added rate limiting + handling error and rate limits + better headers

index.js

@@ -2,16 +2,56 @@ const cheerio = require("cheerio");
 const axios = require("axios");
 const randomUseragent = require("random-useragent");
 
+// Utility functions
+const delay = (ms) => new Promise((resolve) => setTimeout(resolve, ms));
+
+// Cache implementation
+class JobCache {
+  constructor() {
+    this.cache = new Map();
+    this.TTL = 1000 * 60 * 60; // 1 hour
+  }
+
+  set(key, value) {
+    this.cache.set(key, {
+      data: value,
+      timestamp: Date.now(),
+    });
+  }
+
+  get(key) {
+    const item = this.cache.get(key);
+    if (!item) return null;
+    if (Date.now() - item.timestamp > this.TTL) {
+      this.cache.delete(key);
+      return null;
+    }
+    return item.data;
+  }
+
+  clear() {
+    const now = Date.now();
+    for (const [key, value] of this.cache.entries()) {
+      if (now - value.timestamp > this.TTL) {
+        this.cache.delete(key);
+      }
+    }
+  }
+}
+
+const cache = new JobCache();
+
+// Main query function
 module.exports.query = (queryObject) => {
   const query = new Query(queryObject);
-  console.log(query.url(0));
   return query.getJobs();
 };
 
+// Query constructor
 function Query(queryObj) {
   this.host = queryObj.host || "www.linkedin.com";
-  this.keyword = queryObj.keyword?.trim().replace(" ", "+") || "";
+  this.keyword = queryObj.keyword?.trim().replace(/\s+/g, "+") || "";
-  this.location = queryObj.location?.trim().replace(" ", "+") || "";
+  this.location = queryObj.location?.trim().replace(/\s+/g, "+") || "";
   this.dateSincePosted = queryObj.dateSincePosted || "";
   this.jobType = queryObj.jobType || "";
   this.remoteFilter = queryObj.remoteFilter || "";
@@ -22,13 +62,14 @@ function Query(queryObj) {
   this.page = Number(queryObj.page) || 0;
 }
 
+// Query prototype methods
 Query.prototype.getDateSincePosted = function () {
   const dateRange = {
     "past month": "r2592000",
     "past week": "r604800",
     "24hr": "r86400",
   };
-  return dateRange[this.dateSincePosted.toLowerCase()] ?? "";
+  return dateRange[this.dateSincePosted.toLowerCase()] || "";
 };
 
 Query.prototype.getExperienceLevel = function () {
@@ -40,8 +81,9 @@ Query.prototype.getExperienceLevel = function () {
     director: "5",
     executive: "6",
   };
-  return experienceRange[this.experienceLevel.toLowerCase()] ?? "";
+  return experienceRange[this.experienceLevel.toLowerCase()] || "";
 };
+
 Query.prototype.getJobType = function () {
   const jobTypeRange = {
     "full time": "F",
@@ -53,8 +95,9 @@ Query.prototype.getJobType = function () {
     volunteer: "V",
     internship: "I",
   };
-  return jobTypeRange[this.jobType.toLowerCase()] ?? "";
+  return jobTypeRange[this.jobType.toLowerCase()] || "";
 };
+
 Query.prototype.getRemoteFilter = function () {
   const remoteFilterRange = {
     "on-site": "1",
@@ -62,8 +105,9 @@ Query.prototype.getRemoteFilter = function () {
     remote: "2",
     hybrid: "3",
   };
-  return remoteFilterRange[this.remoteFilter.toLowerCase()] ?? "";
+  return remoteFilterRange[this.remoteFilter.toLowerCase()] || "";
 };
+
 Query.prototype.getSalary = function () {
   const salaryRange = {
     40000: "1",
@@ -72,113 +116,193 @@ Query.prototype.getSalary = function () {
     100000: "4",
     120000: "5",
   };
-  return salaryRange[this.salary.toLowerCase()] ?? "";
+  return salaryRange[this.salary] || "";
 };
 
