Note: A fully printable copy of the Talis Version 10.0 edition of this chapter is available in Adobe Acrobat™ (.PDF) format. Click here to view the Version 10.0 edition.
An Author search permits searching for works by or about an Author, Corporate Author or Conference, as well as Editors.
The Author Search Prompt comprises an information window explaining the principles of searching by Author and a prompt bar for user input. Refer to Figure 2.1: Author Search Prompt
Type in the Author's name as fully as possible and press <Enter> to commence the search. Examples:
KENNEDY, JOHN <Enter> GREENE, G <Enter> ROYAL AERONAUTICAL SOCIETY <Enter>
Enter the Author's Surname first, followed by Forenames or Initials. Insert a comma immediately after the Surname. (The space after the comma is optional).
An Author may be either a person or a corporate body such as an organisation or a Government body. Omit initial articles (such as "The") from a corporate body.
To search for a Corporate body (i.e. an organisation or Government department) as the author or issuer of a work, enter the name as fully as possible. It is important to separate subordinate Corporate bodies with a ",", for example:
GREAT BRITAIN, ARMY, MEDICAL CORPS <Enter>
Search restrictions may be useful in order to avoid retrieving too many references in response to a search. The <F10> ("Limit") function is available both at the point of input (pre-restriction) and at the Author Index (post restriction). For further information on restricting a search, refer to Search Restriction.
If the search results in a match, the user will be positioned against that Author in the Author Index. If back browsing is enabled the match will be the second line on the screen, otherwise this will be the top line on screen. Refer to Figure 2.2:The Author Index.
The Author Index always displays, even for an exact match, as this allows the user to see if there are variant forms of the name, or any cross-references.
A count of the number of matching Authors displays at the top right of the screen. The number of references displays in the index table against each Author. This reference count may be incorrect in some circumstances, for example where the Author occurs twice in a Work, or where a Work by an Author has been recently deleted.
If the search does not achieve a match, a dialog box displays, giving the user the option of viewing the closest match in the Author Index or returning to the Author Prompt to refine their search.
"See" and "See also" references appearing in the Author Index will display with the references column blank.
Cross-references can be navigated by pressing the <Enter> key having highlighted the Author. These references will "stack", allowing the user to navigate backwards through the stack by pressing <F12> ("Back"). See Search Stacking.
Any public notes (entered in tag 680 in Authority Control) attached to an Authority cross-reference can be displayed by pressing the <F7> ("Details") key when the highlight bar is against that reference.
Selecting an Author from the Author Index will result in the display of a list of references for that Author. Refer to Figure 2.3: The Author References List. This list will default to Title order, unless an alternative sort order has been specified.
For more information about the Author References List and its composition including sort variables, refer to The References List.
Entries for the Author Index are built from the AUTHOR table. The AUTHOR table is built from MARC tags and subfields, as specified by the list of AUTHOR_TAGS. Weightings in this table govern the order in which Authors are displayed. The AUTHOR_TAG_RULE table specifies rules for concatenating groups of tags. A tag and subfield must appear in both tables in order to be indexed correctly in OPAC. Refer to Chapter 9: Access Points Rules for further details.
Note: It is necessary to run "auto_access_points" to build OPAC entries following any changes to these tables. Inconsistencies in indexing will result if this is not done
Subordinate bodies have entries in the AUTHOR table built under each form of the name. For example:
110 0 10 Great Britain*cParliament*cHouse of Commons *cSocial services committee
An Author entry is built for the following Authors linked to this Work.
The "authority_ap" script is used to build cross-references ("See" and "See Also") into the Author Index in OPAC. It should always be run after "auto_access_points" in order to re-generate the Author cross-references. It can also be run after "update_daily_access_points" to keep the index updated with changes made in Authority Control.
To run this script:
cd /usr/opt/blcmp/talis/access_points <Enter>
authority_ap
Alternatively, a "cron" line can be set up to run it everyday, though be careful to allow sufficient time for "update_daily_access_points" to run.
The "authority_ap" script should take tens of minutes to run, rather than hours. The time taken depends on the number of cross-references set up. Refer to Authority Access Points.
All search terms are converted into uppercase. Certain punctuation is removed, and search weightings added, and the resulting string is then matched against a filing value in the AUTHOR table.
Prefixes are converted to a standard format:
|
User Input |
After Automatic Conversion |
|
MC |
MAC |
|
M' |
MAC |
|
ST |
SAINT |
|
ST. |
SAINT |
|
STE |
SAINTE |
|
STE. |
SAINTE |
Table 2.1: Normalisation of Prefixes
A Title search allows you to search the current Library Catalogue for all records with a specific Title (or partial Title). Any more fields can be designated as the Title if the Library chooses.
The Title Search Prompt comprises an information window explaining the principles of searching by a known Title plus a prompt bar for user input of the search term. Refer to Figure 2.4: Title Search Prompt.
The Title Search Prompt requires input of the first few letters or words of the Title, or the whole Title. Press <Enter> to commence the search. For example:
BLEAK HOUSE <Enter>
You may use upper or lower case characters. You may leave out initial articles, such as "A", "An" and "The" if you wish. The system will strip initial articles from your Title automatically before searching the Title index.
