need help in easytrieve-multiple reading of sequential file

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

i have a situation like, there are 2 files A and B(BOTH ARE PS). For each record in A the FILE B has to be read for the some condition.
for the above logic i tried 2 ways.
1.)job input A
GET B
CHECK LOGIC
PRINT O/P
IF i am doing this then the FILE B is reaching the end of file and giving the logical error.can u help me by saying how can i make the pointer go the first record of B after each record in A is processes.
2.)i tried IF MATCHED(FILEA(KEY) FILEB(KEY))
But it is only reading the file B ONLY ONCE.

CAN AMYONE SUGGEST A SOLUTION FOR THIS.BOTH FILES ARE PS

Manuneedhi K · Posted: Wed Jun 04, 2008 3:54 pm

If FILEB is relatively try declaring it as a TABLE and do a search.

Thanks
Manu

dick scherrer · Posted: Wed Jun 04, 2008 7:58 pm

Hello Chandra and welcome to the forums,

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

hi
thanks for ur reply, and i have some questions about a table which i thought can be helped by u.
1.)my sequential file has around 600 rceords.so its an external table
2.)do i have to get the records sorted in ALPHABATECIAL ORDER IN THE PS to declare it as a table.
3.)file inpfile table {literal-10}
arg 1 16 a desc 17 44 a

this is the decleration i have to give for external table decleration.
if my file size is not variable can i declare it to have an approximate size.

if i am mistaken pls correct me.

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

hi
thanks for the reply.
wat my inclination towards "if matched " was
if FILEA has no dups FILEB has dups and
now for every record in FILEB, FILEA has to be searched.

so

job input(filea(key1) -
fileb(key2))

if matched
{process}
goto job

i was thinking that for if matched the FILEA is not bieng searched every time fileb is read.
our requirement is it should be read multiple times.so after every job input will the filea be read from the top again.

dick scherrer · Posted: Fri Jun 06, 2008 2:09 am

Hello,

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

i tried using "matched" and found that
both the files should be soted before using "matched"

FILEA - HAS NO DUPS(THINK ITS THE MASTER)
FILEB - HAS DUPS

BOTH FILES SHOULD BE SORTED BEFORE USING MATCHED

NOW MY REQUIREMENT IS GETTING SATISFIED WHEN I USE "IF MATCHED (FILEA(KEY1)+
(FILEB(KEY2)) "
FOR EVERY KEY IN FILEB WE HAVE MANY DUP KEYS IN FILEB.
SO ITS ABLE TO MATCH WHEN USING THE ABOVE LOGIC.

EVEN IF THE FILE ORDER IN "IF MATCHED()()" IS CHANGED ITS NOT ABLE TO DO IT.

Manuneedhi K · Posted: Fri Jun 06, 2008 2:55 pm

Can you try using the Search TABLE and then using the IF MATCHED. This might not be efficient method but it works.

Mods,

There are two topics running for the same issue i guess.

Other one is ibmmainframes.com/viewtopic.php?p=136824#136824

Can we merge these two or close one of them

dick scherrer · Posted: Fri Jun 06, 2008 7:43 pm

Hello,

This topic and one titled "Processing table in eastrieve " are the same. I am closing the other topic as we don't need the same discussion 2 places.

I suggest you try the previously mentioned match/merge code. It should only take a short amount of time and you will have another "tool" that you will be able to use over and over for other requirements.

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

thank u for ur suggestion,i will try that

dick scherrer · Posted: Sat Jun 07, 2008 12:34 am

You're welcome - good luck

Someone will be here if there are questions.

d

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

hi this is working gr8

JOB INPUT SAM2
W-COUNTER = 0
MOVE SAM2-FIRST TO W-SAM2-FIRST
SEARCH SAM1 WITH W-SAM2-FIRST GIVING W-DESCRIPTION
IF SAM1
W-FLAG = W-FLAG + 1
DISPLAY 'FOUND'
W-COUNTER = W-COUNTER + 1
DISPLAY W-FLAG
GO TO JOB
END-IF
IF W-COUNTER EQ 0
DISPLAY 'NOTFOUND' W-SAM2-FIRST
END-IF
GOTO JOB
the sequential table is the only thing thats sorted the other sequential file need not be sorted.

thanks

dick scherrer · Posted: Mon Jun 09, 2008 7:36 pm

Hello,

Well, hopefully, the volume will be very small and the process will run very seldom. While that code may work it will waste an incredable amount of system resources (the larger the run, the more waste). I suppose this could be a good thing if your family sells hardware. . .

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

but what can be done if not a table, sequential processing is itself resource consuming.
we have to use vsams instead.but our requirement is not the same.

dick scherrer · Posted: Tue Jun 10, 2008 1:27 pm

Hello,

Have you yet downloaded and used the 2-file match/merge code? Do so and i believe you will see a reduction in resources used.

Reading 2 files sequentionally uses less resources than running back and forth thru some table/array.

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

ya i downloaded that code given and i will see the feasibility of implementing it.

thats a gr8 suggestion

thank u

dick scherrer · Posted: Thu Jun 12, 2008 3:29 am

Hello,

chandrarkar · New User Joined: 27 May 2008 Posts: 24 Location: india

hi

i have a scenario like i have to insert records in a table that has been defined by me

the layout of table is

inpfile
arg 1 21 A
desc 22 44 A

WHEN I AM TRYING TO INSERT RECORDS IN THIS TABLE USING ARG AND DESC ITS SAYING *******B082 NAME IS UNDEFINED - ARG
*******B082 NAME IS UNDEFINED - DESC

dick scherrer · Posted: Fri Jun 13, 2008 1:22 pm

Hello,

Please post the problem code and the associated info from the sysout=.