${database_base}.accents.db ${database_base}.uml.db The database file used for the fuzzy "accents" search algorithm. This database is created by htfuzzy and used by htsearch. en-us en it This attribute allows you to restrict the set of natural languages that are preferred as a response to an HTTP request performed by the digger. This can be done by putting one or more language tags (as defined by RFC 1766) in the preferred order, separated by spaces. By doing this, when the server performs a content negotiation based on the 'accept-language' given by the HTTP user agent, a different content can be shown depending on the value of this attribute. If set to an empty list, no language will be sent and the server default will be returned. true no If set to true, the first occurrence of each matched word in the excerpt will be linked to the closest anchor in the document. This only has effect if the EXCERPT variable is used in the output template and the excerpt is actually going to be displayed. false true If set to true, strings of multiple slashes ('/') in URL paths will be left intact, rather than being collapsed. This is necessary for some search engine URLs which use slashes to separate fields rather than to separate directory components. However, it can lead to multiple database entries refering to the same file, and it causes '/foo//../' to be equivalent to '/foo/', rather than to '/'. search_algorithm search_results_header Allows the specified config file attributes to be specified in search forms as separate fields. This could be used to allow form writers to design their own headers and footers and specify them in the search form. Another example would be to offer a menu of search_algorithms in the form. <SELECT NAME="search_algorithm"> <OPTION VALUE="exact:1 prefix:0.6 synonyms:0.5 endings:0.1" SELECTED>fuzzy <OPTION VALUE="exact:1">exact </SELECT> The general idea behind this is to make an input parameter out of any configuration attribute that's not already automatically handled by an input parameter. You can even make up your own configuration attribute names, for purposes of passing data from the search form to the results output. You're not restricted to the existing attribute names. The attributes listed in the allow_in_form list will be settable in the search form using input parameters of the same name, and will be propagated to the follow-up search form in the results template using template variables of the same name in upper-case. You can also make select lists out of any of these input parameters, in the follow-up search form, using the build_select_lists configuration attribute. false true If set to true, numbers are considered words. This means that searches can be done on number as well as regular words. All the same rules apply to numbers as to words. See the description of valid_punctuation for the rules used to determine what a word is. false true If set to true, htdig will handle URLs that contain embedded spaces. Technically, this is a violation of RFC 2396, which says spaces should be stripped out (as htdig does by default). However, many web browsers and HTML code generators violate this standard already, so enabling this attribute allows htdig to handle these non-compliant URLs. Even with this attribute set, htdig still strips out all white space (leading, trailing and embedded), except that space characters embedded within the URL will be encoded as %20. true false If set to true, htdig will index virtual web sites as expected. If false, all URL host names will be normalized into whatever the DNS server claims the IP address to map to. If this option is set to false, there is no way to index either "soft" or "hard" virtual web sites. body When the first matched word in the excerpt is linked to the closest anchor in the document, this string can be set to specify a target in the link so the resulting page is displayed in the desired frame. This value will only be used if the add_anchors_to_excerpt attribute is set to true, the EXCERPT variable is used in the output template and the excerpt is actually displayed with a link. false yes If set to true, the words in the keywords input parameter in the search form will be joined with logical ORs rather than ANDs, so that any of the words provided will do. Note that this has nothing to do with limiting the search to words in META keywords tags. See the search form documentation for details on this. 1 1 TO BE COMPLETED
See also heading_factor. mypassword This tells htdig to send the supplied username:password with each HTTP request. The credentials will be encoded using the "Basic" authentication scheme. There must be a colon (:) between the username and password.
This attribute can also be specified on htdig's command line using the -u option, and will be blotted out so it won't show up in a process listing. If you use it directly in a configuration file, be sure to protect it so it is readable only by you, and do not use that same configuration file for htsearch. 1000 501.1 This is a weight of "how important" a page is, based on the number of URLs pointing to it. It's actually multiplied by the ratio of the incoming URLs (backlinks) and outgoing URLs (links on the page), to balance out pages with lots of links to pages that link back to them. The ratio gives lower weight to "link farms", which often have many links to them. This factor can be changed without changing the database in any way. However, setting this value to something other than 0 incurs a slowdown on search results. .wav .gz .z .sit .au .zip .tar .hqx .exe .com .gif .jpg .jpeg .aiff .class .map .ram .tgz .bin .rpm .mpg .mov .avi .css .foo .bar .bad This is a list of extensions on URLs which are considered non-parsable. This list is used mainly to supplement the MIME-types that the HTTP server provides with documents. Some HTTP servers do not have a correct list of MIME-types and so can advertise certain documents as text while they are some binary format. If the list is empty, then all extensions are acceptable, provided they pass other criteria for acceptance or rejection. See also valid_extensions. forum=private section=topsecret&passwd=required This is a list of CGI query strings to be excluded from indexing. This can be used in conjunction with CGI-generated portions of a website to control which pages are indexed. ${common_dir}/bad_words ${common_dir}/badwords.txt This specifies a file which contains words which should be excluded when digging or searching. This list should include the most common words or other words that you don't want to be able to search on (things like sex or smut are examples of these.)
The file should contain one word per line. A sample bad words file is located in the contrib/examples directory. BIN_DIR /usr/local/bin This is the directory in which the executables related to ht://Dig are installed. It is never used directly by any of the programs, but other attributes can be defined in terms of this one.

