Uploaded image for project: 'eZ Publish / Platform'
  1. eZ Publish / Platform
  2. EZP-20523

eZFind: object with title in double quotes and containing words included in stopwords.txt not found

    Details

    • Type: Bug Bug
    • Status: Closed
    • Priority: High High
    • Resolution: Won't Fix
    • Affects Version/s: None
    • Fix Version/s: Customer request
    • Component/s: None
    • Labels:
      None
    • Environment:

      eZ Publish 4.7 + eZ Find 2.7

      Description

      When an object has its title in double quotes (e.g. "unser dorf hat zukunft") AND the title contains words that are included on stopwords.txt (e.g. unser, hat), search returns no results.

      Steps to reproduce:

      1. Create a new article with title "Unser Dorf hat Zukunft" (double quotes included);
      2. Modify stopwords.txt and add two new entries, "unser" and "hat";
      3. Restart SOLR and then re-index it. SOLR has to be restarted, re-indexing alone will not activate modifications;
      4. Results:

      unser dorf hat Zukunft : NOT found (issue)
      "unser dorf hat Zukunft" : found (expected behaviour)
      dorf hat Zukunft : found (expected behaviour)
      unser dorf Zukunft : found (expected behaviour)
      unser : NOT found (expected behaviour, as this is a keyword listed in stopwords.txt)
      hat : NOT found (ditto)

        Activity

        Hide
        Paul Borgermans (Inactive) added a comment -

        I cannot reproduce the problem (using ezfind 2.7): query with [unser dorf hat Zukunft] (whole string without brackets) yields the article
        I also verified the index and query time analysis steps which confirm the expected outcome outlined above

        I detected another side effect though in which stopwords are not always removed: the stopword filter should be moved after the ascii folding step in schema.xml

        Show
        Paul Borgermans (Inactive) added a comment - I cannot reproduce the problem (using ezfind 2.7): query with [unser dorf hat Zukunft] (whole string without brackets) yields the article I also verified the index and query time analysis steps which confirm the expected outcome outlined above I detected another side effect though in which stopwords are not always removed: the stopword filter should be moved after the ascii folding step in schema.xml

          People

          • Assignee:
            Unassigned
            Reporter:
            Nuno Oliveira (Inactive)
          • Votes:
            0 Vote for this issue
            Watchers:
            7 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved: