Wrong regex match using cl-ppcre?

775 views Asked by At

Trying to parse the following text file:

prefix1 prefix2 name1(
                 type1 name1,
                 type2 name2
                 );

with following regex:
\\s*prefix1\\s*prefix2\\s*(\\w[\\w\\d_]*).*\\(\\s*([^\\)]*\\))\\s*;\\s*
as a result I get the following two groups (registers):

"name1(
             "

and

"(
             type1 name1,
             type2 name2
             )"

(here quotes limit the string, \n are included)

I cannot get why the first group (\w[\w\d_]*) matches the following .* part. Moreover, I cannot get rid of the unnecessary tail!

What is my mistake?

ADD: The parsed regex:

(cl-ppcre::parse-string "\\s*prefix1\\s*prefix2\\s*(\\w[\\w\\d_]*).*\\(\\s*([^\\)]*\\))\\s*;\\s*")
(:SEQUENCE (:GREEDY-REPETITION 0 NIL :WHITESPACE-CHAR-CLASS) "prefix1"
 (:GREEDY-REPETITION 0 NIL :WHITESPACE-CHAR-CLASS) "prefix2"
 (:GREEDY-REPETITION 0 NIL :WHITESPACE-CHAR-CLASS)
 (:REGISTER
  (:SEQUENCE :WORD-CHAR-CLASS
   (:GREEDY-REPETITION 0 NIL (:CHAR-CLASS :WORD-CHAR-CLASS :DIGIT-CLASS #\_))))
 (:GREEDY-REPETITION 0 NIL :EVERYTHING) #\(
 (:GREEDY-REPETITION 0 NIL :WHITESPACE-CHAR-CLASS)
 (:REGISTER
  (:SEQUENCE (:GREEDY-REPETITION 0 NIL (:INVERTED-CHAR-CLASS #\))) #\)))
 (:GREEDY-REPETITION 0 NIL :WHITESPACE-CHAR-CLASS) #\;
 (:GREEDY-REPETITION 0 NIL :WHITESPACE-CHAR-CLASS))

ADD 2: The complete source:

;; Requirements:
;; cl-ppcre

(defparameter *name-and-parameters-list* (cl-ppcre::create-scanner "\\s*prefix1\\s*prefix2\\s*(\\w[\\w\\d_]*)\\s*\\(\\s*([^\\)]*\\))\\s*;\\s*"))
(defparameter *filename* "c:/pva/home/test.txt")

(defun read-txt-without-comments (file-name)
  "Would epically fail in case the file format changes, because currently it expects
 the \"/*\" and \"*/\" sequences to be on the separate line."
  (let ((fstr (make-array '(0) :element-type 'base-char :fill-pointer 0 :adjustable t)))
    (with-output-to-string (s fstr)
      (let ((comment nil))
    (with-open-file (input-stream file-name :direction :input)
      (do ((line (read-line input-stream nil 'eof) (read-line input-stream nil 'eof)))
          ((eql line 'eof))
        (multiple-value-bind (start-comment-from)
        (cl-ppcre:scan ".*/\\*" line)
          (multiple-value-bind (end-comment-from)
          (cl-ppcre:scan ".*\\*/" line)
        (if start-comment-from
            (setf comment t))
        (if (not comment)
            (format s "~A~%" line))
        (if end-comment-from
            (setf comment nil))))))))
    fstr))

(let* ((string (read-txt-without-comments "c:/pva/home/test.txt")))
  (multiple-value-bind (a b c d) (cl-ppcre::scan *name-and-parameters-list* string)
    (format t "~a ~a ~a ~a~%|~a|~%|~a|~%"
        a b c d
        (subseq string (svref c 0) (svref c 1))
        (subseq string (svref d 0) (svref d 1)))))

ADD 3: The complete input:

prefix1 prefix2 name1(
                 type1 name1,
                 type2 name2
                 );
prefix1 prefix2 name2(  type3 name1, type2 name2  );
1

There are 1 answers

1
AudioBubble On BEST ANSWER

This works for me with the recent cl-ppcre as you would've expected it:

(cl-ppcre:register-groups-bind (name argument)
             ("\\s*prefix1\\s*prefix2\\s*(\\w[\\w\\d_]*).*\\(\\s*([^\\)]*\\))\\s*;\\s*"
              "prefix1 prefix2 name1(
                 type1 name1,
                 type2 name2
                 );" :sharedp t)
           (list name argument))
("name1" "type1 name1,
                 type2 name2
                 )")

Perhaps, show a bit more code?