [1] Too many network errors for this server, skipped
> 
> __
> Do You Yahoo!?
> Yahoo! Auctions - Buy the things you want at great prices.
> http://auctions.yahoo.com/
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Webboard: invoke indexer via HTML (php)

[Caffeinate The World]

if you are running it from a cron job, you can specify the user you
want to run as.

--- Heiko <[EMAIL PROTECTED]> wrote:
> Author: Heiko
> Email: [EMAIL PROTECTED]
> Message:
> Hi,
> 
> I am trying to create a little admin tool for indexing jobs which are
> outside the regular once-a-week crone job. Here my test code:
> ==
> $testcommand = "/path.../indexer -a -u
> http://www.aufbruch.com/newfile.html";
> 
> // $testcommand = "ls -lias";
> 
> $output = array();
> 
> $lastline = exec($testcommand, $output);
> 
> while(list($number,$line) = each($output)){
>   print "$number: $line";
> }
> 
> Nice idea, however, while the commented-out "ls -lias" works fine, 
> the invoking of the indexer does not.
> 
> I know that PHP programs run as user 'nobody', not as the UNIX user. 
> My guess is that the indexer program is being started but has not
> permissions to read or write.  
> 
> But, I am not able to fix it.
> 
> Any idea
> 
> Heiko
> 
> 
> Reply: 
> 
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
> 


__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Too many network errors for this server, skipped

[Caffeinate The World]

after indexing for about 24 hours, i start seeing this error below. but
when i check the url on a browser on the same network, i can load the
page fine. i've seen this in the previous versions but i've not been
able to narrowed it down. upon restarting indexer, things are back to
normal again and the same url get indexed just fine.

some relevant settings:
...
Period 180d
#Tag 
#Category FFAABBCCDD
MaxHops 56
MaxNetErrors 32
ReadTimeOut 120s
DocTimeOut 120s
NetErrorDelayTime 1d
Robots yes
Clones yes
BodyWeight 2
...

[1] Too many network errors for this server, skipped
[1] http://www.e-democracy.org/mn-politics-archive/9802/0662.html
[1] Too many network errors for this server, skipped
[1] http://www.marshallschool.org/school/cours/course/handerson/top.htm
[1] http://www.ci.east-grand-forks.mn.us/water/gsoc.htm
[1] http://www.e-democracy.org/mn-politics-archive/9802/0663.html
[1] Too many network errors for this server, skipped
[1] http://www.e-democracy.org/mn-politics-archive/9802/0664.html
[1] Too many network errors for this server, skipped
[1] http://www.e-democracy.org/mn-politics-archive/9802/0665.html
[1] Too many network errors for this server, skipped
[1] http://www.e-democracy.org/mn-politics-archive/9802/0666.html
[1] Too many network errors for this server, skipped
[1] http://www.e-democracy.org/mn-politics-archive/9802/0667.html
[1] Too many network errors for this server, skipped

__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: invoke indexer via HTML (php)

[Heiko]

Author: Heiko
Email: [EMAIL PROTECTED]
Message:
Hi,

I am trying to create a little admin tool for indexing jobs which are outside the 
regular once-a-week crone job. Here my test code:
==
$testcommand = "/path.../indexer -a -u http://www.aufbruch.com/newfile.html";

// $testcommand = "ls -lias";

$output = array();

$lastline = exec($testcommand, $output);

while(list($number,$line) = each($output)){
print "$number: $line";
}

Nice idea, however, while the commented-out "ls -lias" works fine,  the invoking of 
the indexer does not.

I know that PHP programs run as user 'nobody', not as the UNIX user.  My guess is that 
the indexer program is being started but has not permissions to read or write.  

But, I am not able to fix it.

Any idea

Heiko


Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: DB.robots ?

[Zenon Panoussis]

Author: Zenon Panoussis
Email: [EMAIL PROTECTED]
Message:

v 3.1.12, MySQL, DB created with create.txt only:

Indexer[19721]: indexer from mnogosearch-3.1.12/MySQL started with 
'/usr/local/mn3112-1/etc/indexer.conf'
Indexer[19721]: [1] Error: '#1146: Table 'cachev5.robots' doesn't exist'

Where does this table come from? 

Z


Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: SQL tables with cache mode

[Zenon Panoussis]

Author: Zenon Panoussis
Email: [EMAIL PROTECTED]
Message:

Just checking: in cache mode it doesn't make any difference 
if you use single, multi or CRC-multi tables in the database, 
because all what goes into it are the urls and they go in one 
single table anyway; is that correct? 

Z



Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: Limiting the Number of Documents

[Benjamin Benson]

Author: Benjamin Benson
Email: [EMAIL PROTECTED]
Message:
Is this just a temporary feature, or will this continue to be supported?  I think this 
is a really important feature.


Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: What test error mean with the perl frontend?

[Jason]

Author: Jason
Email: 
Message:


 I do not believe the frontend by Dubun Guillaume works with MySql. I could be 
wrong, but I just compiled it and ran up against the same error. I switched to another 
machine using PgSql and it seemed to work. Mnogosearch should have a disclaimer that 
it is known not to work with MySql.

Laterzzz




Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: space-only line causes error

[Volker Wysk]

Author: Volker Wysk
Email: post   @volker-wysk.de
Message:
Hi.

A line in indexer.conf, which contains only spaces, causes an
"Error in config file".

bye


Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

redirect url too long error

[Caffeinate The World]

[1]
http://minneapolis.about.com/citiestowns/midwestus/minneapolis/local/midwestus/minneapolis/li
brary/weekly/aa092199.htm
[1] Redirect URL too long:
'http://minneapolis.about.com/webapps/whereabout/minneapolis.about.com
/citiestowns/midwestus/minneapolis/local/midwestus/minneapolis/library/weekly/aa092199.htm'
[1] http://www.ams.usda.gov/fv/paca/

__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

DeleteNoServer removing robots.txt

[Caffeinate The World]

...
[1] http://www.southwest.msus.edu/robots.txt
[1] No 'Server' command for url... deleted.
[1] http://www.medformation.com/mf/crspa.nsf/pa/menu8525.htm
...

i've noticed that any url with robots.txt in it will get removed
automatically. i have 'DeleteNoServer no' set. with the above all paths
in 'http://www.southwest.msus.edu/%' were indexed though.

__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Question regarding ftp archiving?

[Smileyq]

I'm attempting to setup a mp3 search for myself and some friends on some
of my current servers. I'm  having some problems with both indexing and
searching with mnogosearch. I'm wondering if anybody has any ideas to
help me out with this. When I index the ftp I get network errors about
3/4 or the way through. It also gives me invalid httpd reponse etc..
etc... Any help would be great thanks for the great program..



--
Smileyq
www.nashlinux.com
www.q3networks.com
www.aboutgoodlettsville.com


___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Php and phrase search

[FL]

Maybe it's a stupid question, but is the PHP frontend supporting the phrase
search mode ?
I've read a message in the archive about it, but it was not clear for me
(was about the mnogo php module )

If it's possible, how does it works ? A simple "Phrase yes" in the template ?

Thanks,
Francois
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: seg fault : nb of urls

[FL]

At 20:25 11/03/01 +0400, you wrote:
>Try to apply this patch:
>
>http://gw.udmsearch.izhnet.ru/~bar/patches/server.c.diff.gz
>

It doesn't work with the patch.
I always have a seg fault

Francois


>FL wrote:
>> 
>> Indexer : 3.1.11
>> Compiled with : pthraeds, mysql, no syslog
>> 
>> MySQL : 3.23.31
>> Mode : crc-multi, cross-word, no ispell, no srv_table, stop words (fr, en,
>> it, de).
>> 
>> The indexer.conf file is one I was using without any problem with old
>> stable version. With 3.1.11 I encoutered a Seg Fault with all the url. But
>> if I just select the first ones, it works fine.
>> 
>> I've read severals messages on the list reporting the same problem. Maybe
>> it's about url lenght ? I'll try to erase the longer Server url I have.
>> 
>> Sorry, but indexer doesn't produce any core file.
>> 
>> Any idea ?
>> 
>> Francois
>> ___
>> If you want to unsubscribe send "unsubscribe general"
>> to [EMAIL PROTECTED]
>___
>If you want to unsubscribe send "unsubscribe general"
>to [EMAIL PROTECTED]
>

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: FYI: ad banner wrong spelling

[Alexander Barkov]

Caffeinate The World wrote:
> 
> the ad banner on search.mnogo.org where it advertises support and
> installation, the word 'suppot' should be 'support'.
> 

Thanks!

We'll fix this tomorrow.
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: Monetize your MnoGo Search Results!

[Drew]

Author: Drew
Email: [EMAIL PROTECTED]
Message:
If you are setting up MnoGo, give us a call and we'll provide you with a direct XML 
feed into MnoGo which will allow you to monetize your search results on each search!

941-747-2160
www.BigWhat.com

Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: ad banner wrong spelling

[Dmitry Tkatchenko]

We're addressing this issue at the moment, thanks 
for reporting!
 
Dmitry

the 
  ad banner on search.mnogo.org where it advertises support andinstallation, 
  the word 'suppot' should be 
  'support'.__Do You 
  Yahoo!?Yahoo! Auctions - Buy the things you want at great prices.http://auctions.yahoo.com/___If 
  you want to unsubscribe send "unsubscribe general"to [EMAIL PROTECTED]

FYI: ad banner wrong spelling

[Caffeinate The World]

the ad banner on search.mnogo.org where it advertises support and
installation, the word 'suppot' should be 'support'.

__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: default follow should be path not site

[Caffeinate The World]

--- Alexander Barkov <[EMAIL PROTECTED]> wrote:
> Caffeinate The World wrote:
> > 
> > --- Alexander Barkov <[EMAIL PROTECTED]> wrote:
> > > Caffeinate The World wrote:
> > > >
> > > > if i insert a bunch of URLs into the db via a text file.. and
> those
> > > > urls have no corresponding server command, by default, it
> should
> > > have a
> > > > Follow of path and NOT site.
> > > >
> > > > here is one example. i pull URLs from a sub section of dmoz
> using
> > > the
> > > > Href directive. Some of those are site URLs and some are a path
> of
> > > a
> > > > site. now, when i pull all the urls from that sub section of
> dmoz,
> > > the
> > > > current setting will index thru all the "whole" sites/servers
> > > instead
> > > > of just the paths.
> > > >
> > > > if default is to follow path, then the whole site would still
> get
> > > > indexed. i hope that made sense. i brought this topic up about
> a
> > > month
> > > > ago and i still see this behavior in -current.
> > >
> > >
> > > When "DeleteNoServer no" is specified, indexer follows whole site
> > > if  there is no "Server" command for some URL.
> > 
> > i understand this, but is this the correct behavior? if it is, i
> can't
> > think of any situation where it's appropriate.
> 
> 
> Probably we'll change this in 3.2.x branch. Currently it requires
> major
> changes in sources. We want to avoid this before stable version
release.

thanks you. i'll wait.


