_getArticleTitle()

in Readability.js [573:661]

• 62 lines of code
• 15 McCabe index (conditional complexity)

  _getArticleTitle() {
    var doc = this._doc;
    var curTitle = "";
    var origTitle = "";

    try {
      curTitle = origTitle = doc.title.trim();

      // If they had an element with id "title" in their HTML
      if (typeof curTitle !== "string") {
        curTitle = origTitle = this._getInnerText(
          doc.getElementsByTagName("title")[0]
        );
      }
    } catch (e) {
      /* ignore exceptions setting the title. */
    }

    var titleHadHierarchicalSeparators = false;
    function wordCount(str) {
      return str.split(/\s+/).length;
    }

    // If there's a separator in the title, first remove the final part
    const titleSeparators = /\|\-–—\\\/>»/.source;
    if (new RegExp(`\\s[${titleSeparators}]\\s`).test(curTitle)) {
      titleHadHierarchicalSeparators = /\s[\\\/>»]\s/.test(curTitle);
      let allSeparators = Array.from(
        origTitle.matchAll(new RegExp(`\\s[${titleSeparators}]\\s`, "gi"))
      );
      curTitle = origTitle.substring(0, allSeparators.pop().index);

      // If the resulting title is too short, remove the first part instead:
      if (wordCount(curTitle) < 3) {
        curTitle = origTitle.replace(
          new RegExp(`^[^${titleSeparators}]*[${titleSeparators}]`, "gi"),
          ""
        );
      }
    } else if (curTitle.includes(": ")) {
      // Check if we have an heading containing this exact string, so we
      // could assume it's the full title.
      var headings = this._getAllNodesWithTag(doc, ["h1", "h2"]);
      var trimmedTitle = curTitle.trim();
      var match = this._someNode(headings, function (heading) {
        return heading.textContent.trim() === trimmedTitle;
      });

      // If we don't, let's extract the title out of the original title string.
      if (!match) {
        curTitle = origTitle.substring(origTitle.lastIndexOf(":") + 1);

        // If the title is now too short, try the first colon instead:
        if (wordCount(curTitle) < 3) {
          curTitle = origTitle.substring(origTitle.indexOf(":") + 1);
          // But if we have too many words before the colon there's something weird
          // with the titles and the H tags so let's just use the original title instead
        } else if (wordCount(origTitle.substr(0, origTitle.indexOf(":"))) > 5) {
          curTitle = origTitle;
        }
      }
    } else if (curTitle.length > 150 || curTitle.length < 15) {
      var hOnes = doc.getElementsByTagName("h1");

      if (hOnes.length === 1) {
        curTitle = this._getInnerText(hOnes[0]);
      }
    }

    curTitle = curTitle.trim().replace(this.REGEXPS.normalize, " ");
    // If we now have 4 words or fewer as our title, and either no
    // 'hierarchical' separators (\, /, > or ») were found in the original
    // title or we decreased the number of words by more than 1 word, use
    // the original title.
    var curTitleWordCount = wordCount(curTitle);
    if (
      curTitleWordCount <= 4 &&
      (!titleHadHierarchicalSeparators ||
        curTitleWordCount !=
          wordCount(
            origTitle.replace(new RegExp(`\\s[${titleSeparators}]\\s`, "g"), "")
          ) -
            1)
    ) {
      curTitle = origTitle;
    }

    return curTitle;
  },