The default value of this attribute is determined at compile time.

and or not et ou non These three strings are used as the keywords used in constructing the LOGICAL_WORDS template variable, and in parsing the words input parameter when the method parameter or match_method attribute is set to boolean. See also the boolean_syntax_errors attribute. Expected 'a search word, a quoted phrase, a boolean expression between ()' 'at the end' 'instead of' 'end of expression' quotes Attendait "un mot" "à la fin" "au lieu de" "fin d'expression" "points de quotation" These six strings are used as the keywords used to construct various syntax error messages for errors encountered in parsing the words input parameter when the method parameter or match_method attribute is set to boolean. They are used in conjunction with the boolean_keywords attribute, and comprise all English-specific parts of these error messages. The order in which the strings are put together may not be ideal, or even gramatically correct, for all languages, but they can be used to make fairly intelligible messages in many languages. MATCH_LIST matchesperpage matches_per_page_list \ 1 1 1 matches_per_page "Previous Amount" \ RESTRICT_LIST,multiple restrict restrict_names 2 1 2 restrict "" \ FORMAT_LIST,radio format template_map 3 2 1 template_name "" This list allows you to define any htsearch input parameter as a select list for use in templates, provided you also define the corresponding name list attribute which enumerates all the choices to put in the list. It can be used for existing input parameters, as well as any you define using the allow_in_form attribute. The entries in this list each consist of an octuple, a set of eight strings defining the variables and how they are to be used to build a select list. The attribute can contain many of these octuples. The strings in the string list are merely taken eight at a time. For each octuple of strings specified in build_select_lists, the elements have the following meaning:
  1. the name of the template variable to be defined as a list, optionally followed by a comma and the type of list, and optional formatting codes
  2. the input parameter name that the select list will set
  3. the name of the user-defined attribute containing the name list
  4. the tuple size used in the name list above
  5. the index into a name list tuple for the value
  6. the index for the corresponding label on the selector
  7. the configuration attribute where the default value for this input parameter is defined
  8. the default label, if not an empty string, which will be used as the label for an additional list item for the current input parameter value if it doesn't match any value in the given list
See the select list documentation for more information on this attribute. 1 1 TO BE COMPLETED
See also heading_factor. true false This specifies whether ht://Dig should consider URLs case-sensitive or not. If your server is case-insensitive, you should probably set this to false. false false Include the modification date of the page in the MD5 hash, to reduce the problem with identical but physically separate pages in different parts of the tree pointing to different pages. false false Uses the MD5 hash of pages to reject aliases, prevents multiple entries in the index caused by such things as symbolic links Note: May not do the right thing for incremental update htdig_docs htdig_bugs This is a list of config file names that are used for searching multiple databases. Simply put, htsearch will loop through the databases specified by each of these config files and present the result of the search on all of the databases. The corresponding config files are looked up in the config_dir directory. Each listed config file must exist, as well as the corresponding databases. COMMON_DIR /tmp Specifies the directory for files that will or can be shared among different search databases. The default value for this attribute is defined at compile time. http:// http://www. ftp:// ftp://ftp. /pub/ .html .htm .gif .jpg .jpeg /index.html /index.htm .com/ .com mailto: //www.htdig.org/ml/ \ .html \ http://dev.htdig.org/ \ http://www.htdig.org/ Sub-strings often found in URLs stored in the database. These are replaced in the database by an internal space-saving encoding. If a string specified in url_part_aliases, overlaps any string in common_url_parts, the common_url_parts string is ignored.
Note that when this attribute is changed, the database should be rebuilt, unless the effect of "changing" the affected URLs in the database is wanted.
 0 6 If specified and the zlib compression library was available when compiled, this attribute controls the amount of compression used in the doc_excerpt file. DEFAULT_CONFIG_FILE Name of configuration file to load. For security reasons, restrictions are placed on the values which can be specified on the command line to htsearch. The default value of this attribute is determined at compile time. CONFIG_DIR /var/htdig/conf This is the directory which contains all configuration files related to ht://Dig. It is never used directly by any of the programs, but other attributes or the include directive can be defined in terms of this one.

