Upload 3 files

README.md

@@ -1,15 +1,42 @@
 ---
-title: CSV Data Anonymizer For GDPR
-emoji: 🐠
-colorFrom: gray
-colorTo: yellow
+title: CSV Data Anonymizer for GDPR
+emoji: 🔒
+colorFrom: blue
+colorTo: purple
 sdk: gradio
-sdk_version: 6.14.0
-python_version: '3.13'
+sdk_version: 4.44.1
 app_file: app.py
 pinned: false
-license: apache-2.0
-short_description: Anonymize sensitive CSV columns (emails, names) for GDPR
+license: mit
+short_description: Anonymize CSV sensitive columns for GDPR compliance
+tags:
+  - privacy
+  - gdpr
+  - data-science
+  - csv
+  - anonymization
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🔒 CSV Data Anonymizer — GDPR Ready
+
+Upload a CSV, select the columns containing sensitive data, and download an anonymized version ready to share with your team or AI tools.
+
+## Features
+
+- **Auto-detection** of data type (email, name, phone, address…)
+- **Deterministic mapping**: the same real value always produces the same fake value, preserving dataset consistency
+- **Supported types**: Email, First name, Last name, Full name, Phone, Address, City, Postal code, Tax ID, Date of birth, Generic
+- **Preview** of the result before downloading
+- **CSV download** ready to use
+
+## How to use
+
+1. Upload your CSV file
+2. Review the auto-detected column types and correct them if needed
+3. Check/uncheck the columns you want to anonymize
+4. Click **"Anonymize"**
+5. Download the anonymized CSV
+
+## Privacy note
+
+The file is processed entirely in memory on the server — no data is saved to disk or transmitted to third parties.

app.py

@@ -0,0 +1,268 @@
+import gradio as gr
+import pandas as pd
+from faker import Faker
+import io
+import re
+import json
+from typing import Optional
+
+fake = Faker(["it_IT", "en_US"])
+Faker.seed(0)
+
+# ── Tipi di anonimizzazione disponibili ────────────────────────────────────────
+ANON_TYPES = {
+    "email":          "📧 Email",
+    "nome":           "👤 Nome",
+    "cognome":        "👤 Cognome",
+    "nome_completo":  "👤 Nome Completo",
+    "telefono":       "📱 Telefono",
+    "indirizzo":      "🏠 Indirizzo",
+    "città":          "🏙️ Città",
+    "cap":            "📮 CAP",
+    "codice_fiscale": "🪪 Codice Fiscale",
+    "data_nascita":   "📅 Data di Nascita",
+    "generico":       "🔒 Generico",
+}
+
+TYPE_LABELS = {v: k for k, v in ANON_TYPES.items()}   # label → chiave interna
+
+# ── Rilevamento automatico del tipo ───────────────────────────────────────────
+def _detect_type(series: pd.Series) -> str:
+    name = (series.name or "").lower()
+    sample = series.dropna().astype(str).head(200)
+
+    # Heuristica sul nome colonna
+    if any(k in name for k in ("email", "mail", "e-mail")):
+        return "email"
+    if any(k in name for k in ("telefon", "phone", "cell", "mobile", "tel")):
+        return "telefono"
+    if any(k in name for k in ("nome", "first", "given")):
+        return "nome"
+    if any(k in name for k in ("cognome", "surname", "last")):
+        return "cognome"
+    if any(k in name for k in ("full_name", "fullname", "nome_completo")):
+        return "nome_completo"
+    if any(k in name for k in ("indirizzo", "address", "via", "street")):
+        return "indirizzo"
+    if any(k in name for k in ("città", "city", "comune", "citta")):
+        return "città"
+    if any(k in name for k in ("cap", "postal", "zip")):
+        return "cap"
+    if any(k in name for k in ("cf", "codice_fiscale", "fiscal", "tax")):
+        return "codice_fiscale"
+    if any(k in name for k in ("nascita", "birth", "dob", "birthday")):
+        return "data_nascita"
+
+    # Heuristica sul contenuto
+    if sample.str.match(r"^[A-Za-z0-9._%+\-]+@[A-Za-z0-9.\-]+\.[A-Za-z]{2,}$").mean() > 0.5:
+        return "email"
+    if sample.str.match(r"^[\+\d\s\-\(\)]{7,15}$").mean() > 0.5:
+        return "telefono"
+
+    return "generico"
+
+# ── Generatori di dati falsi (deterministici per valore) ─────────────────────
+def _make_fake(value: str, anon_type: str, cache: dict) -> str:
+    if value in cache:
+        return cache[value]
+
+    # Seed deterministico: stesso valore → stesso dato finto
+    seed = abs(hash(value)) % (2**32)
