Hello Maria
I got a call just yesterday about 8.1.6 on NT not installing
on P4 machine. I researched metalink and they have some points.
(BTW, we did DLL upgrade. Problem solved).
However, all the fixes are upgrading a DLL or the JInitiator.
So, I do not think that the fix will prevent you from
OS. Tru64 Unix ver 5.1
Oracle 8.1.7.2
m/c GS320 Alpha Server
After 900 User processes are Connected to a Database , sqlplus system/manager@Connect
String
When issued DIRECTLY from the DB Server takes about 1 minute to Reach the SQL prompt
Any parameters , pointers , Docs ?
--
Please see
Hi,
That strange, but this has never come to my mind before.
Ideally, the recovery catalog should be on other than production server (and
it always was for me). And recovery catalog's database should be copied as
well. Perhaps, using production database for recovery catalog.
What if recovery
Probably configuration with shared server. Right?
Maybe after 900 users new server process is starting and this takes time,
but 1 minute looks too much.
Is this only the case when connecting from the server prompt or from
workstations as well?
Alexandre
OS. Tru64 Unix ver 5.1
Oracle 8.1.7.2
Hello VIVEK_SHARMA,
Do you use MTS?
Look at dispatchers load.
Tuesday, May 28, 2002, 3:33:18 PM, you wrote:
V OS. Tru64 Unix ver 5.1
V Oracle 8.1.7.2
V m/c GS320 Alpha Server
V After 900 User processes are Connected to a Database , sqlplus
system/manager@Connect String
V When issued
Does anyone have an online backup script for Oracle 8.1.7 and AIX?
Or can anyone provide a definitive list of the actions required to be taken?
John
--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Author: John Dunn
INET: [EMAIL PROTECTED]
Fat City Network Services--
I have a online backup script which issues the command
ALTER SYSTEM SWITCH LOGFILE;
immediately before copying the archived redo logs.
Does this make sense? I am finding that the ARC process has not finished
archiving the log before I copy the archive logs
John
--
Please see the official
Does anyone have an online backup script for Oracle
8.1.7 and AIX?
Or can anyone provide a definitive list of the
actions required to be taken?
John
John,
Look for the script named hotbackup.sql at http://www.oriole.com/frameindexTK.html.
It's neither 8.1.7 or AIX specific, it uses 'cp'
PERFORMANCE
A Statspack report.txt Covering the problem
period is Very Good Advice .
1) From report.txtpaste only the Section below
"system-wide waits for non-background processes"
Likewise from statspack paste the "top 5 wait" events
section
2) v$session_wait is is
You Should Be Using
Alter System Archive Log Current / ALL
This will wait till the archive Process has written out the archivelog
unlike the Alter System Switch Log file.
HTH
Best Regards,
Ganesh R
Tel : +971 (4) 397 3337 Ext 420
Fax : +971 (4) 397 6262
HP : +971 (50) 745 6019
Hi All,
MY DB is 8.1.7 on AIX, TEMP space is not being
realeasd after sort even after alter tablespace name
coalesce. TEMP space was created as temporary.
We didn't event put ant events to not to clear.
Please let me know if there is any problem.
Cheers,
Ravi.
John,Add a 'sleep' command to your UNIX-side shell script before you copy the logs to ensure the log has finished writing. You need to have that last archivelog - it contains the data regarding datafile header record changes you'll need to apply in the event you have to restore to sync up your
no , i do not .
Thanks
DBAndrey
* 03-9254520
* 058-548133
* mailto:[EMAIL PROTECTED]
-Original Message-
Sent: Tue, May 28, 2002 4:53 AM
To: Multiple recipients of list ORACLE-L
Hello Andrey,
Do you use autonomous transaction?
Monday, May 27, 2002, 7:53:19 PM, you wrote:
AB
Ravi,
Try changing the next extent size to something other than its current
setting, then change it back to what it was, and you should see any/all temp
space not currently in use freed up. Coalesce only merges adjacent space,
and not-in-use temp space is not freed up until someone needs it.
maybe 10i will allow table pt in time recovery :)
[EMAIL PROTECTED] 05/25/02 05:53PM
you mean the export? it's a lot easier to recover a single table from
an export and let everyone else keep working. AFAIK, Oracle still
doesn't do table-level recovery, the lowest granularity is tablespace.
I
Before any public outcry I checked in with, and got permission, from list
moderator to make this posting!!!
Acknowledging we're geographically remote for most members of this list.
