Create zebris_extractor.py

zebris_extractor.py

@@ -0,0 +1,255 @@
+import io
+import numpy as np
+import pandas as pd
+
+
+STANDARD_COLUMNS = [
+    "Nom",
+    "Date",
+    "Poids (kg)",
+    "Vitesse (km/h)",
+    "Cadence (pas/min)",
+    "Contact (%)",
+    "Flight (%)",
+    "Force talon G (N)",
+    "Force talon D (N)",
+    "Force avant-pied G (N)",
+    "Force avant-pied D (N)",
+    "Pression talon G (N/cm²)",
+    "Pression talon D (N/cm²)",
+    "COP G (mm)",
+    "COP D (mm)",
+    "Rotation G (°)",
+    "Rotation D (°)",
+    "Transition G (s)",
+    "Transition D (s)",
+    "Longueur foulée (cm)",
+    "Largeur pas (cm)",
+]
+
+
+def _norm(text):
+    return (
+        str(text)
+        .strip()
+        .lower()
+        .replace("é", "e")
+        .replace("è", "e")
+        .replace("ê", "e")
+        .replace("à", "a")
+        .replace("ù", "u")
+        .replace("ç", "c")
+        .replace("²", "2")
+        .replace("°", "")
+    )
+
+
+COLUMN_CANDIDATES = {
+    "Nom": [
+        ("last name", "first name"),
+        ("nom", "prenom"),
+        ("last name", "first name "),
+    ],
+    "Date": [
+        "date",
+        "recording date",
+        "enregistrement",
+    ],
+    "Poids (kg)": [
+        "weight (kg)",
+        "poids (kg)",
+        "weight",
+        "poids",
+    ],
+    "Vitesse (km/h)": [
+        "speed (km/h)",
+        "vitesse (km/h)",
+        "speed",
+        "vitesse",
+    ],
+    "Cadence (pas/min)": [
+        "cadence (steps/min)",
+        "cadence (pas/min)",
+        "cadence",
+        "cadence, pass/min",
+    ],
+    "Contact (%)": [
+        "total contact (%)",
+        "contact (%)",
+        "total contact",
+    ],
+    "Flight (%)": [
+        "total flight (%)",
+        "flight (%)",
+        "total flight",
+    ],
+    "Force talon G (N)": [
+        "heel (three zones) left force max (n)",
+        "mean force heel left (n)",
+        "heel left force max (n)",
+        "heel left (n)",
+    ],
+    "Force talon D (N)": [
+        "heel (three zones) right force max (n)",
+        "mean force heel right (n)",
+        "heel right force max (n)",
+        "heel right (n)",
+    ],
+    "Force avant-pied G (N)": [
+        "forefoot (three zones) left force max (n)",
+        "mean force forefoot left (n)",
+        "forefoot left force max (n)",
+        "forefoot left (n)",
+    ],
+    "Force avant-pied D (N)": [
+        "forefoot (three zones) right force max (n)",
+        "mean force forefoot right (n)",
+        "forefoot right force max (n)",
+        "forefoot right (n)",
+    ],
+    "Pression talon G (N/cm²)": [
+        "heel (three zones) left pressure max (n/cm2)",
+        "mean pressure heel left (n/cm2)",
+        "heel left pressure max (n/cm2)",
+    ],
+    "Pression talon D (N/cm²)": [
+        "heel (three zones) right pressure max (n/cm2)",
+        "mean pressure heel right (n/cm2)",
+        "heel right pressure max (n/cm2)",
+    ],
+    "COP G (mm)": [
+        "cop length left (mm)",
+        "cop parameters running left length (mm)",
+        "cop left (mm)",
+    ],
+    "COP D (mm)": [
+        "cop length right (mm)",
+        "cop parameters running right length (mm)",
+        "cop right (mm)",
+    ],
+    "Rotation G (°)": [
+        "foot rotation left (deg)",
+        "foot rotation left",
+        "rotation left",
+    ],
+    "Rotation D (°)": [
+        "foot rotation right (deg)",
+        "foot rotation right",
+        "rotation right",
+    ],
+    "Transition G (s)": [
+        "heel to forefoot transition left (s)",
+        "transition left (s)",
+        "instant du passage du talon vers l'avant-pied gauche (s)",
+    ],
+    "Transition D (s)": [
+        "heel to forefoot transition right (s)",
+        "transition right (s)",
+        "instant du passage du talon vers l'avant-pied droite (s)",
+    ],
+    "Longueur foulée (cm)": [
+        "stride length (cm)",
+        "longueur de la foulee (cm)",
+        "longueur de la foulee",
+    ],
+    "Largeur pas (cm)": [
+        "step width (cm)",
+        "largeur du pas (cm)",
+        "largeur du pas",
+    ],
+}
+
+
+def _to_numeric(series):
+    return pd.to_numeric(
+        series.astype(str).str.replace(",", ".", regex=False),
+        errors="coerce"
+    )
+
+
+def _read_csv_flex(uploaded_file):
+    raw = uploaded_file.read()
+    uploaded_file.seek(0)
+
+    for encoding in ["utf-8-sig", "utf-8", "latin1", "cp1252"]:
+        for sep in [";", ",", "\t"]:
+            try:
+                txt = raw.decode(encoding)
+                df = pd.read_csv(io.StringIO(txt), sep=sep)
+                if df.shape[1] > 1:
+                    return df
+            except Exception:
+                pass
+
+    # dernier essai naïf
+    return pd.read_csv(uploaded_file)
+
+
+def _find_column(df, candidates):
+    normalized_cols = {_norm(c): c for c in df.columns}
+    for cand in candidates:
+        key = _norm(cand)
+        if key in normalized_cols:
+            return normalized_cols[key]
+    return None
+
+
+def extract_zebris_csv(uploaded_file):
+    df = _read_csv_flex(uploaded_file)
+
+    out = pd.DataFrame(index=df.index)
+    mapping_debug = {}
+    manquantes = []
+
+    # Nom
+    nom_done = False
+    for cand_pair in COLUMN_CANDIDATES["Nom"]:
+        if isinstance(cand_pair, tuple):
+            c1 = _find_column(df, [cand_pair[0]])
+            c2 = _find_column(df, [cand_pair[1]])
+            if c1 and c2:
+                out["Nom"] = df[c1].astype(str).str.strip() + " " + df[c2].astype(str).str.strip()
+                mapping_debug["Nom"] = [c1, c2]
+                nom_done = True
+                break
+
+    if not nom_done:
+        c = _find_column(df, ["nom"])
+        if c:
+            out["Nom"] = df[c].astype(str)
+            mapping_debug["Nom"] = c
+        else:
+            out["Nom"] = "Inconnu"
+            manquantes.append("Nom")
+
+    # Autres colonnes
+    for target in STANDARD_COLUMNS:
+        if target == "Nom":
+            continue
+
+        candidates = COLUMN_CANDIDATES.get(target, [])
+        col = _find_column(df, candidates)
+
+        if col is None:
+            out[target] = np.nan
+            manquantes.append(target)
+        else:
+            mapping_debug[target] = col
+            if target == "Date":
+                out[target] = df[col]
+            else:
+                out[target] = _to_numeric(df[col])
+
+    # Garde lignes avec allure
+    if "Vitesse (km/h)" in out.columns:
+        out = out[out["Vitesse (km/h)"].notna()].copy()
+
+    out = out.reindex(columns=STANDARD_COLUMNS).reset_index(drop=True)
+
+    debug = {
+        "mapping": mapping_debug,
+        "manquantes": manquantes,
+        "colonnes_csv": list(df.columns),
+    }
+
+    return out, debug