fantasia-archive/src/documents/advancedSearchGuide.md

Advanced search guide

---

Introduction

Fantasia Archive comes with a fairly advanced search-engine present in most of the search fields that look up through either all or at least one type of document - these are for example the multiple and single relationship fields on each document page and the quick-search popup.

---

Intelligent search matching & sorting

The search itself works the following: You can search any amount of words and the software will process them individually as long as they are separated by whitespace.

The search follows these rules

• The search is case-insensitive, which means that you can type everything in UPPER or lower case, it won't matter
• Words can be in any order
  • Example: Dark scary castle will be found even if you type scary castle dark
• Even parts of words will result in successful search
  • Example: Dark scary castle will be found even if you type sca tle ark
• Documents will sort based on the following rules:
  • Direct match has priority over everything else
    • Example: Dark scary castle is a direct match for a search containing dark scary castle
  • Full word match has priority over fragments
    • Each fully matched word counts individually; the more full-matches the document has, the higher it will be in the list
    • Example: Dark scary castle has 2 full word matches from dark scary tle
  • Fragments are at the bottom of the list
    • Each fragment matched counts individually; the more fragments the document has, the higher it will be in the list
    • Example: Dark scary castle has 2 fragment matches from sca tle

---

Filtering

Except for the advanced search functionality, Fantasia Archive also offers instant filtering via multiple attributes for further narrowing search results.

• NOTE: All of the following filter values (including the Full-search filtering in the next section) support matching any part of the search-text with any part of the search-term
  • Example: >nada will match with Continent > North America > Canada > Toronto

The filtering works in the following ways and follows these rules

• Any of the following filter terms will not conflict with the normal word search
• It is possible to only filter by one instance of each of the following filters at once, however, each individual instance may present at the same time
• The filter is case-insensitive, which means that you can type everything in UPPER or lower case, it won't matter
• If your filter-term contained whitespaces, replace them with the - symbol
  • Example: You wish to search for a tag called Player Characters, to fully match this tag, you will need to type #player-characters
• Hierarchical path auto-replaces all > symbols in the path, this result in their omission from the filter string
  • Example: You wish to search for a hierarchical path containing the following USA > Virginia > Richmond, to fully match this hierarchical path, you will need to type >usa-virginia-richmond
• The following filter terms may be used
  • $ - Symbol for document type search
  • # - Symbol for tag search
  • > - Symbol for hierarchical path search
  • ^ - Symbol for switch search (limited values below)
    • ^c - Displays only documents with Is a category ticked on
    • ^d - Displays only documents with Is Dead/Gone/Destroyed ticked on
    • ^f - Displays only documents with Is finished ticked on
    • ^m - Displays only documents with Is a minor document ticked on that are normally invisible and filtered out

Full-search filtering

This feature is meant mostly for those in need of full-scale search that can crawl through any field in any document to match any value field in almost anywhere. Full-search filtering allows the user to narrow down the search marginally by digging through the whole document database and pinpointing exactly what is needed.

A few words of caution

The full-search is a very powerful, but also demanding tool - the more your project will grow, the more demanding it will become. This means that if you for example have 2000+ documents in your project and the search algorithm will have to go crawl through all of them, then the full-search might take a few second to reload your search results - please keep this in mind when using this feature: It can potentially be A LOT of data.

The filtering works in the following ways and follows these rules

• The full-search can be used in combination with any other filters and/or normal search terms
• It is possible to have only a single instance of the full-search present in the search at once
• The filter is case-insensitive, which means that you can type everything in UPPER or lower case, it won't matter
• In the case of lists and multi-relationships, all the entered values get converted to one big text-line for the sake of searching
  • Example with a field called Local currencies:
    • Original values: Canadian Dollar American Dollar Euro Klingon Darsek
    • Converted values: canadian-dollar-american-dollar-euro-klingon-darsek
• The following filter terms must be used inside of the search term
  • % - Symbol for the beginning full-search
  • : - Symbol for the division between the field-name and field value
• If your filter-term contained whitespaces, replace them with the - symbol
  • Example: You wish to search for a field called Local Currencies that contains Canadian Dollars as value, to fully match this tag, you will need to type %local-currencies:canadian-dollars
• It is possible to do a full-text search, checking all fields for the desired text by doing the following: %:canadian-dollars
• A list of fields/field types the full-search doesn't work with:
  • The Break field type (these are the big titles present throughout the document)
  • The Tags field type (this one is covered with a more sophisticated tag filter)
  • The Switch field type (this one doesn't contain any text values to even filter and is partially covered by the switch search option)
  • The Name field (this one is the main concern of the search and the normal search is far more advanced for searching through this one)
  • The Belongs under field (this one is covered by a much more advanced hierarchical path search)