(but maybe some other Irish based folk are here too!)
My company, location North County Dublin, Ireland, has a
Hi everybody,
I've taken responsibility of a couple of Oracle systems on NT (in
addition to my nice Solaris DBs). I'm setting up Legato to do the backups
online using the Legato Networker Oracle Module. After checking the Legato
docs, it seems that the only way of using this module is by
Thank you John,
The things is When we run a partucular query based on
view it takes around 1Gb space from TEMP tablespace
and is not being relased and from that point on words
It give space problem on other txns. that means It
didn't mark fro Free to use does it?
Ravi.
-- Lau, John [EMAIL
Hi Mogens,
What I wouldn't do to be a fly on that wall. Oh,
the interesting discussions to be had! :-)
I too have thought long and hard about this
industry trend, and it has remarkable ramifications that we should all be aware
of.
One implication that you don't mention is the clear
I too am using Legato and Oracle, but haven't ventured to RMAN. I will start using it
soon. What I plan to do:
1) Put RMAN db/catalog on same server as legato backup server.
2) Cold backup RMAN db to disk daily after nightly RMAN backups.
3) Backup RMAN disk backups to tape for offsite.
On Tue, 28 May 2002, [EMAIL PROTECTED] wrote:
Add a 'sleep' command to your UNIX-side shell script before you copy
the logs to ensure the log has finished writing. You need to have
that last archivelog - it contains the data regarding datafile
header record changes you'll need to apply in
isn't that supposed to be flashback query? :)
--- Gene Sais [EMAIL PROTECTED] wrote:
maybe 10i will allow table pt in time recovery :)
[EMAIL PROTECTED] 05/25/02 05:53PM
you mean the export? it's a lot easier to recover a single table from
an export and let everyone else keep working.
there is no problem. Oracle leaves the temp segment allocated (you
should be seeing only one) when your temp tablespace is created
temporary to save allocation time for future sorts.
It will go away when/if you shutdown and restart the database, but it
is SUPPOSED to be there
--- Nalla Ravi
Have you tried the ora_sql_txt event ?
hth
connor
--- david hill [EMAIL PROTECTED] wrote: Hi
Guys, I was wondering if someone could help me
I'm trying to create an AFTER SERVERERROR trigger
that will log all errors
and the SQL that caused the error.
Here's what I have so far
How about DML before (ie before your query starts)?
If there are massive amount of blocks to be cleaned
out, this can lead to ora-1555
hth
connor
--- Andrey Bronfin [EMAIL PROTECTED] wrote:
I meant , no one runs a DML against those tables
during the running of the
program.
DBAndrey
On Mon, 27 May 2002, Nirmal Kumar Muthu Kumaran wrote:
Increase rollback segment size for this transaction and make sure
that the transaction will use the huge rollback segment
I don't think that solution is correct. This is a common
misconception about snapshot too old. Assigning your
oh so tempting but I know England's restrictions on bringing in
pets, what are Ireland's?
--- O'Neill, Sean [EMAIL PROTECTED] wrote:
Before any public outcry I checked in with, and got permission, from
list
moderator to make this posting!!!
Acknowledging we're geographically remote
Title: RE: Covert tnsnames.ora to OID
OID?..
-Original Message-
From: Yechiel Adar [SMTP:[EMAIL PROTECTED]]
Sent: Mon, May 27, 2002 06:58 PM
To: Multiple recipients of list ORACLE-L
Subject: Covert tnsnames.ora to OID
Hello All
How can I covert all the names from
Unfortunately, I, and all of us, have seen the
Pentium
processors in MS-NT/2K.
The general opinion of Pentium/P4 is in the
context
of MS operating systems. And these don't
perform
as well as *nix.
Hemant K Chitale
- Original Message -
From:
Mogens Nørgaard
To:
Thank You !
-Original Message-
Sent: May 27, 2002 11:13 PM
To: Multiple recipients of list ORACLE-L
_tru64_directio_disabled is an NON-Documented Tru64 Unix Specific parameter
To get List of ALL (Documented NON-Documented) parameters :-
set echo on
spool parm1
select
a.ksppinm
It is pretty easy to restore and recover a single table to an
arbitrary point in time from a physical backup. I don't think Oracle
needs to provide an extra feature.
You restore a small subset of the database (system, rollbacks and the
tablespace with the table in it), offline drop the
Should a company be willing to spend a comparable amount annually with
their Linux provider and their hardware provider that they would give to
(for instance) Sun Support, I believe they could easily achieve
comparable levels of hardware and software reliability than any other
commercial
Mike,
We did create another RMAN catalog on a small alternate host.
