DB to CSV

[Mitja ÄŒebokli]

Hello,

i am very new to ruby and i have a major problem i would like to have
some help on...

I would like to execute a query on a database and export the values to
the CSV file. I have tried many different ways and i can't get it to
work.

This is what i have so far:

require 'oci8'
require 'csv'

time = Time.now
rows =[]

connection = OCI8.new("user", "pass", "srv")
dataset = connection.exec("select column from table")
while r = dataset.fetch()
rows << r.join("\n")
end
dataset.close
connection.logoff

CSV.open("c:\\Temp\\test_"+time.strftime("%d-%m-%Y_%H-%M")+".csv","w",:col_sep
=> ";") do |csv|

rows.each do |row|
csv << row
end
end

I am getting this error:
csv.rb:1700 undefined method 'map'

What i want to achieve is pretty simple:
- have a header with names of columns from the database (automatic)
- write a CSV file with each record as a new line
- have a semicolon as a seperator

Thanks for any info on this.
BTW: no RAILS!

 

[Peter Hickman]

[Note: parts of this message were removed to make it a legal post.]

Does something like this work?

require 'csv'

time = Time.now
rows =[]

(1..10).each do |x|
rows << ['row', x, 'cheese']
end

CSV.open("testfile.csv","w",:col_sep => ";") do |csv|
rows.each do |row|
csv << row
end
end

It this works then look at your data.

Also I note that the line

rows << r.join("\n")

is probably not what you want. It will turn an array into a string.

a = ['row',1,'cheese']
a.join("\n") => "row\n1\ncheese"

Are you sure you want to do this?

 

[Peter Hickman]

[Note: parts of this message were removed to make it a legal post.]

Further hacking reveals that this might be a better option to create the
file.

CSV.open("testfile.csv","w", ";") do |csv|

Note we are no longer using the :col_sep => ";" notation but we do get the ;
as a record / column separator.

 

[Mitja ÄŒebokli]

Further hacking reveals that this might be a better option to create the
file.

CSV.open("testfile.csv","w", ";") do |csv|

Note we are no longer using the :col_sep => ";" notation but we do get
the ;
as a record / column separator.

Hi, thanks a lot for your help.

I have changed

while r = dataset.fetch()
rows << r.join("\n")
end

to

while r = dataset.fetch()
rows << r
end

and this now works as expected! No more problems about filling the CSV
file.

On the other hand, not using :col_sep is giving me an error:
can't convert String into Integer (TypeError)

Maybe the formatting of the line is wrong?

One more question on this matter.
How do I extract/use the table columns from the select file or if used
as "select *" to extract all the column names, and than use them as a
CSV header.

Thanks again helping me out, really appreciate it!

BR

 

[Mitja ÄŒebokli]

And one more thing

I have a problem, as it seems, with columns where data are numbers.
For example, row with ID number is returning me something like 0.2001E4

What's with that? Shouldn't the CSV export treat all the data coming
from DB as a string? Maybe there is a problem with wrong oci8 usage? (on
my side of course)

Thxs

 

[Peter Hickman]

On the other hand, not using :col_sep is giving me an error:
can't convert String into Integer (TypeError)

Maybe the formatting of the line is wrong?

Well if :col_sep works for you then use it. Perhaps this is a gem version
issue. It didn't like it for me.

One more question on this matter.

How do I extract/use the table columns from the select file or if used
as "select *" to extract all the column names, and than use them as a
CSV header.

Sorry I have no knowledge of the Oracle db driver so I can't help you with
that.

I have a problem, as it seems, with columns where data are numbers.

For example, row with ID number is returning me something like 0.2001E4

How is the data being returned from the database? If Oracle is returning th=
e
data in this format 0.2001E4 (as a string) then ruby will just copy it as
is. I suspect that Oracle is returning the float as a string because ruby
would display 0.2001E4 as 2001.0 if it was actually given as a number.

You might need to convert the data to get the values displayed in a sensibl=
e
way.

a =3D "0.2001E4" =3D> "0.2001E4"
a.to_f =3D> 2001.0
a.to_f.to_s =3D> "2001.0"

 

[KUBO Takehiro]

And one more thing

I have a problem, as it seems, with columns where data are numbers.
For example, row with ID number is returning me something like 0.2001E4

How did you define the ID number.
If it is defined as NUMBER, the column is not an integer and it may not fit=
to
Float. Thus ruby-oci8 fetch it as BigDecimal by default.
See: http://rubyforge.org/forum/forum.php?thread_id=3D47561&forum_id=3D107=
8

Could you redefine the ID as NUMBER(38) or so?