用戶:Antigng-bot/zhinterwiki

#include <stdio.h>
#include <string.h>
#include <process.h>
#include <windows.h>
#include "mem.h"
#include "network.h"
#include "convert.h"
#include "auth.h"
struct problemlist
{
	unsigned int pageid;
	struct problemlist *next;
};
struct neditargv
{
	const char *id;
	HTTP newtext;
	const char *time;
};
int threadc[1024];
SRWLOCK rwcs;
CRITICAL_SECTION tcs;
CRITICAL_SECTION hcs;
int threadnumber=0;
unsigned char action=0;
struct hashlist *hl=NULL;
struct problemlist *pbl=NULL;

const char *username=NULL;
const char *passwd=NULL;
const char *searchstring=NULL;
const char *ns=NULL;
int maxthread=256;
int doallpage=0;
static void displayerr(unsigned int code)
{
	if(!(code&0x1))
	{
		printf(
			"\tNo username.\n"
			"\t\tA valid username must be specified via \"-u\".\n"
			);
	}
	if(code&0x2)
	{
		printf(
			"\tUsername too long.\n"
			"\t\tThe username should not be longer than 64 bytes.\n"
			);
	}
	if(!(code&0x4))
	{
		printf(
			"\tNo query.\n"
			"\t\tEither a search string (via \"-s\") or \"-a\" should be set.\n"
			);
	}
	if(code&0x8)
	{
		printf(
			"\tPassword too long.\n"
			"\t\tThe password should not be longer than 64 bytes.\n"
			);
	}
	if(!(code&0x10))
	{
		printf(
			"\tNo password.\n"
			"\t\tA valid password must be specified via \"-p\".\n"
			);
	}
	if(code&0x20)
	{
		printf(
			"\tSearch string too long.\n"
			"\t\tThe search string should not be longer than 128 bytes.\n"
			);
	}
	if(code&0x40)
	{
		printf(
			"\tns string too long.\n"
			"\t\tThe ns string should not be longer than 32 bytes.\n"
			);
	}
	return;
}
static int parsearg(int argc,const char *argv[])
{
	int cur_arg=0;
	unsigned int err=0;
	doallpage=0;
	for(cur_arg=1;cur_arg<argc;cur_arg++)
	{
		if(argv[cur_arg][0]=='-'&&((argv[cur_arg+1]&&argv[cur_arg+1][0]!='-')||argv[cur_arg][1]=='a'))
		{
			switch(argv[cur_arg][1])
			{
			case 'u':
				username=G2U(argv[cur_arg+1]);
				if(strlen(username)>64)
				{
					err|=0x2;
				}
				else
				{
					err|=0x1;
				}
				cur_arg++;
				break;
			case 'a':
				err|=0x4;
				doallpage=1;
				break;
			case 'p':
				passwd=argv[cur_arg+1];
				if(strlen(passwd)>64)
				{
					err|=0x8;
				}
				else
				{
					err|=0x10;
				}
				cur_arg++;
				break;
			case 's':
				searchstring=G2U(argv[cur_arg+1]);
				if(strlen(searchstring)>128)
				{
					err|=0x20;
				}
				else
				{
					err|=0x4;
				}
				cur_arg++;
				break;
			case 'n':
				ns=argv[cur_arg+1];
				if(strlen(ns)>32)
				{
					err|=0x40;
				}
				cur_arg++;
				break;
			case 'T':
				maxthread=atoi(argv[cur_arg+1]);
				if(maxthread<1||maxthread>1024) maxthread=32;
				cur_arg++;
				break;
			}
		}
	}
	if(!ns) ns="0";
	if(0x1+0x4+0x10==err) return 0;
	else
	{
		printf("Error code 0x%x:\n",err);
		displayerr(err);
		return 1;
	}
}
static int smartedit(struct neditargv *p,const char *reason, const char *tags)
{
	HTTP res;
	char line[2048],url[4096]={0};
	char reason_e[512];
	char tags_e[256];
	char aft[1024],statusline[128];
	char cur_token[128];
	char err_type[8192];
	char *erm[]={"code"};
	char *erv[1];
	int find=0;
	int has_err=0,token_err=0,filtered=0;
	int retry=0;
	erv[0]=err_type;
	if(reason) URLEncode(reason,strlen(reason),reason_e,510);
	if(tags) URLEncode(tags,strlen(tags),tags_e,254);
