Advanced Search

Using the Advanced Search

When using a list, select from the 'Build list using' options 'Advanced Search.' This will open a new tab in the list window with one text box labeled: 'Querybuilder Predicate Notation.' This is asking for a few lines of code to define search parameters.

Example Code - With Explanations of Results

Searching for Pages

• 'type' defines what sort of object you'll be searching for (it's usually a page).
• 'property' defines what property of the object you'll be filtering by; in this case, its by the title.
• 'property.value' defines your search term.

So this search would be searching all PAGES with the TITLE of PLACES.

So it would find every page titled Places.

*IMPORTANT* Searches by title ARE case-sensitive, so 'Places' is not the same as 'places'.

---

By including 'property.operation=like' and using the '%' wildcard, multiple pages with similar titles can be pulled by the same search.

A wildcard is a character that the program reads as able to stand in for any character in any number (e.g. 'stair%' would get pages named 'stair', 'stairs', and even 'stair case'; and '%stair%' could get a page like 'upstairs', or 'get upstairs').

*REMEMBER* Titles are case-sensitive, but wildcards will search for both cases.

---

Similarly, this search would be searching all PAGES with the TEMPLATE located at /APPS/ODU/TEMPLATES/PLACES.

So it would find every page made with the Places template.

---

By viewing the .5.json information of a page, you can narrow down to almost any page property you want to search by. For example, this search would find every Page that has an advanced search component that is contained in a 'section' component.

In the /adv-search.5.json page, you can see the path to each component, reading like an outline.

If it appears as a continuous line of text, download a json viewer. A good choice for firefox users is: http://jsonview.com/

---

By using 'tagid', you can search by tags (just make sure you're using the full tag), just set the 'tagid' to your desired tag, and set 'tagid.property' to 'jcr:content/cq:tags', which is where the search program will find the tags.

More on the 'path' feature, by including 'path.exact=true' it limits the scope of the search to only that level, not the child directories, meaning it will only search the pages under '' and not, for example, pages under '/elt'. In this case, it lets you find all pages tagged with 'org:acadaffairs/acadenhancement' under '' and didn't pull any of the child pages of the pages it found.

---

You can even search for pages that have certain text on them by using 'fulltext' to define your search. 'Fulltext' searches not just the words on the page, but also for object titles (like images).

In cases like this, and others, it would be good to limit the scope of your search so it won't try searching every page on the site for a string of words. do this by using 'path' to define your search area.

This essentially is checking if a page has the text you're looking for AND is under the defined path.

---

If you want to search under multiple paths, include the line 'group.p.or=true'; it tells the search that the search parameters under 'group' are OR statements. In that, it will be checking if a page has the text you're looking for AND is under path 1 OR 2.

For those who like boolean statements, it's easy to consider it like this:

(fulltext=monarch) AND (path=1_path OR path=2_path)

---

If you want to narrow your search by using two text searches, include the line 'group.p.or=false'; it tells the search that the search parameters under 'group' are AND statements. In that, it will be checking if a page under the given path AND contains text1 AND text2.

For those who like boolean statements, it's easy to consider it like this:

(path=path) AND (fulltext=text1 AND fulltext=text2)

---

Searching for Assets

If you want to narrow your search by excluding a certain path or other search parameter, make a group and use the line 'group.p.not=true', it tells the search that the group will be treated as a 'not' statement. Be sure to include 'group.path.self=true', it tells the search to look in the base node.

---

When searching for assets, use 'type=dam:Asset' and make sure that the path (if used) is pointing to somewhere in the dam.

Defining 'nodename' limits the search to files with the given file extension (eg .pdf .jpg .jpeg .docx .ppt).

---

By including 'property=jcr:content/metadata/dc:title' you can search by title. To search for more than one title use 'property.1_value=title1' and so on for as many titles as needed, otherwise, you can just use 'property.value=title'.

---

By including 'property.operation=like' and using the '%' wildcard, multiple assets with similar titles can be pulled by the same search.

A wildcard is a character that the program reads as able to stand in for any character in any number (e.g. 'stair%' would get assets named 'stair', 'stairs', and even 'stair case'; and '%stair%' could get an asset like 'upstairs', or 'get upstairs').

*REMEMBER* Titles are case-sensitive, but wildcards will search for both cases.

---

By making use of groups, you can search for titles that fulfil multiple requirements. The search on the right will find all documents under the given path that include 'Post" AND '2011' somewhere in the title.

---

Returns all files that begin with 'feature2-lg' and end with '.jpg' in the /content/dam/odu/news/inside/odu/2014/02/27 folder.

This would be useful in creating rules for galleries that would be calling on similarly named images.

'nodename' seems to refer to the actual filename, as opposed to the asset title.