__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: splitter core dump 3.1.12-cvs

[Caffeinate The World]

--- Alexander Barkov <[EMAIL PROTECTED]> wrote:
> Can you put F4D.log somewhere on the net? We'll check
> it on our boxes. Thanks.

here it is:

http://www.minnesota.com/~tom/public_html/f4d.tgz

note, it's was indexed on the DEC/Compag/Alpha

> Caffeinate The World wrote:
> > 
> > i've just gotten my first splitter core since the fix back in 3.1.8
> or
> > .9. i'm running 3.1.12-cvs-2001-03-07. i split about 10 31 mb files
> and
> > things went fine 'til around F4D. i've been splitting about 10 31
> mb
> > files and there has been no core dump 'til now. here are the file
> sizes
> > in './var/splitter/
> > 
> > # ls -l 74*
> > -rw-r--r--  1 root  wheel  101856 Mar 10 12:11 740.log
> > -rw-r--r--  1 root  wheel  596448 Mar 10 12:11 741.log
> > -rw-r--r--  1 root  wheel   57824 Mar 10 12:11 742.log
> > -rw-r--r--  1 root  wheel  119136 Mar 10 12:11 743.log
> > -rw-r--r--  1 root  wheel   77760 Mar 10 12:11 744.log
> > -rw-r--r--  1 root  wheel  882752 Mar 10 12:11 745.log
> > -rw-r--r--  1 root  wheel   32224 Mar 10 12:11 746.log
> > -rw-r--r--  1 root  wheel   45056 Mar 10 12:11 747.log
> > -rw-r--r--  1 root  wheel   72704 Mar 10 12:11 748.log
> > -rw-r--r--  1 root  wheel  124128 Mar 10 12:11 749.log
> > -rw-r--r--  1 root  wheel  141632 Mar 10 12:11 74A.log
> > -rw-r--r--  1 root  wheel   94272 Mar 10 12:11 74B.log
> > -rw-r--r--  1 root  wheel  248448 Mar 10 12:11 74C.log
> > -rw-r--r--  1 root  wheel   16832 Mar 10 12:11 74D.log
> > -rw-r--r--  1 root  wheel   97696 Mar 10 12:11 74E.log
> > -rw-r--r--  1 root  wheel   61216 Mar 10 12:11 74F.log
> > 
> > ...
> > Reading symbols from /usr/lib/libc.so.12...done.
> > Reading symbols from /usr/lib/libresolv.so.1...done.
> > #0  UdmSplitCacheLog (log=984185981) at cache.c:682
> > 682
> > logwords[
> > count+j].wrd_id=table[w].wrd_id;
> > (gdb) bt
> > #0  UdmSplitCacheLog (log=984185981) at cache.c:682
> > warning: Hit heuristic-fence-post without finding
> > warning: enclosing function for address 0x33b093aa991c4
> > This warning occurs if you are debugging a function without any
> symbols
> > (for example, in a stripped executable).  In that case, you may
> wish to
> > increase the size of the search with the `set heuristic-fence-post'
> > command.
> > 
> > Otherwise, you told GDB there was a function where there isn't one,
> or
> > (more likely) you have encountered a bug in GDB.
> > (gdb) p
> > The history is empty.
> > (gdb) l
> > 677
> > elselogwords=(UDM_LOG
> > WORD*)realloc(logwords,(count+r)*sizeof(UDM_LOGWORD));
> > 678
> > for(j=0;j > 679
> > /* Add only those
> >  OLD words which */
> > 680
> > /* do not present
> >  in DEL log  */
> > 681
> > if(!PresentInDelL
> > og(del_buf,del_count,cache[j].url_id)){
> > 682
> > logwords[
> > count+j].wrd_id=table[w].wrd_id;
> > 683
> > logwords[
> > count+j].weight=table[w].weight;
> > 684
> > logwords[
> > count+j].url_id=cache[j].url_id;
> > 685
> > #ifdef UD
> > M_STORE_CACHE_WRDPOS
> > 686
> > logwords[
> > count+j].wrd_pos=cache[j].wrd_pos;
> > (gdb)
> > 
> > upon trying to run splitter again to confirm the core dump:
> > 
> > # splitter -f f4d -t fff
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D0F000 old: 169 new:
>  10
> > total: 179
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D12000 old:   1 new:
>   1
> > total:   2
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D2A000 old:  53 new:
>  46
> > total:  99
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D2C000 old:   6 new:
>   2
> > total:   8
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D33000 old: 544 new:
>  17
> > total: 561
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D35000 old:  22 new:
>   8
> > total:  30
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D46000 old:  12 new:
>   2
> > total:  14
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D4D000 old:1707 new:
> 112
> > total:1819
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D52000 old:1393
> new:1089
> > total:2482
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D5A000 old:  32 new:
>   8
> > total:  40
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D6 old:  33 new:
>   6
> > total:  39
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D66000 old:   1 new:
>   1
> > total:   2
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D6F000 old:  90 new:
>   3
> > total:  93
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D76000 old:  36 new:
>   1
> > total:  37
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D77000 old:   5 new:
>   1
> > total:   6
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D83000 old:  54 new:
>   5
> > total:  59
> > /usr/local/install/mnogosearch/var/tree/F4/D/F4D85000 old:1933 new:
>  18
> > total:1951
> > Segmentation fault - core dumped
> > 
> > __
> > Do You Yahoo!?
> > Yahoo! Auctions - Buy the things you want at great price