	sprintf(url,"/w/api.php?action=edit&pageid=%s&basetimestamp=%s",p->id,p->time);
	find=sprintf(aft,"%s%s&summary=%s&bot=1&minor=1&nocreate=1&format=xml&token=",tags?"&tags=":"",tags?tags_e:"",reason_e);
	if(find<0) return -4;
	do
	{
		res=hopen();;
		while(1)
		{
			AcquireSRWLockShared(&rwcs);
			if(hastoken) break;
			else ReleaseSRWLockShared(&rwcs);
			Sleep(100);
		}		
		aft[find]=0;
		strcat(aft,token);
		ReleaseSRWLockShared(&rwcs);
		hrewind(p->newtext);
		if(smartpost(url,p->newtext,aft,8888,1,res))
		{
			hclose(res);
			return -1;
		}
		hgets(statusline,127,res);
		if(!strstr(statusline," 200"))
		{
			hclose(res);
			return -2;
		}
		skipresponseheader(res);
		filtered=token_err=has_err=0;
		while(!heof(res))
		{
			if(xmlparsetag(res,line)==XML_HAS_VALUE)
			{
				if(!strcmp(line,"error"))
				{
					has_err=1;
					xmlparsearg(res,1,erm,erv);
					if((!strcmp(err_type,"notoken"))||(!strcmp(err_type,"badtoken")))
					{
						token_err=1;
					}
					else if(!strcmp(err_type,"abusefilter-warning"))
					{
						filtered=1;
					}
					break;
				}
			}
		}
		if(token_err)
		{
			AcquireSRWLockExclusive(&rwcs);
			if(!strcmp(aft+find,cur_token)) hastoken=0;
			ReleaseSRWLockExclusive(&rwcs);
		}
		retry++;
		hclose(res);
	}while((token_err||filtered)&&(retry<3));
	if(has_err) return -3;
	else return 0;
}
const char *nsname_and_direct_links[]={"t","talk","user","user_talk","wikipedia","wp","wikipedia_talk","wt","template","template_talk","file","file_talk","portal","portal_talk","s","n","wikt","q"};
const int tokennum=18;
static int removezhinterwiki(const char *source,int length,char *output)
{
	int i=0,j=0,k=0;
	int todo=0,hasw=0;
	int zhpos=0,wpos=0;
	int pos=0;
	char *buf=s_malloc(length+1);
	for(i=0;i<length;i++)
	{
		if(source[i]==':')
		{					
			if(j)
			{
				buf[j]=0;
				if(todo==0&&(!strcmp(buf,"zh")))
				{
					todo=1;
					zhpos=pos;
				}
				else if(hasw==0&&(!strcmp(buf,"w")))
				{
					hasw=1;
					wpos=pos;
				}
				else
				{
					for(k=0;k<tokennum;k++)
					{
						if(!strcmp(buf,nsname_and_direct_links[k])) break;
					}
					if(k==tokennum)
					{
						todo=0;					
						pos=0;
						break;
					}
				}
				j=0;
			}
			pos=i;
		}
		else
		{
			buf[j]=(source[i]>='A'&&source[i]<='Z')?(source[i]+'a'-'A'):((source[i]==' ')?'_':source[i]);
			j++;
		}

	}
	if(todo==1)
	{
		if(hasw==1)
		{
			int minpos,maxpos;
			int minadd,maxadd;
			zhpos<wpos?(minpos=zhpos,minadd=3,maxpos=wpos,maxadd=2):(minpos=wpos,minadd=2,maxpos=zhpos,maxadd=3);
			for(i=0,j=0;i<minpos;i++,j++)
			{
				output[j]=source[i];
			}
			i+=minadd;
			for(;i<maxpos;i++,j++)
			{
				output[j]=source[i];
			}
			i+=maxadd;
			for(;i<length;i++,j++)
			{
				output[j]=source[i];
			}
		}
		else
		{
			for(i=0,j=0;i<zhpos;i++,j++)
			{
				output[j]=source[i];
			}
			i+=3;
			for(;i<length;i++,j++)
			{
				output[j]=source[i];
			}
		}
		output[j]=0;
	}
	else
	{
		for(i=0;i<length;i++)
		{
			output[i]=source[i];
		}
		output[i]=0;
		j=i;
	}
	s_free(buf);
	return j;
}
static int pagecheck(const char *pageid,const char *basetime,HTTP f)
{
	HTTP newtext;
	int error=0,status=0,todo=0;
	int i=0,brac=0,namepos=0,displaypos=0;
	char ch=0,name[4096],name_d[2048],name_e[4096],display[2048],display_e[2048];
	newtext=hopen();
	hputs("&text=",6,newtext);
	while(xmlpulltext(f,&ch)==XML_TEXT_CONTINUE)
