rfcfold is a Bash script that folds or unfolds a text file according to RFC 8792, Handling Long Lines in Content of Internet-Drafts and RFCs. The (un)folding operations are implemented with grep and sed.

Invoking rfcfold with the -h option shows usage information:

Folds or unfolds the input text file according to RFC 8792.

Usage: rfcfold [-h] [-V] [-d] [-q] [-s <strategy>] [-c <col>] [-r] -i <infile> -o <outfile>

  -s: strategy to use, '1' or '2' (default: try 1, else 2)
  -c: column to fold on (default: 69)
  -r: reverses the operation
  -i: the input filename
  -o: the output filename
  -d: show debug messages (unless -q is given)
  -q: quiet (suppress error and debug messages)
  -h: show this message
  -V: print version information

Exit status code: 1 on error, 0 on success, 255 on no-op.

1. Fold the file original.txt and write the result to folded.txt:

   rfcfold -i original.txt -o folded.txt

2. Unfold the file folded.txt and write the result to unfolded.txt:

   rfcfold -r -i folded.txt -o unfolded.txt

Noteworthy changes between versions of rfcfold can be found in the NEWS.md file.

rfcfold has some limitations, because its primary intended use is for text file inclusions in IETF RFC and Internet-Draft (I-D) documents, and because it is implemented as a Bash script relying on grep and sed.

The primary application of rfcfold lies in creating text files that can be included in IETF RFC and I-D documents by ensuring a maximum line length (in characters). Thus the input file is not allowed to contain TAB characters, because they are not allowed for text fragments in an RFC.

Most ASCII control characters and non-ASCII characters are problematic in the context of IETF RFC and I-D documents, and rfcfold emits a warning if it finds them in the input data. IETF RFC and I-D documents allow limited use of UTF-8 encoded Unicode characters, but rfcfold does not take Unicode specifics into account.

Depending on the operating system respectively the available grep and sed implementations, rfcfold may be able to process Unicode text files, but with limitations. The display width of Unicode characters is not taken into consideration when folding. Every character is assumed to have single width. Since printable Unicode characters can be wider than single width, and some implementations even display some single width characters with more than single width, the end result may not look as expected. There are additional complications with Unicode, e.g., combining characters, none of which are taken into account by rfcfold. As such Unicode is not supported by rfcfold.

Usage of both grep and sed conforms to the POSIX specification.

On non-GNU systems, GNU tools are often installed to complement the system-provided ones. On such systems, GNU grep and GNU sed are used if they are available as ggrep and gsed.

Use of both GNU grep and GNU sed may increase the maximum usable value for the folding column over using system-provided grep and sed.