Note: This may also apply to foreign Titles; provided the appropriate initial articles for "Le", "Das" etc. have been defined from within Online Utilities, Access Points Rules. (Refer to Chapter 9: Access Points Rules.
Users are required to enter a minimum of 2 letters of the Title in order to prevent any performance problems. Libraries are able to specify the minimum number of letters required through setting the environment variable TAL_MIN_TITLE=x, where x is the minimum number of letters required. For example:
TAL_MIN_TITLE=3;export TAL_MIN_TITLE
Refer to Chapter 11: OPAC Related Environment Variables.
Search restrictions may be useful in order to avoid retrieving too many references in response to a search. There is no <F10> ("Limit") function available from the Title Index, but an option for post-restriction may be selected from the Options Menu (obtained by pressing <F10> ("Options"). For further information on restricting a search, refer to Search Restriction.
Note: The <F10> ("Options") menu is configurable using the Configurable Menus interface under Utilities, Parameters, Menus. See Variable Menus & Collection Lists in this manual, or System Manager Manual, Volume 2: Application Maintenance, Chapter 42: Interface Configuration for full details.
If the search results in a match, the Title Index displays, with the highlight bar positioned on the closest match. Refer to Figure 2.5: The Title Index (Browsable). If back browsing is enabled the closest match will be the highlighted second line on the screen, otherwise this will be the top line on the screen.
The Title Index always displays, even for an exact match, as this allows the user to see if there are variant forms of the Title.
A count of the number of matching Titles displays at the top right of the screen, with the original search term displayed at the top left.
Entries in the Title Index are sorted alphabetically on the full terms (there is no concept of acronyms in Talis Title searching). Where more than one Title is identical, these will be sorted by date in descending order, (i.e. the most recently published Titles will display first). Any records having no date will display before those that have a date (so that Serial records will file first). The date used is derived from the 008 MARC tag, or if there is no 008, the 260*c will be used.
Note: This may lead to discrepancies: a Title may appear to file out of order if the date in the 008 differs from that in 260*c, as it will be ordered by the 008 value, but the 260*c displays on the reference line.
If the search does not achieve a match, a dialog box will display, giving the user the option of viewing the closest match or returning to the Title Prompt to refine the search.
If a search for a Title containing an Initial Article (for example "The cat sat on the mat") fails to find a match and the user elects to view the references, the initial articles will be replaced and the search carried out.
Note: The Title shown on this line will be from the Title Index; it will not necessarily display the Uniform Title.
Selecting a Title from the Title Index results in the display of a list of items at different locations i.e. the Locations Screen, or if there are no items available for this Title a message box displays saying:
No copies are available for this title. Press any key to return to the list.
Entries for the Title Index are built from the TITLE table. This TITLE table is built from MARC tags and subfields, as specified by the list of TITLE_TAGS. Each entry in the Title References List will be built from the contents of a single MARC tag. It is not currently possible to concatenate tags in any way.
Although it is possible to specify that different tags be used for different types of material (for example, Analytical or Monograph), note that a Serial "parent" will be treated as a Monograph. It is also possible to specify whether entries should be built from Local or Non-Local tags (or both). A maintenance interface (in Online Utilities) is provided to allow users to amend this list. Further information is provided in Chapter 9: Access Points Rules.
Note: It is necessary to run "auto_access_points" to build OPAC entries following any amendments to this table in order for changes to be reflected in the Title Index.
All search terms are converted into uppercase, and certain punctuation and superfluous spaces removed.
Any words at the start of the Title string that match entries in the Initial Articles list(s) are removed (for example, "A", "An" and "The"). Refer to Chapter 9: Access Points Rules.
The English language initial articles will be picked up by default, unless the user is limiting their search to another language. For example, if the user is searching for "Die Welt" and limiting their search by German language then the German initial articles list will be used.
The resulting string is then matched against a filing value in the TITLE table.
A Keyword search is used to search the current Library Catalogue for records containing a particular keyword, or combination of keywords. (In practice other subject-related fields may be searched too. Keywords can be built from any MARC tag the Library chooses).
The Keyword Search Prompt consists of an information window briefly outlining the principles of searching by keyword(s) and a prompt bar for user input. Refer to Figure 2.6: Keyword Search Prompt.
Enter the most specific keyword(s) which describe your subject and which are likely to appear in the library records for relevant works. You may type keywords in upper or lower case. Keywords should be separated with a space character.
Commonly occurring articles and prepositions such as "The", "A", "An", "From" etc. will be stripped from your input automatically to improve the search. Common English plurals will be searched for automatically too.
It is advisable not to search for very common words as these will index large numbers of records which will probably not be relevant. Generally, the more specific the search terms are to your subject, the more relevant the retrieved records are likely to be.
When you are ready to commence your search, press the <Enter> key.
Search restrictions may be useful in order to avoid retrieving too many references in response to a search. The <F10> ("Limit") function will be displayed to the user on the Keyword Prompt Screen if Limit is available. The user must type in the term they wish to limit before pressing <F10> ("Limit") to select the appropriate limiters (search pre-restriction).
The <F10> ("Limit") key is also available from the Keyword Index (post-restriction) if the user chooses to limit their search at this point. For further information on restricting a search, refer to Search Restriction.
