关于「精品店手册:单品页、结账页、私讯入口、会员页与四件套蒸馏规格」,《拇指热区与首屏留存工程》给出的定量结论是什么?

结论:1. `reach`:拇指模型的参数,出自第二章 p21。每 CSS 像素对应的毫米数,390x844 为 0.16565、360x780 为 0.15875;拇指半径 S、M、L 为 64、72、80 毫米;转轴在侧边外 6 毫米、底边上 28 毫米;自然区为 0.45 到 0.80 倍半径,勉强区为 0.25 到 0.45 及 0.80 到 1.00 倍;元素内缩 8 像素、按 4 像素网格取最好的区。 2. `viewport`:`reading_max` 0.15(阅读时被钉住的全宽横条占可见高度的上限),`first_visit_max` 0.35(第一次来访、Cookie 层未处理时的上限)。 3. `contrast.dark_text_max`:18.0,出自第四章的工作上限,只用于暗色主题的正文与小字;长文本目标区间 13.0 到 17.0 作为说明写进令牌注释,不作拒绝条件。 4. `fold`:`hero_kb_max` 185(出自第六章 p61,留 0.5 秒余量的首图预算),`cls_budget` 0.1,`audit_viewport` 取较窄的 360x780。 5. `bar`:底部操作栏高 64 像素,底部内边距取 `env(safe-area-inset-bottom)`。

数据口径:全部品牌与数字都是虚构样本,由书中程序算出;涉及隐私同意与无障碍的法规只写一般框架,不是法律意见。 书中以标注框区分证据、计算、原典与假说,每条假说都写明什么结果会推翻它;原典考辨取《考工记》「天有时,地有气,材有美,工有巧」的良工四条件。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.1, 0.15, 0.15875, 0.16565, 0.25, 0.35

caseopverdictcodehitsdetail
GT01reach_zonePASSOK-right natural / left natural
GT02reach_zoneREFUSET1T1right hard / left hard
GT03reach_zonePASSOK-right natural / left natural
GT04reach_zoneREFUSET1T1right stretch / left stretch
GT05reach_zonePASSOK-right hard / left hard
GT06reach_zoneREFUSEE0E0
GT07viewport_budgetPASSOK-first 0.319 reading 0.139 covered 0.000
GT08viewport_budgetPASSOK-first 0.328 reading 0.148 covered 0.000
GT09viewport_budgetREFUSET2T2first 0.411 reading 0.177 covered 0.000
GT10viewport_budgetREFUSET2T2first 0.439 reading 0.146 covered 0.000
GT11viewport_budgetREFUSET3T3first 0.328 reading 0.148 covered 0.197
GT12viewport_budgetREFUSET9T9first 0.328 reading 0.148 covered 0.000
GT13dark_contrast_capPASSOK-ratio 16.92
GT14dark_contrast_capREFUSET4T4ratio 18.43
GT15dark_contrast_capREFUSET4T4ratio 21.00
GT16dark_contrast_capPASSOK-ratio 8.22
GT17dark_contrast_capREFUSES1S1ratio 3.40
GT18dark_contrast_capPASSOK-ratio 15.79
GT19fold_budgetREFUSET5T5,T6,T7,T8clear to y 508; projected CLS 0.6580
GT20fold_budgetPASSOK-clear to y 468; projected CLS 0.0000
GT21fold_budgetREFUSET6T6clear to y 468; projected CLS 0.0000
GT22fold_budgetREFUSET7T7clear to y 468; projected CLS 0.0000
GT23fold_budgetPASSOK-clear to y 468; projected CLS 0.0573
GT24fold_budgetREFUSET8T8clear to y 468; projected CLS 0.5294
GT25fold_budgetREFUSET5T5clear to y 468; projected CLS 0.0000
25 cases; codes triggered: E0 S1 T1 T2 T3 T4 T5 T6 T7 T8 T9
new tokens: mm per px 390x844 0.16565, 360x780 0.15875; radius S/M/L 64.0/72.0/80.0 mm
查看完整验算程序
# Chapter 7. p73: Reference implementation and golden cases for the four style-kit checks this book adds:
# reach_zone (T1), viewport_budget (T2, T3, T9), dark_contrast_cap (T4), fold_budget (T5 to T8).
# Inputs mirror the book's sample (chapter 1 p10, chapter 3 p30, chapter 6 p60 to p62, chapter 7 p70, p72).
# E0 (malformed input), S1 (text below 4.5) and S2 (large text below 3.0) are the kit's existing codes.
# Thresholds live in TOKENS; boxes are (x0, y0, x1, y1) in screen CSS px, the whole screen as the page.
import math