The default value of this attribute is determined at compile time.

${common_dir}/cookies.txt Specifies the location of the file used for importing cookies for the crawl. These cookies will be preloaded into htdig's in-memory cookie jar, but aren't written back to the file. Cookies are specified according to Netscape's format (tab-separated fields). If this attribute is left blank, no cookie file will be read.

For more information, see the sample cookies.txt file in the ht://Dig source distribution.

false yes If set to true, a file with all the image URLs that were seen will be created, one URL per line. This list will not be in any order and there will be lots of duplicates, so after htdig has completed, it should be piped through sort -u to get a unique list. false yes If set to true, a file with all the URLs that were seen will be created, one URL per line. This list will not be in any order and there will be lots of duplicates, so after htdig has completed, it should be piped through sort -u to get a unique list. ${database_dir}/db ${database_dir}/sales This is the common prefix for files that are specific to a search database. Many different attributes use this prefix to specify filenames. Several search databases can share the same directory by just changing this value for each of the databases. DATABASE_DIR /var/htdig This is the directory which contains all database and other files related to ht://Dig. It is never used directly by any of the programs, but other attributes are defined in terms of this one.

The default value of this attribute is determined at compile time.

0 0.35 This factor, gives higher rankings to newer documents and lower rankings to older documents. Before setting this factor, it's advised to make sure your servers are returning accurate dates (check the dates returned in the long format). Additionally, setting this to a nonzero value incurs a small performance hit on searching. %Y-%m-%d This format string determines the output format for modification dates of documents in the search results. It is interpreted by your system's strftime function. Please refer to your system's manual page for this function, for a description of available format codes. If this format string is empty, as it is by default, htsearch will pick a format itself. In this case, the iso_8601 attribute can be used to modify the appearance of the date. 150 350 Plain old "descriptions" are the text of a link pointing to a document. This factor gives weight to the words of these descriptions of the document. Not surprisingly, these can be pretty accurate summaries of a document's content. See also heading_factor and meta_description_factor. description "description htdig-description" The words in this list are used to search for descriptions in HTML META tags. This list can contain any number of strings that each will be seen as the name for whatever description convention is used. While words in any of the specified description contents will be indexed, only the last meta tag containing a description will be kept as the meta description field for the document, for use in search results. The order in which the names are specified in this configuration attribute is irrelevant, as it is the order in which the tags appear in the documents that matters.
The META tags have the following format:
<META name="somename" content="somevalue">
See also meta_description_factor. true true This option, if set to true, will disable HTTP cookies. ${database_base}.docdb ${database_base}documents.db This file will contain a Berkeley database of documents indexed by document number. It contains all the information gathered for each document, except the document excerpts which are stored in the doc_excerpt file. ${database_base}.excerpts ${database_base}excerpts.db This file will contain a Berkeley database of document excerpts indexed by document number. It contains all the text gathered for each document, so this file can become rather large if max_head_length is set to a large value. The size can be reduced by setting the compression_level, if supported on your system. ${database_base}.docs.index documents.index.db This file contains a mapping of document numbers to URLs and is used by htdig during indexing. It is used on updates if it exists. ${database_base}.docs /tmp/documents.text This file is basically a text version of the file specified in doc_db. Its only use is to have a human readable database of all documents. The file is easy to parse with tools like perl or tcl. 31 Day component of last date allowed as last-modified date of returned docutments. This is most usefully specified as a GCI argument. See also startyear. <strong><code> ...