In a literature review paper, if we consider selected journals and span of publications, is it possible and necessary to consider all the articles published in that time span. If not, whether the review paper gives a rational outcome?
If we consider a time span in a review paper, is it necessary to include all the articles published in that time span?
My suggestion would be do your best to review all the articles during that time span / span of publications or until the state of saturation surfaces. However, you might not need to cite all the articles in your writing especially when there are too many whereby you need to perform some analyses & apply some good judgement calls to provide the significant ones e.g.
the most recent articles include citations & references underpinning on previous / older articles - so you can cite the latest articles especially when they are changes to the latest definitions of terms, ideas, concepts etc.
you can include the oldest articles / books especially when those definitions, ideas, concepts, theories, methods etc. are the classic ones.
after completed your rigorous literature review - you can apply your judgement calls how to illustrate your literature e.g. based on a timeline road map, based on major schools of thoughts, based on hierachical approach, mixed of above etc.
you might need to perform analyses for all the articles reviewed & then synthesize them in different format to present to readers for clarity sake - this can be one of the values of your review paper.
In order to perform the abovementioned, It is also advisable to track the literature you'd reviewed so that you can perform the needed analyses & syntheses etc. I usually use the following tracker template in Excel - see RG link attached.
As dear Professors indicated above. You should not miss any relevant information while relaying on a limited array of publications. Reviews are preferred to be as comprehensive as possible and this is one of the key factors which determine paper's quality.
You should select some parameters to explore the related literature otherwise its very tedious job to end up with the research. Parameters may be time-frame (as u mentioned in your question), metrics (scopus or SCI or SSCI or ABDC etc).