We back up our main catalog. Yes, it is redundant but if you're
going to go to all the work of getting RMAN working, you might as
well go the extra distance and do this so you're covered.
Cherie Machler
Oracle DBA
Gelco
Hi,
Does Oracle reccommend having the initial and next extents set at the same
values???
I am tidying up a current schema and the values are non uniform all the way
through- any comments would be very much appreciated
Rgds
Fawzia
I have put my recovery catalog database in archivelog mode. I do a cold
backup of the database and the archivelogs each Friday. This way I won't
lose it.
Ruth
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Tuesday, May 28, 2002 4:58 AM
Hi,
oh god I was teasing about flashback query! I don't really think it's
robust enough for production recoveries. Technically you don't need
server-managed undo to use flashback query, it will work with manually
managed rollback segments. But you have less of a chance for the data
to still be in the
Title: RE: Covert tnsnames.ora to OID
Oracle
Internet Directory, which is Oracle's
LDAP
product.
Matt Adams - GE Appliances -
[EMAIL PROTECTED]Chaos! Panic! Disaster! (My work here is done)
-Original Message-From: Nirmal Kumar Muthu Kumaran
[mailto:[EMAIL PROTECTED]]Sent:
I need to run SQLLDR73 from a Win 2000 server, loading data to a 7.3.4 Unix
server. I can't load 7.3.4 in the Win 2000 server. What might I need to get
the load to run on the Win 2000 machine?
Ron Smith
--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Author: Smith, Ron L.
Ravi,
1)Temporary solution:
Around off-peak time do the following to get release of temp tablespace
alter tablespace temp offline;
Just after putting offline , run following immediately
alter tablespace temp online;
2) Create temp tablespace using tempfile like
create temporary tablespace temp
Mike - RMAN has two options - the recovery catalog and using the control
file of the target database. If you have only a single system, as I
understand the RMAN philosophy, you should use the control file option.
Also, from everything I've seen, export seems a fine way to keep the RMAN
schema
Fawzia - If you're going to this much effort, consider looking into locally
managed tablespaces (LMT) with uniform extents. Then you won't have to tidy
up again.
Dennis Williams
DBA
Lifetouch, Inc.
[EMAIL PROTECTED]
-Original Message-
Sent: Tuesday, May 28, 2002 11:19 AM
To: Multiple
Recently, I loaded 7.3.4 on XP Pro and it seems to run fine, so I would
expect you should be able to run under Win2000, though these combinations
are not supported of course. Have you tried installing 7.3.4? If so, what
problems are you running into?
If it appears to be a compatibility issue,
Whenever I create or rebuild indexes i have been having the sort_Area_size
as 1 gig,I have never faced any issues,but one thing i want to know why did
u have to bounce the instance for this,whenever i create i add the sort area
size for the session only and not on the instance level.
we were
Mogens
How is Steve Adam himself? Like other listers I am feeling his absence very
much from this list. You may request on my behalf(or on behalf of other
listers like myself) that he must participate in this list...
Regards
Rafiq
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of
No. I can think of several reason when you would not wan them the save.
An example: Think of load a lot of data into a table, and then load on a
very limited basis. This tells me to create a large first extent that
everything can fit into. Example: Say 100M and then you may want to
Arun - Thanks for the tip of using ALTER SESSION. I'll try that next time,
and I hope next time isn't too soon.
Dennis Williams
DBA
Lifetouch, Inc.
[EMAIL PROTECTED]
-Original Message-
Sent: Tuesday, May 28, 2002 12:32 PM
To: Multiple recipients of list ORACLE-L
Whenever I create or
Are u getting that error quite frequently
-Original Message-
Sent: Tuesday, May 28, 2002 1:52 PM
To: Multiple recipients of list ORACLE-L
Arun - Thanks for the tip of using ALTER SESSION. I'll try that next time,
and I hope next time isn't too soon.
Dennis Williams
DBA
Lifetouch, Inc.
Title: RE: how to pick the lastes file using UTL_FILE package ?
Rahul, here is how you'd do it on a Unix box: ls -t|line
ls -t would give you a list of files most recent first, then the line command would give
you back the first filename only. In 2K the first part of that is DIR /O:-D, not
Novice,
There are no privileges associated with a synonym.
Any account that has 'create synonym' privs may create a synonym
for any object in any schema. Scott created a synonym for an object
owned by hlledw.
