Today I read about a blog explaining very detailedly on how to correctly use the key words FOR ALL ENTRIES IN.
The link is : http://www.cnblogs.com/panjun-Donet/archive/2010/10/20/1857100.html.
Recollection: this keywords feature is used when two tables can't be joined to fetch data, such as a cluster table BSEG and a transparent table BSIS, which two are in a sense intersected. Therefore, if we want to fetch data from BSEG and BSIS, we can first put the BSIS data into an internal table called IT_BSIS, then we use SELECT statements to get data from BSEG by using FOR ALL ENTRIES IN IT_BSIS where BUKRS = IT_BSIS-BUKRS and ...
In above situation, we successfully completed the manipulation of fetching data from two intersected tables (Specificly refer to BSEG and BSIS).
Using FOR ALL ENTRIES IN feature is kind of disturbing, or rather wasting memory, sometimes. Especially when there are a large amount of data in the IT_BSIS table, and in this case the actual manipulation is getting all the conditions together and separating them with OR, and the whole condition will seem too abundant. It is better to consider if vital before we directly use this feature.
A necessary reminder is that Check the IT_BSIS table before we use this feature, or else the system might go dumped due to high load of data processing.