Update app.py

app.py

@@ -317,13 +317,59 @@ def find_float_after_label(text: str, label: str, max_numbers: int = 2, window:
             pass
     return out
 
-def extract_pdf_name(text: str):
-    if not text:
+
+
+def extract_name_from_filename(filename: str):
+    """
+    Ex:
+    19850515_ERIC_TEVANE_124522_Analyse de la marche FDM-T_Report (1).pdf
+    -> ERIC TEVANE
+    """
+    if not filename:
         return None
 
-    # Fallback spécifique si le texte contient bien le nom
-    if "ERIC TEVANE" in text.upper():
-        return "ERIC TEVANE"
+    base = filename.rsplit("/", 1)[-1]
+    base = base.rsplit(".", 1)[0]
+
+    parts = base.split("_")
+    name_parts = []
+
+    # on saute le premier token s'il est purement numérique (date)
+    started = False
+    for p in parts:
+        p_clean = p.strip()
+        if not p_clean:
+            continue
+
+        if not started:
+            if p_clean.isdigit():
+                continue
+            if re.fullmatch(r"[A-ZÀ-Ÿ\-]+", p_clean):
+                started = True
+                name_parts.append(p_clean)
+            continue
+
+        # quand on a commencé à lire le nom, on continue tant que c'est alpha MAJ
+        if re.fullmatch(r"[A-ZÀ-Ÿ\-]+", p_clean):
+            name_parts.append(p_clean)
+        else:
+            break
+
+    if name_parts:
+        return " ".join(name_parts)
+
+    return None
+
+
+def extract_pdf_name(text: str, source_pdf: str = None):
+    # 1) fallback robuste via nom de fichier
+    name_from_file = extract_name_from_filename(source_pdf) if source_pdf else None
+    if name_from_file:
+        return name_from_file
+
+    # 2) fallback texte si jamais le PDF est mieux extrait
+    if not text:
+        return None
 
     patterns = [
         r"Personne:\s*([A-Za-zÀ-ÿ\- ]+),\s*\d{2}/\d{2}/\d{4}",
@@ -339,7 +385,12 @@ def extract_pdf_name(text: str):
 
     return None
 
+
 def parse_zebris_pdf(uploaded_pdf):
+    uploaded_pdf.seek(0)
+    raw = uploaded_pdf.read()
+    uploaded_pdf.seek(0)
+
     source_pdf = uploaded_pdf.name
 
     data = {
@@ -368,64 +419,90 @@ def parse_zebris_pdf(uploaded_pdf):
         "attaque_pdf": "indéterminée",
     }
 
-    # -------------------------------------------------
-    # 1) Fallback robuste sur le nom du fichier
-    # ex: 19850515_ERIC_TEVANE_124522_Analyse...
-    # -------------------------------------------------
-    filename_upper = source_pdf.upper()
-
-    m = re.search(r"\d{8}_([A-Z]+)_([A-Z]+)_", filename_upper)
-    if m:
-        first_name = m.group(1).strip()
-        last_name = m.group(2).strip()
-        data["athlete_name"] = f"{first_name} {last_name}"
-
-    # -------------------------------------------------
-    # 2) Cas spécifique du PDF ERIC TEVANE
-    # car l'extraction texte pdfplumber est corrompue
-    # -------------------------------------------------
-    if "ERIC_TEVANE" in filename_upper or "TEVANE_ERIC" in filename_upper:
-        data["athlete_name"] = "ERIC TEVANE"
-
-        data["transition_g"] = 0.01
-        data["transition_d"] = 0.02
-
-        data["fore_force_g"] = 1066.5
-        data["fore_force_d"] = 1040.5
-        data["mid_force_g"] = 342.7
-        data["mid_force_d"] = 306.2
-        data["heel_force_g"] = 141.9
-        data["heel_force_d"] = 181.6
-
-        data["fore_pressure_g"] = 20.4
-        data["fore_pressure_d"] = 20.6
-        data["mid_pressure_g"] = 16.6
-        data["mid_pressure_d"] = 15.2
-        data["heel_pressure_g"] = 6.1
-        data["heel_pressure_d"] = 6.5
-
-        data["fore_peak_time_pct_g"] = 50.4
-        data["fore_peak_time_pct_d"] = 50.0
-        data["mid_peak_time_pct_g"] = 30.9
-        data["mid_peak_time_pct_d"] = 23.6
-        data["heel_peak_time_pct_g"] = 12.2
-        data["heel_peak_time_pct_d"] = 12.2
+    # Nom athlète : priorité au nom de fichier, car le texte extrait est parfois corrompu
+    data["athlete_name"] = extract_pdf_name("", source_pdf=source_pdf)
 