Search limiters are also available as an expert search option,; refer to Expert Searching if further instructions are required.
Keyword searching permits Boolean AND, OR, single and multiple wildcard searching (including embedded wildcards) and automatic stemming.
The Boolean Operators "AND" and "OR" can be used to combine search terms.
AND will look for references that contain all of the terms (two or more) entered. The AND operator is used to narrow a search to those records which include all of the terms combined thereby. For example:
WOMEN AND ALCOHOL <Enter>
will retrieve references containing the word "women" as well as the word "alcohol". Terms separated by AND must both appear in the same record to be retrieved.
OR requests the system to look for catalogue records that contain any of the terms (two or more) entered. The OR operator is used to combine potentially useful terms together. For example:
FOOTBALL OR SOCCER <Enter>
will retrieve references that contain either the word "football" or the term "soccer" or both. Terms combined by OR do not have to be in the same record to be retrieved: they can appear in either or both.
If two or more keywords are entered without an explicit AND or OR to combine them, they will be combined in an implicit AND relationship automatically.
If no references are found as a result of an implicit AND, the user will be given the option of re-doing the search with the terms OR-ed together..
The truncation symbol "*" is used to search for all variant word endings on a common word stem. For example:
CHEM* <Enter>
will retrieve references containing keywords beginning with these letters (the stem), followed by any other letters (such as chemists, chemistry, chemicals etc.).
As little as three characters may be input as a word stem, but users should be wary of over general truncation resulting in large numbers of irrelevant references. "COMP*" for example, would yield records about computers, companies, comprehension and so on.
Truncation characters can be placed anywhere in the search term. For example:
PSYCH*Y <Enter>
will find references about psychology, psychiatry, etc.
While "*" masks for a variable number of characters, "?" masks, or substitutes, for a single character only. For example:
WOM?N <Enter>
will retrieve singular and plural woman/women.
AUTHORI?ATION
will retrieve English and American spellings of authorisation.
Note: A single search cannot include wildcard characters (i.e. "*" or "?") and multiple keywords. For example, "librar*" is an acceptable search, but "librar* information" is not.
If the search results in a match, the user will be positioned on that keyword in the Keyword Index. Refer to Figure 2.7: The Keyword Index. If back browsing is enabled the closest match will be the second line on the screen, otherwise this will be the top line on screen.
Note: Keyword searching uses implicit stemming, so a search on "lib" will automatically place the user on the nearest match to that term, which might be "librarian".
The Keyword Index is always displayed, even for an exact match, as this allows the user to see if there are variant forms of the word.
A count of the number of matching terms displays at the top right of the screen. The number of references displays against each keyword. This reference count may be incorrect in some circumstances, for example where the word occurs twice in a Work, or where a Work containing the word has been recently deleted.
If the search does not achieve a match, a dialog box displays, giving the user the option of viewing the closest match in the Keyword Index, or returning to the Keyword Prompt to refine the search.
Selecting a keyword from the Keyword List results in the display of a list of references for this keyword. This list will default to date order if there are fewer than 18 references, or will not be ordered if there are more than 18 references unless a sort order has been specified. (See The References List).
If the search results in a match, the user will be placed in a list of keyword terms that match the search. The user can select from the list to view the references. It is necessary to enter at least three characters before the wildcard character for it to work. If fewer characters are given, a message will be displayed instructing the user to enter at least three characters before the wildcard.
The user can enter more than one word, in which case the words will be AND-ed together implicitly. If the search results in a match, the user will be positioned in the list of matching references. Refer to Figure 2.8: Keyword References List (Multiple Keywords).
If the search uses Boolean AND (for example, "systems analysis", or "systems AND analysis") and fails to find any matching references, a dialog box displays giving the user the option of expanding the search to Works that include any of the search terms (i.e. Boolean OR), or returning to the prompt to amend their search.
If the search fails to find any hits for any of the search terms entered, a dialog box displays informing the user that:
There are no works indexed by all of the words entered. Do you want to see a list of works indexed by some of the words entered?
Selecting "Yes" will take the user to a list of references indexed by any of the search terms. Selecting "No" will return the user to the search prompt.
Entries in the Keyword Index are built from the KEYWORD table in the database. The KEYWORD table is built from MARC tags and subfields, as specified by the list of KEYWORD_TAGS. A maintenance interface (in Online Utilities) is provided to allow users to amend this list. Refer to Chapter 9: Access Points Rules for further details. Tags can be built from local or non-local tags, or both.
Note: It is necessary to run "auto_access_points" to build OPAC entries following any changes to this table.
All search terms are converted into uppercase and extraneous spacing is removed. Certain punctuation is removed. Checks are made for Boolean operators (AND and OR). Checks are made for keyword stopwords. Plurals are rationalised (see below).
The resulting string is matched against a filing value in the KEYWORD table.
Boolean and wildcard searching are disabled by default. To enable them it is necessary to set up the relevant OPAC General Rules parameters under Online Utilities (accessed by selecting the menu options "Utilities", "Parameters", "Rules", "OPAC" followed by "General Rules". Refer to Chapter 8: OPAC Rules, for further details.
