Update app.py

app.py

@@ -317,6 +317,17 @@ def find_float_after_label(text: str, label: str, max_numbers: int = 2, window:
             pass
     return out
 
+def extract_pdf_name(text: str):
+    patterns = [
+        r"Personne:\s*([A-Za-zÀ-ÿ\- ]+),\s*\d{2}/\d{2}/\d{4}",
+        r"Personne:\s*([A-Za-zÀ-ÿ\- ]+)",
+    ]
+    for pattern in patterns:
+        m = re.search(pattern, text, flags=re.S)
+        if m:
+            return m.group(1).strip()
+    return None
+
 
 def parse_zebris_pdf(uploaded_pdf):
     text = extract_text_from_pdf(uploaded_pdf)
@@ -347,90 +358,71 @@ def parse_zebris_pdf(uploaded_pdf):
         "fore_peak_time_pct_d": np.nan,
     }
 
-    text_flat = re.sub(r"\s+", " ", text)
+    # On normalise le texte pour éviter les problèmes de retours ligne / espaces
+    text_flat = re.sub(r"\s+", " ", text).replace("\xa0", " ").strip()
+
+    num = r"(\d+,\d+|\d+\.\d+|\d+)"
 
     def to_float(x):
         return float(x.replace(",", "."))
 
-    # 1) Transition talon -> avant-pied (s)
-    m = re.search(
-        r"Instant du passage du talon vers l'avant-pied, s .*? Gauche .*? Droite .*? "
-        r"(\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±",
-        text_flat,
-        flags=re.S
-    )
-    if m:
-        data["transition_g"] = to_float(m.group(1))
-        data["transition_d"] = to_float(m.group(2))
-
-    # 2) Force maximale, N
-    m = re.search(
-        r"Force maximale, N .*?"
-        r"Forefoot \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±.*?"
-        r"Midfoot \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±.*?"
-        r"Heel \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±",
-        text_flat,
-        flags=re.S
-    )
-    if m:
-        data["fore_force_g"] = to_float(m.group(1))
-        data["fore_force_d"] = to_float(m.group(2))
-        data["mid_force_g"] = to_float(m.group(3))
-        data["mid_force_d"] = to_float(m.group(4))
-        data["heel_force_g"] = to_float(m.group(5))
-        data["heel_force_d"] = to_float(m.group(6))
-
-    # 3) Pression maximale, N/cm²
-    m = re.search(
-        r"Pression maximale, N/cm² .*?"
-        r"Forefoot \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±.*?"
-        r"Midfoot \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±.*?"
-        r"Heel \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±",
-        text_flat,
-        flags=re.S
-    )
-    if m:
-        data["fore_pressure_g"] = to_float(m.group(1))
-        data["fore_pressure_d"] = to_float(m.group(2))
-        data["mid_pressure_g"] = to_float(m.group(3))
-        data["mid_pressure_d"] = to_float(m.group(4))
-        data["heel_pressure_g"] = to_float(m.group(5))
-        data["heel_pressure_d"] = to_float(m.group(6))
-
-    # 4) Instant pic de force, % de phase d'appui
-    m = re.search(
-        r"Instant pic de force, % de phase d'appui .*?"
-        r"Forefoot \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±.*?"
-        r"Midfoot \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±.*?"
-        r"Heel \(Three zones\) Gauche Droite (\d+,\d+|\d+\.\d+)\±.*?(\d+,\d+|\d+\.\d+)\±",
-        text_flat,
-        flags=re.S
-    )
-    if m:
-        data["fore_peak_time_pct_g"] = to_float(m.group(1))
-        data["fore_peak_time_pct_d"] = to_float(m.group(2))
-        data["mid_peak_time_pct_g"] = to_float(m.group(3))
-        data["mid_peak_time_pct_d"] = to_float(m.group(4))
-        data["heel_peak_time_pct_g"] = to_float(m.group(5))
-        data["heel_peak_time_pct_d"] = to_float(m.group(6))
+    def extract_pair(section_label, sub_label=None):
+        """
+        Extrait une paire Gauche / Droite du type :
+        ... Gauche Droite 1066,5±67,8 1040,5±49,3 ...
+        """
+        if sub_label:
+            pattern = (
+                rf"{re.escape(section_label)}.*?"
+                rf"{re.escape(sub_label)}\s+Gauche\s+Droite\s+"
+                rf"{num}\s*±\s*{num}\s+{num}\s*±\s*{num}"
+            )
+        else:
+            pattern = (
+                rf"{re.escape(section_label)}\s+Gauche\s+Droite\s+"
+                rf"{num}\s*±\s*{num}\s+{num}\s*±\s*{num}"
+            )
 