TOKENS = {
    "reach": {"radius_mm": {"S": 64.0, "M": 72.0, "L": 80.0}, "out_mm": 6.0, "up_mm": 28.0,
              "natural": (0.45, 0.80), "stretch": (0.25, 1.00), "inset_px": 8, "grid_px": 4,
              "mm_per_px": {"390x844": 25.4 / 460 * 3, "360x780": 25.4 / 160}},
    "viewport": {"reading_max": 0.15, "first_visit_max": 0.35},
    "contrast": {"floor_text": 4.5, "floor_large": 3.0, "dark_text_max": 18.0},
    "fold": {"hero_kb_max": 185, "cls_budget": 0.1},
}
SIZE = {"390x844": (390, 844), "360x780": (360, 780)}
RANK = {"natural": 0, "stretch": 1, "hard": 2}


def result(hits, info=""):
    return {"verdict": "REFUSE" if hits else "PASS", "code": hits[0] if hits else "OK", "hits": hits,
            "info": info}


def zone_at(x, y, side, vp, hand):
    t = TOKENS["reach"]
    w, h = SIZE[vp]
    mm = t["mm_per_px"][vp]
    px = (w * mm + t["out_mm"]) if side == "right" else -t["out_mm"]
    d = math.hypot(x * mm - px, y * mm - (h * mm - t["up_mm"])) / t["radius_mm"][hand]
    if t["natural"][0] <= d <= t["natural"][1]:
        return "natural"
    return "stretch" if t["stretch"][0] <= d <= t["stretch"][1] else "hard"


def reach_zone(inp):
    vp, box, hand = inp.get("viewport"), inp.get("box"), inp.get("hand", "M")
    if vp not in SIZE or not box or box[2] - box[0] <= 16 or box[3] - box[1] <= 16:
        return result(["E0"])
    t = TOKENS["reach"]
    zones = {}
    for side in ("right", "left"):
        best = "hard"
        for y in range(box[1] + t["inset_px"], box[3] - t["inset_px"] + 1, t["grid_px"]):
            for x in range(box[0] + t["inset_px"], box[2] - t["inset_px"] + 1, t["grid_px"]):
                z = zone_at(x, y, side, vp, hand)
                best = z if RANK[z] < RANK[best] else best
        zones[side] = best
    bad = inp.get("role") == "key" and (zones["right"] != "natural" or zones["left"] != "natural")
    return result(["T1"] if bad else [], "right {} / left {}".format(zones["right"], zones["left"]))


def viewport_budget(inp):
    t, s = TOKENS["viewport"], inp.get("text_scale", 1.0)
    bar = inp["bar"] * s + (inp["inset"] if inp["bar"] else 0)
    first = (inp["top_first"] * s + inp["cookie"] * s + bar) / inp["visible"]
    reading = (inp["top_reading"] * s + bar) / inp["visible"]
    hits = []
    if first > t["first_visit_max"] or reading > t["reading_max"]:
        hits.append("T2")
    k = inp["key"]
    cov = math.fsum(max(0, min(f[2], k[2]) - max(f[0], k[0])) * max(0, min(f[3], k[3]) - max(f[1], k[1]))
                    for f in inp.get("floats", [])) / ((k[2] - k[0]) * (k[3] - k[1]))
    if cov > 0:
        hits.append("T3")
    if inp["bar"] and inp["inset"] and not inp.get("bar_pads_inset", False):
        hits.append("T9")
    return result(hits, "first {:.3f} reading {:.3f} covered {:.3f}".format(first, reading, cov))