</code></strong> ... When excerpts are displayed in the search output, this string will be appended to the excerpt if there is text following the text displayed. This is just a visual reminder to the user that the excerpt is only part of the complete document. </strong> </font> When excerpts are displayed in the search output, matched words will be highlighted using start_highlight and this string. You should ensure that highlighting tags are balanced, that is, this string should close any formatting tag opened by start_highlight. ${common_dir}/english.aff /var/htdig/affix_rules Specifies the location of the file which contains the affix rules used to create the endings search algorithm databases. Consult the documentation on htfuzzy for more information on the format of this file. ${common_dir}/english.0 /var/htdig/dictionary Specifies the location of the file which contains the dictionary used to create the endings search algorithm databases. Consult the documentation on htfuzzy for more information on the format of this file. ${common_dir}/root2word.db /var/htdig/r2w.db This attributes specifies the database filename to be used in the 'endings' fuzzy search algorithm. The database maps word roots to all legal words with that root. For more information about this and other fuzzy search algorithms, consult the htfuzzy documentation.
Note that the default value uses the common_dir attribute instead of the database_dir attribute. This is because this database can be shared with different search databases. ${common_dir}/word2root.db /var/htdig/w2r.bm This attributes specifies the database filename to be used in the 'endings' fuzzy search algorithm. The database maps words to their root. For more information about this and other fuzzy search algorithms, consult the htfuzzy documentation.
Note that the default value uses the common_dir attribute instead of the database_dir attribute. This is because this database can be shared with different search databases. 12 Month component of last date allowed as last-modified date of returned docutments. This is most usefully specified as a GCI argument. See also startyear. 2002 Year component of last date allowed as last-modified date of returned docutments. This is most usefully specified as a GCI argument. See also startyear. 300 500 This is the maximum number of characters the displayed excerpt will be limited to. The first matched word will be highlighted in the middle of the excerpt so that there is some surrounding context.
The start_ellipses and end_ellipses are used to indicate that the document contains text before and after the displayed excerpt respectively. The start_highlight and end_highlight are used to specify what formatting tags are used to highlight matched words. false yes If set to true, the excerpt of a match will always show the top of the matching document. If it is false (the default), the excerpt will attempt to show the part of the document that actually contains one of the words. myhost.com/mailarchive/ If a URL contains any of the space separated patterns, it will be discarded in the searching phase. This is used to exclude certain URLs from search results. The list can be specified from within the configuration file, and can be overridden with the "exclude" input parameter in the search form. /cgi-bin/ .cgi students.html cgi-bin If a URL contains any of the space separated patterns, it will be rejected. This is used to exclude such common things such as an infinite virtual web-tree which start with cgi-bin. text/html /usr/local/bin/htmlparser \ application/pdf /usr/local/bin/parse_doc.pl \ application/msword->text/plain "/usr/local/bin/mswordtotxt -w" \ application/x-gunzip->user-defined /usr/local/bin/ungzipper This attribute is used to specify a list of content-type/parsers that are to be used to parse documents that cannot by parsed by any of the internal parsers. The list of external parsers is examined before the builtin parsers are checked, so this can be used to override the internal behavior without recompiling htdig.
The external parsers are specified as pairs of strings. The first string of each pair is the content-type that the parser can handle while the second string of each pair is the path to the external parsing program. If quoted, it may contain parameters, separated by spaces.
External parsing can also be done with external converters, which convert one content-type to another. To do this, instead of just specifying a single content-type as the first string of a pair, you specify two types, in the form type1->type2, as a single string with no spaces. The second string will define an external converter rather than an external parser, to convert the first type to the second. If the second type is user-defined, then it's up to the converter script to put out a "Content-Type: type" header followed by a blank line, to indicate to htdig what type it should expect for the output, much like what a CGI script would do. The resulting content-type must be one that htdig can parse, either internally, or with another external parser or converter.
Only one external parser or converter can be specified for any given content-type. However, an external converter for one content-type can be chained to the internal parser for the same type, by appending -internal to the second type string (e.g. text/html->text/html-internal) to perform external preprocessing on documents of this type before internal parsing. There are two internal parsers, for text/html and text/plain.

