zebrash.c

harvest是一个下载html网页得机器人

/* $Id: zebrash.c,v 1.22 2003/09/23 10:09:38 adam Exp $
   Copyright (C) 2002,2003
   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.
*/

/* 
   zebrash.c - command-line interface to zebra API
*/

#include <stdio.h>
#include <stdlib.h>
#include <string.h> 
#include <ctype.h>

#if HAVE_READLINE_READLINE_H
#include <readline/readline.h> 
#endif
#if HAVE_READLINE_HISTORY_H
#include <readline/history.h>
#endif

#include "zebraapi.h"
#include <yaz/log.h>
#include <yaz/proto.h>
#include <yaz/wrbuf.h>

#define MAX_NO_ARGS 32
#define MAX_OUT_BUFF 4096
#define MAX_ARG_LEN 1024
#define PROMPT "ZebraSh>"
#define DEFAULTCONFIG "./zebra.cfg"
#define DEFAULTDATABASE "Default"
#define DEFAULTRESULTSET "MyResultSet"


/**************************************
 * Global variables (yuck!)
 */

ZebraService zs=0;  /* our global handle to zebra */
ZebraHandle  zh=0;  /* the current session */
/* time being, only one session works */
int nextrecno=1;  /* record number to show next */

/**************************************
 * Help functions
 */

 
static int split_args( char *line, char** args )
{ /* splits line into individual null-terminated strings, 
   * returns pointers to them in args */
  /* FIXME - do we need to handle quoted args ?? */
    char *p=line;
    int i=0;
    int n=0;
    args[0]=0; /* by default */
    while (*p==' ' || *p=='\t' || *p=='\n')
	    p++;
    while (*p)
    {
	while (*p==' ' || *p=='\t' || *p=='\n')
	    p++;
	if (*p=='#')  /* skip comments */
	    break;  
	args[i++]=p;
	args[i]=0;
	while (*p && *p!=' ' && *p!='\t' && *p!='\n' && *p!='#')
	    p++;
	*p++='\0';
    }
    n=i;
    while (n<MAX_NO_ARGS)
	args[n++]=0;
    return i;
}

static char *defarg( char *arg, char *def )
{
    if (!arg)
	return def;
    if (!*arg)
	return def;
    return arg;
}
static int defargint( char *arg, int def )
{
    int v=def;
    char *a=defarg(arg,0);
    if (a)
	sscanf(a," %i", &v);
    return v;
}

static char *restargs( char *args[], int n)
{ /* Returns the rest of the arguments, starting at the nth, */
  /* to the end of the command line. Assumes args[0] contains */
  /* the original line, minus the command itself */
    int skiplen= args[n]-args[1];
    if (skiplen > strlen(args[0]))
        return "";
    return args[0]+skiplen;
}

int onecommand( char *line, WRBUF outbuff, const char *prevout); 

/**************************************
 * Simple support commands
 */

int cmd_echo( char *args[], WRBUF outbuff)
{
    wrbuf_printf(outbuff,"%s\n",restargs(args,1));
    return 0;
}
 
int cmd_quit( char *args[], WRBUF outbuff)
{
    if (zs)
    {
        onecommand("zebra_close",outbuff,"");
        zs=0;
    }
    if (zh)
    {
        onecommand("zebra_stop",outbuff,"");
        zh=0;
    }
    wrbuf_puts(outbuff, "bye");
    return -99; /* special stop signal */
}

/**************************************
 * Tests for starting and stopping zebra, etc
 */
 
static int cmd_help( char *args[], WRBUF outbuff); 
 
static int cmd_zebra_start( char *args[], WRBUF outbuff)
{
    char *conf=args[1];
    if (!conf || !*conf) {
	wrbuf_puts(outbuff,"no config file specified, using "
		   DEFAULTCONFIG "\n" );
	conf=DEFAULTCONFIG;
    }
    zs=zebra_start(conf);
    if (!zs) {
	wrbuf_puts(outbuff, "zebra_start failed" );
	return 2;
    }
    return 0; /* ok */
}
 
static int cmd_zebra_stop( char *args[], WRBUF outbuff)
{
    if (!zs)
	wrbuf_puts(outbuff,"zebra seems not to have been started, "
		   "stopping anyway\n");
    zebra_stop(zs);
    zs=0;
    return 0; /* ok */
}

static int cmd_zebra_open( char *args[], WRBUF outbuff)
{
    if (!zs)
	wrbuf_puts(outbuff,"zebra seems not to have been started, "
		   "trying anyway\n");
    zh=zebra_open(zs);
    return 0; /* ok */
}

static int cmd_zebra_close( char *args[], WRBUF outbuff)
{
    if (!zh)
	wrbuf_puts(outbuff,"Seems like you have not called zebra_open,"
		   "trying anyway\n");
    zebra_close(zh);
    return 0; /* ok */
}

static int cmd_quickstart( char *args[], WRBUF outbuff)
{
    char tmp[128];
    int rc=0;
    if (!rc)
        rc=onecommand("yaz_log_file zebrash.log",outbuff,"");
    if (!rc)
        rc=onecommand("yaz_log_prefix ZebraSh", outbuff,"");
    sprintf(tmp, "yaz_log_level 0x%x", LOG_DEFAULT_LEVEL | LOG_APP);
    if (!rc)
        rc=onecommand(tmp,outbuff,"");
    logf(LOG_APP,"quickstart");
    if (!zs)
        if (!rc)
            rc=onecommand("zebra_start",outbuff,"");
    if (!zh)
        if (!rc)
            rc=onecommand("zebra_open",outbuff,"");
    if (!rc)
        rc=onecommand("select_database Default",outbuff,"");
    return rc;
}

