CHN-STUDENT / CSpider

A scalable and convenient crawler framework in C:).

CSpider

A scalable and convenient crawler framework based on C:).
中文文档.

##Mailing list

A micro-community for C/C++ engineers, non-native English speaker friendly, open for everyone.

##Examples Welcome any program using cspider to add links here.

##INSTALL

• Make sure you have installed these libs already.
  • curl
  • libuv
  • libxml2
  • pcre
  • liburi
• Then enter core，and compile:

make

• Move .so and .h files to relevant directory:

make install

• Finally, you could compile your code(such as test.c) using -lcspider:

gcc -o test test.c -lcspider -I /usr/include/libxml2

using -lcspider to link dynamic link libary, and -I /usr/include/libxml2 could let compiler to find libxml2's head files.

##API

###Initial settings

• cspider_t *init_cspider()
  At the beginning of your code, you can get cspider_t by this function. It's essential.

• void cs_setopt_url(cspider_t *, char *)
  Insert urls into work queue when cspider start, passing url's string to second param. url's string could not contain http:// or https://. You could call this function many times to insert many urls.

• void cs_setopt_cookie(cspider_t *, char *)
  Passing cookies to second param, the format is var1=abc; var2=qwe. Optional.

• void cs_setopt_useragent(cspider_t *, char *)
  Passing user agent to second param. Optional.

• void cs_setopt_proxy(cspider_t *, char *)
  Passing proxy to second param. Optional.

• void cs_setopt_timeout(cspider_t *, long)
  Passing timeout value(ms) to second param. Optional.

• void cs_setopt_logfile(cspider_t *, FILE *)
  Passing file pointer to second param, which is used to print log information. Optional.

• void cs_setopt_process(cspider_t *, void (*process)(cspider_t *, char*, char *, void *), void *)
  Passing process function to second param, and user's custom context pointer to third param. You could use this custom context in process function's fourth param and string which is downloaded in second param.

• void cs_setopt_save(cspider_t *, void (*save)(void*, void*), void*)
  Passing data persistence function to second param, and alse user custom context pointer to third param. In this custom function, you can get pointer to prepared data in second param.

• void cs_setopt_threadnum(cspider_t *cspider, int )
  Setting the size of threadpool. The second param could be the size of threadpool you want to set.

• int cs_run(cspider_t *)
  Start cspider. Using this at the end of your code.

###More functions

• void saveString(cspider_t *, void *, int)
  Using this function in custom process function could pass data pointer to custom data persistence function. In the third param, you could use LOCK if you need thread safety. For exmaple, multi-thread write to same file. You could alse use NO_LOCK, if you don't need thread safety.

• void saveStrings(cspider_t *, void **, int, int)
  Using this function could pass array of data pointer to custom data persistence function. Third param is the size of array. Fourth param could be LOCK and NO_LOCK.

• void addUrl(cspider_t *cspider, char *url)
  Using this function in custom process function could insert url to work queue.

• void addUrls(cspider_t *cspider, char *urls, int size)
  Using this function to insert many urls back to work queue. The third param is the number of urls which you want to insert.

• void freeString(char *)
  Using this function to free string.

• void freeStrings(char **, int)
  Using this function to free array of string. Second param is the size of array. You could use this function after regexAll and xpath to free the string array you get.

###Tools

1. Regular expressions:

   • int regexAll(const char *regex, char *str, char **res, int num, int flag);
     regex : regular matching rule
     str : the source string.
     res : array used for saving strings which is matched.
     num : array's size.
     flag : it could be REGEX_ALL and REGEX_NO_ALL, which means whether to return the whole string.
     This function returns the number of matched strings.

   • int match(char *regex, char *str);
     Whether it matches. Return 1 for yes, 0 for no.

2. Using xpath to deal with html and xml:

   • int xpath(char *xml, char *path, char **res, int num);
     xml : prepared to parse.
     path : xpath's rule.
     res : array used for saving strings.
     num : array's size.
     This function returns the number of array which we get.

3. Json:

   cspider contains cJSON. We could use it to parse json data. Usage is here。

4. Uriparser:

   • void joinall(char *baseuri, char **uris, int size); -> join all uris relative to baseuri

   baseuri: Base uri ( char *url in process func ) uris: regex / xpath extracted urls size: length of uris

   • char * join(char *baseuri, char *rel) -> join relative string to the base string

   baseuri: Base uri ( http://test.com/ ) rel: Relative url ( /a/b || ./a/b || ../a/b/./ and ... )

After regexAll and xpath, you should use freeStrings to free the string array which you get.

##Example Print the Github's main page's source code.

#include<cspider/spider.h>
/*
	custom process function
*/
void p(cspider_t *cspider, char *d, char *url, void *user_data) {

  printf("url -> %s\n", url);
  saveString(cspider, d, LOCK);
  
}
/*
	custom data persistence function
*/
void s(void *str, void *user_data) {
  char *get = (char *)str;
  FILE *file = (FILE*)user_data;
  fprintf(file, "%s\n", get);
  return;
}

int main() {
  cspider_t *spider = init_cspider(); 
  char *agent = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:42.0) Gecko/20100101 Firefox/42.0";
  
  cs_setopt_url(spider, "github.com");
  
  cs_setopt_useragent(spider, agent);
  //
  cs_setopt_process(spider, p, NULL);
  cs_setopt_save(spider, s, stdout);
  //set the threadpool's number
  cs_setopt_threadnum(spider, 5);
  
  return cs_run(spider);
}