The parser program takes four command-line parameters, not counting any parameters already given in the command string:
infile content-type URL configuration-file

Parameter Description Example
infile A temporary file with the contents to be parsed. /var/tmp/htdext.14242
content-type The MIME-type of the contents. text/html
URL The URL of the contents. http://www.htdig.org/attrs.html
configuration-file The configuration-file in effect. /etc/htdig/htdig.conf

The external parser is to write information for htdig on its standard output. Unless it is an external converter, which will output a document of a different content-type, then its output must follow the format described here.
The output consists of records, each record terminated with a newline. Each record is a series of (unless expressively allowed to be empty) non-empty tab-separated fields. The first field is a single character that specifies the record type. The rest of the fields are determined by the record type.

Record type Fields Description
w word A word that was found in the document.
location A number indicating the normalized location of the word within the document. The number has to fall in the range 0-1000 where 0 means the top of the document.
heading level A heading level that is used to compute the weight of the word depending on its context in the document itself. The level is in the range of 0-10 and are defined as follows:
0
Normal text
1
Title text
2
Heading 1 text
3
Heading 2 text
4
Heading 3 text
5
Heading 4 text
6
Heading 5 text
7
Heading 6 text
8
unused
9
unused
10
Keywords
u document URL A hyperlink to another document that is referenced by the current document. It must be complete and non-relative, using the URL parameter to resolve any relative references found in the document.
hyperlink description For HTML documents, this would be the text between the <a href...> and </a> tags.
t title The title of the document
h head The top of the document itself. This is used to build the excerpt. This should only contain normal ASCII text
a anchor The label that identifies an anchor that can be used as a target in an URL. This really only makes sense for HTML documents.
i image URL An URL that points at an image that is part of the document.
m http-equiv The HTTP-EQUIV attribute of a META tag. May be empty.
name The NAME attribute of this META tag. May be empty.
contents The CONTENTS attribute of this META tag. May be empty.

See also FAQ questions 4.8 and 4.9 for more examples.

 https /usr/local/bin/handler.pl \ ftp /usr/local/bin/ftp-handler.pl This attribute is a bit like external_parsers since it specifies a list of protocols/handlers that are used to download documents that cannot be retrieved using the internal methods. This enables htdig to index documents with URL schemes it does not understand, or to use more advanced authentication for the documents it is retrieving. This list is checked before HTTP or other methods, so this can override the internal behavior without writing additional code for htdig.
The external protocols are specified as pairs of strings, the first being the URL scheme that the script can handle while the second is the path to the script itself. If the second is quoted, then additional command-line arguments may be given.
If the external protocol does not contain a colon (:), it is assumed to have the standard format "protocol://[usr[:password]@]address[:port]/path". If it ends with a colon, then it is assumed to have the simpler format "protocol:path". If it ends with "://" then the standard form is again assumed.
If the external protocol does not contain a colon (:), it is assumed to have the standard format "protocol://[usr[:password]@]address[:port]/path". If it ends with a colon, then it is assumed to have the simpler format "protocol:path". If it ends with "://" then the standard form is again assumed.
The program takes three command-line parameters, not counting any parameters already given in the command string:
protocol URL configuration-file
Parameter Description Example
protocol The URL scheme to be used. https
URL The URL to be retrieved. https://www.htdig.org:8008/attrs.html
configuration-file The configuration-file in effect. /etc/htdig/htdig.conf

The external protocol script is to write information for htdig on the standard output. The output must follow the form described here. The output consists of a header followed by a blank line, followed by the contents of the document. Each record in the header is terminated with a newline. Each record is a series of (unless expressively allowed to be empty) non-empty tab-separated fields. The first field is a single character that specifies the record type. The rest of the fields are determined by the record type.

