scrape!

by noromanba 2015-12-27 [2015/12/27 11:45:28]
(Forked from scrape by noromanba)

download current html snapshot w/ abs path

@@ -1,8 +1,9 @@
 /*
- * @title scrape
- * @description download current html snapshot
+ * @title scrape!
+ * @description download current html snapshot w/ abs path
  * @include http://*
  * @include https://*
+ * @contributor noromanba http://let.hatelabo.jp/noromanba/let/hJmcu63OhsMl (Fork of)
  * @license MIT License http://opensource.org/licenses/MIT
  * @javascript_url
  */
@@ -31,8 +32,28 @@
     .replace(PUNCT,'_').replace(/(?:_){2,}/g, '_') + '.html';
     pool.download = filename;
 
+    const sham = document.cloneNode(true);
+    Array.from(sham.querySelectorAll([
+        '[src^="/"]:not([src^="//"])',
+        '[src^="."]',
+        '[href^="/"]:not([href^="//"])',
+        '[href^="."]'
+    ])).forEach(node => {
+        // dirty-hack: replace rel-path to abs-path
+        // if you need rel-path by DOM, use node.getAttribute('src') c.f.
+        // http://subtech.g.hatena.ne.jp/secondlife/20090624/1245809935
+        if (node.src) {
+            node.src = node.src;
+            return;
+        }
+        if (node.href) {
+            node.href = node.href;
+            return;
+        }
+    });
+
     // get <!DOCTYPE> et al.
-    let root = document.documentElement;
+    let root = sham.documentElement;
     let buff = [];
     while (root.previousSibling) {
         buff.unshift(root.previousSibling);
@@ -44,11 +65,11 @@
     // http://caniuse.com/#feat=xml-serializer
     const snapshot = buff.map(node => {
       return new XMLSerializer().serializeToString(node);
-    }).join('\n') + document.documentElement.outerHTML;
+    }).join('\n') + sham.documentElement.outerHTML;
 
     const snapURL = window.URL.createObjectURL(new Blob([snapshot], {
         // FIXME always save UTF-8
-        type: document.contentType + ';charset=' + document.characterSet
+        type: sham.contentType + ';charset=' + sham.characterSet
     }));
     pool.href = snapURL;
 

• /*
   * @title scrape!
   * @description download current html snapshot w/ abs path
   * @include http://*
   * @include https://*
   * @contributor noromanba http://let.hatelabo.jp/noromanba/let/hJmcu63OhsMl (Fork of)
   * @license MIT License http://opensource.org/licenses/MIT
   * @javascript_url
   */
  
  // via
  // http://blog.mudatobunka.org/entry/2015/12/23/211425
  // thx id:todays_mitsui
  
  // c.f.
  // http://h.hatena.ne.jp/noromanba/81805704781216737
  
  (() => {
      'use strict';
  
      const pool = document.createElement('a');
      // POSIX class [:punct:] not impl yet, alt expanded ASCII c.f.
      // http://www.regular-expressions.info/posixbrackets.html#class
      // [!"#$%&'()*+,\-.:;<=>?@[\\\]^_`{|}~] minus "-._" plus "\s"
      const PUNCT = /[!"#$%&'()*+,:;<=>?@[\\\]^`{|}~\s]/g;
      const filename = decodeURIComponent([
          location.hostname,
          location.pathname.split('/').slice(1).join('_'),
          location.search,
          location.hash
      ].filter(s => !!s).join('_'))
      .replace(PUNCT,'_').replace(/(?:_){2,}/g, '_') + '.html';
      pool.download = filename;
  
      const sham = document.cloneNode(true);
      Array.from(sham.querySelectorAll([
          '[src^="/"]:not([src^="//"])',
          '[src^="."]',
          '[href^="/"]:not([href^="//"])',
          '[href^="."]'
      ])).forEach(node => {
          // dirty-hack: replace rel-path to abs-path
          // if you need rel-path by DOM, use node.getAttribute('src') c.f.
          // http://subtech.g.hatena.ne.jp/secondlife/20090624/1245809935
          if (node.src) {
              node.src = node.src;
              return;
          }
          if (node.href) {
              node.href = node.href;
              return;
          }
      });
  
      // get <!DOCTYPE> et al.
      let root = sham.documentElement;
      let buff = [];
      while (root.previousSibling) {
          buff.unshift(root.previousSibling);
          root = root.previousSibling;
      }
  
      // dump <!DOCTYPE> et al. and <html> tree, old new thing; c.f.
      // https://developer.mozilla.org/en-US/docs/XMLSerializer
      // http://caniuse.com/#feat=xml-serializer
      const snapshot = buff.map(node => {
        return new XMLSerializer().serializeToString(node);
      }).join('\n') + sham.documentElement.outerHTML;
  
      const snapURL = window.URL.createObjectURL(new Blob([snapshot], {
          // FIXME always save UTF-8
          type: sham.contentType + ';charset=' + sham.characterSet
      }));
      pool.href = snapURL;
  
      pool.click();
      window.URL.revokeObjectURL(snapURL);
  })();
  
  // NOTICE
  // XMLSerializer().serializeToString() sometimes too much escape TextNode e.g.
  // <script> 1 && 1; 1 < 1;</script> -> <script> 1 &amp;&amp; 1 &lt; 1 </script>
  
  

• Permalink RAW Packed Userscript LOADER Metadata ?

  Permalink

  このページへの個別リンクです。

  RAW

  書かれたコードへの直接のリンクです。

  Packed

  文字列が圧縮された書かれたコードへのリンクです。

  Userscript

  Greasemonkey 等で利用する場合の .user.js へのリンクです。

  Loader

  @require やソースコードが長い場合に多段ロードする Loader コミのコードへのリンクです。

  Metadata

  コード中にコメントで @xxx と書かれたメタデータの JSON です。