The only way to remove the added spaces is trimws(x, "r"). However, in some data sets, trailing whitespace at the end of a string may be meaningful.
Since R doesn't allow for NUL characters in strings, I propose defining a new class "padded.character" on top of the raw type and modifying make.fixedwidth() accordingly. Each atomic raw cell should have a width of max(max(nchar(x)), 1) + 1 to keep space for the null-terminating character. By making the field at least 2 bytes long, we can encode NA by placing a non-null character after the first NUL (i.e. '\0') that signals the end of the string.
S3 methods as.character.padded.character() and print.padded.character() should be overridden to recognize the special NA sequence and to remove NUL characters.
The only way to remove the added spaces is
trimws(x, "r"). However, in some data sets, trailing whitespace at the end of a string may be meaningful.Since R doesn't allow for
NULcharacters in strings, I propose defining a new class "padded.character" on top of the raw type and modifyingmake.fixedwidth()accordingly. Each atomic raw cell should have a width ofmax(max(nchar(x)), 1) + 1to keep space for the null-terminating character. By making the field at least 2 bytes long, we can encodeNAby placing a non-null character after the firstNUL(i.e.'\0') that signals the end of the string.S3 methods
as.character.padded.character()andprint.padded.character()should be overridden to recognize the specialNAsequence and to removeNULcharacters.