/**************************************
 * Log file handling
 */

static int cmd_yaz_log_file( char *args[], WRBUF outbuff)
{
    char *fn = defarg(args[1],0);
    wrbuf_printf(outbuff, "sending yaz-log to %s\n",fn);
    yaz_log_init_file(fn);
    return 0; /* ok */
}

static int cmd_yaz_log_level( char *args[], WRBUF outbuff)
{
    int  lev = defargint(args[1],LOG_DEFAULT_LEVEL);
    wrbuf_printf(outbuff, "setting yaz-log to level %d (ox%x)\n",lev,lev);
    yaz_log_init_level(lev);
    return 0; /* ok */
}

static int cmd_yaz_log_prefix( char *args[], WRBUF outbuff)
{
    char *pref = defarg(args[1],"ZebraSh");
    wrbuf_printf(outbuff, "setting yaz-log prefix to %s\n",pref);
    yaz_log_init_prefix(pref);
    return 0; /* ok */
}

static int cmd_logf( char *args[], WRBUF outbuff)
{
    int lev = defargint(args[1],0);
    int i=1;  
    if (lev)
	i=2;
    else
	lev=LOG_LOG; /* this is in the default set!*/
    logf( lev, restargs(args,i));
    return 0; /* ok */
}
 
/****************
 * Error handling 
 */
static int cmd_err ( char *args[], WRBUF outbuff)
{
    wrbuf_printf(outbuff, "errCode: %d \nerrStr:  %s\nerrAdd:  %s \n",
		 zebra_errCode (zh),
		 zebra_errString (zh),  
		 zebra_errAdd (zh) );
    return 0; /* ok */
}
static int cmd_errcode ( char *args[], WRBUF outbuff)
{
    wrbuf_printf(outbuff, "errCode: %d \n",
		 zebra_errCode (zh));
    return 0; /* ok */
}
static int cmd_errstr ( char *args[], WRBUF outbuff)
{
    wrbuf_printf(outbuff, "errStr:  %s\n",
		 zebra_errString (zh));
    return 0; /* ok */
}
static int cmd_erradd ( char *args[], WRBUF outbuff)
{
    wrbuf_printf(outbuff, "errAdd:  %s \n",
		 zebra_errAdd (zh) ); 
    return 0; /* ok */
}

/**************************************
 * Admin commands
 */

static int cmd_init ( char *args[], WRBUF outbuff)
{
    zebra_init(zh);
    return 0; /* ok */
}

static int cmd_select_database ( char *args[], WRBUF outbuff)
{
    char *db=defarg(args[1],DEFAULTDATABASE);
	wrbuf_puts(outbuff,"Selecting database 'Default'\n");
    return zebra_select_database(zh, db);
}
 
static int cmd_create_database( char *args[], WRBUF outbuff)
{
    char *db=defarg(args[1],DEFAULTDATABASE);
    wrbuf_printf(outbuff,"Creating database '%s'\n",db);
    
    return zebra_create_database(zh, db);
}

static int cmd_drop_database( char *args[], WRBUF outbuff)
{
    char *db=args[1];
    if (!db)
        db="Default";
    wrbuf_printf(outbuff,"Dropping database '%s'\n",db);
    return zebra_drop_database(zh, db);
}

static int cmd_begin_trans( char *args[], WRBUF outbuff)
{
    int rw=0;
    if (args[1] && ( (args[1][0]=='1') || (args[1][0]=='w') ))
        rw=1;
    return zebra_begin_trans(zh,rw);
}

static int cmd_end_trans( char *args[], WRBUF outbuff)
{
    return zebra_end_trans(zh);
}
/*************************************
 * Inserting and deleting
 */

static int cmd_record_insert( char *args[], WRBUF outbuff)
{
    int sysno=0;
    int rc;
    char *rec=restargs(args,1);
    
    rc=zebra_record_insert(zh,rec, strlen(rec), &sysno);
    if (0==rc)
    {
        wrbuf_printf(outbuff,"ok sysno=%d\n",sysno);
    }
    return rc;
}


static int cmd_exchange_record( char *args[], WRBUF outbuff)
{
    char *base=args[1];
    char *id = args[2];
    char *action = args[3];
    int rc;
    char *rec=restargs(args,4);
    if (!(base && id && action && args[4] ))
    {
	wrbuf_puts(outbuff,"Missing arguments!\n");
	onecommand("help exchange_record", outbuff, "");
	return -90;
    }
    rc=zebra_admin_exchange_record(zh, base, rec, strlen(rec),
        id, strlen(id), atoi(action));
    return rc;
}

/**********************************
 * Searching and retrieving
 */

static int cmd_search_pqf( char *args[], WRBUF outbuff)
{
    int hits=0;
    char *set=args[1];
    char *qry=restargs(args,2);
    int rc;
    rc=zebra_search_PQF(zh, qry, set, &hits);
    if (0==rc)
        wrbuf_printf(outbuff,"%d hits found\n",hits);