w-dwdata-hub 1.0.1 → 1.0.2

package/src/fetchRSS.mjs

@@ -1,10 +1,13 @@
 import Parser from 'rss-parser'
+import fetchWebByCurl from 'w-fetch-web/src/fetchWebByCurl.mjs'
 import get from 'lodash-es/get.js'
+import isbol from 'wsemi/src/isbol.mjs'
 import isestr from 'wsemi/src/isestr.mjs'
 import isearr from 'wsemi/src/isearr.mjs'
 import fetchWithRetry from './fetchWithRetry.mjs'
 import getOptFetch from './getOptFetch.mjs'
 import toDatetimeUTC8 from './toDatetimeUTC8.mjs'
+import stripHtml from './stripHtml.mjs'
 
 
 //本函數之預設抓取設定, 因YouTube等站台之RSS會回傳暫時性404故將404納入重試範圍
@@ -19,20 +22,84 @@ let DFLT = {
 }
 
 
+/**
+ * 以curl取回XML
+ *
+ * 委派w-fetch-web之fetchWebByCurl, 失敗時拋出帶statusCode之錯誤(與fetchWithRetry之錯誤形狀一致, 供上層統一處理)。
+ * 注意: fetchWebByCurl之參數名為timeoutMs(非timeout), 且其重試退避為內建之3至15秒線性,
+ * 故opt之baseDelayMs、maxDelayMs與retryStatus於curl模式不生效
+ *
+ * @param {String} rssUrl 輸入RSS網址字串
+ * @param {Object} cfg 輸入已正規化之抓取設定物件
+ * @returns {Promise} 回傳Promise，resolve回傳XML字串
+ */
+async function getXmlByCurl(rssUrl, cfg) {
+
+    let r = await fetchWebByCurl(rssUrl, {
+        timeoutMs: cfg.timeout,
+        maxRetries: cfg.maxRetries,
+    })
+    if (r.status !== 'success') {
+        let err = new Error(r.message || 'curl fetch failed')
+        err.statusCode = r.httpCode
+        err.reason = r.reason
+        err.url = rssUrl
+        throw err
+    }
+
+    return r.html
+}
+
+
+/**
+ * 依method分派取回XML
+ *
+ * @param {String} rssUrl 輸入RSS網址字串
+ * @param {String} method 輸入抓取方式字串，可為'fetch'、'curl'、'auto'
+ * @param {Object} cfg 輸入已正規化之抓取設定物件
+ * @returns {Promise} 回傳Promise，resolve回傳XML字串
+ */
+async function getXml(rssUrl, method, cfg) {
+
+    if (method === 'curl') {
+        return await getXmlByCurl(rssUrl, cfg)
+    }
+
+    if (method === 'auto') {
+        try {
+            return await getXmlByCurl(rssUrl, cfg)
+        }
+        catch (err) {
+            //curl不可用(系統無curl)或該站擋curl時退回fetch
+            return await fetchWithRetry(rssUrl, cfg)
+        }
+    }
+
+    return await fetchWithRetry(rssUrl, cfg)
+}
+
+
 /**
  * 抓取RSS訂閱源
  *
  * 支援RSS與Atom, 並轉為統一格式, 時間欄位一律轉為UTC+8之'YYYY-MM-DD HH:mm:ss'字串;
- * 因YouTube等站台之RSS會回傳暫時性404, 故404亦納入重試範圍
+ * 因YouTube等站台之RSS會回傳暫時性404, 故404亦納入重試範圍(僅fetch模式)。
+ *
+ * 部分站台(如以TLS指紋阻擋非瀏覽器用戶端者)對node內建fetch回403、對curl回200,
+ * 此時可指定opt.method為'curl'改走系統curl, 或'auto'先curl失敗再退回fetch;
+ * curl模式需系統PATH內有curl, 且重試退避由w-fetch-web內建(3至15秒線性),
+ * opt之baseDelayMs、maxDelayMs與retryStatus於curl模式不生效, 極短之feed(小於100字元)會被curl模式判為empty-response
  *
  * @param {String} rssUrl 輸入RSS網址字串，須為http或https
  * @param {Object} [opt={}] 輸入設定物件，預設{}
+ * @param {String} [opt.method='fetch'] 輸入抓取方式字串，可為'fetch'(node內建fetch)、'curl'(系統curl)、'auto'(curl優先失敗退回fetch)，預設'fetch'
+ * @param {Boolean} [opt.withContent=false] 輸入是否於各項目附加content欄布林值，true時依content:encoded、content、summary、description之優先序取feed內嵌全文並去除HTML標籤與實體，預設false
  * @param {Integer} [opt.timeout=30000] 輸入單次請求逾時毫秒整數，預設30000
  * @param {Integer} [opt.maxRetries=5] 輸入最大重試次數整數，含初始共執行maxRetries+1次，預設5
- * @param {Integer} [opt.baseDelayMs=3000] 輸入重試之線性退避基礎毫秒整數，預設3000
- * @param {Integer} [opt.maxDelayMs=15000] 輸入重試之線性退避上限毫秒整數，預設15000
+ * @param {Integer} [opt.baseDelayMs=3000] 輸入重試之線性退避基礎毫秒整數(僅fetch模式生效)，預設3000
+ * @param {Integer} [opt.maxDelayMs=15000] 輸入重試之線性退避上限毫秒整數(僅fetch模式生效)，預設15000
  * @param {Boolean} [opt.showLog=true] 輸入是否顯示重試訊息布林值，預設true
- * @returns {Promise} 回傳Promise，resolve回傳項目物件陣列，各物件為{url,time,title,description,from}，網址無效或抓取失敗時reject回傳錯誤物件
+ * @returns {Promise} 回傳Promise，resolve回傳項目物件陣列，各物件為{url,time,title,description,from}，withContent為true時另含content欄，網址無效或抓取失敗時reject回傳錯誤物件
  * @example
  *
  * import fetchRSS from './src/fetchRSS.mjs'
@@ -43,6 +110,11 @@ let DFLT = {
  *     console.log(rs.length, rs[0])
  *     // => 20 { url: 'https://...', time: '2026-08-07 09:02:03', title: '...', description: '...', from: '...' }
  *
+ *     //擋node fetch之站台改走curl, 並取feed內嵌全文
+ *     let rs2 = await fetchRSS('https://alphaarchitect.com/feed', { method: 'curl', withContent: true })
+ *     console.log(rs2.length, rs2[0].content.length)
+ *     // => 5 2093
+ *
  * }
  * await test()
  *     .catch((err) => {
@@ -57,8 +129,20 @@ async function fetchRSS(rssUrl, opt = {}) {
         throw new Error('rssUrl 須為有效的 http/https RSS 網址')
     }
 