Rationalisation of plurals works as follows. First the word is checked against a list of plural stopwords (words that should not be rationalised) as defined in Online Utilities. This list is set up under Utilities, accessed by selecting the menu options "Utilities", "Parameters", "Rules", "Access Points" followed by "Plurals". Refer to Chapter 9: Access Points Rules, for further details.
If the word occurs in this list then the system will search for it as entered. If it does not occur in the list then it will be treated as follows:
Every keyword term is checked against a list of Stopwords. Any words occurring in this list will be removed. This list is set up under Utilities, accessed by selecting the menu options "Utilities", "Parameters", "Rules", "Access Points" followed by "Stopwords". Refer to Chapter 9: Access Points Rules, for further instructions.
The English language stopwords will be picked up by default, unless the user is limiting the search to another language. For example, if entering a search with the French language limiter, the French stopwords will be used.
Searches for words containing hyphens have the hyphen removed and the words closed-up. Where a term has been entered with a hyphen, each term is indexed separately, in addition to being indexed as a single term (with the hyphen removed) so that such terms will be retrieved irrespective of whether the user enters them as separate words, hyphenated or as a single word.
An Author/Title search allows the current Library Catalogue to be searched for either all of the records by, or about, a specific Author or all records with a specific Title (or partial Title) as well as for a specific Author and Title combination.
The Author/Title Search Prompt includes an information window explaining the principles of searching by known Author and/or Title with a double prompt bar for input of the search term(s). Refer to Figure 2.9: Author/Title Search Prompt.
If searching just for an Author, the search is exactly the same as described in Author Searching. If searching solely for a Title, the search is exactly the same as explained in Title Searching. This chapter describes how the combined search operates.
If the Author is not known, or if you are interested in all Works with a certain Title, leave the Author input field blank and press <Enter> to move down to the Title prompt.
Similarly, if the Title is not known, or if you are interested in all Works by a certain Author, leave the Title input field blank and press <Enter> to commence the search.
If both Author and Title are known, the user enters an Author name and the first few words of the title and presses <Enter> to commence the search.
The search performed will be truncated; for example, a search for Author = "Shake", Title = "King" will search for all Authors beginning with "Shake*" and all Titles beginning with "King*".
Search restrictions may be useful in order to avoid retrieving too many references in response to a search. The <F10> ("Limit") function key is available to users at the point of search input (pre-restriction).
The <F10> ("Limit") key or <F10> ("Options") menu is also available from the List of References, the Author Index or the Title Index depending on search results (post-restriction). For further information on restricting a search, refer to Search Restriction.
Note: The <F10> ("Options") menu is configurable using the Configurable Menus interface under Utilities, Parameters, Menus. See See Variable Menus & Collection Lists in this manual, or System Manager Manual, Volume 2: Application Maintenance, Chapter 42: Interface Configuration for full details.
If the search finds a match, the list of matching references will display. Refer to Figure 2.10: The Author/Title References List.
This list will be in random order unless a sort order has been specified. (Refer to Sort Variables). For more information about the References List and its composition, including sort variables, refer to The References List.
A count of the number of matching Titles will display at the top right of the screen, with the original search terms displayed at the top left. The user's current position in the list is also displayed.
If the combined Author/Title search fails to find matching records in the catalogue, but there is a matching Author, the user is given the option of viewing the Author Index or returning to the prompt.
If the combined Author/Title search fails to find matching records in the catalogue, as there are no matching Authors but there are matching Titles, the user is given the option of viewing the Title Index or returning to the prompt to refine the search.
Should the combined search fail to locate any matching Authors or Titles, the user is prompted to perform a Keyword search.
The search uses the same indexes as are used for the Author and Title searches. There is no separate author/title index. The way these indexes are built is detailed within their respective chapters.
The treatment of search terms is the same as for their respective searches.
An Author/Keyword search allows the current Library Catalogue to be searched for either all records by or about a specific Author or all of the records containing a particular keyword (or a combination of keywords) as well as for a specific Author or Keyword combination. The principles of the Author component of a search are explained in Author Searching. The principles of the Keyword component of a search are explained above in Keyword Searching.
The Author/Keyword Search Prompt contains an information window explaining the principles of searching by Author and/or Keyword with a double prompt bar for input of search term(s). Refer to Figure 2.11: Author/Keyword Search Prompt.
If the Author is not known, or if you are interested in all works containing a certain keyword (or combination of keywords), leave the Author input field blank and press <Enter> to move down to the Keyword prompt.
If one or more keyword(s) are not known, or if you are interested in all Work record containing a specific Author, leave the Keyword prompt blank and press <Enter>.
If both the Author and Keyword(s) are known the user inputs an Author and keyword and presses <Enter> to commence the search.
Search restrictions may be useful in order to avoid retrieving too many references in response to a search. The <F10> ("Limit") function key is available to users at the point of input (pre-restriction).
The <F10> ("Limit") or <F10> ("Options") key is also available from the list of references, from the Author Index or Keyword Index depending on the search results (post-restriction).
For further information on restricting a search, refer to Search Restriction.
Note: The <F10> ("Options") menu is configurable using the Configurable Menus interface under Utilities, Parameters, Menus. See See Variable Menus & Collection Lists in this manual, or System Manager Manual, Volume 2: Application Maintenance, Chapter 42: Interface Configuration for full details.