	{
		switch(status)
		{
		case 0:
			if(ch=='[')
			{
				status=1;
				brac=1;
			}
			smartURLEncode(ch,newtext);
			break;
		case 1:
			if(ch!='[')
			{
				brac=0;
				status=0;
			}
			else brac++;
			if(brac==2)
			{
				brac=0;
				status=2;
				namepos=0;
			}
			smartURLEncode(ch,newtext);
			break;
		case 2:
			if(ch=='|')
			{
				name[namepos]=0;
				status=3;
				displaypos=0;
			}
			else if(ch==']')
			{
				name[namepos]=0;
				if((i=removezhinterwiki(name,namepos,name_d))!=namepos) todo=1;
				i=URLEncode(name_d,i,name,4095);
				hputs(name,i,newtext);
				status=0;
				smartURLEncode(']',newtext);
			}
			else if(ch=='[')
			{
				name[namepos]=0;
				i=URLEncode(name,namepos,name_e,4095);
				hputs(name_e,i,newtext);
				status=1;
				brac=1;
				smartURLEncode(ch,newtext);
			}
			else if((namepos>1024)||(ch=='<')||(ch=='>')||(ch=='{')||(ch=='}'))
			{
				name[namepos]=0;
				i=URLEncode(name,namepos,name_e,4095);
				hputs(name_e,i,newtext);
				status=0;
				smartURLEncode(ch,newtext);
			}
			else
			{
				name[namepos++]=ch;
			}
			break;
		case 3:
			if(ch==']')
			{
				display[displaypos]=0;
				if((i=removezhinterwiki(name,namepos,name_d))!=namepos) todo=1;
				i=URLEncode(name_d,i,name,4095);
				hputs(name,i,newtext);
				if((*name_d==':'?strcmp(name_d+1,display):strcmp(name_d,display)))
				{
					smartURLEncode('|',newtext);
					i=URLEncode(display,displaypos,display_e,2047);
					hputs(display_e,i,newtext);
				}
				smartURLEncode(']',newtext);
				status=0;
			}
			else if(ch=='[')
			{
				display[displaypos]=0;
				if((i=removezhinterwiki(name,namepos,name_d))!=namepos) todo=1;
				i=URLEncode(name_d,i,name,4095);
				hputs(name,i,newtext);
				smartURLEncode('|',newtext);
				i=URLEncode(display,displaypos,display_e,2047);
				hputs(display_e,i,newtext);
				status=1;
				brac=1;
				smartURLEncode(ch,newtext);
			}
			else if(displaypos>512)
			{
				display[displaypos]=0;
				if((i=removezhinterwiki(name,namepos,name_d))!=namepos) todo=1;
				i=URLEncode(name_d,i,name,4095);
				hputs(name,i,newtext);
				smartURLEncode('|',newtext);
				i=URLEncode(display,displaypos,display_e,2047);
				hputs(display_e,i,newtext);
				status=0;
				smartURLEncode(ch,newtext);
			}
			else
			{
				display[displaypos++]=ch;
			}
			break;
		}
	}
	if(xmlpulltext(f,&ch)!=XML_TEXT_END) error=1;
	if(todo&&!status&&!error)
	{
		struct neditargv point;
		point.newtext=newtext;
		point.id=pageid;
		point.time=basetime;
		smartedit(&point,"bot: cleanup zh interwiki",NULL);
	}
	hclose(newtext);
	return 0;
}
static int proceedchild(const char *ids)
{
	char url[4096];
	char buf[8192];
	char pageid[256];
	char timestamp[256];
	char contentmodel[64],contentformat[64];
	const char *ttm[]={"pageid"};
	const char *tmm[]={"timestamp"};
	const char *cmm[]={"contentmodel","contentformat"};
	char *ttv[1];
	char *tmv[1];
	char *cmv[2];
	int result;
	int status;
	HTTP h;
	ttv[0]=pageid;
	tmv[0]=timestamp;
	cmv[0]=contentmodel;
	cmv[1]=contentformat;
	if(!ids)
	{
		return -1;
	}
	sprintf(url,"/w/api.php?action=query&format=xml&prop=revisions&rvprop=content|timestamp&pageids=%s&rvslots=main",ids);
	h=hopen();
	if(get(url,8888,1,h))
	{
		hclose(h);
		return -2;
	}
	hgets(buf,4096,h);
	if(!strstr(buf," 200"))
	{
		hclose(h);
		return -3;
	}