+    //method
+    let method = get(opt, 'method')
+    if (method !== 'curl' && method !== 'auto') {
+        method = 'fetch' //預設維持原行為
+    }
+
+    //withContent
+    let withContent = get(opt, 'withContent')
+    if (!isbol(withContent)) {
+        withContent = false
+    }
+
     //xml
-    let xml = await fetchWithRetry(rssUrl, getOptFetch(opt, DFLT))
+    let xml = await getXml(rssUrl, method, getOptFetch(opt, DFLT))
 
     //feed
     let parser = new Parser()
@@ -74,13 +158,27 @@ async function fetchRSS(rssUrl, opt = {}) {
     }
 
     return items.map((item) => {
-        return {
+
+        let r = {
             url: String(get(item, 'link', '') || '').trim(),
             time: toDatetimeUTC8(get(item, 'isoDate', '') || get(item, 'pubDate', '')),
             title: String(get(item, 'title', '') || '').trim(),
             description: String(get(item, 'contentSnippet', '') || get(item, 'summary', '') || '').trim(),
             from: String(get(item, 'creator', '') || feedFrom).trim(),
         }
+
+        //content, 取feed內嵌全文, 依content:encoded、content、summary、description優先序
+        //content:encoded為RSS content模組之全文欄(如WordPress系), arXiv系全文則在content(源自description)
+        if (withContent) {
+            r.content = stripHtml(
+                get(item, 'content:encoded', '') ||
+                get(item, 'content', '') ||
+                get(item, 'summary', '') ||
+                get(item, 'description', '')
+            )
+        }
+
+        return r
     })
 }
 

