The logo of the Melammu Project

The Melammu Project

The Heritage of Mesopotamia and the Ancient Near East


  The Melammu Project
  
   General description
   Search string
   Browse by topic
   Search keyword
   Submit entry
  
   About
   Open search
   Thematic search
   Digital Library
   Submit item
  
   Ancient texts
   Dictionaries
   Projects
   Varia
   Submit link
  FAQ
  Contact us
  About

  The Newsletter
  To Project Information >

 

Help for String Searches

General

The string search engine searches the selected fields of all database entries for the search strings entered by the user. The results of the search will be sorted by the number of occurences of the search strings first, after that alphabetically by the headings. The search strings can be either literal strings or Perl regular expressions, but do not use html-code. See below for a full explanation of the way the string search works. If you have any questions or remarks, please do not hesitate contact us.


Multiple search terms

If your search string contains multiple search terms, always seperate them with a space, whether you are using flags or quotation marks or not. If you do not use a space (e.g. =greek+gods instead of =greek +gods), the search engine will treat the non-separated terms as one term and will not produce the results you are looking for.


Flagging search terms

Search strings can be flagged with "+", "-" or "=" to ensure that a string appears among the selected fields of an entry or to exclude entries that contain that string in the selected fields. The most effective use of the search form is for finding correlations among multiple strings.

The "+" sign
The "+" sign placed immediately before a search string will require that string to be present for a valid match. For example, searching for +god +Greeks will return only those entries that contain both "god" and "Greeks" in the selected fields.

The "-" sign
The "-" sign placed immediately before a search string will cause any entry that has that term in the selected fields to be rejected. Prefixing a "-" sign to all the strings in a search will return all the entries that do not contain the search strings. Searching for +god -Greeks will return all entries that contain "god" but do not contain "Greeks" in the selected fields.

The "=" sign
The "=" sign placed immediately before a search string functions as 'or'. The search engine will return entries that contain one of the search strings. So =gods =Greek will return all entries that contain either "gods" or "Greek" or both in the selected fields. It may seem useless to include unflagged search strings in a search that already contains terms flagged with "+", but because the list of results is sorted by the number of times the search strings occur, it is not. This means that +god +Greeks and +god +Greeks =storm will return the same entries, but with the latter, entries containing "storm" in the selected fields will be put on top of the list of results.

Neither "+" nor "-" nor "="
Without flags, the search engine will treat a search string as if it had a "+".


Using quotation marks

Multiword strings must be enclosed in quotation marks or else the search engine will treat them as alternatives to be recognized separately. Searching for Greek god will make the search engine search for both "Greek" and "god", while searching for "Greek god" makes it search for the combination "Greek god".

Multiword strings enclosed in quotation marks can be flagged with the "+", "-" or "=" sign. The flag may come either before or after the opening quotation mark.


Wildcards

"." can be used as a wildcard. This means that searching for gree. with "Match whole words only" turned off will return entries which contain "greek", "greec", greed", etc. in the selected fields. You can use multiple wildcards in a row, but remember that spaces will also count as one character (i.e., gree.. will return "greeks" and "greece", but also "greek " and "greed "). See also below ("Using regular expressions to extend matches").


Ignore case

The default setting for the search is to ignore case. Thus a search string god will find both "god" and "God" (or any other combination of upper- and lowercase letters in this word. This setting can be deselected by clicking on the checkbox in the search parameters. In this event, the search will find only the exact string entered. The default can be restored by clicking on the checkbox again.


Match whole words only

The default setting for the search is to match whole words only. The search string god will find the string as a separate word. This setting can be deselected by clicking on the checkbox in the search parameters. In this event, the search will find the string regardless of where it occurs. A search for god will find "god", "goddess", and "ungodly" (or any other word that contains these three letters in this order). The default can be restored by clicking on the checkbox again.


'Headings' and 'Headings, Summaries and Texts'

Since the fields 'Headins', 'Summaries' and 'Texts' often complement each other, it is pointless to search one of them separately. Therefore, they can only be searched together. However, since the heading of each entry is unique and thus the best means to find back a specific entry, the field 'Headings' is searchable separately as well. This has also been done because of the way the database is compiled, which makes linking to the html-files of specific entries impossible.


Choosing from the lists

The topics, keywords, sources, periods, channels, bibliographical entries and contributors that appear in the database entries have all been indexed. Clicking on "list" will take you to such an index and allows you to enter a search string exactly as it appears in the database.


Using regular expressions to extend matches

Those who are familiar with Perl regular expressions (regex) can use them in searches to extend matches without turning off the "Match whole words only" default. Searching for gree with the "Match whole words only" feature turned off will return all entries that contain "Greece", "Greek" or "Greeks" in the selected fields. However, it will also find "greed", "agree", "agreed", "agreement", "degree", and any other sequence of "gree" that it may encounter. Using the regex gree[ck][es]? with "Match whole words only" turned on will also return all entries that contain "Greece", "Greek" or "Greeks", but without the noise of unwanted sequences of "gree". "." can be used as a wildcard: searching for gree. will match any five-letter word starting with "gree".

Do not place "/" around your search string to indicate that it is a regular expression. The search engine will consider any string entered as a regex whether it contains metacharacters or not.


^
T
O
P