At one point or another of any such retrieval, you must identify the positions and reading frames of the coding regions. This can be done:

1) from existing annotations (perhaps you can get a BED or GFF3 file from the same source as the sequences), use getfasta from bedtools, but many sources, such as Ensembl! or UCSC Genome Browser will let you download just the coding sequences with proper settings

2) by similarity to known transcripts (maybe use tblastx?) or proteins (use blastx from the BLAST software suite) then transform to BED or GFF3 or parse the BLAST output in some other way

3) use gene finder software to annotate the sequence and get BED or GFF3 files, then proceed as above

You can use BLAST to look for cDNA for predicted coding regions.  Keep in mind these are just a prediction and need to be verified at the bench.

simply go to the following page.. paste your sequence and submit. Rest is self explanatory..

https://www.ncbi.nlm.nih.gov/orffinder/

Best of luck..

SD