package/src/stripHtml.mjs

@@ -0,0 +1,58 @@
+import isstr from 'wsemi/src/isstr.mjs'
+import decodeEntities from './decodeEntities.mjs'
+
+
+/**
+ * 去除HTML標籤與實體, 保留段落換行
+ *
+ * 與htmlToMarkdown不同, 本函數輸出純文字(不產生Markdown標記), 供feed內嵌全文等
+ * 「只要內文、不要格式」之場景使用;
+ * 先移除script與style, 再將段落類閉合標籤與br轉為換行, 最後清除殘留標籤並解碼實體
+ *
+ * @param {String} html 輸入HTML字串
+ * @returns {String} 回傳純文字字串，輸入非字串時回傳''
+ * @example
+ *
+ * import stripHtml from './src/stripHtml.mjs'
+ *
+ * console.log(stripHtml('<p>第一段<b>粗體</b></p><p>第二段&amp;實體</p>'))
+ * // => '第一段粗體\n第二段&實體'
+ *
+ */
+function stripHtml(html) {
+
+    //check
+    if (!isstr(html)) {
+        return ''
+    }
+
+    let s = html
+
+    //移除script與style
+    s = s.replace(/<script[\s\S]*?<\/script>/gi, '')
+    s = s.replace(/<style[\s\S]*?<\/style>/gi, '')
+
+    //段落類閉合標籤與br轉為換行
+    s = s.replace(/<\/(?:p|div|li|h[1-6])>/gi, '\n')
+    s = s.replace(/<br\s*\/?>/gi, '\n')
+
+    //清除殘留標籤
+    s = s.replace(/<[^>]+>/g, '')
+
+    //解碼實體
+    s = decodeEntities(s)
+
+    //decodeEntities把&nbsp;解為U+00A0, 統一為半形空白(必須用顯式跳脫u00A0, 不可貼字面字元)
+    s = s.replace(/\u00A0/g, ' ')
+
+    //行內空白標準化
+    s = s.split('\n').map((l) => l.replace(/[ \t]+/g, ' ').trim()).join('\n')
+
+    //多重空行收斂
+    s = s.replace(/\n{3,}/g, '\n\n')
+
+    return s.trim()
+}
+
+
+export default stripHtml

package/src/toDatetimeUTC8.mjs

@@ -1,6 +1,5 @@
 import ot from 'dayjs'
 import utc from 'dayjs/plugin/utc.js'
-import isnum from 'wsemi/src/isnum.mjs'
 import isestr from 'wsemi/src/isestr.mjs'
 
 
@@ -16,13 +15,22 @@ let TZ_OFFSET_HOUR = 8
 let FORMAT = 'YYYY-MM-DD HH:mm:ss'
 
 
+//無時間部分之日期字串(如'2024'、'20260807'、'2026-08-07'), 即不含'T'與':'者
+let REG_DATE_ONLY = /^[^T:]*$/
+
+
 /**
  * 時間轉UTC+8之日期時間字串
  *
- * 可接受ISO時間字串、RFC822時間字串、unix秒數(非毫秒)或Date物件, 一律輸出台灣時間'YYYY-MM-DD HH:mm:ss';
- * 以dayjs之utc外掛指定固定偏移+8, 不受執行環境時區影響(在UTC伺服器上結果相同)
+ * 可接受ISO時間字串、RFC822時間字串、純日期字串、unix秒數(Number, 非毫秒)或Date物件,
+ * 一律輸出台灣時間'YYYY-MM-DD HH:mm:ss', 且不受執行環境時區影響。
+ *
+ * 兩類輸入之時區語意不同, 故分開處理：
+ * 帶有時刻資訊者(ISO、RFC822、unix秒數、Date)代表一個絕對時點, 換算為UTC+8之當地時刻；
+ * 純日期字串本身不含時點, 直接視為UTC+8當日之00:00:00, 不因執行環境時區而位移日期
+ * (否則在UTC+14環境下'2024'會被解為2023-12-31)。
  *
- * @param {String|Number|Date} value 輸入ISO或RFC822時間字串、unix秒數或Date物件
+ * @param {String|Number|Date} value 輸入ISO或RFC822時間字串、純日期字串、unix秒數或Date物件
  * @returns {String} 回傳UTC+8之日期時間字串，格式為'YYYY-MM-DD HH:mm:ss'，輸入為空或無法解析時回傳''
  * @example
  *
@@ -34,6 +42,9 @@ let FORMAT = 'YYYY-MM-DD HH:mm:ss'
  * console.log(toDatetimeUTC8(1786000000))
  * // => '2026-08-06 15:06:40'
  *
+ * console.log(toDatetimeUTC8('2024'), toDatetimeUTC8('20260807'))
+ * // => '2024-01-01 00:00:00' '2026-08-07 00:00:00'
+ *
  * console.log(toDatetimeUTC8('abc'), toDatetimeUTC8(''), toDatetimeUTC8(null))
  * // => '' '' ''
  *
@@ -42,8 +53,13 @@ function toDatetimeUTC8(value) {
 
