how to create mate data in talend

[Anonymous]

Hi Team,

I have txt file, i want to load in oracle table. Below is file format. It is possible to create meta data from below file contents in talend.

Report:212
date:21-02-2019
------------------------
no name col3
------------------------
02 sanajay sagar
03 xxx yyy

Report:213
date:21-02-2019
------------------------
no name col3
------------------------
05 zxz sa
06 ew qa

ID: 121
Row count: 4

Thanks

Shree

[Anonymous]

Hi,

What does your expected output look like?

In talend, you can centralize file metadata in repository. Here is online document about:TalendHelpCenter:Centralizing File Delimited metadata

Best regards

Sabrina

[Anonymous]

Hi Sabira.

Thanks for your input.

Please find the attached file.

How to create metadata for file, This file is system generated am sharing sample file contents.

Thanks

Shree

sample_text.txt

[Anonymous]

Hello,

What does your expected output look like?

no;name;col3
02;sanajay;sagar
03;xxx;yyy
05;zxz;sa
06;ew;qa

Best regards

Sabrina

[Anonymous]

Hi Sabira,

Yes, Below expected output.

no;name;col3
02;sanajay;sagar
03;xxx;yyy
05;zxz;sa
06;ew;qa

Kindly suggest how to implement or how to achieve this scenario.

Thanks

Shree

[Anonymous]

Hi Shree
No a component can be used to extract the data directly from this file. If the format is fixed, I would suggest you to read the file line by line with tFileInputRaw component, then remove the useless lines like starting with "Report", "date", "---", ..., finally, extract the fields with tExtractDelimitedFields, the job looks like:
tFileInputRaw--main--tJavaRow--main--tExtractDelimitedFields--main--tOracleOutput

Hope it helps you!

Regards
Shong

[Anonymous]

Sure Shong,

Let me try as mentioned flow. Let you know if i face any issues.

Thanks

Shree

[Anonymous]

Hi Shong,

Can you give sample example to remove "Report", "date", "---", ..., using tJavaRow. 

I don't know java. May be i am asking simple/ silly question here.

Thanks

Shree

[Anonymous]

Sorry, it was a mistake on my previous post, I meant to use tFileInputFullRow to read the file line by line, and use tFilterRow to remove the useless lines,
tFileInputFullRow--main--tFiterRow--main--tExtractDelimitedFields--tLogRow

on tFilterRow, check the 'use advanced mode' checkbox, and write these Java codes:
(!input_row.line.startsWith("Report"))&&(!input_row.line.startsWith("no name col3"))&&(!input_row.line.startsWith("------"))&&(!input_row.line.startsWith("date"))&&(!input_row.line.startsWith("ID"))&&(!input_row.line.startsWith("Row count"))&&(!input_row.line.startsWith(" "))

Regards
Shong

[Anonymous]

Hi Shong,

No problem,  let me try this code.

Thanks

shree