 Query.prototype.getPage = function () {
   return this.page * 25;
 };
+
 Query.prototype.url = function (start) {
   let query = `https://${this.host}/jobs-guest/jobs/api/seeMoreJobPostings/search?`;
-  if (this.keyword !== "") query += `keywords=${this.keyword}`;
+
-  if (this.location !== "") query += `&location=${this.location}`;
+  const params = new URLSearchParams();
-  if (this.getDateSincePosted() !== "")
+
-    query += `&f_TPR=${this.getDateSincePosted()}`;
+  if (this.keyword) params.append("keywords", this.keyword);
-  if (this.getSalary() !== "") query += `&f_SB2=${this.getSalary()}`;
+  if (this.location) params.append("location", this.location);
-  if (this.getExperienceLevel() !== "")
+  if (this.getDateSincePosted())
-    query += `&f_E=${this.getExperienceLevel()}`;
+    params.append("f_TPR", this.getDateSincePosted());
-  if (this.getRemoteFilter() !== "") query += `&f_WT=${this.getRemoteFilter()}`;
+  if (this.getSalary()) params.append("f_SB2", this.getSalary());
-  if (this.getJobType() !== "") query += `&f_JT=${this.getJobType()}`;
+  if (this.getExperienceLevel())
-  if (this.getPage() !== "") query += `&start=${start + this.getPage()};`;
+    params.append("f_E", this.getExperienceLevel());
-  if (this.sortBy == "recent" || this.sortBy == "relevant") {
+  if (this.getRemoteFilter()) params.append("f_WT", this.getRemoteFilter());
-    let sortMethod = "R";
+  if (this.getJobType()) params.append("f_JT", this.getJobType());
-    if (this.sortBy == "recent") sortMethod = "DD";
+
-    query += `&sortBy=${sortMethod}`;
+  params.append("start", start + this.getPage());
-  }
+
-  return encodeURI(query);
+  if (this.sortBy === "recent") params.append("sortBy", "DD");
+  else if (this.sortBy === "relevant") params.append("sortBy", "R");
+
+  return query + params.toString();
 };
 
 Query.prototype.getJobs = async function () {
+  let allJobs = [];
+  let start = 0;
+  const BATCH_SIZE = 25;
+  let hasMore = true;
+  let consecutiveErrors = 0;
+  const MAX_CONSECUTIVE_ERRORS = 3;
+
   try {
-    let parsedJobs,
+    // Check cache first
-      resultCount = 1,
+    const cacheKey = this.url(0);
-      start = 0,
+    const cachedJobs = cache.get(cacheKey);
-      jobLimit = this.limit,
+    if (cachedJobs) {
-      allJobs = [];
+      console.log("Returning cached results");
+      return cachedJobs;
+    }
 
-    while (resultCount > 0) {
+    while (hasMore) {
-      const userAgent = randomUseragent.getRandom();
+      try {
+        const jobs = await this.fetchJobBatch(start);
 
-      const { data } = await axios.get(this.url(start), {
+        if (!jobs || jobs.length === 0) {
-        headers: {
+          hasMore = false;
-          "User-Agent": userAgent,
+          break;
-          Accept:
+        }
-            "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8",
-          "Accept-Language": "en-US,en;q=0.9",
-          "Accept-Encoding": "gzip, deflate, br",
-          Connection: "keep-alive",
-          "Upgrade-Insecure-Requests": "1",
-        },
-      });
 
-      const $ = cheerio.load(data);
+        allJobs.push(...jobs);
-      const jobs = $("li");
+        console.log(`Fetched ${jobs.length} jobs. Total: ${allJobs.length}`);
-      resultCount = jobs.length;
-      console.log("I got ", jobs.length, " jobs");
 
-      parsedJobs = parseJobList(data);
+        if (this.limit && allJobs.length >= this.limit) {
-      allJobs.push(...parsedJobs);
+          allJobs = allJobs.slice(0, this.limit);
+          break;
+        }
 
-      start += 25;
+        // Reset error counter on successful fetch
+        consecutiveErrors = 0;
+        start += BATCH_SIZE;
 