     let dt = null
 
-    //數值一律視為unix秒數, dayjs之ot(number)為毫秒故須改用ot.unix
-    if (isnum(value)) {
+    //keepLocalTime, 為true時保留原始年月日時分秒僅改標時區偏移, 供純日期字串使用
+    let keepLocalTime = false
+
+    //Number型別一律視為unix秒數, dayjs之ot(number)為毫秒故須改用ot.unix
+    //型別分派不可用wsemi之isnum, 其對數值字串亦回true, 會使'2024'、'20260807'等日期字串
+    //被誤當unix秒數而解成1970年
+    if (typeof value === 'number') {
         if (value === 0) {
             return ''
         }
@@ -54,6 +70,7 @@ function toDatetimeUTC8(value) {
     }
     else if (isestr(value)) {
         dt = ot(value)
+        keepLocalTime = REG_DATE_ONLY.test(value)
     }
     else {
         return ''
@@ -64,7 +81,7 @@ function toDatetimeUTC8(value) {
         return ''
     }
 
-    return dt.utcOffset(TZ_OFFSET_HOUR).format(FORMAT)
+    return dt.utcOffset(TZ_OFFSET_HOUR, keepLocalTime).format(FORMAT)
 }
 
 

package/test/WDwdataHub.test.mjs

@@ -54,7 +54,7 @@ describe('WDwdataHub', function() {
     })
 
     it('不匯出內部共用函數', function() {
-        let ks = ['fetchWithRetry', 'getOptFetch', 'isYmd', 'toRocDate', 'toDatetimeUTC8', 'decodeEntities', 'htmlToMarkdown', 'extractDivContent', 'parseIntComma']
+        let ks = ['fetchWithRetry', 'getOptFetch', 'isYmd', 'toRocDate', 'toDatetimeUTC8', 'decodeEntities', 'htmlToMarkdown', 'stripHtml', 'extractDivContent', 'parseIntComma']
         let r = ks.map((k) => Object.prototype.hasOwnProperty.call(WDwdataHub, k))
         let rr = ks.map(() => false)
         assert.strict.deepEqual(r, rr)

package/test/fetchRSS.test.mjs

@@ -82,4 +82,82 @@ describe('fetchRSS', function() {
         assert.strict.deepEqual(r, rr)
     })
 
+    it('預設不附content欄, 回傳形狀維持五欄不變', async function() {
+        let t = await fetchRSS(svr.url('/rss/feed-content.xml'), { showLog: false })
+        let r = Object.keys(t[0])
+        let rr = ['url', 'time', 'title', 'description', 'from']
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('withContent為true時附content欄, 取content:encoded去標籤與實體且顯著長於description', async function() {
+        let t = await fetchRSS(svr.url('/rss/feed-content.xml'), { withContent: true, showLog: false })
+        let r = [
+            Object.keys(t[0]),
+            t[0].description,
+            t[0].content,
+            t[0].content.length > t[0].description.length * 4,
+            t[0].content.includes('<'),
+            t[0].content.includes('bad()'),
+        ]
+        let rr = [
+            ['url', 'time', 'title', 'description', 'from', 'content'],
+            '摘要開頭',
+            '第一段完整全文，含粗體與&實體。\n第二段 內容，長度遠大於摘要，供斷言withContent顯著長於description。',
+            true,
+            false,
+            false,
+        ]
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('withContent對無content:encoded之feed(arXiv系)退取content, 全文與description一致', async function() {
+        let t = await fetchRSS(svr.url('/rss/feed-desc.xml'), { withContent: true, showLog: false })
+        let r = [t[0].content, t[0].content === t[0].description]
+        let rr = ['arXiv:2608.01234 Announce Type: new Abstract: 這是一段完整摘要，不含HTML標籤，長度足以代表全文內容本體。', true]
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('method為curl時走系統curl抓取, 解析結果與fetch相同', async function() {
+        let t = await fetchRSS(svr.url('/rss/feed.xml'), { method: 'curl', maxRetries: 0, showLog: false })
+        let r = [t.length, t[0].title, t[0].from, t[1].from]
+        let rr = [2, 'RSS標題一', '作者甲', '測試訂閱源']
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('curl模式對極短feed判empty-response而reject, 證明分派確實走curl路徑', async function() {
+        //同一網址fetch模式可正常解析(見下一測試), 故此reject即為curl路徑之行為證據
+        let r = null
+        await fetchRSS(svr.url('/rss/feed-tiny.xml'), { method: 'curl', maxRetries: 0, showLog: false })
+            .then(() => {
+                r = 'resolved'
+            })
+            .catch((err) => {
+                r = err.reason
+            })
+        let rr = 'empty-response'
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('fetch模式對極短feed正常解析為0項, 與curl模式行為對照', async function() {
+        let r = await fetchRSS(svr.url('/rss/feed-tiny.xml'), { maxRetries: 0, showLog: false })
+        let rr = []
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('auto模式於curl失敗時退回fetch, 極短feed不reject而回0項', async function() {
+        let r = await fetchRSS(svr.url('/rss/feed-tiny.xml'), { method: 'auto', maxRetries: 0, showLog: false })
+        let rr = []
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('method非法值一律退回預設fetch行為', async function() {
+        let r = []
+        for (let method of ['FETCH', 'wget', 123, null, {}]) {
+            let t = await fetchRSS(svr.url('/rss/feed.xml'), { method, showLog: false })
+            r.push(t.length)
+        }
+        let rr = [2, 2, 2, 2, 2]
+        assert.strict.deepEqual(r, rr)
+    })
+
 })