Hlledw may view scott.edw_week_lvl, as it is a synonym pointing to an
object owned
Arun - No, I just needed to rebuild our largest table this weekend and
re-create the indexes afterward. Fortunately with SORT_AREA_SIZE at
150-meg., the indexes build in a reasonable amount of time, so it was merely
an inconvenience. I was able to enjoy one day of the weekend. I just
wondered for
An example: Think of load a lot of data into a table, and then load
on a very limited basis. This tells me to create a large first
extent that everything can fit into.
why? there is no benefit to that
--- basher 59 [EMAIL PROTECTED] wrote:
No. I can think of several reason when you
Hi,
Asked you sys admin to boost the maximum memory
allowable for a process.
I had that error on a hp box. I do not remember the
name of the parameters. Once the sys admin had changed
the memory process parameters, processes were able to
use way more than 150M of ram.
You can increase the
Are you using PQO?
-Original Message-
Sent: Tuesday, May 28, 2002 12:54 PM
To: Multiple recipients of list ORACLE-L
This weekend I was rebuilding a large table and when it came time to rebuild
the indexes, I cranked the SORT_AREA_SIZE up to 1/2 gig. It bombed off with
an ORA-04030. I
Hi Larry,
I sure wasn't disappointed to receive this note from a CBO/SQL guru such as
yourself. Thanks for the ideas.
My question here is would this database and your findings be
applicable to
other databases and the nature of their data? I wouldn't
The nature of the data and the
Hi Fazia,
I would recommend the following white paper, it advocates using the SAFE
methodology.
HOW TO STOP DEFRAGMENTIING AND START LIVING
Bhaskar Himatsingka, Oracle Corporation
Juan Loaiza, Oracle Corporation
You can find the paper on Cary Millsap's site. www.hotsos.com
Sincerely,
Hey all,
I'm using DBMS_STATS (despite it's bugs) on 8.1.7.2.0 / HPUX 11.0. When I
specify degree = 2 in the parms for either GATHER_SCHEMA_STATS or
GATHER_TABLE_STATS, I notice that 4 p processes are kicked off (e.g.
ora_p001_sid). Since this is on a test system, there is only one
processor.
LOL!
Igor Neyman, OCP DBA
[EMAIL PROTECTED]
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Tuesday, May 28, 2002 2:58 PM
I have to respectfully disagree. Creating everything
in one huge extent and so called defragmentation are
extremely
Ian,
John are you saying to create histograms on all indexed
columns, or just the ones with distributions which are skewed
and also for ones which although symmetric in distribution
have some values much more prevalent than others?
To keep this simplistic, I wouldn't use Histograms
Hi
Which init.ora parameter allows a SORT operation to bypass the buffer cache?
tHX
-sEEMA
_
Send and receive Hotmail on your mobile device: http://mobile.msn.com
--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Hey gang;
I have an 8.0.5.0 database running on a Solaris platform.
The developer is running a simple query which returns the error
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 7, block # 191659)
ORA-01110: data file 7: '/u08/oradata/TTCT/dynamici01.dbf'
So I
John,
I know in a previous job, we determined that histograms where not worth it. The
following is from a test that we performed:
***
Table- F_tab Uniform DistributionMax Distribution
Field- P_code
Looking for a way in Unix to pass a variable into a Sqlloader control, and
then use that as a constant. Database version is 7.3.4
I tried something like:
control file ---
$1 = filename
.
.
.
.
.
SOURCE_FILE CONSTANT $SOURCE_FILE
.
.
.
When running, I
One more best way also is to create those indexes with nologging along with
ur sort_area_size,this i think u must be knowing, if that database has a
standby then copy this file which has the index data
Cause as u know only the structure of the index goes to the dictionary and
not the index data
Run following query and check whether relevent table/index has degree 1
select index_name,degree from dba_indexes where degree 1
/
select table_name,degree from dba_tables where degree 1
/
If degree 1 then it will use nymber of PQ process equivalent to number od
degree
Regards
Rafiq
Kevin Lange wrote:
Hey gang;
I have an 8.0.5.0 database running on a Solaris platform.
