adding ats suggestion code(basic)

app.py

@@ -2,20 +2,28 @@ from fastapi import FastAPI, UploadFile, File, Form
 from fastapi.middleware.cors import CORSMiddleware
 from ats_core import extract_pdf_text, ats_score
 from role_templates import ROLE_TEMPLATES
-
+from ai_suggestions import generate_suggestions
 app = FastAPI()
 
-app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"],
-    allow_methods=["*"],
-    allow_headers=["*"],
-)
+app.add_middleware(CORSMiddleware, allow_origins=["*"], allow_methods=["*"], allow_headers=["*"],)
 
 @app.get("/health")
 def health():
     return {"status": "ok"}
 
+# @app.post("/ats-score")
+# async def score( resume: UploadFile = File(...), role: str = Form(...), job_description: str = Form("") ):
+#     pdf_bytes = await resume.read()
+#     resume_text = extract_pdf_text(pdf_bytes)
+
+#     role = role.lower().strip()
+#     jd = job_description if job_description.strip() else ROLE_TEMPLATES.get(role, "")
+
+#     if not jd:
+#         return {"error": "Invalid role"}
+
+#     return ats_score(resume_text, jd, role)
+
 @app.post("/ats-score")
 async def score(
     resume: UploadFile = File(...),
@@ -31,4 +39,9 @@ async def score(
     if not jd:
         return {"error": "Invalid role"}
 
-    return ats_score(resume_text, jd, role)
+    ats_result = ats_score(resume_text, jd, role)
+
+    ai_text = generate_suggestions(resume_text, jd, ats_result)
+
+    ats_result["ai_suggestions"] = ai_text
+    return ats_result

ats_core.py

@@ -12,25 +12,12 @@ from nltk.corpus import stopwords
 from nltk import pos_tag
 from nltk.tokenize import wordpunct_tokenize  # ✅ SAFE TOKENIZER
 
-# ----------------------------
-# Logging
-# ----------------------------
 logging.getLogger("pdfminer").setLevel(logging.ERROR)
 
-# ----------------------------
-# REQUIRED NLTK ASSETS
-# (download ONLY in Docker build)
-# ----------------------------
 STOPWORDS = set(stopwords.words("english"))
 
-# ----------------------------
-# Embedding model
-# ----------------------------
 model = SentenceTransformer("all-MiniLM-L6-v2")
 
-# ----------------------------
-# Helpers
-# ----------------------------
 def f(x):
     return float(round(x, 2))
 
@@ -39,9 +26,6 @@ def clean(text: str) -> str:
     text = re.sub(r"[^a-z0-9\s\-]", " ", text)
     return re.sub(r"\s+", " ", text).strip()
 
-# ----------------------------
-# Resume text extraction
-# ----------------------------
 def extract_pdf_text(file_bytes: bytes) -> str:
     text = ""
     with pdfplumber.open(BytesIO(file_bytes)) as pdf:
@@ -50,9 +34,6 @@ def extract_pdf_text(file_bytes: bytes) -> str:
                 text += page.extract_text() + "\n"
     return text.strip()
 
-# ----------------------------
-# Contact info extraction
-# ----------------------------
 def extract_email(text: str):
     m = re.search(r"[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}", text)
     return m.group(0) if m else None
@@ -84,9 +65,6 @@ def extract_name(text: str):
                 return line
     return None
 
-# ----------------------------
-# NLP scoring
-# ----------------------------
 def embed_sim(a: str, b: str) -> float:
     emb = model.encode([a, b])
     return float(cosine_similarity([emb[0]], [emb[1]])[0][0])
@@ -131,9 +109,6 @@ def generic_penalty(text: str) -> float:
     penalty = sum(0.05 for g in GENERIC if g in text)
     return min(penalty, 0.20)
 
-# ----------------------------
-# Verdict + calibration
-# ----------------------------
 def verdict(score: float) -> str:
     if score < 35: return "Poor Fit"
     if score < 50: return "Below Average"
@@ -147,9 +122,6 @@ def calibrate(raw: float) -> float:
     scaled = (raw - RAW_MIN) / (RAW_MAX - RAW_MIN)
     return 20 + scaled * 70  # → [20, 90]
 
-# ----------------------------
-# MAIN ENTRY
-# ----------------------------
 def ats_score(resume_text: str, jd_text: str, role: str) -> Dict[str, Any]:
     resume_clean = clean(resume_text)
     jd_clean = clean(jd_text)