def lum(hx):
    c = [int(hx[i:i + 2], 16) / 255 for i in (1, 3, 5)]
    lin = [v / 12.92 if v <= 0.03928 else ((v + 0.055) / 1.055) ** 2.4 for v in c]
    return math.fsum(w * v for w, v in zip((0.2126, 0.7152, 0.0722), lin))


def dark_contrast_cap(inp):
    t = TOKENS["contrast"]
    la, lb = lum(inp["fg"]), lum(inp["bg"])
    r = (max(la, lb) + 0.05) / (min(la, lb) + 0.05)
    hits = []
    if r < (t["floor_text"] if inp["use"] == "text" else t["floor_large"]):
        hits.append("S1" if inp["use"] == "text" else "S2")
    elif inp["theme"] == "dark" and inp["use"] == "text" and r > t["dark_text_max"]:
        hits.append("T4")
    return result(hits, "ratio {:.2f}".format(r))


def fold_budget(inp):
    t = TOKENS["fold"]
    vis, w = inp["visible"], SIZE[inp["viewport"]][0]
    clear = vis - inp["cookie"] - inp["bar"]
    hits = []
    price_ok = inp["price_in_bar"] or inp["price_box"][3] <= clear
    key_ok = inp["key_in_bar"] or inp["key_box"][3] <= clear
    if not (price_ok and key_ok):
        hits.append("T5")
    hero = inp["hero"]
    if hero["kb"] > t["hero_kb_max"]:
        hits.append("T6")
    if hero["loading"] == "lazy" or hero["found_by"] not in ("html", "preload"):
        hits.append("T7")
    late = list(inp.get("late", []))
    if not hero["sized"]:
        late.append((hero["top"], hero["height"]))
    cls = math.fsum(((vis - y) / vis) * (moved / max(w, vis)) for y, moved in late if y < vis)
    if cls > t["cls_budget"]:
        hits.append("T8")
    return result(hits, "clear to y {:.0f}; projected CLS {:.4f}".format(clear, cls))