The developer is running a simple query which returns the error
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 7, block # 191659)
ORA-01110: data file 7:
of course, I was selfishly thinking of this only from the perspective
of the DBA and not the pocketbooks of the storage manufacturers
--- Gogala, Mladen [EMAIL PROTECTED] wrote:
I have to respectfully disagree. Creating everything
in one huge extent and so called defragmentation are
Rachel Carmichael wrote:
An example: Think of load a lot of data into a table, and then load
on a very limited basis. This tells me to create a large first
extent that everything can fit into.
why? there is no benefit to that
Rachel,
No benefit when the data is loaded, but
John,
Use ECHO commands to generate your control-file from within the shell
script. You can then use the environmentals created by your input params.
Something like:
$1 = filename
echo SOURCE_FILECONSTANT $1
control.ctl
echo SECOND_LINE
Hi,
Find the corrupted object by the following statement:
Select segment_name, segment_type
from dba_extents
where file_id=your_file# and
your_block# between block_id and block_id + blocks -1
- If it's index, recreate it.
- If not, recover the datafile
- if no backup, tell us
regards...
I have had some really good experiences with using histograms.
They didn't always produce the improvements that I expected
but in many cases, I saw 10 times, 100 times, or even 1,000 faster
execution times after adding histograms.
I don't have the specifics, but these were cases where the data
I didn't think 7.3.4 supported multiple homes. I wouldn't want to screw up
the registry entries.
Ron
-Original Message-
Sent: Tuesday, May 28, 2002 12:26 PM
To: Multiple recipients of list ORACLE-L
Recently, I loaded 7.3.4 on XP Pro and it seems to run fine, so I would
expect you
One item to remember on this note of using alter session is that if a pile of
your SQL*Plus users have the alter session privilege and set sort_area_size to a
large value at the same time, you can soon run out of memory. At which time all
of the above listed users will come down complaining
There are limits to amount of memory a process can attach to. I have hit
that wall with the size of the SGA from time to time. I would think that
limit would be associated with the sort area also, sort_area + sga if you
will. Take a look at document
tom,
This is interesting. How did you determine max
distribution and uniform distribution? Did you use
standard deviation and variance?
regards,
jack silvey
--- Terrian, Tom [EMAIL PROTECTED] wrote:
John,
I know in a previous job, we determined that
histograms where not worth it. The
Arun - Yes, I have discovered the nologging option. Also someone asked
whether I was using parallel, and the answer is no.
Dennis Williams
DBA
Lifetouch, Inc.
[EMAIL PROTECTED]
-Original Message-
Sent: Tuesday, May 28, 2002 3:01 PM
To: Multiple recipients of list ORACLE-L
One more
Hi All,
There has been some discussion in the past regarding moving
the audit tables out of the SYSTEM table space to avoid fragmentation.
Is this a good idea and if so does any have a procedure for doing so?
Are there any problems to watch out for?
O7345 - WinNT - RAID 0+1
--
Chris J. Guidry
Typically 1578s are due to hardware hickups or a known bug. Dbverify
will not capture all types of data corruption. The only truly safe way
to capture physical corruption of a datablock is to get it into the
buffer cache with a statement that exercises the datablock.
The easiest way to determine
I really did not think you could recreate an object in the same tablespace
once corrupted blocks were found I was wrong. It was a group of 20
indexes.Rachel told me offline to just drop and recreate the indexes.
I did that and things worked just fine.
The only other time I had this
but if the load is a one time thing (as he described) then the
allocation hit happens only once and I still don't see a benefit and in
fact can see how it might hurt -- tablespace fragmentation etc
I'd rather see large extents but more of them
--- Stephane Faroult [EMAIL PROTECTED] wrote:
Hi Chris,
go to the following link you'll find a paper about
that:
http://www.oracle.com/oramag/oracle/99-Nov/index.html?69dbajj.html
kader
--- Guidry, Chris [EMAIL PROTECTED]
wrote:
Hi All,
There has been some discussion in the past regarding
moving
the audit tables out of the SYSTEM
From Metalink:
Bookmark
Default Font
Go to End
Doc ID:
Note:1019377.6
Subject:
Script
Nope. All tables are a degree of 1. And according to the docs,
specifying the degree parameter when calling DBMS_STATS overrides the
table default.
Rich Jesse System/Database Administrator
[EMAIL PROTECTED] Quad/Tech International, Sussex, WI USA
Yep, and here's how (for 8i).
alter table sys.aud$ move tablespace tools
storage (initial 128K next 128K);
alter index sys.i_aud1 rebuild tablespace tools
storage (initial 128K next 128K);
Guidry, Chris wrote:
Hi All,
There has been some discussion in the past
Chris:
It can be done, but Oracle, from my understanding, does not officially
support this. Remember, in their eyes, disk is cheap. Here is a synopsis
of what is needed.