	if(skipresponseheader(h))
	{
		hclose(h);
		return -3;
	}
	status=0;
	while(!heof(h))
	{
		result=xmlparsetag(h,buf);
		if(result==XML_HAS_VALUE)
		{
			switch(status)
			{
			case 0:
				if(!strcmp(buf,"page")&&(xmlparsearg(h,1,ttm,ttv)==XML_HAS_VALUE))
				{
					if(atoi(pageid)>0) status=1;
				}
				break;
			case 1:
				if(!strcmp(buf,"rev"))
				{
					if(xmlparsearg(h,1,tmm,tmv)!=XML_HAS_VALUE)
					{
						status=3;
						goto _cleanup;
					}
					status=2;
				}
				else goto _cleanup;
				break;
			case 2:
				if(!strcmp(buf,"slot"))
				{
					if(xmlparsearg(h,2,cmm,cmv)!=XML_HAS_VALUE)
					{
						status=3;
						goto _cleanup;
					}
					if(!strcmp(contentmodel,"wikitext")&&!strcmp(contentformat,"text/x-wiki"))
					{
						pagecheck(pageid,timestamp,h);
					}
					status=0;
				}
				else goto _cleanup;
				break;
			}
		}
		else if(result==XML_PARSE_ERROR)
		{
			status=3;
			goto _cleanup;
		}
	}
_cleanup:
	hclose(h);
	return status?-4:0;
}
static void threadfunc(void *c)
{
	int i=*(int *)c;
	int ext=0;
	char pageid[10][64];
	int count=0;
	char ids[4096];
	int result=0;
	struct problemlist *temp;
	while(!action) Sleep(1);
	while(1)
	{
		EnterCriticalSection(&hcs);
		for(count=0;count<10;count++)
		{
			if(pbl)
			{
				temp=pbl;
				pbl=pbl->next;
				sprintf(pageid[count],"%d",temp->pageid);
				free(temp);
			}
			else
			{
				ext=(count==0?1:2);
				count++;
				break;				
			}
		}
		count--;
		if(count>=0)
		{
			strcpy(ids,pageid[count]);
			count--;
			while(count>=0)
			{
				strcat(ids,"|");
				strcat(ids,pageid[count]);
				count--;
			}
		}
		LeaveCriticalSection(&hcs);
		if(ext==1) break;
		else
		{
			result=proceedchild(ids);
			if(ext==2) break;
		}
	}
	EnterCriticalSection(&tcs);
	threadnumber--;
	LeaveCriticalSection(&tcs);
	return ;
} 
static int threadini(int count)
{
	int i=0;
	int flag=0;
	threadnumber=0;
	for(i=0;i<count;i++)
	{
		threadc[i]=i;
		flag=_beginthread(threadfunc,0,(void *)(threadc+i));
		if(flag>0) threadnumber++;
	}
	return 0;
}
static int query(const char *target,const char *ns)
{
	HTTP f;
	char target_e[512],line[2048]={0},url[4096]={0},snd[4096]={0},id[512]={0},title[512]={0},sroffset[2048]={0},offseto[512]={0};
	char statusline[128]={0};
	int next=0,retry=0,pageid=0;
	struct problemlist *temp=0;
	char *ctm[]={"gsroffset"};
	char *ctv[1];
	char *idm[]={"pageid","title"};
	char *idv[2];
	ctv[0]=offseto;
	idv[0]=id;
	idv[1]=title;
	if(strlen(target)>128)
	{
		printf("Search string too long!\n");
		return -1;
	}
	URLEncode(target,strlen(target),target_e,511);
	sprintf(url,"/w/api.php?action=query&format=xml&generator=search&prop=info&gsrlimit=500&gsrnamespace=%s&gsrsearch=%s",ns,target_e);
	do
	{
		strcpy(snd,url);
		if(next)
		{
			strcat(snd,"&gsroffset=");
			strcat(snd,sroffset);
		}
		f=hopen();
		retry=0;
		do
		{
			if(get(snd,8888,1,f))
			{
				hclose(f);
				f=hopen();
			}
			else
			{
				hgets(statusline,127,f);
				if(strstr(statusline,"200")) break;
				else
				{
					hclose(f);
					f=hopen();
				}
			}
			retry++;
		}while(retry<20);
		if(retry==20)
 		{
			hclose(f);
			return 1;
		}
		skipresponseheader(f);
		next=0;
		do
		{
			xmlparsetag(f,line);
			if(!next)
			{
				if(!strcmp(line,"continue"))
				{
					xmlparsearg(f,1,ctm,ctv);
					URLEncode(offseto,strlen(offseto),sroffset,990);
