voidarc/omnisearch
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fix: refactored scraping components

Browse Source
This commit is contained in:
frosty
2026-03-17 13:51:12 -04:00
parent 8c6632502f
commit c7b95d0571
8 changed files with 816 additions and 700 deletions
Download Patch File Download Diff File Expand all files Collapse all files

603
src/Scraping/Scraping.c
View File

@@ -1,395 +1,20 @@
#include "Scraping.h"
#include "../Cache/Cache.h"
#include "../Proxy/Proxy.h"
#include "../Utility/Unescape.h"
#include "../Utility/XmlHelper.h"
#include "Config.h"
#include <curl/curl.h>
#include <libxml/HTMLparser.h>
#include <libxml/xpath.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>
static size_t WriteMemoryCallback(void *contents, size_t size, size_t nmemb,
void *userp) {
size_t realsize = size * nmemb;
MemoryBuffer *mem = (MemoryBuffer *)userp;
if (mem->size + realsize + 1 > mem->capacity) {
size_t new_cap =
mem->capacity == 0 ? INITIAL_BUFFER_SIZE : mem->capacity * 2;
while (new_cap < mem->size + realsize + 1)
new_cap *= 2;
char *ptr = (char *)realloc(mem->memory, new_cap);
if (!ptr) {
return 0;
}
mem->memory = ptr;
mem->capacity = new_cap;
}
memcpy(&(mem->memory[mem->size]), contents, realsize);
mem->size += realsize;
mem->memory[mem->size] = 0;
return realsize;
}
static const char *get_random_user_agent(void) {
static const char *agents[] = {
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, "
"like Gecko) Chrome/120.0.0.0 Safari/537.36",
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 "
"(KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36",
"Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like "
"Gecko) "
"Chrome/120.0.0.0` Safari/537.36",
"Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:109.0) Gecko/20100101 "
"Firefox/121.0",
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 "
"(KHTML, like Gecko) Version/17.2 Safari/605.1.15"};
return agents[rand() % 5];
}
static int parse_ddg_lite(const char *engine_name, xmlDocPtr doc,
SearchResult **out_results, int max_results) {
(void)engine_name;
int found_count = 0;
xmlXPathContextPtr xpathCtx = xmlXPathNewContext(doc);
if (!xpathCtx) {
int check_cache_for_job(ScrapeJob *job) {
if (get_cache_ttl_search() <= 0)
return 0;
}
xmlXPathObjectPtr xpathObj = xml_xpath_eval(
xpathCtx, "//tr[not(contains(@class, "
"'result-sponsored'))]//a[@class='result-link']");
if (!xpathObj || !xpathObj->nodesetval || xpathObj->nodesetval->nodeNr == 0) {
if (xpathObj)
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return 0;
}
int num_links = xpathObj->nodesetval->nodeNr;
*out_results = xml_result_alloc(num_links, max_results);
if (!*out_results) {
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return 0;
}
for (int i = 0; i < num_links && found_count < max_results; i++) {
xmlNodePtr linkNode = xpathObj->nodesetval->nodeTab[i];
char *title = xml_node_content(linkNode);
char *url = (char *)xmlGetProp(linkNode, (xmlChar *)"href");
char *snippet_text = NULL;
xmlNodePtr current = linkNode->parent;
while (current && xmlStrcasecmp(current->name, (const xmlChar *)"tr") != 0)
current = current->parent;
if (current && current->next) {
xmlNodePtr snippetRow = current->next;
while (snippetRow &&
xmlStrcasecmp(snippetRow->name, (const xmlChar *)"tr") != 0)
snippetRow = snippetRow->next;
if (snippetRow) {
xpathCtx->node = snippetRow;
xmlXPathObjectPtr sObj =
xml_xpath_eval(xpathCtx, ".//td[@class='result-snippet']");
if (sObj && sObj->nodesetval && sObj->nodesetval->nodeNr > 0) {