+    fake_local = Faker(["it_IT", "en_US"])
+    fake_local.seed_instance(seed)
+
+    generators = {
+        "email":          lambda: fake_local.email(),
+        "nome":           lambda: fake_local.first_name(),
+        "cognome":        lambda: fake_local.last_name(),
+        "nome_completo":  lambda: fake_local.name(),
+        "telefono":       lambda: fake_local.phone_number(),
+        "indirizzo":      lambda: fake_local.street_address(),
+        "città":          lambda: fake_local.city(),
+        "cap":            lambda: fake_local.postcode(),
+        "codice_fiscale": lambda: _fake_cf(fake_local),
+        "data_nascita":   lambda: fake_local.date_of_birth(minimum_age=18, maximum_age=80).strftime("%d/%m/%Y"),
+        "generico":       lambda: f"ANONIMIZZATO_{abs(hash(value)) % 100000:05d}",
+    }
+
+    result = generators.get(anon_type, generators["generico"])()
+    cache[value] = result
+    return result
+
+def _fake_cf(f: Faker) -> str:
+    """Genera un codice fiscale plausibile (16 char alfanumerici)."""
+    letters = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+    digits = "0123456789"
+    cf = (
+        "".join(f.random_choices(letters, length=3))
+        + "".join(f.random_choices(letters, length=3))
+        + "".join(f.random_choices(digits, length=2))
+        + f.random_choices(letters, length=1)[0]
+        + "".join(f.random_choices(digits, length=2))
+        + f.random_choices(letters, length=1)[0]
+        + "".join(f.random_choices(digits, length=3))
+        + f.random_choices(letters, length=1)[0]
+    )
+    return cf
+
+# ── Logica principale ─────────────────────────────────────────────────────────
+def load_csv(file_obj) -> tuple:
+    """Carica il CSV e restituisce anteprima + configurazione colonne."""
+    if file_obj is None:
+        return None, None, gr.update(visible=False), "Carica un file CSV per iniziare."
+
+    try:
+        df = pd.read_csv(file_obj.name, encoding="utf-8", on_bad_lines="skip")
+    except UnicodeDecodeError:
+        df = pd.read_csv(file_obj.name, encoding="latin-1", on_bad_lines="skip")
+    except Exception as e:
+        return None, None, gr.update(visible=False), f"Errore nel leggere il file: {e}"
+
+    if df.empty or len(df.columns) == 0:
+        return None, None, gr.update(visible=False), "Il file CSV sembra vuoto o mal formattato."
+
+    # Costruisce la configurazione colonne (nome, tipo rilevato, selezionata)
+    col_config = pd.DataFrame({
+        "Colonna":   df.columns.tolist(),
+        "Tipo":      [ANON_TYPES[_detect_type(df[c])] for c in df.columns],
+        "Anonimizza": [True] * len(df.columns),
+    })
+
+    preview = df.head(5)
+    msg = f"✅ Caricato: **{len(df)} righe × {len(df.columns)} colonne**. Controlla i tipi rilevati e premi **Anonimizza**."
+    return df.to_json(orient="split"), col_config, gr.update(visible=True), msg
+
+
+def anonymize(df_json: Optional[str], col_config_df: Optional[pd.DataFrame]) -> tuple:
+    """Anonimizza il DataFrame secondo la configurazione."""
+    if df_json is None:
+        return None, None, "Prima carica un CSV."
+    if col_config_df is None or len(col_config_df) == 0:
+        return None, None, "Configurazione colonne mancante."
+
+    df = pd.read_json(io.StringIO(df_json), orient="split")
+    result = df.copy()
+
+    processed = 0
+    for _, row in col_config_df.iterrows():
+        col_name  = row["Colonna"]
+        do_anon   = row["Anonimizza"]
+        type_label = row["Tipo"]
+
+        if not do_anon or col_name not in result.columns:
+            continue
+
+        anon_type = TYPE_LABELS.get(type_label, "generico")
+        cache: dict = {}
+
+        result[col_name] = result[col_name].apply(
+            lambda v: _make_fake(str(v), anon_type, cache) if pd.notna(v) and str(v).strip() != "" else v
+        )
+        processed += 1
+
+    # Esporta in CSV
+    csv_bytes = result.to_csv(index=False).encode("utf-8")
+    csv_path = "/tmp/anonimizzato.csv"
+    with open(csv_path, "wb") as f:
+        f.write(csv_bytes)
+
+    msg = f"✅ Anonimizzate **{processed} colonne** su {len(df.columns)} totali."