If the search finds a match, the list of matching references displays. Refer to Figure 2.12: The Author/Keyword References List. This list will be in random order unless a sort order has been specified. For more information about the References List and its composition including sort variables, refer to The References List.
A count of the number of matching references displays at the top right of the screen, with the original search terms displayed at the top left. The users current position in the list is also displayed.
If the combined Author/Keyword search fails to find any matching records in the catalogue, the user is given the option of viewing the references that match the Keyword component(s) of the search, or returning to the prompt to refine their search.
If the combined Author/Keyword search fails to find any matching Keywords, the user is prompted to refine the search.
The search uses the same indexes as are used for the Author and Keyword searches. There is no separate Author/Keyword index. The way these indexes are built is detailed within Author Searching and Keyword Searching respectively.
The treatment of search terms is the same as for their respective searches, with Boolean keyword and wildcards being allowed within the combined search. The Author search will effectively be wildcarded, for example a search for Author = "Shake", Keyword = "King" will search for all Authors beginning with "Shake*" and Works containing the keyword "King*".
Subject search permits searching for Works by Subject terms, as defined by the Library. Refer to Figure 2.13: Subject Search Prompt.
The Subject Search Prompt contains an information window briefly outlining the principles of searching by subject(s) and a prompt bar for user input.
You may type Subject Headings in upper or lower case. Adjacent terms should be separated with a space character. If the Subject Headings List used by your Library specifies certain punctuation between terms, you must enter the terms and punctuation exactly as found in your Library's authority list of Subject Headings.
Examples:
Determinism Philosophy : Determinism Behaviour-Patterns
Press the <Enter> key to commence the search.
If the search results in a match (exactly or partially) the Subject Index displays with the highlight bar positioned on the closest match. Refer to Figure 2.14: The Subject Index.
Individual Subject Indexes can be linked to different Collections, as explained in How a Subject Search Works The Subject Index actually searched is the one linked to the Collection currently in use. The name of the Subject Index attached to the Catalogue currently selected is displayed at the top of the search results list.
If Back Browsing is enabled the closest match will be the second line on the screen, otherwise it will be on the top line of the screen.
A count of the number of matching entries displays at the top right of the screen, with the original search term displayed at the top left.
If the search does not achieve a match, a dialog box displays, giving the user the option of viewing the closest match in the Subject Index or returning to the Subject Prompt to refine the search.
The Subject Index also provides the facility to set up and navigate to "See" and "See also" references. Selecting a "See" or "See also" cross-reference from the Subject Index will place the user in the Index at the term referred to. It is then possible to select that term, or use <F12> ("Back") to go back to the original term.
The Subject Index always displays. Selecting an entry from the Subject Index results in the display of the Classification Index for the Collection currently in use. This performs a Classification search, as illustrated in Figure 2.15: Example Classification Search from the Subject Index. This is dealt with more fully in Figure 2.15: Example Classification Search from the Subject Index.
If there is no Classification Number corresponding to the Class Number assigned to the Subject term, the user will be given the option of viewing similar Classmarks.
Libraries may have multiple Subject Indexes and an interface in Online Utilities is available for linking Subject Indexes to Collections. Refer to System Manager Manual, Volume 2: Application Maintenance, Section 7.14: Linking Subject Indexes to Collections.
The "subject_build" script, as documented in System Manager Manual, Volume 2: Application Maintenance, Chapter 21: Building Subject Indexes, is used to build a Subject Index from an ASCII file.
Entries for the Subject Index are built from the SUBJECT table, which is built using the subject builder.
Any ASCII file can be used to build a Subject authority file which will then use the BCP (Bulk Copy Program) facility to build the Subject table.
The Subject Index handles "See" / "See also" references, and Classmarks.
The Subject file can be built using a "vi" edited file which should include information in the format shown below. The data entered will always have a Display as term and will have either a Classmark or a Reference. It is advisable to enter sample test data with both, neither etc. to test the Subject Builder and the report file.
The input convention is:
Display as term| Classmark term| Reference term
Note: ":" is the only valid delimiter within Display as terms.
If the following is input:
Information technology: Libraries| 123.45|
this will result in a normal Subject Classmark entry.
If the following is supplied:
Maths| | Mathematics
this will build a See reference term.
Spaces may be entered as required; leading and trailing spaces will be stripped from the data in each field, but otherwise your input will be used as entered.
Further examples of contents in the Subject authority file are shown below.
Bricks and mortar|691| Building |690| Building : Bibliographies|016.69| Building:History|690.09| Building:Research|690.072| Building components||Bricks and mortar Building construction drawing ||Architectural design Building construction practices|692| Building failures||Rotunda Building industry:Economic organisation|338.476904872910018374848490| Building industry:Economic organisation:management|658.99| Building societies:financial economics|332.32| Buildings: Photography|778.94|
Each Subject Index has to be built separately. To generate a Subject Index, run the Subject Builder i.e. the "subject_build" script. The script is able to handle multiple Subject Indexes using switches specified at the command line.
Note: This script should always be followed by a "full_dbdump" as it sets options on the database which effect transaction logging.
