Function processDoc

src/archivist.js:878–920

({documents, strings})

  }

  function processDoc({documents, strings}) {
    /* 
      Info
      Implementation Notes 

      1. Code uses spec at: 
        https://chromedevtools.github.io/devtools-protocol/tot/DOMSnapshot/#type-NodeTreeSnapshot

      2. Note that so far the below will NOT produce text for and therefore we will NOT
      index textarea or input elements. We can access those by using the textValue and
      inputValue array properties of the doc, if we want to implement that.
    */
       
    const texts = [];
    for( const doc of documents) {
      const textIndices = doc.nodes.nodeType.reduce((Indices, type, index) => {
        if ( type === TEXT_NODE ) {
          const parentIndex = doc.nodes.parentIndex[index];
          const forbiddenParent = parentIndex >= 0 && 
            FORBIDDEN_TEXT_PARENT.has(strings[
              doc.nodes.nodeName[
                parentIndex
              ]
            ])
          if ( ! forbiddenParent ) {
            Indices.push(index);
          }
        }
        return Indices;
      }, []);
      textIndices.forEach(index => {
        const stringsIndex = doc.nodes.nodeValue[index];
        if ( stringsIndex >= 0 ) {
          const text = strings[stringsIndex];
          texts.push(text);
        }
      });
    }

    const pageText = texts.filter(t => t.trim()).join(' ');
    DEBUG.verboseSlow && console.log('Page text>>>', pageText);
    return pageText;
  }

  async function isReady() {
    return await untilTrue(() => Status.loaded);