LAE

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: phudson

Hi, how do I extract Name and corresponding Service Profile Name from below file format? Note each customer has 10 rows of data.

Service ID : 1
Name : kennethpatnett
Subscriber Name : BTL_BMP
Service Profile Name : BMAX_512
SU MAC Address : 00-10-e7-8a-90-92
Hybrid VLAN Mode : On
VLAN Classification Mode : Off
Access Vlan : 243
Admin Status : Enabled
Operation Status : Up
-------------------------------------
Service ID : 2
Name : Lo Chih En
Subscriber Name : BTL_BMP
Service Profile Name : BMAX_1024
SU MAC Address : 00-10-e7-8a-88-df
Hybrid VLAN Mode : On
VLAN Classification Mode : Off
Access Vlan : 245
Admin Status : Enabled
Operation Status : Up
-------------------------------------
Service ID : 3
Name : Test unit
Subscriber Name : BTL_BMP
Service Profile Name : BMAX_1024
SU MAC Address : 00-10-e7-8a-8f-23
Hybrid VLAN Mode : On
VLAN Classification Mode : Off
Access Vlan : 244
Admin Status : Enabled
Operation Status : Up

I need this result:

Name, Service Profile Name
kennethpatnett, BMAX_512
Lo Chih En, BMAX_1024
Test unit, BMAX_1024

Any suggestions? See attachment.
Attachments:
BMP 11102011.TXT

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: mgajdosik

Hi,

for that you need kind of specific reader, I have created one very simple for you, you probably may need to clean the code a bit, in case of some additional items will appear in your datafile.

Of course there are thousands of different ways how to handle reading such files, this is only one of them.

It extracts all the values from your file, not only the ones you specified, you can exclude them in next filter.

If you have any questions, please let me know.
Attachments:
Reader.brg

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: phudson

Pivot node giving this error: Error while starting execution of node ''
No executor for type: core::Pivot - Data To Names

I think it might be an issue with the version of LAE I am using.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: mgajdosik

Hi,

I guess you are not at 4.5 version, so that is why it does not work for you. So instead you can use rotate node, so just replace the last node in the graph I created.

and put in following params:

NameField = Field
ValueField = Value
PivotField = SID
IgnoreExtraNames = True

It should work almost the same way as the previous with small difference. If there is certain field appearing only for some SIDs, then this field will not be in the output. In this case it is VLAN field, which is popullated only for one customer I think.

once you upgrade to 4.5.X just replace the rotate by the orginal node I sent and you are still ok.

Let me know if you have any further issues or questions.

Marek

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: henk01

Hello,

Actually, all can be done is a single aggregation node. The key is to keep retrieving values, until a �group� separator is found (here a line with �------�), or when reaching the end of the file.

When a �group� separator is found, simply output the current values, and reset the values for the next section (after the emit).

Please find the BRG attached, please note:
Record.split(" : ")[0] is the same as Record.split(" : ").getItem(0)
Attachments:
reader_test_.brg

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: henk01

Hello,

Just to clarify, I did mention the �coding� option to show another option which may be of interest. I also assumed that the record does consist of 10 rows.

Of couse Marek�s solution is a very good one, as no hard-coding of column names is required. No maintenance is required if fields are added in future.

So just for completeness, including the first solution (Marek�s) as well but using the static data node to produce 10 rows for each record (and using rotate).

Best regards,

Henk
Attachments:
reader_test_2.brg

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: phudson

Thanks Henk. Works like a charm!!!

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: phudson

I have a similar scenario below. I need to be able extract:

IMSI, 1STPROFILE, 2NDPROFILE, 3RDPROFILE
702670160001001, PROFILE1 1, PROFILE4 4
702670160001003, PROFILE1 1, PROFILE4 4
702670160001010, PROFILE1 1, PROFILE4 4, PROFILE5 5

from:
MCC MNC MSIN PDPLIST
--------------------
702 67 0160001001
(PROFILE1 1 $)
(PROFILE4 4 $) $
702 67 0160001003
(PROFILE1 1 $)
(PROFILE4 4 $) $
702 67 0160001010
(PROFILE1 1 $)
(PROFILE4 4 $)
(PROFILE5 5 $) $

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: rboccuzzi

I'm not sure this will work, but on the small set you showed it should....I assume that I can clobber "(", ")", and "$" from your data everywhere. If that is true, I think this works. Note, it needs 4.5.x to work, as I use the new pivot node.

Cheers
Rich
Attachments:
reader_test_3.brg