snippet_text = xml_node_content(sObj->nodesetval->nodeTab[0]);
}
if (sObj)
xmlXPathFreeObject(sObj);
xpathCtx->node = NULL;
}
}
(*out_results)[found_count].url = unescape_search_url(url);
(*out_results)[found_count].title = strdup(title ? title : "No Title");
(*out_results)[found_count].snippet =
strdup(snippet_text ? snippet_text : "");
found_count++;
if (title)
xmlFree(title);
if (url)
xmlFree(url);
if (snippet_text)
xmlFree(snippet_text);
}
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return found_count;
}
static int parse_startpage(const char *engine_name, xmlDocPtr doc,
SearchResult **out_results, int max_results) {
(void)engine_name;
int found_count = 0;
xmlXPathContextPtr xpathCtx = xmlXPathNewContext(doc);
if (!xpathCtx) {
return 0;
}
xmlXPathObjectPtr xpathObj =
xml_xpath_eval(xpathCtx, "//div[contains(@class, 'result')]");
if (!xpathObj || !xpathObj->nodesetval || xpathObj->nodesetval->nodeNr == 0) {
if (xpathObj)
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return 0;
}
int num_results = xpathObj->nodesetval->nodeNr;
*out_results = xml_result_alloc(num_results, max_results);
if (!*out_results) {
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return 0;
}
for (int i = 0; i < num_results && found_count < max_results; i++) {
xmlNodePtr resultNode = xpathObj->nodesetval->nodeTab[i];
xpathCtx->node = resultNode;
xmlXPathObjectPtr linkObj =
xml_xpath_eval(xpathCtx, ".//a[contains(@class, 'result-link')]");
char *url =
(linkObj && linkObj->nodesetval && linkObj->nodesetval->nodeNr > 0)
? (char *)xmlGetProp(linkObj->nodesetval->nodeTab[0],
(xmlChar *)"href")
: NULL;
xmlXPathObjectPtr titleObj =
xml_xpath_eval(xpathCtx, ".//h2[contains(@class, 'wgl-title')]");
char *title =
(titleObj && titleObj->nodesetval && titleObj->nodesetval->nodeNr > 0)
? xml_node_content(titleObj->nodesetval->nodeTab[0])
: NULL;
xmlXPathObjectPtr snippetObj =
xml_xpath_eval(xpathCtx, ".//p[contains(@class, 'description')]");
char *snippet_text =
(snippetObj && snippetObj->nodesetval &&
snippetObj->nodesetval->nodeNr > 0)
? xml_node_content(snippetObj->nodesetval->nodeTab[0])
: NULL;
if (url && title) {
(*out_results)[found_count].url = strdup(url);
(*out_results)[found_count].title = strdup(title);
(*out_results)[found_count].snippet =
strdup(snippet_text ? snippet_text : "");
found_count++;
}
if (title)
xmlFree(title);
if (url)
xmlFree(url);
if (snippet_text)
xmlFree(snippet_text);
if (linkObj)
xmlXPathFreeObject(linkObj);
if (titleObj)
xmlXPathFreeObject(titleObj);
if (snippetObj)
xmlXPathFreeObject(snippetObj);
}
xpathCtx->node = NULL;
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return found_count;
}
static int parse_yahoo(const char *engine_name, xmlDocPtr doc,
SearchResult **out_results, int max_results) {
(void)engine_name;
int found_count = 0;
xmlXPathContextPtr xpathCtx = xmlXPathNewContext(doc);
if (!xpathCtx) {
return 0;
}
xmlXPathObjectPtr xpathObj =
xml_xpath_eval(xpathCtx, "//div[contains(@class, 'algo-sr')]");
if (!xpathObj || !xpathObj->nodesetval || xpathObj->nodesetval->nodeNr == 0) {
if (xpathObj)
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return 0;
}
int num_results = xpathObj->nodesetval->nodeNr;
*out_results = xml_result_alloc(num_results, max_results);
if (!*out_results) {
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return 0;
}
for (int i = 0; i < num_results && found_count < max_results; i++) {
xmlNodePtr resultNode = xpathObj->nodesetval->nodeTab[i];
xpathCtx->node = resultNode;
xmlXPathObjectPtr linkObj = xml_xpath_eval(