The script is located in the $TALIS_HOME/utils/bin directory, typically /usr/opt/blcmp/talis/utils/bin. To run this script:
cd /usr/opt/blcmp/talis/bin <Enter>
subject_build [arguments]
The "subject_build" script can accept a number of mandatory and optional switches:
subject_build [d(database_name)] [h] [i(filename)] [l(lock_database_name)] n(subject_index_code) [r(directory)] [s(directory)]
The switches used by "subject_build" are explained fully in System Manager Manual, Volume 2: Application Maintenance, Chapter 21: Building Subject Indexes. The Subject Index Code entered against the "n" switch must be a code that has been defined in Online Utilities interface, as explained in System Manager Manual, Volume 2: Application Maintenance, Section 2.12: Subject Index Names.
The "subject_build" script cannot be used to add to or amend an existing Subject Index. This can be done online using the functionality described in Chapter 19: Subject Index Editor of the Cataloguing Reference Manual.
All input terms have superfluous spaces and all punctuation except ":" (colon) removed. The terms are uppercased. Spaces between terms are replaced with a "$" (dollar sign), and colons replaced by space (for searching purposes i.e. to match the filing values). The resulting string is then matched against a filing value in the SUBJECT table.
Classification search permits searching for Works by Dewey or Library of Congress (LC) classification number. It will search for an exact or partial match. Refer to Figure 2.16: The Classmark Search Prompt.
The Classmark Search Prompt comprises an information window briefly outlining the principles of searching by classification number and a prompt bar for the user's input.
Type in the Classmark of your choice. For Example:
330.1 <Enter> GT4965 <Enter>
Be sure to type the Classmark with any punctuation exactly as shown in the Library's Subject Index.
Press the <Enter> key to commence a search of the Classmark Index for the Collection currently in use.
Note: If you do not know the classification number you may access the relevant Classmark(s) directly by performing a Subject search, as explained in Subject Searching.
Search restrictions may be useful in order to avoid retrieving too many references in response to a search. The <F10> ("Limit") function is available to users at the point of input (pre-restriction) only for Dewey Class Number searching. It is also possible to limit at the Subject Index (post restriction) using the <F10> ("Options") menu. For further information on restricting a search, refer to Search Restriction.
Note: The <F10> ("Options") menu is configurable. Refer to Variable Menus & Collection Lists in this manual, or System Manager Manual, Volume 2: Application Maintenance, Chapter 42: Interface Configuration for full details.
If the search results in a match, the Class Index displays for the Collection currently in use, with the highlight bar positioned on the closest match. Refer to Figure 2.17: References in the Class Index (Dewey).
If back browsing is enabled the closest match will be the second line on the screen, otherwise this will be the top line on screen. Back browsing is not possible for Library of Congress Class Number searching, so the highlight bar will always be positioned on the top line on the screen.
A count of the number of similar Class Numbers will be displayed in the top right of the screen, with the original search term displayed at the top left.
Entries in the Class Index are grouped by Class Number. They will not be sorted within the same Class Number unless a sort order has been defined. For more information about sort variables, refer to The References List.
If the search does not achieve a match, a dialog box will be displayed, giving the user the option of viewing the closest match in the Class Index for the Collection currently in use, or returning to the Classmark Prompt to refine the search.
The system can handle searches by both Dewey Decimal and/or Library of Congress classification numbers. This section covers the complex rules which govern:
Any Classification search will be treated as a Dewey Number search, unless any environment variables are used to specify otherwise (see LC Searching (TAL_LC_ON) below).
The environment variable used to indicate that a collection follows the Dewey Decimal scheme is:
TAL_DEWEY_ON=YES;export TAL_DEWEY_ON <Enter>
This tells OPAC to normalise the class number input as a Dewey classmark, and to try to get a hit on that type of search first. If the user's input cannot be found as a Dewey number, it will try again by treating it as a LC number. If no matches are found, the user will be placed at the nearest equivalent Dewey number.
To permit searching for Library of Congress (LC) class numbers the following environment variables need to be set:
TAL_LC_ON=yes;export TAL_LC_ON <Enter>
If this is set then the search will be treated as an LC class search. This tells OPAC to normalise Class Number input as a Library of Congress (LC) classmark, and to search for a hit on that first. If the input cannot be found as an LC Number, the system will try again, this time searching it as a Dewey Number.
The following variable also needs to be set, if it has not already been set in the LC_CLASS_LEN attribute of the LIBPARAM table to control the length of the filing key for LC Classmarks:
TAL_LC_LENGTH=nn;export TAL_LC_LENGTH <Enter>
where "nn" is the length of filing key used (normally not more than 36 characters). It will default to 256, if not set. This variable is used by Cataloguing and OPAC.
If the length set differs, for any reason, between OPAC and Cataloguing then searches in OPAC for LC classmarks will fail.
Note: Libraries not using LC do not need to set this variable - it has no effect on Dewey numbers.
The LC_CLASS_LEN attribute of the LIBPARAM table can only be checked using "sql" as no interface is available to amend this table currently; call the Help Desk if you require assistance.
Alternatively, the TAL_LC_LENGTH environment variable can be set in the .profile of the logon used by Cataloguers, as well as that used by OPACs, in Libraries which use LC.
Classmark entries are built from the CLASS table in the database. The CLASS table is made up of all the Classification Numbers entered for all Works; both Main and Added Class Numbers are used.
