@@ -0,0 +1,341 @@
|
||||
/**
|
||||
* 概念/行业分析 — 数据适配层
|
||||
*
|
||||
* 处理两种扩展数据结构：
|
||||
* - 结构 A（个股维度）：每行一只股票，维度字段（如 concept）存该股票所属的概念/行业
|
||||
* - 结构 B（板块维度）：每行一个概念/行业，带成分股列表（如 constituents: [...]）
|
||||
*
|
||||
* 两种结构统一输出为 DimensionGroup[]，供页面组件消费。
|
||||
*/
|
||||
|
||||
import type { ExtDataConfig, ExtDataField, ExtDataRowsResult } from '@/lib/api'
|
||||
|
||||
// ===== 公共类型 =====
|
||||
|
||||
export interface StockRow {
|
||||
symbol: string
|
||||
code?: string
|
||||
name?: string
|
||||
[key: string]: unknown
|
||||
}
|
||||
|
||||
export interface DimensionGroup {
|
||||
/** 维度名称（概念名/行业名） */
|
||||
key: string
|
||||
/** 成分股数量 */
|
||||
count: number
|
||||
/** 成分股原始行 */
|
||||
stocks: StockRow[]
|
||||
/** 聚合指标（如涨跌幅均值等） */
|
||||
metrics: Record<string, number | null>
|
||||
}
|
||||
|
||||
export interface ResolvedDimension {
|
||||
/** 是否成功解析 */
|
||||
ok: boolean
|
||||
/** 数据结构类型 */
|
||||
structure: 'per_stock' | 'per_dimension' | 'unknown'
|
||||
/** 维度字段名 */
|
||||
dimensionField: string
|
||||
/** 所有解析出的分组 */
|
||||
groups: DimensionGroup[]
|
||||
/** 原始全部行（结构 A 下为原始行，结构 B 下展平后的全部成分股） */
|
||||
allStocks: StockRow[]
|
||||
/** 解析提示 */
|
||||
hint?: string
|
||||
}
|
||||
|
||||
// ===== 结构探测 =====
|
||||
|
||||
const SEPARATORS = /[、,，;；|/\s]+/
|
||||
const CONSTITUENT_KEYS = [
|
||||
'constituents', '成分股', 'stocks', 'members', 'codes', 'list',
|
||||
'symbol_list', 'stock_list', 'member_list',
|
||||
]
|
||||
const DIMENSION_NAME_KEYS = [
|
||||
'name', '概念名称', '概念', '行业名称', '行业', '板块名称', '板块',
|
||||
'concept', 'industry', 'sector', 'theme', 'title', 'label',
|
||||
]
|
||||
|
||||
/** 检测行是否是"板块维度"结构（含成分股列表字段） */
|
||||
function detectConstituentField(fields: ExtDataField[]): string | null {
|
||||
return fields.find(f =>
|
||||
CONSTITUENT_KEYS.some(k => f.name.toLowerCase() === k.toLowerCase() || f.label?.includes(k))
|
||||
)?.name ?? null
|
||||
}
|
||||
|
||||
/** 检测维度名称字段 */
|
||||
function detectDimensionNameField(fields: ExtDataField[]): string | null {
|
||||
return fields.find(f =>
|
||||
DIMENSION_NAME_KEYS.some(k => f.name.toLowerCase() === k.toLowerCase() || f.label?.includes(k))
|
||||
)?.name ?? null
|
||||
}
|
||||
|
||||
/** 从候选名中选取最佳维度字段（结构 A） */
|
||||
export function pickDimensionField(
|
||||
fields: ExtDataField[],
|
||||
candidates: string[],
|
||||
): string {
|
||||
const nonMeta = fields.filter(f =>
|
||||
!['symbol', 'code', 'name', '股票简称', '股票代码', 'date'].includes(f.name)
|
||||
)
|
||||
for (const c of candidates) {
|
||||
const m = nonMeta.find(f =>
|
||||
f.name.toLowerCase().includes(c.toLowerCase()) ||
|
||||
f.label?.toLowerCase().includes(c.toLowerCase())
|
||||
)
|
||||
if (m) return m.name
|
||||
}
|
||||
// 回退：第一个非数值字段
|
||||
return nonMeta.find(f => f.dtype !== 'int' && f.dtype !== 'float')?.name ?? nonMeta[0]?.name ?? ''
|
||||
}
|
||||
|
||||
/** 判断字段是否为数值类型 */
|
||||
function isNumericField(f: ExtDataField): boolean {
|
||||
return f.dtype === 'int' || f.dtype === 'float'
|
||||
}
|
||||
|
||||
// ===== 结构 A 解析：个股维度 =====
|
||||
|
||||
function parsePerStock(
|
||||
rows: Record<string, any>[],
|
||||
dimensionField: string,
|
||||
numericFields: string[],
|
||||
): DimensionGroup[] {
|
||||
const map = new Map<string, StockRow[]>()
|
||||
|
||||
for (const row of rows) {
|
||||
const raw = row[dimensionField]
|
||||
if (raw == null) continue
|
||||
const text = String(raw).trim()
|
||||
if (!text) continue
|