xpathCtx, ".//div[contains(@class, 'compTitle')]//a[@target='_blank']");
char *url =
(linkObj && linkObj->nodesetval && linkObj->nodesetval->nodeNr > 0)
? (char *)xmlGetProp(linkObj->nodesetval->nodeTab[0],
(xmlChar *)"href")
: NULL;
xmlXPathObjectPtr titleObj =
xml_xpath_eval(xpathCtx, ".//h3[contains(@class, 'title')]");
char *title =
(titleObj && titleObj->nodesetval && titleObj->nodesetval->nodeNr > 0)
? xml_node_content(titleObj->nodesetval->nodeTab[0])
: NULL;
xmlXPathObjectPtr snippetObj =
xml_xpath_eval(xpathCtx, ".//div[contains(@class, 'compText')]//p");
char *snippet_text =
(snippetObj && snippetObj->nodesetval &&
snippetObj->nodesetval->nodeNr > 0)
? xml_node_content(snippetObj->nodesetval->nodeTab[0])
: NULL;
if (url && title) {
(*out_results)[found_count].url = unescape_search_url(url);
(*out_results)[found_count].title = strdup(title);
(*out_results)[found_count].snippet =
strdup(snippet_text ? snippet_text : "");
found_count++;
}
if (title)
xmlFree(title);
if (url)
xmlFree(url);
if (snippet_text)
xmlFree(snippet_text);
if (linkObj)
xmlXPathFreeObject(linkObj);
if (titleObj)
xmlXPathFreeObject(titleObj);
if (snippetObj)
xmlXPathFreeObject(snippetObj);
}
xpathCtx->node = NULL;
xmlXPathFreeObject(xpathObj);
xmlXPathFreeContext(xpathCtx);
return found_count;
}
const SearchEngine ENGINE_REGISTRY[] = {
{.name = "DuckDuckGo Lite",
.base_url = "https://lite.duckduckgo.com/lite/?q=",
.host_header = "lite.duckduckgo.com",
.referer = "https://lite.duckduckgo.com/",
.page_param = "s",
.page_multiplier = 30,
.page_base = 0,
.parser = parse_ddg_lite},
{.name = "Startpage",
.base_url = "https://www.startpage.com/sp/search?q=",
.host_header = "www.startpage.com",
.referer = "https://www.startpage.com/",
.page_param = "page",
.page_multiplier = 1,
.page_base = 1,
.parser = parse_startpage},
{.name = "Yahoo",
.base_url = "https://search.yahoo.com/search?p=",
.host_header = "search.yahoo.com",
.referer = "https://search.yahoo.com/",
.page_param = "b",
.page_multiplier = 10,
.page_base = 1,
.parser = parse_yahoo}};
const int ENGINE_COUNT = sizeof(ENGINE_REGISTRY) / sizeof(SearchEngine);
#define CURL_TIMEOUT 15L
#define CURL_DNS_TIMEOUT 300L
static void configure_curl_handle(CURL *curl, const char *full_url,
MemoryBuffer *chunk,
struct curl_slist *headers) {
curl_easy_setopt(curl, CURLOPT_URL, full_url);
curl_easy_setopt(curl, CURLOPT_HTTPHEADER, headers);
curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, WriteMemoryCallback);
curl_easy_setopt(curl, CURLOPT_WRITEDATA, (void *)chunk);
curl_easy_setopt(curl, CURLOPT_USERAGENT, get_random_user_agent());
curl_easy_setopt(curl, CURLOPT_HTTP_VERSION, CURL_HTTP_VERSION_2_0);
curl_easy_setopt(curl, CURLOPT_ACCEPT_ENCODING, "");
curl_easy_setopt(curl, CURLOPT_DNS_CACHE_TIMEOUT, CURL_DNS_TIMEOUT);
curl_easy_setopt(curl, CURLOPT_FOLLOWLOCATION, 1L);
curl_easy_setopt(curl, CURLOPT_TIMEOUT, CURL_TIMEOUT);
curl_easy_setopt(curl, CURLOPT_SSL_VERIFYPEER, 1L);
curl_easy_setopt(curl, CURLOPT_COOKIEFILE, "");
apply_proxy_settings(curl);
}
static char *build_search_url(const char *base_url, const char *page_param,
int page_multiplier, int page_base,
const char *encoded_query, int page) {
int page_value = (page < 1 ? 1 : page - 1) * page_multiplier + page_base;
char *url = malloc(BUFFER_SIZE_LARGE);
if (!url) {
return NULL;
}
snprintf(url, BUFFER_SIZE_LARGE, "%s%s&%s=%d", base_url, encoded_query,
page_param, page_value);
return url;
}
static struct curl_slist *build_request_headers(const char *host_header,
const char *referer) {
struct curl_slist *headers = NULL;