1. After the database is built, create a copy of the
$ORACLE_HOME/rdbms/admin/sql.bsq file. This contains the create
Raj,
Yours is one of many license monitoring questions that keep haunting the
list from time to time I regret that we seldomly get a decent answer. That
being the case the fact that we've just gone through a visit from Oracle
License Management Services I asked the LMS rep if he would
I meant , no one runs a DML against those tables during
the running of the program.
Right. If there is no DML while your program is running, then the snapshot
error is due to DML that was run earlier. The solution is simple.
In the programs that do large batch inserts or updates on these
Hi folks -
I have received a stream of requests from developers/production support (
yep, same group, dont ask ) to do ad-hoc data massaging in the production
databases. Since I don't know the applications that well, it's hard for me
to push back these requests when told that if the script don't
Cherie,
In my experience, histograms seem a bit hit or miss but in the
cases where they've worked, the performance improvement
has been good or even fantastic. In the cases where they haven't
helped, I've simply removed them.
As I observed before, histograms help only when literal
I'd be interested to know of the bugs you've found on DBMS_STATS; most of
the bugs I've seen logged against it were created due to differences with
ANALYZE, and in the end it was determined that ANALYZE produced the wrong
result, not DBMS_STATS...
Anyway, the symptoms you describe match those
Hi Mike,
I've been using the Legato/RMAN combination for a couple of years now. The
recovery catalog resides on the Legato Backup server. Once all the database
backups finish, we then do a cold backup of the recovery catalog database
straight to tape.
I don't know what others do, but this if
SELECT USERNAME FROM DBA_USERS WHERE USERNAME = 'LBACSYS';
if returns row then Label security installed.
naren
-Original Message-
Sent: Wednesday, 29 May 2002 10:25
To: Multiple recipients of list ORACLE-L
Hi List,
How can I find out Oracle Label Security is installed or NOT?
List, slightly off topic but
Unix OS
I need to join lines/ words in a file.
So that it must be in a readable Oracle format.
They are seperated by a newline.
Here is a snippet of what the file looks like.
FILE1
delete from JDAPROD.HBI_LOST_SALES where SKU_TECHNICAL_KEY = 1410 and
STORE_TECH
$ perl -e 'print join \n, ' myfile;
-- Suhen Pather [EMAIL PROTECTED]
List, slightly off topic but
Unix OS
I need to join lines/ words in a file.
So that it must be in a readable Oracle format.
They are seperated by a newline.
Here is a snippet of what the file looks like.
FILE1
Dennis,
If management is Ok with this ( have you asked? ) you need to
take some steps to protect your database, your job and your reputation.
'cuz the duhvelopers will do their best to destroy all three.
1. You need a test database with a reasonable amount of test data
2. Your duhvelopers
oops, wrong direction -- you wanted to take out the newlines.
use chomp and print the resulting array with $, left at the
default value:
$ perl -e 'chomp (my @a = ); print @a' myfile [anotherfile ...];
i.e., read from ARGV, put it all in an array, slice off the
input record separators;
Here's a perl one liner:
perl -ne 'chomp; print; print qq{\n} if /\;$/' file1.txt newfile.txt
If isn't perfect. An 'and' at the end of the line will be joined with
the beginning of the next line, which is not right.
I use the following two regular expressions to create executable SQL
from
Flex, Bison some programming will probably do the trick.
There is a nice O'Reilly book dealing with Lex Yacc wnd even
nicer book dealing with the C programming language.
The Good Book is: Brian Kernighan and Dennis Ritchie: The C Programming
Language.
You should get the King James (ANSII)
List, slightly off topic but
Unix OS
I need to join lines/ words in a file.
So that it must be in a readable Oracle format.
They are seperated by a newline.
Here is a snippet of what the file looks like.
I am trying using sed but cant seem to work it out.
I could use tr but if there are
-- Jared Still [EMAIL PROTECTED]
If isn't perfect. An 'and' at the end of the line will be joined with
the beginning of the next line, which is not right.
Don't strip the newlines, replace them with white space:
perl -e 'undef $/; ($a=ARGV) =~ s/\n+/ /g; print $a' \
[file
Hey this is a not a trick question, without buying the books
listed below is there a way using the std awk, sed, tr *nix
utilities.
Suhen
Flex, Bison some programming will probably do the trick.
There is a nice O'Reilly book dealing with Lex Yacc wnd even
nicer
1 - 100 of 103 matches
Mail list logo