Note: It is not possible to search on Feature Headings or Class Suffixes, or for Class Numbers added to Items.
Note: The "/" (slash) cannot be used as a valid character in a Dewey Classification Number because this is the syntax used by Talis for expert searching and search limiters, hence it cannot be searched literally.
Processing of a Library of Congress Number Search: All input terms are converted as follows. Leading and trailing spaces are removed. Characters are converted to uppercase. All punctuation is removed, except "-" (hyphen) "&" (ampersand) and "/" (forward slash).
Note: The "/" (slash) character is valid in an LC Class Number. For this reason it is not possible to perform expert searches (including search limit) when searching for a Library of Congress Class Number.
If the Library uses both Dewey and LC classification schemes, then the TAL_DEWEY_ON environment variable should be set (in addition to TAL_LC_ON):
If both schemes are set the system will make a judgement on how to treat the search, based on whether the search string begins with a letter or a number; if it begins with alphabetic characters then it will attempt initially to search for a Library of Congress Number, otherwise it will be presumed Dewey.
If the term input is treated as an LC number, but cannot be found as an LC number the system will try again, this time treating the search as a Dewey number. If no matches are found the user will be placed at the nearest equivalent LC number. A search for a term treated as a Dewey number will be treated likewise.
Number search permits searching for Works by Control Number. It looks only for an exact match. Refer to Figure 2.18: The Control Number Search Prompt.
The Control Number Search Prompt consists of an information window briefly outlining the principles of searching by Control Number and a prompt bar for the user's input.
Number searching is advantageous if you know the Control Number for a Work and wish to search for Items with this number (i.e. to see whether the Library has copies of a specific Work).
Any recognised Work identification numbers are searchable under a Control Number search. The options, and a summary of the required search term formats, follows:
Type in 10 characters, followed by <Enter>. For example:
0306405490 <Enter>
The last character is a check digit. If it is a letter "X" you may type this in upper or lower case.
Type in 8 numeric characters followed by <Enter>. Omit any hyphens (for example 89016828). Insert "0" pad characters if required and prefix with a space. For example:
89016828 <Enter>
Enter a lower case "b" and 7 numerics followed by <Enter>. (The 2nd and 4th characters may be alphanumerics). For example:
b6521130 <Enter>
Type in 4 numeric characters, followed by a hyphen and the remaining 4 numeric characters. Press <Enter> to commence your search. For example:
00312290 <Enter>
Type in an upper case "G" and 8 numeric characters, followed by <Enter>. Alternatively, you may need to type in an upper case "GC" and 7 numerics, followed by <Enter>. Examples:
G00235603 <Enter> GC6705463H <Enter>
Type in 6 numeric characters, where the last character is a check digit, followed by <Enter>. For example:
997139 <Enter>
Type in a lower case "mmu" and 7 numeric characters, followed by <Enter>. For example:
mmu1234567 <Enter>
Type in lower case "vz" and 8 numerics. For example:
vz00066230 <Enter>
Always prefixed the number you require with an upper case "R". Type in a combination of alphabetic and numeric characters, up to 10 characters maximum. The number is constructed from the manufacturer's number: for example, "CDL52084" should be entered as "RCDL52084". Press <Enter> to commence your search. For example:
RCDL52084 <Enter>
One exception is when the number would be constructed with 2 alphabetic characters followed by 7 numerics, such as RC8834561. In such a case, add another upper case "R" to the end of the number: i.e. "RC8834561R".
The <F10> ("Limit") function is not available for Number searching as a Control Number search should only obtain one hit.
If the search results in a match, the Number Index displays with the highlight bar positioned on the closest match. Refer to Figure 2.19: The Number Index. If back browsing is enabled this will be the second line on the screen, otherwise it will be the top line on screen.
A count of the number of matching numbers displays in the top right of the screen, with the original search term displayed at the top left.
If the search does not achieve a match, a dialog box will be displayed, giving the user the option of viewing the closest match in the Number Index as illustrated in Figure 2.19: The Number Index, or returning to the Number Prompt to refine the search.
Control Number entries are built from the NUMBER table in the database. The NUMBER table is compiled from Control Numbers added to Works, which includes both Main and Added numbers (including link numbers).
All input terms have superfluous spaces and all punctuation, except "-" (hyphen) and "." (full stop) removed. The resulting string is then matched against a filing value in the NUMBER table.
The standard Control Number validation is not applied to number search in order to permit Libraries to search for local invalid Control Numbers. Also, letters are searched in the case in which they were entered, they are not converted to uppercase.
Reading Lists may have been created for certain courses. A Reading List comprises a list of all Works which are recommended reading for a particular topic.
Titles in a Reading List can be viewed and reservations placed as elsewhere in OPAC.
Selecting the Reading Lists option from the main OPAC Search Menu results in presentation of a sub-menu, as shown in Figure 2.20: The Reading Lists Menu.
Reading List permits searching for Reading Lists by:
Reading List searching takes no account of the Catalogue which the user is currently using.
Note: It is not possible to limit Reading List searches, as Reading Lists are static lists.
This option permits users to search for the owners of Reading Lists (i.e. a lecturer's name typically). Refer to Figure 2.21: The Owner Search Prompt.