package/test/stripHtml.test.mjs

@@ -0,0 +1,58 @@
+import assert from 'assert'
+import stripHtml from '../src/stripHtml.mjs'
+
+
+describe('stripHtml', function() {
+
+    it('去除標籤輸出純文字, 段落閉合轉為換行', function() {
+        let r = stripHtml('<p>第一段<b>粗體</b></p><p>第二段</p>')
+        let rr = '第一段粗體\n第二段'
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('與htmlToMarkdown不同, 不產生Markdown標記', function() {
+        let r = stripHtml('<h1>標題</h1><p><a href="https://a.example/">連結</a>與<em>斜體</em></p>')
+        let rr = '標題\n連結與斜體'
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('移除script與style', function() {
+        let r = stripHtml('<p>前</p><script>var a=1</script><style>.a{}</style><p>後</p>')
+        let rr = '前\n後'
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('br與li閉合轉為換行, 開標籤不產生換行', function() {
+        let r = [
+            stripHtml('一<br/>二<br>三'),
+            stripHtml('<ul><li>三</li><li>四</li></ul>'),
+        ]
+        let rr = ['一\n二\n三', '三\n四']
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('解碼實體且nbsp統一為半形空白', function() {
+        let t = stripHtml('<p>a&amp;b&nbsp;c&hellip;</p>')
+        let r = [t, t.includes('\u00A0')]
+        let rr = ['a&b c…', false]
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('多重空行收斂且前後修剪', function() {
+        let r = stripHtml('<p></p><p></p><p>內文</p><p></p>')
+        let rr = '內文'
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('非字串回傳空字串', function() {
+        let r = [
+            stripHtml(null),
+            stripHtml(undefined),
+            stripHtml(123),
+            stripHtml({}),
+        ]
+        let rr = ['', '', '', '']
+        assert.strict.deepEqual(r, rr)
+    })
+
+})

package/test/toDatetimeUTC8.test.mjs

@@ -48,6 +48,38 @@ describe('toDatetimeUTC8', function() {
         assert.strict.deepEqual(r, rr)
     })
 
+    it('數值字串為日期字串而非unix秒數, 不得解成1970年', function() {
+        //wsemi之isnum對數值字串亦回true, 若以其做型別分派會使日期字串被誤當unix秒數
+        let r = [
+            toDatetimeUTC8('2024'),
+            toDatetimeUTC8('20260807'),
+            toDatetimeUTC8('2026-08-07'),
+        ]
+        let rr = ['2024-01-01 00:00:00', '2026-08-07 00:00:00', '2026-08-07 00:00:00']
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('純日期字串視為UTC+8當日00:00:00, 不受執行環境時區位移日期', function() {
+        //在UTC+14環境下若以當地時區解析, '2024'會位移為2023-12-31而跳年
+        let r = [
+            toDatetimeUTC8('2024'),
+            toDatetimeUTC8('20260807'),
+            toDatetimeUTC8('2026/08/07'),
+        ]
+        let rr = ['2024-01-01 00:00:00', '2026-08-07 00:00:00', '2026-08-07 00:00:00']
+        assert.strict.deepEqual(r, rr)
+    })
+
+    it('帶時刻資訊者仍為絕對時點換算, 不套用保留當地時間', function() {
+        let r = [
+            toDatetimeUTC8('2026-08-07T01:02:03Z'),
+            toDatetimeUTC8('Fri, 07 Aug 2026 01:02:03 GMT'),
+            toDatetimeUTC8(1786000000),
+        ]
+        let rr = ['2026-08-07 09:02:03', '2026-08-07 09:02:03', '2026-08-06 15:06:40']
+        assert.strict.deepEqual(r, rr)
+    })
+
     it('無法解析或為空時回傳空字串', function() {
         let r = [
             toDatetimeUTC8('not-a-date'),