-    data["attaque_pdf"] = estimate_attack_from_pdf(data)
-    return data
+        m = re.search(pattern, text_flat, flags=re.S)
+        if not m:
+            return np.nan, np.nan
 
+        # groupes 1 et 3 = valeurs moyennes G / D
+        return to_float(m.group(1)), to_float(m.group(3))
 
-def extract_pdf_name(text: str):
-    patterns = [
-        r"Personne:\s*([A-Za-zÀ-ÿ\- ]+),\s*\d{2}/\d{2}/\d{4}",
-        r"Personne:\s*([A-Za-zÀ-ÿ\- ]+)",
-    ]
+    # 1) Transition (section spécifique)
+    tg, td = extract_pair("Instant du passage du talon vers l'avant-pied, s")
+    data["transition_g"], data["transition_d"] = tg, td
 
-    for pattern in patterns:
-        m = re.search(pattern, text)
-        if m:
-            return m.group(1).strip()
+    # 2) Forces max
+    fg, fd = extract_pair("Force maximale, N", "Forefoot (Three zones)")
+    mg, md = extract_pair("Force maximale, N", "Midfoot (Three zones)")
+    hg, hd = extract_pair("Force maximale, N", "Heel (Three zones)")
 
-    return None
+    data["fore_force_g"], data["fore_force_d"] = fg, fd
+    data["mid_force_g"], data["mid_force_d"] = mg, md
+    data["heel_force_g"], data["heel_force_d"] = hg, hd
+
+    # 3) Pressions max
+    fpg, fpd = extract_pair("Pression maximale, N/cm²", "Forefoot (Three zones)")
+    mpg, mpd = extract_pair("Pression maximale, N/cm²", "Midfoot (Three zones)")
+    hpg, hpd = extract_pair("Pression maximale, N/cm²", "Heel (Three zones)")
+
+    data["fore_pressure_g"], data["fore_pressure_d"] = fpg, fpd
+    data["mid_pressure_g"], data["mid_pressure_d"] = mpg, mpd
+    data["heel_pressure_g"], data["heel_pressure_d"] = hpg, hpd
 
+    # 4) Timing pic de force
+    ftg, ftd = extract_pair("Instant pic de force, % de phase d'appui", "Forefoot (Three zones)")
+    mtg, mtd = extract_pair("Instant pic de force, % de phase d'appui", "Midfoot (Three zones)")
+    htg, htd = extract_pair("Instant pic de force, % de phase d'appui", "Heel (Three zones)")
+
+    data["fore_peak_time_pct_g"], data["fore_peak_time_pct_d"] = ftg, ftd
+    data["mid_peak_time_pct_g"], data["mid_peak_time_pct_d"] = mtg, mtd
+    data["heel_peak_time_pct_g"], data["heel_peak_time_pct_d"] = htg, htd
+
+    data["attaque_pdf"] = estimate_attack_from_pdf(data)
+    return data
 
 def estimate_attack_from_pdf(pdf_data: dict):
     heel_peak_t = avg(pdf_data.get("heel_peak_time_pct_g"), pdf_data.get("heel_peak_time_pct_d"))
@@ -1216,6 +1208,7 @@ if uploaded_pdfs:
     for pdf in uploaded_pdfs:
         try:
             pdfs_data.append(parse_zebris_pdf(pdf))
+            st.write(pdfs_data[-1])
         except Exception as e:
             st.warning(f"{pdf.name} : erreur lecture PDF ({e})")