Configure multiple LINE_BREAKER strings?

[woodcock]

I have line breaks signified by 2 different strings.

This works (keeping BK1 text as part of next event):

LINE_BREAKER = ([\r\n]+)(BK1)

This works (discarding BK2 text as part of breaker):

LINE_BREAKER = ([\r\n]+BK2)

But this does not work:

LINE_BREAKER = (?:([\r\n]+)(BK1))|(?:([\r\n]+BK2))

I assume this doesn't work because the 3 captures are taken as $1, $2 and $3 (and $3 is undefined as far as LINE_BREAKER REGEX goes) instead of as $1, $2v1, $2v2.

[Ayn]

I find that dealing with LINE_BREAKER settings tends to get messy when often it's simpler and equally effective to achieve the same thing by configuring proper line merging settings instead. Have you investigated that option?

[woodcock]

Using LINE_BREAKER if it is a valid option, is always the most efficient impact on the Indexer, so I always start there.

[fairoldi]

You can try:

LINE_BRAKER=([\r\n]+(?=BK1)|[\r\n]+BK2)

However this is subject to lookahead limitations: if I remember correctly this means you can not use +,? and * in BK1 (not sure about {n})

[_d_]

Give this a try:

LINE_BREAKER = ([\r\n]+BK2)|([\r\n]+)(?=BK1)

Hope this helps.

> please upvote and accept answer if you find it useful - thanks!

[woodcock]

No worky; this completely ate the records starting with BK1 (for the same $1/$2/$3 reason/problem).