by what tool I know the genome coverage (X) of assembly contigs.
Without knowing more information, see this post.
Thank you for your reply. but I can't use the calculated theoretical C = R x L / G, because I have the reads length 32-250 (so it is not a single value)
and I can't use bedtools because I want to know the fold-covrage from assembly of Novo result (not from mapping).
If you want to know the coverage of your denovo assembly you could try mapping the reads to your assembly and then calculating coverage as suggested by the following answers. A not-so-correct way is to see how much of data you have (in total bases) in your assembly and use this as the denominator to divide the total data you have (i.e in place of G in the above formula). This post might help -
Counting Number Of Bases In A Fastq File
But you would obviously be over-expecting coverage since there would be missing data due to a fragmented assembly. So you definitely need to map back your reads to your contigs (keeping these as a reference) to understand how much of coverage you might have.
Googe is full of answers:
Tools To Calculate Average Coverage For A Bam File?
I want the fold-coverage from contigs (fatsa file) not from mapping (bam file)
Do you have a reference genome available? If not, what "reference" (would that be the de novo assembly you seem to refer to in the post above) are you planning to use to calculate fold coverage?
I made a de novo assembly of a bacterial strain and I want to know their genome coverage from fasta file (contigs) ie how many times the sequenced reads but not according to a reference
In that case your contigs are going to be the "reference". You would need to align your reads to that "reference" and then calculate coverage using one of the methods @dago noted above.