char host_buf[BUFFER_SIZE_MEDIUM], ref_buf[BUFFER_SIZE_MEDIUM];
snprintf(host_buf, sizeof(host_buf), "Host: %s", host_header);
snprintf(ref_buf, sizeof(ref_buf), "Referer: %s", referer);
headers = curl_slist_append(headers, host_buf);
headers = curl_slist_append(headers, ref_buf);
headers = curl_slist_append(
headers,
"Accept: "
"text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8");
headers = curl_slist_append(headers, "Accept-Language: en-US,en;q=0.5");
headers = curl_slist_append(headers, "DNT: 1");
return headers;
}
static int check_cache_for_job(ScrapeJob *job) {
if (get_cache_ttl_search() <= 0) {
return 0;
}
char *key = cache_compute_key(job->query, job->page, job->engine->name);
if (!key) {
if (!key)
return 0;
}
char *cached_data = NULL;
size_t cached_size = 0;
@@ -414,27 +39,31 @@ static int check_cache_for_job(ScrapeJob *job) {
return 0;
}
static void process_job_response(ScrapeJob *job, CURL *handle, CURLMsg *msg) {
if (msg->data.result == CURLE_OK && job->response.size > 0) {
char *key = cache_compute_key(job->query, job->page, job->engine->name);
if (key && get_cache_ttl_search() > 0) {
cache_set(key, job->response.memory, job->response.size);
free(key);
}
void parse_and_cache_response(ScrapeJob *job) {
if (job->response.size == 0) {
job->results_count = 0;
return;
}
xmlDocPtr doc = htmlReadMemory(
job->response.memory, job->response.size, NULL, NULL,
HTML_PARSE_RECOVER | HTML_PARSE_NOERROR | HTML_PARSE_NOWARNING);
char *key = cache_compute_key(job->query, job->page, job->engine->name);
if (key && get_cache_ttl_search() > 0)
cache_set(key, job->response.memory, job->response.size);
free(key);
if (doc) {
job->results_count = job->engine->parser(
job->engine->name, doc, job->out_results, job->max_results);
xmlFreeDoc(doc);
}
xmlDocPtr doc = htmlReadMemory(
job->response.memory, job->response.size, NULL, NULL,
HTML_PARSE_RECOVER | HTML_PARSE_NOERROR | HTML_PARSE_NOWARNING);
if (doc) {
job->results_count = job->engine->parser(
job->engine->name, doc, job->out_results, job->max_results);
xmlFreeDoc(doc);
} else {
job->results_count = 0;
}
}
void cleanup_job_handle(ScrapeJob *job, CURL *handle) {
struct curl_slist *headers = NULL;
curl_easy_getinfo(handle, CURLINFO_PRIVATE, &headers);
if (headers)
@@ -444,67 +73,112 @@ static void process_job_response(ScrapeJob *job, CURL *handle, CURLMsg *msg) {
job->response.memory = NULL;
}
void process_response(ScrapeJob *job, CURL *handle, CURLMsg *msg) {
if (msg->data.result == CURLE_OK)
parse_and_cache_response(job);
else
job->results_count = 0;
cleanup_job_handle(job, handle);
}
int setup_job(ScrapeJob *job, CURLM *multi_handle) {
if (job->handle)
curl_easy_cleanup(job->handle);
if (job->response.memory)
free(job->response.memory);
if (check_cache_for_job(job)) {
job->results_count = job->results_count > 0 ? job->results_count : 0;
return 0;
}
char *encoded_query = curl_easy_escape(NULL, job->query, 0);
if (!encoded_query)
return -1;
char *full_url =
build_search_url(job->engine->base_url, job->engine->page_param,
job->engine->page_multiplier, job->engine->page_base,
encoded_query, job->page);
free(encoded_query);
if (!full_url)
return -1;
job->handle = curl_easy_init();
if (!job->handle) {
free(full_url);
return -1;
}
job->response.memory = (char *)malloc(INITIAL_BUFFER_SIZE);
job->response.size = 0;
job->response.capacity = INITIAL_BUFFER_SIZE;
struct curl_slist *headers =
build_request_headers(job->engine->host_header, job->engine->referer);
configure_curl_handle(job->handle, full_url, &job->response, headers);