Re: default follow should be path not site

[Alexander Barkov]

Caffeinate The World wrote:
> 
> --- Alexander Barkov <[EMAIL PROTECTED]> wrote:
> > Caffeinate The World wrote:
> > >
> > > if i insert a bunch of URLs into the db via a text file.. and those
> > > urls have no corresponding server command, by default, it should
> > have a
> > > Follow of path and NOT site.
> > >
> > > here is one example. i pull URLs from a sub section of dmoz using
> > the
> > > Href directive. Some of those are site URLs and some are a path of
> > a
> > > site. now, when i pull all the urls from that sub section of dmoz,
> > the
> > > current setting will index thru all the "whole" sites/servers
> > instead
> > > of just the paths.
> > >
> > > if default is to follow path, then the whole site would still get
> > > indexed. i hope that made sense. i brought this topic up about a
> > month
> > > ago and i still see this behavior in -current.
> >
> >
> > When "DeleteNoServer no" is specified, indexer follows whole site
> > if  there is no "Server" command for some URL.
> 
> i understand this, but is this the correct behavior? if it is, i can't
> think of any situation where it's appropriate.


Probably we'll change this in 3.2.x branch. Currently it requires major
changes in sources. We want to avoid this before stable version release.
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: default follow should be path not site

[Caffeinate The World]

--- Alexander Barkov <[EMAIL PROTECTED]> wrote:
> Caffeinate The World wrote:
> > 
> > if i insert a bunch of URLs into the db via a text file.. and those
> > urls have no corresponding server command, by default, it should
> have a
> > Follow of path and NOT site.
> > 
> > here is one example. i pull URLs from a sub section of dmoz using
> the
> > Href directive. Some of those are site URLs and some are a path of
> a
> > site. now, when i pull all the urls from that sub section of dmoz,
> the
> > current setting will index thru all the "whole" sites/servers
> instead
> > of just the paths.
> > 
> > if default is to follow path, then the whole site would still get
> > indexed. i hope that made sense. i brought this topic up about a
> month
> > ago and i still see this behavior in -current.
> 
> 
> When "DeleteNoServer no" is specified, indexer follows whole site
> if  there is no "Server" command for some URL.

i understand this, but is this the correct behavior? if it is, i can't
think of any situation where it's appropriate.

__
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

3.1.12 announce

[Alexander Barkov]

Hello!

3.1.12 is now available from our site http://search.mnogo.ru/

This is mostly bug fixes and minor improvements version.



  From ChangeLog:


* cachelogd now accepts port to listen from -pXXX command line argument. 
* Crash in UdmAddURL() function has been fixed. 
* CPU usage overhead bug in cachelogd has been fixed. 
* A bug in search.cgi in IspelMode db has been fixed. 
* Ispell data loading speed-up. 
* Some built-in database bugs have been fixed. Thanks Darko Koruga
  <[EMAIL PROTECTED]> 
* Added search template variable $L - language limit selection. 
* A minor bug in navigator bar has been fixed. 
* Some memory leaks have been eliminated. 
* Stopwords are not checked anymore in the case of "substring match"
searches. 
* A bug in Include indexer.conf command has been fixed. Indexer didn't
stop when
  included file had wrong syntax. 
* A minor bug that URLs were not unescaped before parsing (when non-zero
  URLWeight) has been fixed. 


  Regards.
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: patch: limit MySQL insert packets

[Alexander Barkov]

It seems to be MySQL error. Please follow MySQL
documentation.


