resuml 1.12.0 → 1.12.2

package/dist/index.cjs

@@ -762,7 +762,206 @@ var init_en = __esm({
         "process",
         "robust",
         "consistent",
-        "operations"
+        "operations",
+        // URL/email/domain fragments
+        "http",
+        "https",
+        "www",
+        "com",
+        "org",
+        "net",
+        "mailto",
+        // Resume/YAML schema field names (in case raw YAML is pasted)
+        "name",
+        "keywords",
+        "highlights",
+        "startdate",
+        "enddate",
+        "website",
+        "profiles",
+        "basics",
+        "position",
+        "institution",
+        "studytype",
+        "fluency",
+        "issuer",
+        "network",
+        "username",
+        "countrycode",
+        "region",
+        // Generic nouns that aren't skills
+        "product",
+        "company",
+        "service",
+        "services",
+        "platform",
+        "solutions",
+        "ability",
+        "opportunity",
+        "candidate",
+        "applicant",
+        "position",
+        "salary",
+        "compensation",
+        "benefits",
+        "perks",
+        "bonus",
+        "development",
+        "management",
+        "knowledge",
+        "modern",
+        "advanced",
+        "practices",
+        "nice",
+        "technologies",
+        "technology",
+        "frameworks",
+        "framework",
+        "tools",
+        "data",
+        "based",
+        "contribute",
+        "contributions",
+        "migration",
+        "leading",
+        "source",
+        "visit",
+        // Common verbs & verb forms (not technical skills, supplement action verbs list)
+        "collaborate",
+        "collaborating",
+        "collaboratively",
+        "communicate",
+        "communicating",
+        "contributing",
+        "coordinate",
+        "coordinating",
+        "demonstrate",
+        "demonstrating",
+        "design",
+        "designing",
+        "designed",
+        "develop",
+        "developing",
+        "developed",
+        "drive",
+        "driving",
+        "driven",
+        "enable",
+        "enabling",
+        "evaluate",
+        "evaluating",
+        "execute",
+        "executing",
+        "facilitate",
+        "facilitating",
+        "identify",
+        "identifying",
+        "influence",
+        "influencing",
+        "interact",
+        "interacting",
+        "lead",
+        "leverage",
+        "leveraging",
+        "manage",
+        "managing",
+        "mentor",
+        "mentoring",
+        "operate",
+        "operating",
+        "optimize",
+        "optimizing",
+        "participate",
+        "participating",
+        "report",
+        "reporting",
+        "solve",
+        "solving",
+        "understand",
+        "understanding",
+        // Common adjectives & descriptors (not technical skills)
+        "fluent",
+        "proficient",
+        "deep",
+        "solid",
+        "proven",
+        "hands-on",
+        "detail-oriented",
+        "results-driven",
+        "self-motivated",
+        "proactive",
+        "creative",
+        "innovative",
+        "dynamic",
+        "strategic",
+        "analytical",
+        "collaborative",
+        "effective",
+        "efficient",
+        "reliable",
+        "flexible",
+        "adaptable",
+        "motivated",
+        "dedicated",
+        "capable",
+        "qualified",
+        "diverse",
+        "inclusive",
+        "global",
+        "local",
+        "remote",
+        "hybrid",
+        "onsite",
+        "full-time",
+        "part-time",
+        "contract",
+        "permanent",
+        // Role titles & department names (not skills themselves)
+        "designer",
+        "designers",
+        "developer",
+        "developers",
+        "engineer",
+        "engineers",
+        "manager",
+        "managers",
+        "director",
+        "analyst",
+        "analysts",
+        "architect",
+        "architects",
+        "consultant",
+        "consultants",
+        "specialist",
+        "specialists",
+        "coordinator",
+        "lead",
+        "principal",
+        "staff",
+        "junior",
+        "mid",
+        "department",
+        "organization",
+        "division",
+        "stakeholder",
+        "stakeholders",
+        "client",
+        "clients",
+        "customer",
+        "customers",
+        // Date & time words
+        "date",
+        "dates",
+        "month",
+        "months",
+        "week",
+        "weeks",
+        "daily",
+        "weekly",
+        "monthly",
+        "quarterly",
+        "annual",
+        "annually"
       ]
     };
     en_default = en;
@@ -1333,8 +1532,20 @@ var init_genericChecks = __esm({
 });
 