-      if (jobLimit != 0 && allJobs.length > jobLimit) {
+        // Add reasonable delay between requests
-        while (allJobs.length != jobLimit) allJobs.pop();
+        await delay(2000 + Math.random() * 1000);
-        return allJobs;
+      } catch (error) {
+        consecutiveErrors++;
+        console.error(
+          `Error fetching batch (attempt ${consecutiveErrors}):`,
+          error.message
+        );
+
+        if (consecutiveErrors >= MAX_CONSECUTIVE_ERRORS) {
+          console.log("Max consecutive errors reached. Stopping.");
+          break;
+        }
+
+        // Exponential backoff
+        await delay(Math.pow(2, consecutiveErrors) * 1000);
       }
     }
+
+    // Cache results if we got any
+    if (allJobs.length > 0) {
+      cache.set(this.url(0), allJobs);
+    }
+
     return allJobs;
   } catch (error) {
-    console.error(error);
+    console.error("Fatal error in job fetching:", error);
+    throw error;
+  }
+};
+
+Query.prototype.fetchJobBatch = async function (start) {
+  const headers = {
+    "User-Agent": randomUseragent.getRandom(),
+    Accept: "application/json, text/javascript, */*; q=0.01",
+    "Accept-Language": "en-US,en;q=0.9",
+    "Accept-Encoding": "gzip, deflate, br",
+    Referer: "https://www.linkedin.com/jobs",
+    "X-Requested-With": "XMLHttpRequest",
+    Connection: "keep-alive",
+    "Sec-Fetch-Dest": "empty",
+    "Sec-Fetch-Mode": "cors",
+    "Sec-Fetch-Site": "same-origin",
+    "Cache-Control": "no-cache",
+    Pragma: "no-cache",
+  };
+
+  try {
+    const response = await axios.get(this.url(start), {
+      headers,
+      validateStatus: function (status) {
+        return status === 200;
+      },
+      timeout: 10000,
+    });
+
+    return parseJobList(response.data);
+  } catch (error) {
+    if (error.response?.status === 429) {
+      throw new Error("Rate limit reached");
+    }
+    throw error;
   }
 };
 
 function parseJobList(jobData) {
-  const $ = cheerio.load(jobData);
+  try {
-  const jobs = $("li");
+    const $ = cheerio.load(jobData);
+    const jobs = $("li");
 
-  const jobObjects = jobs
+    return jobs
-    .map((index, element) => {
+      .map((index, element) => {
-      const job = $(element);
+        try {
-      const position = job.find(".base-search-card__title").text().trim() || "";
+          const job = $(element);
-      const company =
+          const position = job.find(".base-search-card__title").text().trim();
-        job.find(".base-search-card__subtitle").text().trim() || "";
+          const company = job.find(".base-search-card__subtitle").text().trim();
-      const location =
+          const location = job.find(".job-search-card__location").text().trim();
-        job.find(".job-search-card__location").text().trim() || "";
+          const dateElement = job.find("time");
-      const date = job.find("time").attr("datetime") || "";
+          const date = dateElement.attr("datetime");
-      const salary =
+          const salary = job
-        job
+            .find(".job-search-card__salary-info")
-          .find(".job-search-card__salary-info")
+            .text()
-          .text()
+            .trim()
-          .trim()
+            .replace(/\s+/g, " ");
-          .replace(/\n/g, "")
+          const jobUrl = job.find(".base-card__full-link").attr("href");
-          .replaceAll(" ", "") || "";
+          const companyLogo = job
-      const jobUrl = job.find(".base-card__full-link").attr("href") || "";
+            .find(".artdeco-entity-image")
-      const companyLogo =
+            .attr("data-delayed-url");
-        job.find(".artdeco-entity-image").attr("data-delayed-url") || "";
+          const agoTime = job.find(".job-search-card__listdate").text().trim();
-      const agoTime =
-        job.find(".job-search-card__listdate").text().trim() || "";
-      return {
-        position: position,
-        company: company,
-        companyLogo: companyLogo,
-        location: location,
-        date: date,
-        agoTime: agoTime,
-        salary: salary,
-        jobUrl: jobUrl,
-      };
-    })
-    .get();
 
-  return jobObjects;
+          // Only return job if we have at least position and company
+          if (!position || !company) {
+            return null;
+          }
+
+          return {
+            position,
+            company,
+            location,
+            date,
+            salary: salary || "Not specified",
+            jobUrl: jobUrl || "",
+            companyLogo: companyLogo || "",
+            agoTime: agoTime || "",
+          };
+        } catch (err) {
+          console.warn(`Error parsing job at index ${index}:`, err.message);
+          return null;
+        }
+      })
+      .get()
+      .filter(Boolean);
+  } catch (error) {
+    console.error("Error parsing job list:", error);
+    return [];
+  }
 }
+
+// Export additional utilities for testing and monitoring
+module.exports.JobCache = JobCache;
+module.exports.clearCache = () => cache.clear();
+module.exports.getCacheSize = () => cache.cache.size;

package.json

@@ -1,6 +1,6 @@
 {
   "name": "linkedin-jobs-api",
-  "version": "1.0.5",
+  "version": "1.0.6",
   "description": "advanced node.js package for getting job listings from LinkedIn",
   "main": "index.js",
   "scripts": {