+    return result.head(10), csv_path, msg
+
+# ── UI ────────────────────────────────────────────────────────────────────────
+DESCRIPTION = """
+# 🔒 Data Anonymizer — GDPR Ready
+
+Carica un CSV, controlla le colonne rilevate come sensibili, e scarica la versione anonimizzata.
+Il mapping è **deterministico**: lo stesso valore produce sempre lo stesso dato finto, preservando la consistenza del dataset.
+"""
+
+with gr.Blocks(theme=gr.themes.Soft(), title="Data Anonymizer") as demo:
+    # Stato interno
+    df_state = gr.State(None)
+
+    gr.Markdown(DESCRIPTION)
+
+    with gr.Row():
+        with gr.Column(scale=1):
+            file_input = gr.File(
+                label="📂 Carica CSV",
+                file_types=[".csv"],
+                type="filepath",
+            )
+            status_box = gr.Markdown("Carica un file CSV per iniziare.")
+
+        with gr.Column(scale=2):
+            col_editor = gr.Dataframe(
+                label="⚙️ Configurazione colonne (modifica Tipo e Anonimizza)",
+                headers=["Colonna", "Tipo", "Anonimizza"],
+                datatype=["str", "str", "bool"],
+                col_count=(3, "fixed"),
+                visible=False,
+                interactive=True,
+                wrap=True,
+            )
+
+    anon_btn = gr.Button("🔒 Anonimizza", variant="primary", size="lg")
+
+    gr.Markdown("### Anteprima risultato (prime 10 righe)")
+    preview_out = gr.Dataframe(label="Anteprima anonimizzata", interactive=False, wrap=True)
+
+    download_out = gr.File(label="⬇️ Scarica CSV anonimizzato", visible=False)
+    result_status = gr.Markdown("")
+
+    # ── Event handlers ──────────────────────────────────────────────────────
+    file_input.change(
+        fn=load_csv,
+        inputs=[file_input],
+        outputs=[df_state, col_editor, col_editor, status_box],
+    )
+
+    def run_anonymize(df_json, col_config):
+        preview, path, msg = anonymize(df_json, col_config)
+        file_update = gr.update(value=path, visible=path is not None)
+        return preview, file_update, msg
+
+    anon_btn.click(
+        fn=run_anonymize,
+        inputs=[df_state, col_editor],
+        outputs=[preview_out, download_out, result_status],
+    )
+
+    gr.Examples(
+        examples=[["example.csv"]],
+        inputs=[file_input],
+        label="📋 Prova con il CSV di esempio",
+        run_on_click=False,
+    )
+
+    gr.Markdown(
+        """
+---
+*Il file viene elaborato interamente in memoria — nessun dato viene salvato su disco o trasmesso a terzi.*
+        """
+    )
+
+# ── CSV di esempio (generato al primo avvio) ──────────────────────────────────
+import os
+
+def _create_example():
+    if os.path.exists("example.csv"):
+        return
+    sample = pd.DataFrame({
+        "id":       range(1, 6),
+        "nome":     ["Mario", "Lucia", "Giovanni", "Sofia", "Marco"],
+        "cognome":  ["Rossi", "Bianchi", "Ferrari", "Esposito", "Romano"],
+        "email":    ["mario.rossi@gmail.com", "lucia.b@yahoo.it", "g.ferrari@outlook.com",
+                     "sofia.e@libero.it", "m.romano@hotmail.com"],
+        "telefono": ["+39 333 1234567", "02 9876543", "+39 347 9988776",
+                     "080 5551234", "+39 320 4567890"],
+        "città":    ["Roma", "Milano", "Napoli", "Bari", "Torino"],
+        "note":     ["cliente premium", "newsletter sì", "B2B", "trial", "enterprise"],
+    })
+    sample.to_csv("example.csv", index=False)
+
+_create_example()
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,3 @@
+gradio==4.44.1
+pandas==2.2.2
+faker==26.0.0