Record type Fields Description
s status code An HTTP-style status code, e.g. 200, 404. Typical codes include:
200
Successful retrieval
304
Not modified (for example, if the document hasn't changed)
301
Redirect (to another URL)
401
Not authorized
404
Not found
r reason A text string describing the status code, e.g "Redirect" or "Not Found."
m status code The modification time of this document. While the code is fairly flexible about the time/date formats it accepts, it is recommended to use something standard, like RFC1123: Sun, 06 Nov 1994 08:49:37 GMT, or ISO-8601: 1994-11-06 08:49:37 GMT.
t content-type A valid MIME type for the document, like text/html or text/plain.
l content-length The length of the document on the server, which may not necessarily be the length of the buffer returned.
u url The URL of the document, or in the case of a redirect, the URL that should be indexed as a result of the redirect.
_ These characters are considered part of a word. In contrast to the characters in the valid_punctuation attribute, they are treated just like letter characters.
Note that the locale attribute is normally used to configure which characters constitute letter characters. false true This option works only if we take advantage of persistent connections (see persistent_connections attribute). If set to true an HTTP/1.1 HEAD call is made in order to retrieve header information about a document. If the status code and the content-type returned let the document be parsable, then a following 'GET' call is made. 5 20 This is a factor which will be used to multiply the weight of words between <h1> and </h1> tags, as well as headings of levels <h2> through <h6>. It is used to assign the level of importance to headings. Setting a factor to 0 will cause words in these headings to be ignored. The number may be a floating point number. See also author_factor backlink_factor caps_factor date_factor description_factor keywords_factor meta_description_factor text_factor title_factor url_text_factor ${common_dir}/notify_prefix.txt Specifies the file containing text to be inserted in each mail message sent by htnotify before the list of expired webpages. If omitted, nothing is inserted. design-group@foo.com This specifies the email address that htnotify email messages include in the Reply-to: field. webmaster@www bigboss@yourcompany.com This specifies the email address that htnotify email messages get sent out from. The address is forged using /usr/lib/sendmail. Check htnotify/htnotify.cc for detail on how this is done. ${common_dir}/notify_suffix.txt Specifies the file containing text to be inserted in each mail message sent by htnotify after the list of expired webpages. If omitted, htnotify will insert a standard message. ht://Dig Notification Service Notification Service This provides a name for the From field, in addition to the email address for the email messages sent out by htnotify. 3128 When this attribute is set, all HTTP document retrievals will be done using the HTTP-PROXY protocol. The URL specified in this attribute points to the host and port where the proxy server resides.
The use of a proxy server greatly improves performance of the indexing process. mypassword This tells htdig to send the supplied username:password with each HTTP request, when using a proxy with authorization requested. The credentials will be encoded using the "Basic" authentication scheme. There must be a colon (:) between the username and password. //intranet.foo.com/ When this is set, URLs matching this will not use the proxy. This is useful when you have a mixture of sites near to the digging server and far away. false true If set, this causes the text of the ALT field in an <IMG...> tag not to be indexed as part of the text of the document, nor included in excerpts. true false Determines whether htdig will continue to index URLs from a server after an attempted connection to the server fails as "no host found" or "host not found (port)." If set to false, htdig will try every URL from that server. ${database_base}.images allimages This is the file that a list of image URLs gets written to by htdig when the create_image_list is set to true. As image URLs are seen, they are just appended to this file, so after htdig finishes it is probably a good idea to run sort -u on the file to eliminate duplicates from the file. IMAGE_URL_PREFIX /images/htdig This specifies the directory portion of the URL used to display star images. This attribute isn't directly used by htsearch, but is used in the default URL for the star_image and star_blank attributes, and other attributes may be defined in terms of this one.