"Briggs, Gary" wrote:
> 
> Fabulous!
> 
> Well, it hasn't broken again yet, but If it does...
> 
> Hmmm.
> 
> At the same time, though, I'm getting the following error:
> 
> 
> Can't create a new thread (errno 11). If you are not out of available
> memory, you can consult the manual for a possible OS-dependent bug
> 
> 
> The machine I'm running this on has 4G of virtual memory. I'm NOT out of
> available memory. And it's solaris 2.6.
> 
> My my.cnf file contains the following:
> 
> [client]
> port= 54000
> socket  = /tmp/mysql.sock
> 
> [mysqld]
> port= 54000
> socket  = /tmp/mysql.sock
> #set-variable   = max_allowed_packet=4M
> set-variable= wait_timeout=120
> set-variable= flush_time=200
> set-variable= key_buffer=124M
> set-variable= table_cache=512
> set-variable= sort_buffer=16M
> set-variable= record_buffer=12M
> 
> [isamchk]
> set-variable= key_buffer=16M
> 
> Am I missing something really obvious?
> 
> Thank-you very much,
> Gary (=
> 
> > -Original Message-
> > From: Shane Wegner [SMTP:[EMAIL PROTECTED]]
> > Sent: Thursday, March 08, 2001 10:56 PM
> > To:   [EMAIL PROTECTED]
> > Subject:  patch: limit MySQL insert packets
> >
> > Hi,
> >
> > The default MySQL max packet length is 1mb.  Often this
> > length can be exceeded as Mnogosearch is now using extended
> > inserts.  This patch limits insert queries into dict tables
> > to 512k.  Applies against Mnogosearch 3.11.
> >
> > --- sql.c.origThu Mar  8 13:48:58 2001
> > +++ sql.c Thu Mar  8 14:10:46 2001
> > @@ -2251,6 +2251,18 @@
> >
> > sprintf(qe,"(%d,%d,%d)",url_id,UdmStrCRC32(Indexer->Word[i].word),Indexer-
> > >Word[i].count);
> >   }
> >   qe=qe+strlen(qe);
> > + /* Insert 512k packets max to stay *
> > +  * under MySQL's default limit. */
> > + if (strlen(qb) >= 512000){
> > + sql_query(Indexer,qb);
> > +
> > if(UdmDBErrorCode(Indexer->db))return(IND_ERROR);
> > +
> > if(Indexer->Conf->DBMode==UDM_DBMODE_MULTI)
> > + sprintf(qb,"INSERT INTO %s
> > (url_id,word,intag) VALUES ",tbl_nm);
> > + else
> > + sprintf(qb,"INSERT INTO %s
> > (url_id,word_id,intag) VALUES ",tbl_nm);
> > + qe=qb+strlen(qb);
> > + have_words = 0;
> > + }
> >   }
> >   }
> >   if(have_words){
> >
> >
> > --
> > Shane Wegner: [EMAIL PROTECTED]
> >   http://www.cm.nu/~shane/
> > PGP:  1024D/FFE3035D
> >   A0ED DAC4 77EC D674 5487
> >   5B5C 4F89 9A4E FFE3 035D
> > ___
> > If you want to unsubscribe send "unsubscribe general"
> > to [EMAIL PROTECTED]
> >
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Why do i get only relevance=1?

[Alexander Barkov]

Div wrote:
> 
> Pe data de Monday 12 March 2001 11:57, ai scris:
> > search.cgi displays the number of unique words found in a document.
> > However sorting is done taking in account word weight,
> >
> > Div wrote:
> > > I have the following problem: if i search anything with search.cgi it
> > > shows me the results allright, but all of them have the relevance 1. What
> > > is going on? I used also the bodyweight, titleweight, etc. but without
> > > succes!
> > >
> 
> I see. In documentation i have read that $DR holds the document rating which
> i thought it was in fact the relevance. Is there a way i could also show the
> relevance (the number of unique words is also usefull) - maybe by adding a
> new meta-symbol?
> Thank you!


We'll do it in 3.2.x branch.
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: Miem type does not match

[Alexander Barkov]

Author: Alexander Barkov
Email: [EMAIL PROTECTED]
Message:
Hello!  Mime command understands case insensitive string match
with ? and * signs. You may use:

  Mime application/pdf*  

Dmitry, please add this into documentation.


> Hi
> 
> It seems like when the HTTP server returns the mime type 
> "application/pdf; charset=iso-8859-1", that in a "Mime" command
> in indexer.conf, the type "application/pdf" won't match.
> 
> This can be considered a bug, or not. At least, it should be 
> documented.
> 
> bye
> 

Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: indexing file names

[Alexander Barkov]

Author: Alexander Barkov
Email: [EMAIL PROTECTED]
Message:
Hello!

Thanks, fixed!

> Hi
> 
> I've indexed 
> http://localhost/Texte/A%20Gentle%20Introduction%20to%20Tex/GentleIntro.ps
> and got the following:
> 
> mysql> select * from dict where url_id=35693; 
> +++---+
> | url_id | word   | intag |
> +++---+
> |  35693 | http   | 65537 |
> |  35693 | localhost  | 65537 |
> |  35693 | text   | 65537 |
> |  35693 | 20gentle   | 65537 |
> |  35693 | 20introduction | 65537 |
> |  35693 | 20to   | 65537 |
> |  35693 | 20tex  | 65537 |
> |  35693 | gentleintro| 65537 |
> |  35693 | ps | 65537 |
> +++---+
> 
> - The URL is indexed in its URL-encoded form (the "20"s). It
>   should be decoded first.
> - The protocol and probably host parts should not get indexed.
> - The suffix ("ps") should probably not get indexed.
> 
> (Moreover, the content should have been indexed, but that's another
> story; still configuring.)
> 
> I'm using mnogosearch-3.1.11.
> 
> bye!
> 
> 

Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: What test error mean with the perl frontend?

[Bob Randall]

Author: Bob Randall
Email: 
Message:
> linux/auto/Mnogosearch/Mnogosearch.so: undefined symbol: 
> mysql_escape_string at /usr/lib/perl5/5.00503/i386-

Looks like whoever wrote the code didn't think it was important to escape characters 
to mysql and mysql is aborting. The code needs to add escape to the string and you 
won't see this. This is something the author of the perl frontend should quickly 
correct since I doubt very few people will be able to use it as it with the latest 
mysql release.  

Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: patch: limit MySQL insert packets

[Alexander Barkov]

Thanks!

We've applied this into 3.1.12 source (with minor modifications).



