This is file is both documentation and a test file, showing how some txt2regex features work in practice, with the command line options required to trigger them and their expected result.
The clitest tool can identify and run all the commands listed here and check if their actual output matches the expected one. Just run clitest tests/features.md
.
Make sure all the commands use the same Bash version and the same txt2regex file.
$ txt2regex() { bash ./txt2regex.sh "$@"; }
$
When informing numbers and non-numbers (a5!6
in this test) when prompted for a number, the non-numbers are silently removed.
$ txt2regex --prog egrep --history '215¤a5!6'
Regex egrep: .{56}
$
When informing literal characters to be put inside a []
list, txt2regex will deduplicate those characters, because the repetition in this case is not meaningful ([aabbcc]
is the same as [abc]
).
$ txt2regex --prog egrep --history '24¤aabbbcab'
Regex egrep: [abc]
$
When informing literal characters to be put inside a []
list, some special cases have to be handled:
^
must not be the first char, otherwise it would mean a negated list-
must not be between two other chars, otherwise it would mean a range.]
must be the very first char, otherwise it would end the list prematurely.[
is not special since the list is already opened, nothing to be done in this case.
$ txt2regex --prog egrep --history '24¤^abc' # move ^ to the last position
Regex egrep: [abc^]
$ txt2regex --prog egrep --history '24¤a^bc' # ^ is not special in the 2nd position
Regex egrep: [a^bc]
$ txt2regex --prog egrep --history '24¤a-bc' # move - to the last position
Regex egrep: [abc-]
$ txt2regex --prog egrep --history '24¤-abc' # - is not special in the 1st position
Regex egrep: [-abc]
$ txt2regex --prog egrep --history '24¤a]bc' # move ] to the 1st position
Regex egrep: []abc]
$ txt2regex --prog egrep --history '24¤a[bc' # [ is not special
Regex egrep: [a[bc]
$ txt2regex --prog egrep --history '24¤^a[b-c]' # everything together
Regex egrep: []a[bc^-]
$
In some programs, it's required to escape the \
character when using it inside []
lists, making it \\
or even \\\\
.
$ txt2regex --all --history '241¤\'
Regex awk : [\\]
Regex chicken : [\\\\]
Regex ed : [\]
Regex egrep : [\]
Regex emacs : [\\\\]
Regex expect : [\\]
Regex find : [\]
Regex gawk : [\\]
Regex grep : [\]
Regex javascript: [\\]
Regex lex : [\\]
Regex mawk : [\\]
Regex mysql : [\\\\]
Regex perl : [\\]
Regex php : [\\\\]
Regex postgres : [\\]
Regex procmail : [\]
Regex python : [\\]
Regex sed : [\]
Regex tcl : [\\]
Regex vi : [\]
Regex vim : [\\]
$
The user has typed .*+?[]{}()|^$\
as a literal string.
Every metacharacter should be escaped so it will match as a literal character.
$ txt2regex --all --history '23¤.*+?[]{}()|^$\'
Regex awk : \.\*\+\?\[]{}\(\)\|\^\$\\
Regex chicken : \\.\\*\\+\\?\\[]{}\\(\\)\\|\\^\\$\\\\
Regex ed : \.\*+?\[]{}()|^$\\
Regex egrep : \.\*\+\?\[]\{}\()\|\^\$\\
Regex emacs : \\.\\*\\+\\?\\[]{}()|^$\\\\
Regex expect : \.\*\+\?\[]\{\}\(\)\|\^\$\\
Regex find : \.\*\+\?\[]\{}\()\|\^\$\\
Regex gawk : \.\*\+\?\[]{}\()\|\^\$\\
Regex grep : \.\*+?\[]{}()|^$\\
Regex javascript: \.\*\+\?\[]{}\(\)\|\^\$\\
Regex lex : \.\*\+\?\[]\{\}\(\)\|^$\\
Regex mawk : \.\*\+\?\[]{}\(\)\|\^\$\\
Regex mysql : \\.\\*\\+\\?\\[]{}\\()\\|\\^\\$\\\\
Regex perl : \.\*\+\?\[]\{}\(\)\|\^\$\\
Regex php : \\.\\*\\+\\?\\[]\\{}\\(\\)\\|\\^\\$\\\\
Regex postgres : \.\*\+\?\[]{}\(\)\|\^\$\\
Regex procmail : \.\*\+\?\[]{}\(\)\|\^\$\\
Regex python : \.\*\+\?\[]\{}\(\)\|\^\$\\
Regex sed : \.\*+?\[]{}()|^$\\
Regex tcl : \.\*\+\?\[]\{\}\(\)\|\^\$\\
Regex vi : \.\*+?\[]{}()|^$\\
Regex vim : \.\*+?\[]{}()|^$\\
$
Now try some Bash special chars to make sure nothing will break. Those chars should not be escaped since they are not metacharacters.
$ txt2regex --prog egrep --history '23¤#!&;/`"%>'
Regex egrep: #!&;/`"%>
$
Turn on all the options under the "a special combination" menu.
$ txt2regex --prog sed --history '26¤:012345'
Regex sed: [A-Za-z0-9_ \t]
$
Turn on all the options under the "a POSIX combination (locale aware)" menu.
$ txt2regex --prog egrep --history '27¤:01234567'
Regex egrep: [[:alpha:][:lower:][:upper:][:digit:][:alnum:][:xdigit:][:blank:][:graph:]]
$
If the program does not support POSIX character classes, a !!
string is shown instead.
$ txt2regex --all --history '27¤:0'
Regex awk : [[:alpha:]]
Regex chicken : [[:alpha:]]
Regex ed : [[:alpha:]]
Regex egrep : [[:alpha:]]
Regex emacs : [[:alpha:]]
Regex expect : [[:alpha:]]
Regex find : [[:alpha:]]
Regex gawk : [[:alpha:]]
Regex grep : [[:alpha:]]
Regex javascript: !!
Regex lex : [[:alpha:]]
Regex mawk : !!
Regex mysql : [[:alpha:]]
Regex perl : [[:alpha:]]
Regex php : [[:alpha:]]
Regex postgres : [[:alpha:]]
Regex procmail : !!
Regex python : !!
Regex sed : [[:alpha:]]
Regex tcl : [[:alpha:]]
Regex vi : [[:alpha:]]
Regex vim : [[:alpha:]]
$
If the program does not support using \t
inside []
lists to represent a tab character, a <TAB>
string is shown instead.
This is a reminder for the user that this string must be replaced by a literal tab to be able to use the regex.
$ txt2regex --all --history '26¤:5'
Regex awk : [\t]
Regex chicken : [\t]
Regex ed : [<TAB>]
Regex egrep : [<TAB>]
Regex emacs : [\t]
Regex expect : [\t]
Regex find : [<TAB>]
Regex gawk : [\t]
Regex grep : [<TAB>]
Regex javascript: [\t]
Regex lex : [\t]
Regex mawk : [\t]
Regex mysql : [\t]
Regex perl : [\t]
Regex php : [\t]
Regex postgres : [\t]
Regex procmail : [<TAB>]
Regex python : [\t]
Regex sed : [\t]
Regex tcl : [\t]
Regex vi : [<TAB>]
Regex vim : [\t]
$