||||
|
||||
// 支持多值分隔（如 "人工智能,芯片,5G"）
|
||||
const values = text.split(SEPARATORS).map(s => s.trim()).filter(Boolean)
|
||||
const stock: StockRow = { ...row, symbol: row.symbol ?? row.code ?? '' }
|
||||
|
||||
for (const v of values) {
|
||||
const list = map.get(v) ?? []
|
||||
list.push(stock)
|
||||
map.set(v, list)
|
||||
}
|
||||
}
|
||||
|
||||
return [...map.entries()]
|
||||
.map(([key, stocks]) => ({
|
||||
key,
|
||||
count: stocks.length,
|
||||
stocks,
|
||||
metrics: computeMetrics(stocks, numericFields),
|
||||
}))
|
||||
.sort((a, b) => b.count - a.count)
|
||||
}
|
||||
|
||||
// ===== 结构 B 解析：板块维度 =====
|
||||
|
||||
function parsePerDimension(
|
||||
rows: Record<string, any>[],
|
||||
constituentField: string,
|
||||
nameField: string,
|
||||
numericFields: string[],
|
||||
): DimensionGroup[] {
|
||||
const allStocks: StockRow[] = []
|
||||
|
||||
const groups = rows.map(row => {
|
||||
const key = String(row[nameField] ?? row[constituentField] ?? '').trim()
|
||||
if (!key) return null
|
||||
|
||||
// 成分股可能是字符串数组、对象数组、逗号分隔字符串
|
||||
const rawList = row[constituentField]
|
||||
const stocks = parseConstituents(rawList)
|
||||
|
||||
stocks.forEach(s => { if (s.symbol) allStocks.push(s) })
|
||||
|
||||
// 维度自身的数值指标也保留
|
||||
const metrics = computeMetrics(stocks, numericFields)
|
||||
// 补上行级别的数值
|
||||
for (const f of numericFields) {
|
||||
if (typeof row[f] === 'number') {
|
||||
metrics[`__dim_${f}`] = row[f]
|
||||
}
|
||||
}
|
||||
|
||||
return { key, count: stocks.length, stocks, metrics } as DimensionGroup
|
||||
}).filter((g): g is DimensionGroup => g !== null && g.key !== '')
|
||||
|
||||
return groups.sort((a, b) => b.count - a.count)
|
||||
}
|
||||
|
||||
/** 解析成分股字段（支持多种格式） */
|
||||
function parseConstituents(raw: unknown): StockRow[] {
|
||||
if (raw == null) return []
|
||||
if (typeof raw === 'string') {
|
||||
// 逗号/分隔符分隔的股票代码字符串
|
||||
return raw.split(SEPARATORS).map(s => s.trim()).filter(Boolean).map(s => ({
|
||||
symbol: normalizeSymbol(s),
|
||||
code: s,
|
||||
}))
|
||||
}
|
||||
if (Array.isArray(raw)) {
|
||||
return raw.map(item => {
|
||||
if (typeof item === 'string') {
|
||||
return { symbol: normalizeSymbol(item), code: item }
|
||||
}
|
||||
if (typeof item === 'object' && item !== null) {
|
||||
const obj = item as Record<string, any>
|
||||
return {
|
||||
symbol: obj.symbol ?? obj.code ?? obj.股票代码 ?? '',
|
||||
code: obj.code ?? obj.symbol ?? '',
|
||||
name: obj.name ?? obj.股票简称 ?? obj.名称 ?? '',
|
||||
...obj,
|
||||
}
|
||||
}
|
||||
return { symbol: String(item) }
|
||||
})
|
||||
}
|
||||
return []
|
||||
}
|
||||
|
||||
function normalizeSymbol(s: string): string {
|
||||
// 尝试补全为 6 位代码
|
||||
if (/^\d{6}$/.test(s)) return s
|
||||
return s
|
||||
}
|
||||
|
||||
// ===== 聚合指标计算 =====
|
||||
|
||||
function computeMetrics(
|
||||
stocks: StockRow[],
|
||||
numericFields: string[],
|
||||
): Record<string, number | null> {
|
||||
const result: Record<string, number | null> = {}
|
||||
for (const f of numericFields) {
|
||||
const vals = stocks
|
||||
.map(s => s[f])
|
||||
.filter((v): v is number => typeof v === 'number' && Number.isFinite(v))
|
||||
if (vals.length === 0) {
|
||||
result[f] = null
|
||||
} else {
|
||||
result[f] = vals.reduce((a, b) => a + b, 0) / vals.length
|
||||
}
|
||||
}
|
||||
return result
|
||||
}
|
||||
|
||||
// ===== 主入口：自动探测 + 解析 =====
|
||||
|
||||
export function resolveDimension(
