On 06/06/2014 13:37, ph10@???<
mailto:ph10@hermes.cam.ac.uk> wrote:
On Fri, 6 Jun 2014, GAUTIER Herve wrote:
> In my opinion, no. As I have used the pcretest program in a xterm on
> Linux, only a \n is added when I press the return key
Then why did you configure with
PCRE_CONFIGURE_OPTS += --enable-newline-is-crlf
?? On Linux, as you say, newlines are just LF, but that configuration
specifies that lines are terminated with two characters (CR, LF). So a
single LF character will not be treated as indicating a newline.
Sure.
I need --enable-newline-is-crlf because the data that I match is a protocol where the end of line is CR LF.
But my regex editor is on Linux thus end of line are LF.
I was thinking (I don't know why...) that --enable-newline-is-crlf affects only the data, not the definition of the regex, and particularly when using /x modifier with comments in regex.
Maybe I should use --enable-newline-is-anycrlf or "automagically" add CR before LF in my regex definition.
How test with pcretest using my configuration ? Can I put CR in my RE ? Is pcretest take in account the --enable-newline-is-* while entering the RE ?
$ pcretest
PCRE version 8.34 2013-12-15
re> /^\r
> [+-]? # test\r
> \d+\.\d+\r
> |\d+\.\r
> |\.\d+\r
> )
> ([eE][+-]?\d+)?\r
> $/xm\r
** Unknown modifier '\'
re> /^\r
> [+-]? # test\r
> \d+\.\d+\r
> |\d+\.\r
> |\.\d+\r
> )\r
> ([eE][+-]?\d+)?\r
> $/xm
data> START\r\n1.2E3\r\nEND
No match
Thank for your help and time Philip !
Rv
--
Hervé GAUTIER