Fix experience section parsing

src/main/java/com/resumeparser/processor/EntityExtractor.java

@@ -310,10 +310,10 @@ public class EntityExtractor {
             "CERTIFICATIONS", "LANGUAGES", "SUMMARY", "ACHIEVEMENTS"
         );
 
-        Pattern datePattern = Pattern.compile("(\\b\\d{4}\\b|\\b[A-Z][a-z]{2,8}\\s+\\d{4}\\b)(\\s*[-—–]\\s*(\\b\\d{4}\\b|\\b[A-Z][a-z]{2,8}\\s+\\d{4}\\b|Present))?");
+        Pattern datePattern = Pattern.compile("(?i)(\\b[A-Za-z]{3,9}\\.?\\s+\\d{4}|\\b\\d{4})\\s*(?:[-—–]|to)\\s*(\\b[A-Za-z]{3,9}\\.?\\s+\\d{4}|\\b\\d{4}|Present)");
 
         // Split by double newline
-        String[] parts = text.split("\\n\\n+");
+        String[] parts = text.split("\\n\\s*\\n+");
         for (String part : parts) {
             String trimmed = part.trim();
             if (trimmed.isEmpty()) continue;
@@ -333,24 +333,109 @@ public class EntityExtractor {
             }
             if (isBlacklisted) continue;
 
-            // Simple heuristic to split title/dates/description
-            // Input: Title + Dates (Line 1), Company + Location (Line 2), Description (Lines 3+)
-            String titleAndDates = firstLine;
-            String companyInfo = lines.length > 1 ? lines[1].trim() : "";
-            StringBuilder descBuilder = new StringBuilder();
-            for (int i = 2; i < lines.length; i++) {
-                descBuilder.append(lines[i].trim()).append("\n");
+            // Pre-process: join lines that are split mid-sentence or mid-date
+            List<String> cleanLines = new ArrayList<>();
+            for (int i = 0; i < lines.length; i++) {
+                String current = lines[i].trim();
+                if (current.isEmpty()) continue;
+                
+                boolean endsWithMonth = current.matches(".*\\b(?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)[a-z]*\\.?$");
+                boolean endsWithDash = current.endsWith("-") || current.endsWith("–") || current.endsWith("—") || current.toLowerCase().endsWith("to");
+                
+                if ((endsWithMonth || endsWithDash) && i + 1 < lines.length) {
+                    String next = lines[i + 1].trim();
+                    current = current + " " + next;
+                    i++;
+                }
+                cleanLines.add(current);
             }
 
-            String title = titleAndDates;
-            String dates = "";
-            Matcher m = datePattern.matcher(titleAndDates);
-            if (m.find()) {
-                dates = m.group().trim();
-                title = titleAndDates.replace(dates, "").trim().replaceAll("\\s*[-—–]\\s*$", "").trim();
+            if (cleanLines.isEmpty()) continue;
+
+            String titleLine = cleanLines.get(0);
+            String companyInfo = cleanLines.size() > 1 ? cleanLines.get(1) : "";
+            
+            String startDate = "";
+            String endDate = "";
+            String datesStr = "";
+
+            // Find dates in titleLine or companyInfo or any of the lines
+            boolean dateFound = false;
+            for (int i = 0; i < cleanLines.size(); i++) {
+                String line = cleanLines.get(i);
+                Matcher dateMatcher = datePattern.matcher(line);
+                if (dateMatcher.find()) {
+                    datesStr = dateMatcher.group(0).trim();
+                    startDate = dateMatcher.group(1).trim();
+                    endDate = dateMatcher.group(2).trim();
+                    
+                    cleanLines.set(i, line.replace(dateMatcher.group(0), "").trim());
+                    dateFound = true;
+                    break;
+                }
+            }
+
+            titleLine = cleanLines.get(0).replaceAll("(?i)^\\s*(?:Link|Present|Date|Dates|Timeline|Timeline:)?\\s*[|:\\-–—]+\\s*", "")
+                                         .replaceAll("(?i)\\s*[|:\\-–—]+\\s*(?:Link|Present|Date|Dates|Timeline|Timeline:)?\\s*$", "")
+                                         .trim();
+            
+            if (cleanLines.size() > 1) {
+                companyInfo = cleanLines.get(1).replaceAll("(?i)^\\s*(?:Link|Present|Date|Dates|Timeline|Timeline:)?\\s*[|:\\-–—]+\\s*", "")
+                                               .replaceAll("(?i)\\s*[|:\\-–—]+\\s*(?:Link|Present|Date|Dates|Timeline|Timeline:)?\\s*$", "")
+                                               .trim();
             }
 
-            experience.add(new Candidate.Experience(companyInfo, title, dates, descBuilder.toString().trim()));
+            int descStartIndex = 2;
+            if (titleLine.isEmpty() && cleanLines.size() > 1) {
+                titleLine = companyInfo;
+                companyInfo = cleanLines.size() > 2 ? cleanLines.get(2) : "";
+                descStartIndex = 3;
+            }
+
+            // Extract location from companyInfo
+            String companyName = companyInfo;
+            String location = "";
+            if (companyInfo.contains(",")) {
+                int commaIdx = companyInfo.lastIndexOf(",");
+                String beforeComma = companyInfo.substring(0, commaIdx).trim();
+                String afterComma = companyInfo.substring(commaIdx + 1).trim();
+                
+                int lastSpace = beforeComma.lastIndexOf(" ");
+                if (lastSpace >= 0) {
+                    location = beforeComma.substring(lastSpace + 1).trim() + ", " + afterComma;
+                    companyName = beforeComma.substring(0, lastSpace).trim();
+                } else {
+                    location = beforeComma + ", " + afterComma;
+                    companyName = "";
+                }
+            }
+
+            // Extract description & bullets
+            List<String> bulletsList = new ArrayList<>();
+            StringBuilder descBuilder = new StringBuilder();
+
+            for (int i = descStartIndex; i < cleanLines.size(); i++) {
+                String line = cleanLines.get(i).trim();
+                if (line.isEmpty()) continue;
+
+                if (line.contains("mailto:") || line.startsWith("http") || line.contains("github.com") || line.contains("linkedin.com")) {
+                    continue; // skip noise lines at the end
+                }
+
+                descBuilder.append(line).append("\n");
+                String cleanBullet = line.replaceAll("^[\\s•\\-*▪]+", "").trim();
+                if (!cleanBullet.isEmpty()) {
+                    bulletsList.add(cleanBullet);
+                }
+            }
+
+            Candidate.Experience exp = new Candidate.Experience(companyName, titleLine, datesStr, descBuilder.toString().trim());
+            exp.setFromDate(startDate);
+            exp.setToDate(endDate);
+            exp.setLocation(location);
+            exp.setBullets(bulletsList);
+
+            experience.add(exp);
         }
         return experience;
     }