|
||||
data: ExtDataRowsResult | null | undefined,
|
||||
config: ExtDataConfig | null | undefined,
|
||||
candidateFields: string[],
|
||||
): ResolvedDimension {
|
||||
if (!data || !config || !data.rows.length) {
|
||||
return { ok: false, structure: 'unknown', dimensionField: '', groups: [], allStocks: [] }
|
||||
}
|
||||
|
||||
const fields = data.fields ?? config.fields
|
||||
const rows = data.rows
|
||||
const numericFields = fields.filter(f => isNumericField(f)).map(f => f.name)
|
||||
|
||||
// 先检测是否为结构 B（板块维度）
|
||||
const constituentField = detectConstituentField(fields)
|
||||
if (constituentField) {
|
||||
const nameField = detectDimensionNameField(fields) ?? 'name'
|
||||
const groups = parsePerDimension(rows, constituentField, nameField, numericFields)
|
||||
const allStocks = groups.flatMap(g => g.stocks)
|
||||
return {
|
||||
ok: true,
|
||||
structure: 'per_dimension',
|
||||
dimensionField: nameField,
|
||||
groups,
|
||||
allStocks,
|
||||
hint: `检测到板块维度结构（成分股字段: ${constituentField}）`,
|
||||
}
|
||||
}
|
||||
|
||||
// 结构 A（个股维度）
|
||||
const dimensionField = pickDimensionField(fields, candidateFields)
|
||||
if (!dimensionField) {
|
||||
return {
|
||||
ok: false,
|
||||
structure: 'unknown',
|
||||
dimensionField: '',
|
||||
groups: [],
|
||||
allStocks: rows as StockRow[],
|
||||
hint: '未找到合适的维度字段',
|
||||
}
|
||||
}
|
||||
|
||||
const groups = parsePerStock(rows, dimensionField, numericFields)
|
||||
const allStocks = rows as StockRow[]
|
||||
|
||||
return {
|
||||
ok: true,
|
||||
structure: 'per_stock',
|
||||
dimensionField,
|
||||
groups,
|
||||
allStocks,
|
||||
hint: `按 ${dimensionField} 分组，共 ${groups.length} 个维度`,
|
||||
}
|
||||
}
|
||||
|
||||
// ===== 行情数据关联 =====
|
||||
|
||||
export interface QuoteMap {
|
||||
symbol: string
|
||||
price?: number
|
||||
pct?: number
|
||||
change_pct?: number
|
||||
name?: string
|
||||
[key: string]: unknown
|
||||
}
|
||||
|
||||
/** 构建 symbol → quote 的快速查找 */
|
||||
export function buildQuoteMap(quotes: QuoteMap[]): Map<string, QuoteMap> {
|
||||
const map = new Map<string, QuoteMap>()
|
||||
for (const q of quotes) {
|
||||
if (q.symbol) map.set(q.symbol, q)
|
||||
// 也用纯数字代码做索引
|
||||
const code = q.symbol?.replace(/\.\w+$/, '')
|
||||
if (code) map.set(code, q)
|
||||
}
|
||||
return map
|
||||
}
|
||||
|
||||
/** 为分组计算行情聚合指标 */
|
||||
export function computeQuoteMetrics(
|
||||
stocks: StockRow[],
|
||||
quoteMap: Map<string, QuoteMap>,
|
||||
): {
|
||||
avgPct: number | null
|
||||
upCount: number
|
||||
downCount: number
|
||||
flatCount: number
|
||||
totalVolume: number
|
||||
} {
|
||||
let up = 0, down = 0, flat = 0, totalVol = 0
|
||||
let sumPct = 0, countPct = 0
|
||||
|
||||
for (const s of stocks) {
|
||||
const sym = String(s.symbol ?? '')
|
||||
const q = quoteMap.get(sym) ?? quoteMap.get(sym.replace(/\.\w+$/, ''))
|
||||
if (!q) continue
|
||||
const pct = q.pct ?? q.change_pct
|
||||
if (pct != null && typeof pct === 'number' && Number.isFinite(pct)) {
|
||||
sumPct += pct
|
||||
countPct++
|
||||
if (pct > 0) up++
|
||||
else if (pct < 0) down++
|
||||
else flat++
|
||||
}
|
||||
totalVol += (typeof q.price === 'number' ? 1 : 0) // 简化计数
|
||||
}
|
||||
|
||||
return {
|
||||
avgPct: countPct > 0 ? sumPct / countPct : null,
|
||||
upCount: up,
|
||||
downCount: down,
|
||||
flatCount: flat,
|
||||
totalVolume: totalVol,
|
||||
}
|
||||
}
|
||||
Reference in New Issue
Block a user