curl_easy_setopt(job->handle, CURLOPT_PRIVATE, headers);
free(full_url);
curl_multi_add_handle(multi_handle, job->handle);
return 0;
}
int handle_responses(CURLM *multi_handle, ScrapeJob *jobs, int num_jobs) {
CURLMsg *msg;
int msgs_left;
while ((msg = curl_multi_info_read(multi_handle, &msgs_left))) {
if (msg->msg != CURLMSG_DONE)
continue;
CURL *handle = msg->easy_handle;
for (int i = 0; i < num_jobs; i++) {
if (jobs[i].handle && jobs[i].handle == handle) {
process_response(&jobs[i], handle, msg);
curl_multi_remove_handle(multi_handle, handle);
curl_easy_cleanup(handle);
jobs[i].handle = NULL;
break;
}
}
}
return 0;
}
int should_retry(ScrapeJob *jobs, int num_jobs) {
if (proxy_count <= 0)
return 0;
for (int i = 0; i < num_jobs; i++) {
if (jobs[i].results_count == 0 && jobs[i].response.size == 0)
return 1;
}
return 0;
}
int scrape_engines_parallel(ScrapeJob *jobs, int num_jobs) {
int retries = 0;
retry:
CURLM *multi_handle = curl_multi_init();
if (!multi_handle) {
if (!multi_handle)
return -1;
}
for (int i = 0; i < num_jobs; i++) {
ScrapeJob *job = &jobs[i];
if (job->handle) {
curl_easy_cleanup(job->handle);
job->handle = NULL;
if (setup_job(&jobs[i], multi_handle) != 0 && jobs[i].handle) {
curl_multi_remove_handle(multi_handle, jobs[i].handle);
curl_easy_cleanup(jobs[i].handle);
jobs[i].handle = NULL;
}
if (job->response.memory) {
free(job->response.memory);
}
if (check_cache_for_job(job)) {
job->results_count = job->results_count > 0 ? job->results_count : 0;
continue;
}
char *encoded_query = curl_easy_escape(NULL, job->query, 0);
if (!encoded_query) {
continue;
}
char *full_url =
build_search_url(job->engine->base_url, job->engine->page_param,
job->engine->page_multiplier, job->engine->page_base,
encoded_query, job->page);
free(encoded_query);
if (!full_url) {
continue;
}
job->handle = curl_easy_init();
if (!job->handle) {
free(full_url);
continue;
}
job->response.memory = (char *)malloc(INITIAL_BUFFER_SIZE);
job->response.size = 0;
job->response.capacity = INITIAL_BUFFER_SIZE;
struct curl_slist *headers =
build_request_headers(job->engine->host_header, job->engine->referer);
configure_curl_handle(job->handle, full_url, &job->response, headers);
curl_easy_setopt(job->handle, CURLOPT_PRIVATE, headers);
free(full_url);
curl_multi_add_handle(multi_handle, job->handle);
}
usleep(100000 + (rand() % 100000));
http_delay();
int still_running = 0;
curl_multi_perform(multi_handle, &still_running);
@@ -512,50 +186,17 @@ retry:
do {
int numfds = 0;
CURLMcode mc = curl_multi_wait(multi_handle, NULL, 0, 1000, &numfds);
if (mc != CURLM_OK) {
if (mc != CURLM_OK)
break;
}
curl_multi_perform(multi_handle, &still_running);
} while (still_running);
CURLMsg *msg;
int msgs_left;
while ((msg = curl_multi_info_read(multi_handle, &msgs_left))) {
if (msg->msg == CURLMSG_DONE) {
CURL *handle = msg->easy_handle;
for (int i = 0; i < num_jobs; i++) {
if (jobs[i].handle && jobs[i].handle == handle) {
ScrapeJob *job = &jobs[i];
process_job_response(job, handle, msg);
curl_multi_remove_handle(multi_handle, handle);
if (handle)
curl_easy_cleanup(handle);
job->handle = NULL;
break;
}
}
}
}
handle_responses(multi_handle, jobs, num_jobs);
curl_multi_cleanup(multi_handle);
if (retries < max_proxy_retries && proxy_count > 0) {
int any_failed = 0;
for (int i = 0; i < num_jobs; i++) {
if (jobs[i].results_count == 0 && jobs[i].response.size == 0) {
any_failed = 1;
break;
}
}
if (any_failed) {
retries++;
goto retry;
}
if (retries < max_proxy_retries && should_retry(jobs, num_jobs)) {
retries++;
goto retry;
}
return 0;