+    def to_float(x):
+        return float(x.replace(",", "."))
+
+    with pdfplumber.open(io.BytesIO(raw)) as pdf:
+        page_texts = []
+        for page in pdf.pages:
+            txt = page.extract_text() or ""
+            txt = txt.replace("\xa0", " ")
+            page_texts.append(txt)
+
+    full_text = "\n".join(page_texts)
+
+    # fallback si le nom de fichier n'a pas suffi
+    if not data["athlete_name"]:
+        data["athlete_name"] = extract_pdf_name(full_text, source_pdf=source_pdf)
+
+    # ---------------------------------------------------
+    # PAGE 8 : "Analyse du pieds en trois zones"
+    # On parse les paires Gauche / Droite dans l'ordre.
+    # ---------------------------------------------------
+    zone_page_text = None
+    for txt in page_texts:
+        if "Analyse du pieds en trois zones" in txt:
+            zone_page_text = txt
+            break
+
+    if zone_page_text is None:
         data["attaque_pdf"] = estimate_attack_from_pdf(data)
         return data
 
-    # -------------------------------------------------
-    # 3) Sinon, tentative standard sur le texte PDF
-    # -------------------------------------------------
-    try:
-        text = extract_text_from_pdf(uploaded_pdf)
-    except Exception:
-        return data
+    # normalisation
+    zone_page_text = zone_page_text.replace("\xa0", " ")
+    zone_page_text = re.sub(r"\s+", " ", zone_page_text).strip()
+
+    # Dans TON extraction pdfplumber, "Droite" est souvent corrompu en "oeitDr"
+    # et l'écart-type sort avec un tiret/en dash, ex: 1066,5–67,8
+    # On récupère seulement la valeur moyenne gauche puis droite.
+    pair_pattern = re.compile(
+        r"Gauche\s+(\d+,\d+|\d+\.\d+|\d+)\s*[–-]\s*(\d+,\d+|\d+\.\d+|\d+).*?"
+        r"(?:Droite|oeitDr)\s+(\d+,\d+|\d+\.\d+|\d+)\s*[–-]\s*(\d+,\d+|\d+\.\d+|\d+)",
+        flags=re.S
+    )
 
-    if not data["athlete_name"]:
-        extracted_name = extract_pdf_name(text)
-        if extracted_name:
-            data["athlete_name"] = extracted_name
+    matches = pair_pattern.findall(zone_page_text)
+
+    # Ordre attendu sur la page 8 Zebris :
+    # 0 transition s
+    # 1 transition %
+    # 2 fore force
+    # 3 mid force
+    # 4 heel force
+    # 5 fore pressure
+    # 6 mid pressure
+    # 7 heel pressure
+    # 8 fore peak time
+    # 9 mid peak time
+    # 10 heel peak time
+    #
+    # Sur ton PDF, cet ordre est bien visible dans la page "Analyse du pieds en trois zones". :contentReference[oaicite:1]{index=1}
+
+    values = []
+    for m in matches:
+        g_mean = to_float(m[0])
+        d_mean = to_float(m[2])
+        values.append((g_mean, d_mean))
+
+    if len(values) >= 11:
+        data["transition_g"], data["transition_d"] = values[0]
+        # values[1] = transition en %, non utilisé ici
+
+        data["fore_force_g"], data["fore_force_d"] = values[2]
+        data["mid_force_g"], data["mid_force_d"] = values[3]
+        data["heel_force_g"], data["heel_force_d"] = values[4]
+
+        data["fore_pressure_g"], data["fore_pressure_d"] = values[5]
+        data["mid_pressure_g"], data["mid_pressure_d"] = values[6]
+        data["heel_pressure_g"], data["heel_pressure_d"] = values[7]
+
+        data["fore_peak_time_pct_g"], data["fore_peak_time_pct_d"] = values[8]
+        data["mid_peak_time_pct_g"], data["mid_peak_time_pct_d"] = values[9]
+        data["heel_peak_time_pct_g"], data["heel_peak_time_pct_d"] = values[10]
 
     data["attaque_pdf"] = estimate_attack_from_pdf(data)
     return data
@@ -1214,7 +1291,6 @@ if uploaded_pdfs:
     for pdf in uploaded_pdfs:
         try:
             pdfs_data.append(parse_zebris_pdf(pdf))
-            st.write("PDF parsé :", pdfs_data[-1])
         except Exception as e:
             st.warning(f"{pdf.name} : erreur lecture PDF ({e})")