Initial project baseline

backend/core/weights.py

@@ -0,0 +1,101 @@
+import math
+
+import numpy as np
+import pandas as pd
+from PIL import Image, ImageDraw
+
+from . import config
+from .fonts import get_cached_font
+from .layout import normalize_relative_scores
+
+
+def get_stroke_complexity_batch(names, font_p, test_size=64):
+    font = get_cached_font(font_p, test_size)
+    img = Image.new("L", (test_size, test_size), 255)
+    draw = ImageDraw.Draw(img)
+    char_complexity_cache = {}
+    weights = {}
+    all_chars = set("".join(names))
+    for char in all_chars:
+        draw.rectangle([0, 0, test_size, test_size], fill=255)
+        draw.text((0, 0), char, font=font, fill=0)
+        char_complexity_cache[char] = np.sum(np.array(img) < 200)
+
+    unique_names = set(names)
+    for name in unique_names:
+        if not name:
+            weights[name] = 10
+            continue
+        complexities = [char_complexity_cache.get(c, 10) for c in name]
+        weights[name] = max(complexities)
+    return weights
+
+
+def extract_weights_from_df(df, names):
+    series = None
+
+    if config.WEIGHT_COL_NAME is not None:
+        if config.WEIGHT_COL_NAME in df.columns:
+            series = df[config.WEIGHT_COL_NAME]
+        else:
+            config._warn(f"权重列名不存在: {config.WEIGHT_COL_NAME}，尝试使用权重列索引")
+    if series is None and config.WEIGHT_COL_INDEX is not None:
+        if 0 <= config.WEIGHT_COL_INDEX < len(df.columns):
+            series = df.iloc[:, config.WEIGHT_COL_INDEX]
+        else:
+            config._warn(f"权重列索引越界: {config.WEIGHT_COL_INDEX}，将回退到笔画权重")
+
+    if series is None:
+        return {}
+
+    name_series = df.iloc[:, config.DATA_COL_INDEX]
+    numeric = pd.to_numeric(series, errors='coerce')
+    pairs = pd.DataFrame({"name": name_series, "weight": numeric})
+    pairs = pairs[pairs["name"].notna()]
+    pairs["name"] = pairs["name"].astype(str)
+    pairs = pairs[pairs["weight"].notna() & (pairs["weight"] > 0)]
+
+    if pairs.empty:
+        config._warn("Excel 权重列没有可用正数，全部回退到笔画权重")
+        return {}
+
+    if config.REMOVE_DUPLICATES:
+        grouped = pairs.groupby("name", as_index=False)["weight"].max()
+        return dict(zip(grouped["name"], grouped["weight"]))
+
+    valid_name_set = set(names)
+    pairs = pairs[pairs["name"].isin(valid_name_set)]
+    if pairs.empty:
+        config._warn("Excel 权重与名称列未形成有效映射，全部回退到笔画权重")
+        return {}
+
+    grouped = pairs.groupby("name", as_index=False)["weight"].max()
+    return dict(zip(grouped["name"], grouped["weight"]))
+
+
+def calculate_font_by_area_model(mask, names, weights_map, fill_ratio, size_ratio, packing_efficiency, n_rep):
+    free_area = int(np.sum(mask == 0))
+    if free_area <= 0:
+        free_area = int(mask.size)
+
+    effective_fill = fill_ratio if fill_ratio > 0 else max(config.MIN_ACCEPT_FILL_RATIO, 0.82)
+    target_area = free_area * effective_fill * packing_efficiency
+
+    weights = [max(float(weights_map.get(name, 10)), 1.0) for name in names]
+    if not weights:
+        return max(config.MIN_FONT_SIZE, 10), max(config.MIN_FONT_SIZE + 4, 20)
+
+    log_scores = normalize_relative_scores([math.log1p(weight) for weight in weights])
+    char_mass = 0.0
+    for name, score in zip(names, log_scores):
+        length = max(1, len(name))
+        char_mass += length * (0.9 + 0.9 * score)
+
+    char_mass *= max(1, n_rep)
+    if char_mass <= 0:
+        return max(config.MIN_FONT_SIZE, 10), max(config.MIN_FONT_SIZE + 4, 20)
+
+    nominal_size = math.sqrt(target_area / char_mass)
+    min_f = max(config.MIN_FONT_SIZE, int(nominal_size * 0.72))
+    max_f = max(min_f + 1, int(min_f * max(1.4, size_ratio)))
+    return min_f, max_f