@qcplay/cli 1.0.15 → 1.0.16

package/lib/content-rules.js

@@ -14,6 +14,10 @@ const RULE_ACTIONS = new Set([
   "remove_text_block",
   "replace_text",
   "remove_image",
+  "remove_first_image",
+  "remove_last_image",
+  "remove_images_between_text",
+  "remove_first_video",
   "replace_image",
   "remove_image_caption"
 ]);
@@ -168,7 +172,15 @@ function validateAction(action, location) {
   const imageFields = new Set(["src", "alt", "title", "caption", "nearby_text"]);
   const removableImageFields = new Set([...imageFields, "sha256"]);
   const textFields = new Set(["text"]);
-  if (action.type === "remove_image_caption") {
+  if (["remove_first_image", "remove_last_image", "remove_first_video"].includes(action.type)) {
+    if (action.match !== undefined) ruleError(location, `${action.type} 不支持 match`);
+  } else if (action.type === "remove_images_between_text") {
+    validateMatch(action.match, `${location}.match`, textFields);
+    validateMatch(action.end_match, `${location}.end_match`, textFields);
+    if (action.match.text === undefined || action.end_match.text === undefined) {
+      ruleError(location, "remove_images_between_text 必须提供起止 text");
+    }
+  } else if (action.type === "remove_image_caption") {
     validateMatch(action.match || {}, `${location}.match`, imageFields);
   } else if (action.type === "remove_image" || action.type === "replace_image") {
     validateMatch(action.match, `${location}.match`, action.type === "remove_image" ? removableImageFields : imageFields);
@@ -343,7 +355,7 @@ function docxImageHashes(buffer) {
     const commentLength = buffer.readUInt16LE(offset + 32);
     const localHeaderOffset = buffer.readUInt32LE(offset + 42);
     const fileName = buffer.subarray(offset + 46, offset + 46 + fileNameLength).toString("utf8");
-    if (/^word\/media\//i.test(fileName)) {
+    if (/^word\/media\/[^/]+$/i.test(fileName)) {
       if (buffer.readUInt32LE(localHeaderOffset) !== 0x04034b50) throw new Error("DOCX 条目无效");
       const localNameLength = buffer.readUInt16LE(localHeaderOffset + 26);
       const localExtraLength = buffer.readUInt16LE(localHeaderOffset + 28);
@@ -368,10 +380,9 @@ function parseRuleText(text, file, imageHashes = []) {
   const fields = { "规则名称": "name", "项目": "project", "地区": "region", "平台": "platform", "来源": "source", "阶段": "phase" };
   const ignored = /^(?:#|\/\/|说明[：:]?|其他要求[：:]?)\s*/;
   const unsupported = [];
-  let hasImageHashRule = false;
   for (const rawLine of trimmed.split(/\r?\n/)) {
     const line = rawLine.trim().replace(/^(?:[-*•]\s*|\d+[.、]\s*)/, "");
-    if (!line || ignored.test(line) || /(?:样式全部保留|样式.*保留不变|区分.*b站.*非b站|可以有两篇文章|^文章.*规则(?:修改)?$)/i.test(line)) continue;
+    if (!line || ignored.test(line) || /(?:样式全部保留|样式.*保留不变|区分.*b站.*非b站|可以有两篇文章|^文章.*规(?:则)?(?:修改)?$)/i.test(line)) continue;
     if (line === "最强蜗牛") {
       rule.project = line;
       continue;
@@ -412,21 +423,51 @@ function parseRuleText(text, file, imageHashes = []) {
     }
     const imageSection = /^(?:b站的内容要删除如下图片|删除图片)\s*[：:]?\s*$/i.test(line);
     if (imageSection) {
-      if (imageHashes.length > 0 && !hasImageHashRule) {
-        rules.push({
-          name: `${rule.name}-B站图片`,
-          enabled: true,
-          platform: "bilibili",
-          phase: "import",
-          actions: imageHashes.map(sha256 => ({ type: "remove_image", match: { sha256 } }))
-        });
-        hasImageHashRule = true;
-      }
+      rule.actions.push(...imageHashes.map(sha256 => ({ type: "remove_image", match: { sha256 } })));
+      continue;
+    }
+    if (/^删除第一张图片\s*[：:]?\s*$/i.test(line)) {
+      rule.actions.push({ type: "remove_first_image" });
+      continue;
+    }
+    if (/^删除最后一张图片\s*[：:]?\s*$/i.test(line)) {
+      rule.actions.push({ type: "remove_last_image" });
+      continue;
+    }
+    const imagesBetween = line.match(/^删除(?:从)?\s*(?!图片[：:])(.+?)\s*(?:到|至)\s*(.+?)\s*中间的图片$/);
+    if (imagesBetween) {
+      rule.actions.push({
+        type: "remove_images_between_text",
+        match: { text: imagesBetween[1].trim(), operator: "contains" },
+        end_match: { text: imagesBetween[2].trim(), operator: "contains" }
+      });
       continue;
     }
     const image = line.match(/^删除图片[：:]\s*(.+)$/);
     if (image) {
-      rule.actions.push({ type: "remove_image", match: { nearby_text: image[1].trim(), operator: "contains" } });
+      const instruction = image[1].trim();
+      if (/^最后一张图片$/i.test(instruction)) {
+        rule.actions.push({ type: "remove_last_image" });
+        continue;
+      }
+      if (/^第一张图片$/i.test(instruction)) {
+        rule.actions.push({ type: "remove_first_image" });
+        continue;
+      }
+      const nestedImagesBetween = instruction.match(/^删除(?:从)?\s*(.+?)\s*(?:到|至)\s*(.+?)\s*中间的图片$/);
+      if (nestedImagesBetween) {
+        rule.actions.push({
+          type: "remove_images_between_text",
+          match: { text: nestedImagesBetween[1].trim(), operator: "contains" },
+          end_match: { text: nestedImagesBetween[2].trim(), operator: "contains" }
+        });
+        continue;
+      }
+      rule.actions.push({ type: "remove_image", match: { nearby_text: instruction, operator: "contains" } });
+      continue;
+    }
+    if (/^删除视频[：:]\s*第一个视频$/i.test(line)) {
+      rule.actions.push({ type: "remove_first_video" });
       continue;
     }
     unsupported.push(rawLine.trim());
@@ -693,6 +734,22 @@ function replaceTextNodes($, match, replacement) {
   return changed;
 }
 
