data1.h

harvest是一个下载html网页得机器人

/* $Id: data1.h,v 1.8 2003/06/23 15:36:11 adam Exp $
   Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002
   Index Data Aps

This file is part of the Zebra server.

Zebra is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free
Software Foundation; either version 2, or (at your option) any later
version.

Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
for more details.

You should have received a copy of the GNU General Public License
along with Zebra; see the file LICENSE.zebra.  If not, write to the
Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
02111-1307, USA.
*/

#ifndef DATA1_H
#define DATA1_H

#define ENHANCED_XELM 1

#include <stdio.h>

#include <yaz/nmem.h>
#include <yaz/oid.h>
#include <yaz/proto.h>

#include <d1_attset.h>
#include <d1_map.h>
#include <yaz/yaz-util.h>
#include <yaz/wrbuf.h>
#include <dfa.h>         /* pop */
#include <zebra_xpath.h> /* pop */

#define d1_isspace(c) strchr(" \r\n\t\f", c)
#define d1_isdigit(c) ((c) <= '9' && (c) >= '0')

YAZ_BEGIN_CDECL

#define data1_matchstr(s1, s2) yaz_matchstr(s1, s2)

#define DATA1_MAX_SYMBOL 31

typedef struct data1_name
{
    char *name;
    struct data1_name *next;
} data1_name;

typedef struct data1_absyn_cache_info *data1_absyn_cache;
typedef struct data1_attset_cache_info *data1_attset_cache;

typedef enum data1_datatype
{
    DATA1K_unknown,
    DATA1K_structured,
    DATA1K_string,
    DATA1K_numeric,
    DATA1K_bool,
    DATA1K_oid,
    DATA1K_generalizedtime,
    DATA1K_intunit,
    DATA1K_int,
    DATA1K_octetstring,
    DATA1K_null
} data1_datatype;

typedef struct data1_marctab
{
    char *name;
    oid_value reference;

    char record_status[2];
    char implementation_codes[5];
    int  indicator_length;
    int  identifier_length;
    char user_systems[4];

    int  length_data_entry;
    int  length_starting;
    int  length_implementation;
    char future_use[2];

    int  force_indicator_length;
    int  force_identifier_length;
    char leader[24]; /* Fixme! Need linear access to LEADER of MARC record */  
    struct data1_marctab *next;
} data1_marctab;

typedef struct data1_esetname
{
    char *name;
    Z_Espec1 *spec;
    struct data1_esetname *next;
} data1_esetname;

/*
 * Variant set definitions.
 */

typedef struct data1_vartype
{
    char *name;
    struct data1_varclass *zclass;
    int type;
    data1_datatype datatype;
    struct data1_vartype *next;
} data1_vartype;

typedef struct data1_varclass
{
    char *name;
    struct data1_varset *set;
    int zclass;
    data1_vartype *types;
    struct data1_varclass *next;
} data1_varclass;

typedef struct data1_varset
{
    char *name;
    oid_value reference;
    data1_varclass *classes;
} data1_varset;

/*
 * Tagset definitions
 */

struct data1_tagset;

typedef struct data1_tag
{
    data1_name *names;
#define DATA1T_numeric 1
#define DATA1T_string 2
    int which;
    union
    {
	int numeric;
	char *string;
    } value;
    data1_datatype kind;

    struct data1_tagset *tagset;
    struct data1_tag *next;
} data1_tag;

typedef struct data1_tagset data1_tagset;

struct data1_tagset
{
    int type;                        /* type of tagset in current context */
    char *name;                      /* symbolic name */
    oid_value reference;
    data1_tag *tags;                 /* tags defined by this set */
    data1_tagset *children;          /* children */
    data1_tagset *next;              /* sibling */
};

typedef struct data1_termlist
{
    data1_att *att;
    char *structure;
    char *source;
    struct data1_termlist *next;
} data1_termlist;

/*
 * abstract syntax specification
 */

typedef struct data1_element
{
    char *name;
    data1_tag *tag;
    data1_termlist *termlists;
    char *sub_name;
    struct data1_element *children;
    struct data1_element *next;
} data1_element;

typedef struct data1_sub_elements {
    char *name;
    struct data1_sub_elements *next;
    data1_element *elements;
} data1_sub_elements;

/* pop */
typedef struct data1_xpelement
{
    char *xpath_expr;
#ifdef ENHANCED_XELM 
    struct xpath_location_step xpath[XPATH_STEP_COUNT];
    int xpath_len;
#endif
    struct DFA *dfa;  
    data1_termlist *termlists;
    struct data1_xpelement *next;
} data1_xpelement;

typedef struct data1_xattr {
    char *name;
    char *value;
    struct data1_xattr *next;
    unsigned short what;  /* DATA1I_text, .. see data1_node.u.data */
} data1_xattr;

#if 0
typedef struct data1_absyn data1_absyn;
#else
typedef struct data1_absyn
{
    char *name;
    oid_value reference;
    data1_tagset *tagset;
    data1_attset *attset;
    data1_varset *varset;
    data1_esetname *esetnames;
    data1_maptab *maptabs;
    data1_marctab *marc;
    data1_sub_elements *sub_elements;
    data1_element *main_elements;
    data1_xpelement *xp_elements; /* pop */
    struct data1_systag *systags;
    char *encoding;
    int  enable_xpath_indexing;
} data1_absyn;
#endif
/*
 * record data node (tag/data/variant)
 */

typedef struct data1_node
{
    /* the root of a record (containing global data) */
#define DATA1N_root 1 
    /* a tag */
#define DATA1N_tag  2       
    /* some data under a leaf tag or variant */
#define DATA1N_data 3
    /* variant specification (a triple, actually) */
#define DATA1N_variant 4
    /* comment (same as data) */
#define DATA1N_comment 5
    /* preprocessing instruction */
#define DATA1N_preprocess 6
    int which;
    union
    {
	struct
	{
	    char *type;
	    data1_absyn *absyn;  /* abstract syntax for this type */
	} root;