OPS = {"reach_zone": reach_zone, "viewport_budget": viewport_budget, "dark_contrast_cap": dark_contrast_cap,
       "fold_budget": fold_budget}
VB_A = dict(visible=664, top_first=92, top_reading=92, cookie=120, bar=0, inset=0, key=(16, 432, 374, 484))
VB_B = dict(visible=664, top_first=48, top_reading=0, cookie=72, bar=64, inset=34, bar_pads_inset=True,
            key=(158, 574, 374, 622))
FB_B = dict(viewport="360x780", visible=628, cookie=72, bar=88, price_in_bar=True, key_in_bar=True,
            price_box=(16, 456, 200, 492), key_box=(158, 548, 344, 596),
            hero=dict(kb=96, loading="eager", found_by="html", sized=True, top=48, height=360))
FB_A = dict(viewport="360x780", visible=628, cookie=120, bar=0, price_in_bar=False, key_in_bar=False,
            price_box=(16, 726, 200, 766), key_box=(16, 838, 344, 890),
            hero=dict(kb=420, loading="eager", found_by="script", sized=False, top=124, height=450),
            late=[(0, 36), (124, 20)])
CASES = [
    ("GT01", "reach_zone", dict(viewport="390x844", box=(158, 754, 374, 802), role="key")),
    ("GT02", "reach_zone", dict(viewport="390x844", box=(302, 56, 374, 88), role="key")),
    ("GT03", "reach_zone", dict(viewport="360x780", box=(158, 700, 344, 748), role="key")),
    ("GT04", "reach_zone", dict(viewport="390x844", box=(318, 738, 374, 794), role="key")),
    ("GT05", "reach_zone", dict(viewport="390x844", box=(338, 8, 382, 52), role="tap")),
    ("GT06", "reach_zone", dict(viewport="414x896", box=(158, 754, 374, 802), role="key")),
    ("GT07", "viewport_budget", VB_A),
    ("GT08", "viewport_budget", VB_B),
    ("GT09", "viewport_budget", dict(VB_B, text_scale=1.3)),
    ("GT10", "viewport_budget", dict(VB_A, visible=628, top_first=156)),
    ("GT11", "viewport_budget", dict(VB_B, floats=[(314, 588, 374, 648)])),
    ("GT12", "viewport_budget", dict(VB_B, bar_pads_inset=False)),
    ("GT13", "dark_contrast_cap", dict(fg="#F2F0EA", bg="#0E0E10", use="text", theme="dark")),
    ("GT14", "dark_contrast_cap", dict(fg="#F2F0EA", bg="#000000", use="text", theme="dark")),
    ("GT15", "dark_contrast_cap", dict(fg="#FFFFFF", bg="#000000", use="text", theme="dark")),
    ("GT16", "dark_contrast_cap", dict(fg="#C9A45C", bg="#0E0E10", use="large", theme="dark")),
    ("GT17", "dark_contrast_cap", dict(fg="#7A6438", bg="#0E0E10", use="text", theme="dark")),
    ("GT18", "dark_contrast_cap", dict(fg="#1A1A1F", bg="#F7F4EE", use="text", theme="light")),
    ("GT19", "fold_budget", FB_A),
    ("GT20", "fold_budget", FB_B),
    ("GT21", "fold_budget", dict(FB_B, hero=dict(FB_B["hero"], kb=240))),
    ("GT22", "fold_budget", dict(FB_B, hero=dict(FB_B["hero"], loading="lazy"))),
    ("GT23", "fold_budget", dict(FB_B, late=[(0, 36)])),
    ("GT24", "fold_budget", dict(FB_B, hero=dict(FB_B["hero"], sized=False))),
    ("GT25", "fold_budget", dict(FB_B, price_in_bar=False)),
]
print("case  op                 verdict  code  hits         detail")
seen = set()
for cid, op, inp in CASES:
    out = OPS[op](inp)
    seen |= set(out["hits"])
    print("{}  {:<18} {:<7}  {:<4}  {:<12} {}".format(cid, op, out["verdict"], out["code"],
                                                       ",".join(out["hits"]) or "-", out["info"]))
print("{} cases; codes triggered: {}".format(len(CASES), " ".join(sorted(seen))))
r, v, c, f = TOKENS["reach"], TOKENS["viewport"], TOKENS["contrast"], TOKENS["fold"]
print("new tokens: mm per px 390x844 {:.5f}, 360x780 {:.5f}; radius S/M/L {}/{}/{} mm".format(
    r["mm_per_px"]["390x844"], r["mm_per_px"]["360x780"], *r["radius_mm"].values()))
print("   natural {:.2f}-{:.2f} R; stretch {:.2f}-{:.2f} R; element inset {} px, grid {} px".format(
    *r["natural"], *r["stretch"], r["inset_px"], r["grid_px"]))
print("   reading_max {}, first_visit_max {}, dark_text_max {}, hero_kb_max {}, cls_budget {}".format(
    v["reading_max"], v["first_visit_max"], c["dark_text_max"], f["hero_kb_max"], f["cls_budget"]))
print("carried from chapters 4 and 5 (not checked here): long-text band 13.0-17.0; 'do not tweak' flag")
print("   when a pair's margin over its floor is below 1.15; gold lightness dark >= 0.43 (aim 0.50-0.65),")
print("   light <= 0.36; settle easing (0.2, 0, 0, 1) over 240 ms; wait tiers 100 / 1000 / 10000 ms")

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《移动端拇指热区动力学、暗黑奢品微交互与首屏视觉留存工程》第 7 章「精品店手册:单品页、结账页、私讯入口、会员页与四件套蒸馏规格」(版本 0.1.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-137 · v0.1.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026137,
  author    = {Fan, Yafu},
  title     = {Mobile Thumb-Zone Kinetics, Dark-Luxury Micro-Interactions and Above-the-Fold Visual Retention Engineering},
  year      = {2026},
  publisher = {5D Library},
  edition   = {0.1.0},
  note      = {DOI pending}
}

本书其他章节结论