Wrong documentation about erb -P?

ruby-talk
1

according to "man erb", the -P options: Evaluates lines starting with %
as Ruby code and removes the tailing EOLs.

However, it seems to be the opposite: by default, erb evaluates lines
starting with % as Ruby code;
and by specifying the -P option, erb does.

$ echo "hello" | erb -x --
#coding:ASCII-8BIT
_erbout = ''; _erbout.concat "hello\n"
; _erbout.force_encoding(__ENCODING__)

$ echo "%hello" | erb -x --
#coding:ASCII-8BIT
_erbout = ''; hello <<<< ruby code
_erbout.force_encoding(__ENCODING__)

$ echo "hello" | erb -P -x --
#coding:ASCII-8BIT
_erbout = ''; _erbout.concat "hello\n"
; _erbout.force_encoding(__ENCODING__)

$ echo "%hello" | erb -P -x --
#coding:ASCII-8BIT
_erbout = ''; _erbout.concat "%hello\n" <<<< no ruby code
; _erbout.force_encoding(__ENCODING__)

is it a bug in the "man erb" documentation,
or I am misunderstanding it?

···

--
Posted via http://www.ruby-forum.com/.

2

Looks like the man page for erb is just wrong:

$ erb --help
print this help
erb [switches] [inputfile]
   -x print ruby script
   -n print ruby script with line number
   -v enable verbose mode
   -d set $DEBUG to true
   -r library load a library
   -S safe_level set $SAFE (0..4)
   -E ex[:in] set default external/internal encodings
   -U set default encoding to UTF-8.
   -T trim_mode specify trim_mode (0..2, -)
   -P ignore lines which start with "%"

erb (the commandline tool) has always defaulted to interpreting lines beginning with % as Ruby code: add bin/erb · ruby/ruby@4de16df · GitHub

So just a documentation issue in the man page which has also been there forever. File a bug?

-Justin

···

On 11/22/2013 02:34 AM, David P. wrote:

according to "man erb", the -P options: Evaluates lines starting with %
as Ruby code and removes the tailing EOLs.

However, it seems to be the opposite: by default, erb evaluates lines
starting with % as Ruby code;
and by specifying the -P option, erb does.

$ echo "hello" | erb -x --
#coding:ASCII-8BIT
_erbout = ''; _erbout.concat "hello\n"
; _erbout.force_encoding(__ENCODING__)

$ echo "%hello" | erb -x --
#coding:ASCII-8BIT
_erbout = ''; hello <<<< ruby code
_erbout.force_encoding(__ENCODING__)

$ echo "hello" | erb -P -x --
#coding:ASCII-8BIT
_erbout = ''; _erbout.concat "hello\n"
; _erbout.force_encoding(__ENCODING__)

$ echo "%hello" | erb -P -x --
#coding:ASCII-8BIT
_erbout = ''; _erbout.concat "%hello\n" <<<< no ruby code
; _erbout.force_encoding(__ENCODING__)

is it a bug in the "man erb" documentation,
or I am misunderstanding it?