+function documentOrder($) {
+  const nodes = [];
+  const visit = node => {
+    nodes.push(node);
+    for (const child of node.children || []) visit(child);
+  };
+  for (const node of $.root().contents().toArray()) visit(node);
+  return nodes;
+}
+
+function firstMatchingContentBlock($, match) {
+  return $(BLOCK_SELECTOR)
+    .toArray()
+    .find(node => matchesValue($(node).text().trim(), match.text, match));
+}
+
 function applyAction($, action) {
   if (action.type === "remove_section_until_next_heading") {
     let changed = false;
@@ -713,6 +770,44 @@ function applyAction($, action) {
   if (action.type === "replace_text") {
     return replaceTextNodes($, action.match, action.replacement);
   }
+  if (action.type === "remove_first_image") {
+    const firstImage = $("img").first();
+    if (!firstImage.length) return false;
+    removeImage($, firstImage);
+    return true;
+  }
+  if (action.type === "remove_last_image") {
+    const lastImage = $("img").last();
+    if (!lastImage.length) return false;
+    removeImage($, lastImage);
+    return true;
+  }
+  if (action.type === "remove_first_video") {
+    const firstVideo = $("video,mp-common-videosnap").first();
+    if (!firstVideo.length) return false;
+    const container = firstVideo.closest("p,figure,div,section").first();
+    if (container.length && !container.text().trim() && container.find("video,mp-common-videosnap").length === 1) {
+      container.remove();
+    } else {
+      firstVideo.remove();
+    }
+    return true;
+  }
+  if (action.type === "remove_images_between_text") {
+    const startNode = firstMatchingContentBlock($, action.match);
+    const endNode = firstMatchingContentBlock($, action.end_match);
+    if (!startNode || !endNode) return false;
+    const nodes = documentOrder($);
+    const start = nodes.indexOf(startNode);
+    const end = nodes.indexOf(endNode);
+    if (start < 0 || end < 0 || start >= end) return false;
+    const images = $("img").toArray().filter(node => {
+      const position = nodes.indexOf(node);
+      return position > start && position < end;
+    });
+    images.forEach(node => removeImage($, $(node)));
+    return images.length > 0;
+  }
   const images = $("img").toArray().filter(node => imageMatches($, $(node), action.match || {}));
   if (action.type === "remove_image") {
     images.forEach(node => removeImage($, $(node)));
@@ -756,6 +851,10 @@ export function applyContentRules(article, entry = {}, config = entry.contentRul
     "remove_text_block",
     "replace_text",
     "remove_image",
+    "remove_first_image",
+    "remove_last_image",
+    "remove_images_between_text",
+    "remove_first_video",
     "replace_image",
     "remove_image_caption"
   ]);
@@ -767,6 +866,10 @@ export function applyContentRules(article, entry = {}, config = entry.contentRul
   if (options.phase === "import") {
     const priority = {
       remove_image: 0,
+      remove_first_image: 0,
+      remove_last_image: 0,
+      remove_images_between_text: 0,
+      remove_first_video: 0,
       replace_image: 0,
       remove_image_caption: 0,
       remove_section_until_next_heading: 1,

package/lib/wechat-article.js

@@ -1386,6 +1386,14 @@ export async function convertWechatHtml(html, pageUrl, options = {}) {
       const key = assertWechatImageUrl(rawSource, sourceUrl).toString();
       return uploaded.get(key);
     });
+    const retainedSources = entry.sources.filter(rawSource => {
+      const key = assertWechatImageUrl(rawSource, sourceUrl).toString();
+      return !ignored.has(key);
+    });
+    if (retainedSources.length === 0) {
+      $(entry.element).css("background-image", "none");
+      return;
+    }
     $(entry.element).attr("data-qcplay-background-processed", "1");
     $(entry.element).attr("data-qcplay-background-srcs", urls.filter(Boolean).join("|"));
   });

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@qcplay/cli",
-  "version": "1.0.15",
+  "version": "1.0.16",
   "description": "QCPlay CLI",
   "type": "module",
   "bin": {