I am writing a batch job for Postgres for first time. I have return ".sh" file, which has a command with out any out put in the log or console.
Code
export PGPASSWORD=<password>
psql -h <host> -p <port> -U <user> -d <database> --file cleardata.sql > log\cleardata.log 2>&1
What I did at cammond line
su postgres
and run ./cleardatasetup.sh
Nothing is happening.
Please note : When I try psql command in Unix command line, I am getting message as some SQL exception which is valid.
Can any one please help me in this regard.
You probably wanted to create log/cleardata.log but you have a backslash where you need a slash. You will find that the result is a file named log\cleardata.log instead.
The backslash is just a regular character in the file's name, but it's special to the shell, so you'll need to quote or escape it to (unambiguously) manipulate it from the shell;
ls -l log\\cleardata.log # escaped
mv 'log\cleardata.log' log/cleardata.log # quoted
Related
I am trying to find a way to pass a file to psql while using '\copy'. There is a question posted here use sql file in "\copy" psql command line asking a similar thing, but the accepted solution doesn't actually pass a file to psql, it passes the contents of the file, so this cannot be done when the contents of the sql file exceeds the maximum allowed length of the psql command.
e.g. something like this psql -c data_base "\copy \file_path <file.sql> To './test.csv' With CSV"
Using copy rather than \copy, you can do it like this:
(echo "copy ("; cat file.sql ; echo ") to STDOUT with CSV")| psql -X > ./test.csv
I'm dealing with psql for the first time and I need a command that executes a .sql file and after executing, it should make a .log file with the script output. I'm looking for something similar to this other command that I use with SQL Server:
sqlcmd -U userid -P password -S serveraddress -i path_to_the_sql_file -o path_where_to_save_log_file
Can you help me, please? :-)
I would spend some time at the psql page.
A quick example:
psql -U userid -h serveraddress -d some_db -f path_to_the_sql_file -L=path_where_to_save_log_file
With Postgres you need to connect to a database with a client. There are other options for inputting commands and capturing output at the link above.
In postgresql 9.3.1, when interactively developing a query using the psql command, the end result is sometimes to write the query results to a file:
boron.production=> \o /tmp/output
boron.production=> select 1;
boron.production=> \o
boron.production=> \q
$ cat /tmp/output
?column?
----------
1
(1 row)
This works fine. But how can I get the query itself to be written to the file along with the query results?
I've tried giving psql the --echo-queries switch:
-e, --echo-queries
Copy all SQL commands sent to the server to standard output as well.
This is equivalent to setting the variable ECHO to queries.
But this always echoes to stdout, not to the file I gave with the \o command.
I've tried the --echo-all switch as well, but it does not appear to echo interactive input.
Using command editing, I can repeat the query with \qecho in front of it. That works, but is tedious.
Is there any way to direct an interactive psql session to write both the query and the query output to a file?
You can try redirecting the stdout to a file directly from your shell (Win or Linux should work)
psql -U postgres -c "select 1 as result" -e nomedb >> hello.txt
This has the drawback of not letting you see the output interactively. If that's a problem, you can either tail the output file in a separate terminal, or, if in *nix, use the tee utility:
psql -U postgres -c "select 1 as result" -e nomedb | tee hello.txt
Hope this helps!
Luca
I know this is an old question, but at least in 9.3 and current versions this is possible using Query Buffer meta-commands shown in the documentation or \? from the psql console: https://www.postgresql.org/docs/9.3/static/app-psql.html
\w or \write filename
\w or \write |command
Outputs the current query buffer to the file filename or pipes it to the shell command command.
Please try this format as I got the output from the same:
psql -h $host -p $port -q -U $user -d $Dbname -c "SELECT \"Employee-id\",\"Employee-name\" FROM Employee_table" >> Employee_Date.csv
I need the output in a CSV file.
I am trying to perform a postgres dump of a specific table using -t. However, the table has a capital letter in it and I get a "No matching tables were found." I tried using quotations and double quotations around the table name but they did not work. How can I get pg to recognize the capitals? Thanks!
pg_dump -h hostname dbname -t tableName > pgdump.sql
Here is the complete command to dump your table in plain mode:
pg_dump --host localhost --port 5432 --username "postgres" --role "postgres" --format plain --file "complete_path_file" --table "schema_name.\"table_name\"" "database_name"
OR you can just do:
pg_dump -t '"tablename"' database_name > data_base.sql
Look to the last page here: Documentation
The above solutions do not work for me under Windows 7 x64. PostgreSQL 9.4.5. But this does, at last (sigh):
-t "cms.\"FooContents\""
either...
pg_dump.exe -p 8888 --username=user -t "cms.\"FooContents\"" basdb
...or...
pg_dump.exe -p 8888 --username=user -table="cms.\"FooContents\"" basdb
Inside a cmd window, I had to put three (!) double quotes around the table name if it countains upper case letters.
