mlex.h File Reference

Markup Language EXpression. More...

Go to the source code of this file.

Data Structures
struct	chunk_t
	the struct that is used to identify a piece of string More...
Functions
list_t *	mlmatch (char *data, char *exp, char *ret)
	Finds Markup Language chuks matching exp.
void	mlmatch_print_results (list_t *res, char *str)
	debug functions that prints the resul matrix
void	mlmatch_free_results (list_t *res)
	free the list of lists returned by mlmatch
char *	mlmatch_get_result (int x, int y, list_t *res, char *s)
	gets a cell from the result matrix

---

Detailed Description

Markup Language EXpression.

Author:

Enrico Tassi   <>

Definition in file mlex.h.

---

Function Documentation

list_t* mlmatch	(	char *	data,
		char *	exp,
		char *	ret
	)

Finds Markup Language chuks matching exp.

What is an ml-expression?
Simply a regular expression with some more infos about murkups.
Grammar:

• MLEX := MREGEX | MTAGEX | MLEX MLEX | ''
• MTAGEX := '{'REGEX'}' | '<'REGEX'>'
• MREGEX := '['REGEX']' | REGEX
• REGEX := regular expression

Example:

• ".*<b>([0-9]*(Kb|Mb))</b>"
  This matches a generic size in bold.
• ".*<(b|i)>([0-9]*(Kb|Mb))</(b|i)>"
  This matches a generic size in bold or italics, obviously it doesn't check if it opens with a b and closes with a /i.
• "a<b>[c]{d}e{f}[g]<h>"
  This matches abdefgh, abeh and othe strigs created considering optionals the tags/strings between {} and []

Limitation:

• You can use regular expressions inside tags or outside tags,but you can't use regexp with tags. For example it is impossible to specify an arbutrary number of "<b>".
• A string, say an MREGEX not optional, cant start with [ since it is reserver for optional strings. You must put the expression into round brackets to avoid this.
• The parser is not really smart. It always alternates a string with a tag, so an xpression "<a><b>" is interpreted as this sequence of tokens: "","<a>","","<b>".

What is an ml-get-expression?
It is the counterpart of a ml-expression. It selects what is important and what not.
Grammar:

• MLGEX := REGGEX TAGGEX | MLGEX MLGEX | ''
• TAGGEX := '<'EX'>' | '{'EX'}'
• REGGEX := EX | '['EX']'
• EX := 'X' | 'O'

Example:

• If the ml-expression is ".*<b>.*<.*img.*src.*>.*</b>"
  and the ml-get-expression is "O<O>O<X>X<O>"
  and data is "<tt><b><img src="nice.jpg">hello</b>"
  mlmatch returns a list of length 2 (read: the nember of "X") the first defining "img src="nice.jpg"" and the second defining "hello".

Remembre that if an optional string/tag is used in the ml-expression, the corrspong optional string/tag signature must be used in the ml-get-expression.

A short explanation of how the engine works (considering the prevoius example):

1. tokenize the strings:
   • "<tt><b><img src="nice.jpg">hello</b>" becames "","<tt>","","<b>","","<img src="nice.jpg">","hello","</b>"
   • ".*<b>.*<.*img.*src.*>.*</b>" becames ".*","<b>",".*","<.*img.*src.*>",".*","</b>"
   • "O<O>O<X>X<O>" becames "O","<O>","O","<X>","X","<O>"
2. The ml-expression matches perfectly the data starting from the third token, since each regexp matches the corresponding token. so we obtain this sub-list of tokens "","<b>","","<img src="nice.jpg">","hello","</b>"
3. The sublist has the same length of the ret expression and selecting only the tokens with a corresponding X we obtain {"img src="nice.jpg"","hello".}

Notes:

• data, exp and ret MUST be modifyable. they will not be altered, but during processing they may be accessed in write.

Parameters:

	data	is a Markup Language file like an html page (must be modifyable)
	exp	is the ml-expression (must be modifyable)
	ret	is the ml-get-expression (must be modifyable)

Returns:

a list of list of chunk_t

void mlmatch_free_results

(

list_t *

res

)

free the list of lists returned by mlmatch

char* mlmatch_get_result	(	int	x,
		int	y,
		list_t *	res,
		char *	s
	)

gets a cell from the result matrix

mlmatch returns a list of lists. this is a matrix. each line is the list of X fields. Example:

• src := "<b>hello</b> bad <i>guys</b>"
exp := "<.*>.*</b>"
ret := "<X>X<O>"
  calling
  rc = mlmatch(src,exp,ret);
  will return
  {{"b","hello"},
 {"i","guys"} }
  and the respective coordinates are from 0,0 to 1,1. For example "hello" is 1,0. The returned poiter must be freed by the caller.

Parameters:

	x	column
	y	row
	res	returned by mlmatch
	s	the src string

Returns:

a strdup of s chunked in the right position or NULL

void mlmatch_print_results	(	list_t *	res,
		char *	str
	)

debug functions that prints the resul matrix