Shane Wegner wrote:
> 
> Hi,
> 
> The default MySQL max packet length is 1mb.  Often this
> length can be exceeded as Mnogosearch is now using extended
> inserts.  This patch limits insert queries into dict tables
> to 512k.  Applies against Mnogosearch 3.11.
> 
> --- sql.c.orig  Thu Mar  8 13:48:58 2001
> +++ sql.c   Thu Mar  8 14:10:46 2001
> @@ -2251,6 +2251,18 @@
> 
>sprintf(qe,"(%d,%d,%d)",url_id,UdmStrCRC32(Indexer->Word[i].word),Indexer->Word[i].count);
> }
> qe=qe+strlen(qe);
> +   /* Insert 512k packets max to stay *
> +* under MySQL's default limit. */
> +   if (strlen(qb) >= 512000){
> +   sql_query(Indexer,qb);
> +   
>if(UdmDBErrorCode(Indexer->db))return(IND_ERROR);
> +   
>if(Indexer->Conf->DBMode==UDM_DBMODE_MULTI)
> +   sprintf(qb,"INSERT INTO %s 
>(url_id,word,intag) VALUES ",tbl_nm);
> +   else
> +   sprintf(qb,"INSERT INTO %s 
>(url_id,word_id,intag) VALUES ",tbl_nm);
> +   qe=qb+strlen(qb);
> +   have_words = 0;
> +   }
> }
> }
> if(have_words){
> 
> --
> Shane Wegner: [EMAIL PROTECTED]
>   http://www.cm.nu/~shane/
> PGP:  1024D/FFE3035D
>   A0ED DAC4 77EC D674 5487
>   5B5C 4F89 9A4E FFE3 035D
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: $Q related question

[Alexander Barkov]

Author: Alexander Barkov
Email: [EMAIL PROTECTED]
Message:
Hello!

I would prefer to do that at index time. This will allow
more quick search and will eliminate for example boolean
related problems. Is there a way to detect that words are in
reverse order? What HTTP headers or META or something else are responsible for this?

>  since there are two standarts of writing hebrew, logical and visual, I'm faced with 
>a problem 
> that when a user types in a query it get searched in the same order the characters 
>were typed in 
> (obviously) and not all of the pages which might otherwise be displayed get to be 
>displayed in the 
> results.
>  What I mean is, that if a user types in "Joseph" only the word 
>"Joseph" gets to be searched for 
> however since in hebrew we have two standarts the word "hpesoJ" will also 
>yield a valid results 
> since the browser (or at least those browsers that support this feature) 
>automatically reverse the 
> order of the characters appearing on screen.
>  Is there a way I can send both "Joseph" and "hpesoJ" to the 
>search.cgi and get a concatenated 
> report of both quesries as a result? Without interfacing the search.cgi with a 
>wrapper that is.
> 
> 
>  Regards,
> 
>  Nir
> 

Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Why do i get only relevance=1?

[Div]

Pe data de Monday 12 March 2001 11:57, ai scris:
> search.cgi displays the number of unique words found in a document.
> However sorting is done taking in account word weight,
>
> Div wrote:
> > I have the following problem: if i search anything with search.cgi it
> > shows me the results allright, but all of them have the relevance 1. What
> > is going on? I used also the bodyweight, titleweight, etc. but without
> > succes!
> >

I see. In documentation i have read that $DR holds the document rating which 
i thought it was in fact the relevance. Is there a way i could also show the 
relevance (the number of unique words is also usefull) - maybe by adding a 
new meta-symbol?
Thank you!
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: default follow should be path not site

[Alexander Barkov]

Caffeinate The World wrote:
> 
> if i insert a bunch of URLs into the db via a text file.. and those
> urls have no corresponding server command, by default, it should have a
> Follow of path and NOT site.
> 
> here is one example. i pull URLs from a sub section of dmoz using the
> Href directive. Some of those are site URLs and some are a path of a
> site. now, when i pull all the urls from that sub section of dmoz, the
> current setting will index thru all the "whole" sites/servers instead
> of just the paths.
> 
> if default is to follow path, then the whole site would still get
> indexed. i hope that made sense. i brought this topic up about a month
> ago and i still see this behavior in -current.


When "DeleteNoServer no" is specified, indexer follows whole site
if  there is no "Server" command for some URL.
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Why do i get only relevance=1?

[Alexander Barkov]

search.cgi displays the number of unique words found in a document.
However sorting is done taking in account word weight,


Div wrote:
> 
> I have the following problem: if i search anything with search.cgi it shows
> me the results allright, but all of them have the relevance 1. What is going
> on? I used also the bodyweight, titleweight, etc. but without succes!
> 
> I am using the latest MnogoSearch from CVS, in cache mode, using the
> following indexer.conf:
> 
> DBAddr  mysql://xxx:xxx@localhost/udmsearch1/
> DBMode cache
> LogdAddr xxx.localdomain:7000
> DeleteNoServer No
> #BodyWeight 2
> #TitleWeight 4
> #KeywordWeight 8
> #DescWeight 16
> #CrossWeight 32
> #UrlWeight 5
> #UrlHostWeight 4
> #UrlPathWeight 3
> #UrlFileWeight 1
> Robots Yes
> Period 20d
> MaxHops 99
> DeleteBad yes
> Follow world
> Proxy xxx.localdomain:3128
> Realm http://*.ro/*
> URL http://www.xxx.ro/
> 
> What is wrong?
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: indexer sped up with Posgresql

[Alexander Barkov]

Thanks!

I've added this into pgsql create script.