 // src/ats/jdMatcher.ts
+function stripNoise(text) {
+  return text.replace(/https?:\/\/[^\s]+/gi, " ").replace(/www\.[^\s]+/gi, " ").replace(/[\w.+-]+@[\w.-]+\.[a-z]{2,}/gi, " ").replace(/(?:^|\s)\/[\w/.-]+/g, " ").replace(/\b[a-z]+[A-Z][a-zA-Z]*\b/g, (match2) => {
+    return match2.replace(/([a-z])([A-Z])/g, "$1 $2");
+  });
+}
 function tokenize(text, stopWords) {
-  return text.toLowerCase().replace(/[^a-zA-Z0-9äöüßÄÖÜàáâãéèêëíìîïóòôõúùûüñç\s/+-]/g, " ").split(/\s+/).filter((word) => word.length > 2 && !stopWords.has(word));
+  return text.toLowerCase().replace(/[^a-zA-Z0-9äöüßÄÖÜàáâãéèêëíìîïóòôõúùûüñç\s/+-]/g, " ").split(/\s+/).filter((word) => {
+    if (word.length <= 2) return false;
+    if (stopWords.has(word)) return false;
+    if (word.startsWith("//") || word.startsWith("http")) return false;
+    if (/^\d+$/.test(word)) return false;
+    if (/^[/+-]+$/.test(word)) return false;
+    return true;
+  });
 }
 function simpleStem(word, language) {
   if (language === "de") {
@@ -1491,17 +1702,84 @@ function extractBrandNames(text) {
     "firefox"
   ];
   for (const b of knownBrands) brands.add(b);
+  const locations = [
+    "zurich",
+    "z\xFCrich",
+    "berlin",
+    "london",
+    "paris",
+    "amsterdam",
+    "munich",
+    "m\xFCnchen",
+    "new york",
+    "san francisco",
+    "seattle",
+    "austin",
+    "boston",
+    "chicago",
+    "toronto",
+    "vancouver",
+    "singapore",
+    "tokyo",
+    "sydney",
+    "dublin",
+    "bangalore",
+    "hyderabad",
+    "remote",
+    "hybrid",
+    "onsite",
+    "switzerland",
+    "germany",
+    "france",
+    "spain",
+    "italy",
+    "netherlands",
+    "united states",
+    "united kingdom",
+    "canada",
+    "australia",
+    "india",
+    "japan",
+    "china",
+    "brazil",
+    "israel",
+    "sweden",
+    "norway",
+    "denmark",
+    "finland",
+    "austria",
+    "belgium",
+    "portugal",
+    "ireland",
+    "poland"
+  ];
+  for (const loc of locations) {
+    brands.add(loc);
+    for (const part of loc.split(/\s+/)) {
+      if (part.length > 2) brands.add(part);
+    }
+  }
   return brands;
 }
-function extractKeywords(text, language, maxKeywords = 30) {
+function extractKeywords(text, language, maxKeywords = 30, extraStopWords) {
   const langData = getLanguageData(language);
-  const stopWords = new Set(langData.stopWords);
-  const compoundTerms = extractCompoundTerms(text);
+  const stopWords = /* @__PURE__ */ new Set([
+    ...langData.stopWords,
+    ...langData.actionVerbs,
+    ...langData.actionVerbs.map((v) => simpleStem(v, language)),
+    ...extraStopWords || []
+  ]);
+  const cleanText = stripNoise(text);
+  const compoundTerms = extractCompoundTerms(cleanText);
   const brandNames = extractBrandNames(text);
-  const { requirementText, otherText } = splitJdSections(text);
-  const reqTokens = tokenize(requirementText, stopWords).filter((t) => !brandNames.has(t));
-  const otherTokens = tokenize(otherText, stopWords).filter((t) => !brandNames.has(t));
-  const allTokens = [...reqTokens, ...reqTokens, ...reqTokens, ...otherTokens];
+  const { requirementText } = splitJdSections(cleanText);
+  const hasRequirementSections = requirementText.trim().length > 0;
+  let allTokens;
+  if (hasRequirementSections) {
+    allTokens = tokenize(requirementText, stopWords).filter((t) => !brandNames.has(t));
+  } else {
+    allTokens = tokenize(cleanText, stopWords).filter((t) => !brandNames.has(t));
+  }
   const stemmed = allTokens.map((t) => simpleStem(t, language));
   const tf = buildTfMap(stemmed);
   const stemToOriginal = /* @__PURE__ */ new Map();
@@ -1515,7 +1793,7 @@ function extractKeywords(text, language, maxKeywords = 30) {
   const compoundWordSet = new Set(compoundsFlat);
   const singleKeywords = [...tf.entries()].filter(([stem]) => stem.length > 2).filter(([stem]) => {
     const original = stemToOriginal.get(stem) || stem;
-    if (compoundWordSet.has(original) && !reqTokens.includes(original)) {
+    if (compoundWordSet.has(original) && !allTokens.includes(original)) {
       return false;
     }
     return true;
@@ -1533,7 +1811,19 @@ function extractKeywords(text, language, maxKeywords = 30) {
 function matchJobDescription(resume, jobDescription, language = "en") {
   const langData = getLanguageData(language);
   const stopWords = new Set(langData.stopWords);
-  const jdKeywords = extractKeywords(jobDescription, language);
+  const nameStopWords = /* @__PURE__ */ new Set();
+  if (resume.basics?.name) {
+    for (const part of resume.basics.name.toLowerCase().split(/\s+/)) {
+      if (part.length > 2) nameStopWords.add(part);
+    }
+  }
+  if (resume.basics?.location?.city) {
+    nameStopWords.add(resume.basics.location.city.toLowerCase());
+  }
+  if (resume.basics?.location?.region) {
+    nameStopWords.add(resume.basics.location.region.toLowerCase());
+  }
+  const jdKeywords = extractKeywords(jobDescription, language, 30, nameStopWords);
   const resumeText = extractResumeText(resume);
   const resumeTokens = tokenize(resumeText, stopWords);
   const resumeStems = new Set(resumeTokens.map((t) => simpleStem(t, language)));