package/test/tools/fixtures.mjs

@@ -40,6 +40,46 @@ let rssXml = `<?xml version="1.0" encoding="UTF-8"?>
 </rss>`
 
 
+//含content:encoded內嵌全文之feed(WordPress系), 供斷言withContent取全文且顯著長於description;
+//全文內含粗體、實體、&nbsp;與script, 供斷言stripHtml各轉換
+let rssContentXml = `<?xml version="1.0" encoding="UTF-8"?>
+<rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/">
+<channel>
+<title>內容測試源</title>
+<link>https://rss.example/</link>
+<item>
+<title>含全文項目</title>
+<link>https://rss.example/c1</link>
+<pubDate>Fri, 07 Aug 2026 01:02:03 GMT</pubDate>
+<description>摘要開頭</description>
+<content:encoded><![CDATA[<p>第一段完整全文，含<b>粗體</b>與&amp;實體。</p><p>第二段&nbsp;內容，長度遠大於摘要，供斷言withContent顯著長於description。</p><script>bad()</script>]]></content:encoded>
+</item>
+</channel>
+</rss>`
+
+
+//全文位於description之feed(arXiv系), rss-parser將其映為item.content,
+//供斷言withContent之content fallback(無content:encoded時取content)
+let rssDescOnlyXml = `<?xml version="1.0" encoding="UTF-8"?>
+<rss version="2.0">
+<channel>
+<title>arXiv式測試源</title>
+<link>https://rss.example/</link>
+<item>
+<title>論文一</title>
+<link>https://rss.example/p1</link>
+<pubDate>Fri, 07 Aug 2026 01:02:03 GMT</pubDate>
+<description>arXiv:2608.01234 Announce Type: new Abstract: 這是一段完整摘要，不含HTML標籤，長度足以代表全文內容本體。</description>
+</item>
+</channel>
+</rss>`
+
+
+//合法但小於100字元之feed, 低於fetchWebByCurl之MIN_HTML_LENGTH,
+//供斷言curl模式判empty-response、auto模式退回fetch
+let rssTinyXml = '<rss version="2.0"><channel><title>t</title></channel></rss>'
+
+
 //===== Hacker News =====
 
 
@@ -494,6 +534,9 @@ let guanchaArticleHtml = `<!DOCTYPE html><html><head><title>OPEC+宣布增产_
 export {
     aiNews,
     rssXml,
+    rssContentXml,
+    rssDescOnlyXml,
+    rssTinyXml,
     hnNewStories,
     hnItems,
     cnyesPage1,

package/test/tools/serverForTest.mjs

@@ -126,6 +126,18 @@ function serverForTest() {
                 sendText(200, fx.rssXml, 'application/rss+xml; charset=utf-8')
                 return
             }
+            if (pathname === '/rss/feed-content.xml') {
+                sendText(200, fx.rssContentXml, 'application/rss+xml; charset=utf-8')
+                return
+            }
+            if (pathname === '/rss/feed-desc.xml') {
+                sendText(200, fx.rssDescOnlyXml, 'application/rss+xml; charset=utf-8')
+                return
+            }
+            if (pathname === '/rss/feed-tiny.xml') {
+                sendText(200, fx.rssTinyXml, 'application/rss+xml; charset=utf-8')
+                return
+            }
 
             //===== Hacker News =====
 

package/toolg/gDistRollup.mjs

@@ -12,6 +12,7 @@ rollupFiles({
     fdTar,
     nameDistType: 'kebabCase', //直接由hookNameDist給予
     globals: {
+        'child_process': 'child_process',
         'util': 'util',
         'path': 'path',
         'module': 'module',
@@ -24,11 +25,11 @@ rollupFiles({
         'cheerio': 'cheerio',
         'rss-parser': 'rss-parser',
         'playwright': 'playwright',
-        'w-fetch-web': 'w-fetch-web',
         '@mozilla/readability': '@mozilla/readability',
         'jsdom': 'jsdom',
     },
     external: [
+        'child_process',
         'util',
         'path',
         'module',
@@ -41,7 +42,6 @@ rollupFiles({
         'cheerio',
         'rss-parser',
         'playwright',
-        'w-fetch-web',
         '@mozilla/readability',
         'jsdom',
     ],