The Owner Search screen contains an information window explaining the principles of searching by Owner and a prompt bar for user input.
This works in the same way as an Author search; in that the user enters the name; surname first, then a comma followed by forename or initials. (Refer to Author Searching if further details are required).
All input terms are converted into uppercase. Certain punctuation is removed and then the search term is compared with a value based on the Borrower Name for that lecturer.
This option permits users to search for Reading Lists by name; this might be a Library-assigned name or the name allocated to the Reading List by the owner/lecturer. Refer to Figure 2.22: The Reading List Name Search Prompt.
The Reading List Name Search screen consists of an information window briefly explaining the principles of searching by name of Reading List and a prompt bar for user input. This works in the same way as a Title search (as explained in Title Searching), in that the user enters the name - or as much of the name as they remember - and the system matches this against the Reading List Names Index.
This option permits users to search for Reading Lists by the name of the course. Refer to Figure 2.23: The Reading Course Search Prompt.
The Reading List Course Search screen consists of an information window briefly explaining the principles of searching by Reading List Course and a prompt bar for user input. This works in the same way as a Title search (as explained in Title Searching), in that the user enters the name - or as much of the name as he/she can remember - and the system matches this against the Reading List Courses Index
If an Owner, Name or Course search results in a match, the user will be positioned at the correct point in the relevant Reading List Index.
Back browsing is always possible in Reading Lists, irrespective or whether back browse has been enabled.
If the search does not achieve a match, a dialog box will display, giving the user the option of viewing the closest match in the Index or returning to the previous search prompt to refine their search.
The Index always displays, even for an exact match, as this allows the user to see if there are variants of the list available. Refer to Figure 2.24: Example Reading List Index (Owners).
A count of the number of matching entries will be displayed in the top right of the screen.
Selecting a Reading List from this Index will result in the display of a list of the references on this list.
Libraries are now able to modify the order of references on a Reading List, in a manner similar to how search results can be sorted elsewhere in Talis OPAC / TalisWeb OPAC.
Search results are sorted using a new TAL_OPAC_RL_SORT environment variable. This variable accepts three parameters separated by the pipe character "|". The parameters are:
SORT_LENGTH | MAX_SORT_ROWS | SORT_ORDER
This specifies the length of sort key data which will be returned and sorted upon. If set to "0" then the Reading List references are not sorted. The SORT_LENGTH is set to "0" by default. The lower the SORT_LENGTH figure, the lower will be the impact on client performance. The recommended length is approximately "10".
This specifies the maximum number of rows that the system will sort. This parameter can be used to adjust search performance. If the number of Works in the Reading List is greater than this value, they will not be sorted. MAX_SORT_ROWS is set to 500 by default.
If MAX_SORT_ROWS is set to "-1" then sorting always occurs regardless of the number of references returned.
Note: The value defined in MAX_SORT_ROWS should be set to around the same value (or less) than that specified against the TAL_MAX_BROWSE environment variable; otherwise sorting may not always occur where you would expect. (For example, if MAX_SORT_ROWS is set to sort 250 Works, but the value of the TAL_MAX_BROWSE environment variable is set to 200, then only Reading Lists containing less than 200 rows would be sorted).
This parameter specifies the fields which determine the sort order:
|
Sort Key |
Sort Order for References |
|
"A" |
Sort by Author. |
|
"T" |
Sort by Title. |
|
"D" |
Sort by publication date. |
Table 2.2: The Sort Order for Reading List References
It is possible to use a combination of these fields. The SORT_ORDER parameter defaults to "ATD".
To sort on the first 25 characters and sort up to 500 rows by the concatenated values of Title, Author then Date, specify:
TAL_OPAC_RL_SORT="25|500|TAD"; export TAL_OPAC_RL_SORT
The TAL_OPAC_RL_SORT environment variable can be set globally in the "talis" script or it can be set on an individual user basis by amending the user's .profile.
Libraries have the option to include or exclude access to Borrower details from specified Sites, thereby making personal data secure from Library staff and users. This is a particular requirement for Prison Libraries, where special Data Protection measures apply to prison inmates, requiring personal data to be secure. Prison inmates are unable to access details of any other Borrower on the database.
Information about excluded Borrowers / Reading List Owners is masked with asterisks on the Reading List Indexes. Refer to Figure 2.25: Example Index of Courses (Masked Owners).
Inclusion and exclusion of access to Borrower details for specified Sites are handled by two environment variables:
A "TAL_BOR_IN" environment variable is used to determine whether Borrower information for a specified Site will be visible onscreen.
If "TAL_BOR_IN" is set to Site1, Site 3, etc. then Borrowers are excluded from display and transactions are prevented if their Home Site does not match on one of these Sites. For example:
TAL_BOR_IN=CL,N,W,NW,P
A "TAL_BOR_OUT" environment variable is used to determine whether Borrower information for a specified Site will be excluded in the display and transactions prevented.
If "TAL_BOR_OUT" is set to Site2, Site4, etc. then Borrowers are excluded and transactions involving them are prevented if their Home Site matches on one of the Sites specified.
The search type options available from the Search Menu are as follows:
Copyright 2000 © Talis Information Ltd. All Rights Reserved.