					next=1;
				}
			}
			if(!strcmp(line,"page"))
			{
				xmlparsearg(f,2,idm,idv);								
				if((pageid=atoi(id))>0)
				{
					temp=(struct problemlist *)s_malloc(sizeof(struct problemlist));
					temp->pageid=pageid;
					temp->next=pbl;
					pbl=temp;
				}
			}
		}while(!heof(f));
		hclose(f);
	}while(next);
	return 0;
}
static int allpagequery(const char *ns)
{
	HTTP f;
	char line[2048]={0},url[4096]={0},snd[4096]={0},id[512]={0},title[512]={0},sroffset[2048]={0},offseto[512]={0};
	char statusline[128];
	int next=0,retry=0,pageid=0;
	struct problemlist *temp=0;
	char *ctm[]={"apcontinue"};
	char *ctv[1];
	char *idm[]={"pageid","title"};
	char *idv[2];
	ctv[0]=offseto;
	idv[0]=id;
	idv[1]=title;
	sprintf(url,"/w/api.php?action=query&format=xml&list=allpages&apnamespace=%s&aplimit=5000",ns);
	do
	{
		strcpy(snd,url);
		if(next)
		{
			strcat(snd,"&apcontinue=");
			strcat(snd,sroffset);
		}
		f=hopen();
		for(retry=0;retry<20;retry++)
		{
			if(get(snd,8888,1,f))
			{
				hclose(f);
				f=hopen();
			}
			else
			{
				hgets(statusline,127,f);
				if(strstr(statusline,"200")) break;
				else
				{
					hclose(f);
					f=hopen();
				}
			}
		}
		if(retry==20)
		{
			hclose(f);
			return 1;
		}
		skipresponseheader(f);
		next=0;
 		do
		{
			xmlparsetag(f,line);
			if(!next)
			{
				if(!strcmp(line,"continue"))
				{
					xmlparsearg(f,1,ctm,ctv);
					URLEncode(offseto,strlen(offseto),sroffset,990);
					next=1;
				}
			}
			if(!strcmp(line,"p"))
			{
				xmlparsearg(f,2,idm,idv);
				if((pageid=atoi(id))>0)
				{
					temp=(struct problemlist *)s_malloc(sizeof(struct problemlist));
					temp->pageid=pageid;
					temp->next=pbl;
					pbl=temp;
				}
			}
		}while(!heof(f));
		hclose(f);
	}while(next);
	return 0;
}
int main(int argc,char *argv[])
{
	int count=0;
	HANDLE tk_thread;
	if(parsearg(argc,argv))
	{
		printf("usage: -u username -p passwd [-T concurrency -s searchstring -a allpagequery -n namespace]\n");
		return -1;
	}
	InitializeSRWLock(&rwcs);
	InitializeCriticalSection(&tcs);
	InitializeCriticalSection(&hcs);
	buckini(20);
	if(login(username,passwd))
	{
		printf("Login error!\n");
		return -2;
	}
	hastoken=0;
	printf("Login complete.\n");
	fflush(stdout);
	tk_thread=(HANDLE)_beginthread(tokenmanage,0,0);
	if(doallpage)
	{
		allpagequery(ns);
	}
	else query(searchstring,ns);
	if(pbl==NULL)
	{
		printf("No page!\n");
		return -3;
	}
	printf("Query complete.\n");
	action=0;
	threadini(maxthread);
	action=1;
	while(1)
	{
		EnterCriticalSection(&hcs);
		if(pbl!=NULL)
		{
			LeaveCriticalSection(&hcs);
			Sleep(1000);
		}
		else
		{
			LeaveCriticalSection(&hcs);
			break;
		}
	}
	count=0;
	while(count<150)
	{
		count++;
		EnterCriticalSection(&tcs);
		if(threadnumber>0)
		{
			printf("Waiting for all threads to exit. Current thread number: %d\n",threadnumber);
			LeaveCriticalSection(&tcs);
		}
		else
		{
			LeaveCriticalSection(&tcs);
			break;
		}
		fflush(stdout);
		Sleep(1000);
	}
	if(!threadnumber)
	{
		printf("Cleanup..\n");
		DeleteCriticalSection(&tcs);
		DeleteCriticalSection(&hcs);
		AcquireSRWLockExclusive(&rwcs);
		hastoken=-1;
		ReleaseSRWLockExclusive(&rwcs);
		WaitForSingleObject(tk_thread,INFINITE);
		buckdestroy();
	}
	printf("---------------Ok done.---------------\n");
	fflush(stdout);
	return 0;
}