Example
pg_dump -t """Colors""" database > database.colors.psql
This worked for me:
pg_dump -f file.sql -U user -t 'schema.\"Table\"' database
As part of a node script I had to surround with single and double quotes, e.g.
` ... --table 'public."IndexedData"'`
The accepted solution worked in a bash console, but not as part of a node script, only the single quote approach.
Thanks to #Dirk Zabel suggestion, the following worked for me:
Windows 10 CMD
pg_dump -d "MyDatabase" -h localhost -p 5432 -U postgres --schema=public -t """TableName""" > TableName.sql
Bash
pg_dump -d "MyDatabase" -h localhost -p 5432 -U postgres --schema=public -t "\"TableName\"" > TableName.sql
Powershell
the good (shortest)
& 'C:\Program Files\PostgreSQL\12\bin\pg_dump.exe' -d db_name -t '\"CasedTableName\"'
the bad (requires --%)
& 'C:\Program Files\PostgreSQL\12\bin\pg_dump.exe' --% -d db_name -t "\"CasedTableName\""
the ugly (requires `")
& 'C:\Program Files\PostgreSQL\12\bin\pg_dump.exe' -d db_name -t "\`"CasedTableName\`""
The main point of confusion for me was the absolute necessity of having \" in there. I assumed that maybe there was a weird bug in the way powershell or psql was parsing the arguments, but it turns out it's explained in the docs:
Some native commands expect arguments that contain quote characters. Normally, PowerShell's command line parsing removes the quote character you provided. The parsed arguments are then joined into a single string with each parameter separated by a space. This string is then assigned to the Arguments property of a ProcessStartInfo object. Quotes within the string must be escaped using extra quotes or backslash (\) characters.
And of course ProcessStartInfo.Arguments Remarks tells us:
To include quotation marks in the final parsed argument, triple-escape each mark.
what is wrong with this command:
pg_dump -U postgres -W admin --disable-triggers -a -t employees -f D:\ddd.txt postgres
This is giving error of too many command-line arguments
Looks like its the -W option. There is no value to go with that option.
-W, --password force password prompt (should happen automatically)
If you want to run the command without typing is a password, use a .pgpass file.
http://www.postgresql.org/docs/9.1/static/libpq-pgpass.html
For posterity, note that pg_dump and pg_restore (and many other commands) cannot process long hyphens that word processors create. If you are cut-pasting command lines from a word processor, be sure it hasn't converted your hyphens to something else in editing. Else you will get command lines that look correct but hopelessly confuse the argument parsers in these tools.
pg_dump and pg_restore need to ask password on commandline, if you put it command, they always give "too many command-line arguments" error. You can use below for setting related environment variable in commandline or batch file:
"SET PGPASSWORD=<password>"
so that you are not asked to enter password manually in your batch file. They use given environment variable.
Instead of passing password with -W flag start with setting temporary variable for postgres:
PGPASSWORD="mypass" pg_dump -U postgres--disable-triggers -a -t employees -f D:\ddd.txt postgres
-W -> will prompt for a password
to take full DB dump
use some thing like
pg_dump -h 192.168.44.200 -p 5432 -U postgres -W -c -C -Fc -f C:\MMM\backup10_3.backup DATABASE_NAME
I got this from copy-pasting, where 1 of the dashes were different.
Was: –-host= (first dash i a "long" dash)
Corrected to --host= solved it
Another option is to add ~/.pgpass file with content like this:
hostname:port:database:username:password
read more here
Additionally, if you don't want password prompt, use connection string directly.
pg_dump 'postgresql://<username>:<password>#localhost:5432/<dbname>'
So, combination with options in original question,
pg_dump 'postgresql://postgres:<password>#localhost:5432/postgres' --table='"employees"' --format='t' --file='D:\ddd.txt' --data-only --disable-triggers
(Don't forget to use quotes when you have letter-casing issues)
reference:
https://www.postgresql.org/docs/current/app-pgdump.html
Postgres dump specific table with a capital letter
2021-11-30, pg v12, windows 10
pg_dump -U postgres -W -F t postgres > C:\myfolder\pg.tar
-U "postgres" as username,
-W to prompt for psd,
-F t means format is .tar,
> C:\myfolder\pg.tar is the destination path and filename