Caffeinate The World wrote:
> 
> mnogosearch-cvs, postgresql-cvs
> 
> i just wanted to let everyone who uses postgresql with mnogosearch how
> i got this incredible speed increase. I had about 500,000 URLs and when
> i started indexing, it took for ever between each URL. per alex's
> suggestion, i turned on debugging for sql.c and found where it took the
> most time. a query like below took about 38 secs before i created the
> index, now takes less than 1 sec between urls.
> 
> [13289] SQL 0.26s: SELECT rec_id FROM url WHERE crc32=40757896 AND
> status=200 AND docsize=10927
> 
> i created the index:
> 
> CREATE INDEX url_crc32_status_docsize ON url (crc32,status,docsize)
> 
> perhaps this can be added to the create.txt file for postgresql.
> 
> __
> Do You Yahoo!?
> Get email at your own domain with Yahoo! Mail.
> http://personal.mail.yahoo.com/
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

RE: patch: limit MySQL insert packets

[Briggs, Gary]

Fabulous!

Well, it hasn't broken again yet, but If it does...

Hmmm.

At the same time, though, I'm getting the following error:


Can't create a new thread (errno 11). If you are not out of available
memory, you can consult the manual for a possible OS-dependent bug


The machine I'm running this on has 4G of virtual memory. I'm NOT out of
available memory. And it's solaris 2.6.

My my.cnf file contains the following:


[client]
port= 54000
socket  = /tmp/mysql.sock

[mysqld]
port= 54000
socket  = /tmp/mysql.sock
#set-variable   = max_allowed_packet=4M
set-variable= wait_timeout=120
set-variable= flush_time=200
set-variable= key_buffer=124M
set-variable= table_cache=512
set-variable= sort_buffer=16M
set-variable= record_buffer=12M


[isamchk]
set-variable= key_buffer=16M

Am I missing something really obvious?

Thank-you very much,
Gary (=



> -Original Message-
> From: Shane Wegner [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, March 08, 2001 10:56 PM
> To:   [EMAIL PROTECTED]
> Subject:  patch: limit MySQL insert packets
> 
> Hi,
> 
> The default MySQL max packet length is 1mb.  Often this
> length can be exceeded as Mnogosearch is now using extended
> inserts.  This patch limits insert queries into dict tables
> to 512k.  Applies against Mnogosearch 3.11.
> 
> --- sql.c.origThu Mar  8 13:48:58 2001
> +++ sql.c Thu Mar  8 14:10:46 2001
> @@ -2251,6 +2251,18 @@
>  
> sprintf(qe,"(%d,%d,%d)",url_id,UdmStrCRC32(Indexer->Word[i].word),Indexer-
> >Word[i].count);
>   }
>   qe=qe+strlen(qe);
> + /* Insert 512k packets max to stay *
> +  * under MySQL's default limit. */
> + if (strlen(qb) >= 512000){
> + sql_query(Indexer,qb);
> +
> if(UdmDBErrorCode(Indexer->db))return(IND_ERROR);
> +
> if(Indexer->Conf->DBMode==UDM_DBMODE_MULTI)
> + sprintf(qb,"INSERT INTO %s
> (url_id,word,intag) VALUES ",tbl_nm);
> + else
> + sprintf(qb,"INSERT INTO %s
> (url_id,word_id,intag) VALUES ",tbl_nm);
> + qe=qb+strlen(qb);
> + have_words = 0;
> + }
>   }
>   }
>   if(have_words){
> 
> 
> -- 
> Shane Wegner: [EMAIL PROTECTED]
>   http://www.cm.nu/~shane/
> PGP:  1024D/FFE3035D
>   A0ED DAC4 77EC D674 5487
>   5B5C 4F89 9A4E FFE3 035D
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
> 
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: splitter core dump 3.1.12-cvs

[Alexander Barkov]

Can you put F4D.log somewhere on the net? We'll check
it on our boxes. Thanks.



Caffeinate The World wrote:
> 
> i've just gotten my first splitter core since the fix back in 3.1.8 or
> .9. i'm running 3.1.12-cvs-2001-03-07. i split about 10 31 mb files and
> things went fine 'til around F4D. i've been splitting about 10 31 mb
> files and there has been no core dump 'til now. here are the file sizes
> in './var/splitter/
> 
> # ls -l 74*
> -rw-r--r--  1 root  wheel  101856 Mar 10 12:11 740.log
> -rw-r--r--  1 root  wheel  596448 Mar 10 12:11 741.log
> -rw-r--r--  1 root  wheel   57824 Mar 10 12:11 742.log
> -rw-r--r--  1 root  wheel  119136 Mar 10 12:11 743.log
> -rw-r--r--  1 root  wheel   77760 Mar 10 12:11 744.log
> -rw-r--r--  1 root  wheel  882752 Mar 10 12:11 745.log
> -rw-r--r--  1 root  wheel   32224 Mar 10 12:11 746.log
> -rw-r--r--  1 root  wheel   45056 Mar 10 12:11 747.log
> -rw-r--r--  1 root  wheel   72704 Mar 10 12:11 748.log
> -rw-r--r--  1 root  wheel  124128 Mar 10 12:11 749.log
> -rw-r--r--  1 root  wheel  141632 Mar 10 12:11 74A.log
> -rw-r--r--  1 root  wheel   94272 Mar 10 12:11 74B.log
> -rw-r--r--  1 root  wheel  248448 Mar 10 12:11 74C.log
> -rw-r--r--  1 root  wheel   16832 Mar 10 12:11 74D.log
> -rw-r--r--  1 root  wheel   97696 Mar 10 12:11 74E.log
> -rw-r--r--  1 root  wheel   61216 Mar 10 12:11 74F.log
> 
> ...
> Reading symbols from /usr/lib/libc.so.12...done.
> Reading symbols from /usr/lib/libresolv.so.1...done.
> #0  UdmSplitCacheLog (log=984185981) at cache.c:682
> 682
> logwords[
> count+j].wrd_id=table[w].wrd_id;
> (gdb) bt
> #0  UdmSplitCacheLog (log=984185981) at cache.c:682
> warning: Hit heuristic-fence-post without finding
> warning: enclosing function for address 0x33b093aa991c4
> This warning occurs if you are debugging a function without any symbols
> (for example, in a stripped executable).  In that case, you may wish to
> increase the size of the search with the `set heuristic-fence-post'
> command.
> 
> Otherwise, you told GDB there was a function where there isn't one, or
> (more likely) you have encountered a bug in GDB.
> (gdb) p
> The history is empty.
> (gdb) l
> 677
> elselogwords=(UDM_LOG
> WORD*)realloc(logwords,(count+r)*sizeof(UDM_LOGWORD));
> 678
> for(j=0;j 679
> /* Add only those
>  OLD words which */
> 680
> /* do not present
>  in DEL log  */
> 681
> if(!PresentInDelL
> og(del_buf,del_count,cache[j].url_id)){
> 682
> logwords[
> count+j].wrd_id=table[w].wrd_id;
> 683
> logwords[
> count+j].weight=table[w].weight;
> 684
> logwords[
> count+j].url_id=cache[j].url_id;
> 685
> #ifdef UD
> M_STORE_CACHE_WRDPOS
> 686
> logwords[
> count+j].wrd_pos=cache[j].wrd_pos;
> (gdb)
> 
> upon trying to run splitter again to confirm the core dump:
> 
> # splitter -f f4d -t fff
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D0F000 old: 169 new:  10
> total: 179
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D12000 old:   1 new:   1
> total:   2
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D2A000 old:  53 new:  46
> total:  99
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D2C000 old:   6 new:   2
> total:   8
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D33000 old: 544 new:  17
> total: 561
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D35000 old:  22 new:   8
> total:  30
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D46000 old:  12 new:   2
> total:  14
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D4D000 old:1707 new: 112
> total:1819
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D52000 old:1393 new:1089
> total:2482
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D5A000 old:  32 new:   8
> total:  40
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D6 old:  33 new:   6
> total:  39
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D66000 old:   1 new:   1
> total:   2
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D6F000 old:  90 new:   3
> total:  93
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D76000 old:  36 new:   1
> total:  37
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D77000 old:   5 new:   1
> total:   6
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D83000 old:  54 new:   5
> total:  59
> /usr/local/install/mnogosearch/var/tree/F4/D/F4D85000 old:1933 new:  18
> total:1951
> Segmentation fault - core dumped
> 
> __
> Do You Yahoo!?
> Yahoo! Auctions - Buy the things you want at great prices.
> http://auctions.yahoo.com/
> ___
> If you want to unsubscribe send "unsubscribe general"
> to [EMAIL PROTECTED]
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Re: Big trouble and minor typos in indexer.conf - mnogosearch-3.1.11

[Alexander Barkov]

Joerg Behrens wrote:
> 
> Hi,
> i found some minor typos in the configfile 'indexer.conf'.
> 
> line ~182 - change prase to Phrase
> line ~806 - the config paramter calls 'IndexMP3TagOnly' an not '
> CheckID3TagOnly'

Thanks, fixed.


> After successfully compiled mnogosearch-3.1.11 i'll recieve a error during
> runing ./indexer.
> 
> Indexer[8514]: indexer from mnogosearch-3.1.11/MySQL started with
> '/usr/local/mnogosearch/etc/indexer.conf'
> Indexer[8514]: [1] Error: '#1054: Unknown column 'msg_id' in 'field list''
> 
> Whats going wrong? I have create all tables like the definition i found in
> /create/mysql/ . Btw. i think some are broken .. i 'DROP TABLE IF EXISTS
> ' works better as 'DROP TABLE '.
> 
> Using mysql 3.22.32 , linux 2.2.14-SMP and php4.0.5dev. ( i would like to
> take a look to the php mnogosearch extension).
> 


You compiled msearch with  --enable-news-extensions.
This key required special search front-end as well as
tables structure. Take either Perl front-end or PHP front-end with 
NEWS EXTENSIONS support from our download page. Those packages have
CREATE TABLE scripts suitable for news extensions.
___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Webboard: Language based search with PHP front end

[csraje]

Author: csraje
Email: [EMAIL PROTECTED]
Message:

Hi,

1) I used php front end and i indexed two sites(one german site and one english site. 
when i searched with certain words(not all) ex:test,insist, i got the below error.

Warning: REG_EMPTY in /usr/local/apache_1.3.17/htdocs/mnofrontend/ispell.inc on line 
146

Warning: REG_EMPTY in /usr/local/apache_1.3.17/htdocs/mnofrontend/ispell.inc on line 
146

Warning: REG_EMPTY in /usr/local/apache_1.3.17/htdocs/mnofrontend/ispell.inc on line 
146


How to avaid this? 

2)
what should i do to do language based search? even though i indexed a  german language 
site,I selected the german language in the language selection combo,and typed a german 
word int he text box,the out put is"no results found" . But when i select English in 
the language combo, and i searched the same german word, results are displayed. Please 
explain how to install language based search in detail.

Regards,
jarch

Reply: 

___
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]