The default value of this attribute is determined at compile time.

${config_dir}/htdig.conf This is not quite a configuration attribute, but rather a directive. It can be used within one configuration file to include the definitions of another file. The last definition of an attribute is the one that applies, so after including a file, any of its definitions can be overridden with subsequent definitions. This can be useful when setting up many configurations that are mostly the same, so all the common attributes can be maintained in a single configuration file. The include directives can be nested, but watch out for nesting loops. false true This sets whether dates should be output in ISO 8601 format. For example, this was written on: 1998-10-31 11:28:13 EST. See also the date_format attribute, which can override any date format that htsearch picks by default.
This attribute also affects the format of the date htnotify expects to find in a htdig-notification-date field. documentation Keywords which must be found on all pages returned, even if the "or" ("Any") method is selected. 100 12 This is a factor which will be used to multiply the weight of words in the list of keywords of a document. The number may be a floating point number. See also the heading_factorattribute. keywords htdig-keywords keywords description The words in this list are used to search for keywords in HTML META tags. This list can contain any number of strings that each will be seen as the name for whatever keyword convention is used.
The META tags have the following format: <META name="somename" content="somevalue"> //www.mydomain.com This specifies a set of patterns that all URLs have to match against in order for them to be included in the search. Unlike the limit_urls_to attribute, this is done after the URL is normalized and the server_aliases attribute is applied. This allows filtering after any hostnames and DNS aliases are resolved. Otherwise, this attribute is the same as the limit_urls_to attribute. ${start_url} .sdsu.edu kpbs [.*\.html] This specifies a set of patterns that all URLs have to match against in order for them to be included in the search. Any number of strings can be specified, separated by spaces. If multiple patterns are given, at least one of the patterns has to match the URL.
Matching, by default, is a case-insensitive string match on the URL to be used, unless the case_sensitive attribute is set. The match will be performed after the relative references have been converted to a valid URL. This means that the URL will always start with http://.
Granted, this is not the perfect way of doing this, but it is simple enough and it covers most cases. index.html default.html default.htm index.html index.htm Set this to the default documents in a directory used by the server. This is used for local filesystem access to translate URLs like http://foo.com/ into something like /home/foo.com/index.html
The list should only contain names that the local server recognizes as default documents for directory URLs, as defined by the DirectoryIndex setting in Apache's srm.conf, for example. As of version 3.1.5, this can be a string list rather than a single name, and htdig will use the first name that works. Since this requires a loop, setting the most common name first will improve performance. Special characters can be embedded in these names using %xx hex encoding. //www.foo.com/=/usr/www/htdocs/ Set this to tell ht://Dig to access certain URLs through local filesystems. At first ht://Dig will try to access pages with URLs matching the patterns through the filesystems specified. If it cannot find the file, or if it doesn't recognize the file name extension, it will try the URL through HTTP instead. Note the example--the equal sign and the final slashes in both the URL and the directory path are critical.
The fallback to HTTP can be disabled by setting the local_urls_only attribute to true. To access user directory URLs through the local filesystem, set local_user_urls. The only file name extensions currently recognized for local filesystem access are .html, .htm, .txt, .asc, .ps, .eps and .pdf. For anything else, htdig must ask the HTTP server for the file, so it can determine the MIME content-type of it. As of version 3.1.5, you can provide multiple mappings of a given URL to different directories, and htdig will use the first mapping that works. Special characters can be embedded in these names using %xx hex encoding. For example, you can use %3D to embed an "=" sign in an URL pattern. false true Set this to tell ht://Dig to access files only through the local filesystem, for URLs matching the patterns in the local_urls or local_user_urls attribute. If it cannot find the file, it will give up rather than trying HTTP or another protocol. //www.my.org/=/home/,/www/ Set this to access user directory URLs through the local filesystem. If you leave the "path" portion out, it will look up the user's home directory in /etc/password (or NIS or whatever). As with local_urls, if the files are not found, ht://Dig will try with HTTP or the appropriate protocol. Again, note the example's format. To map http://www.my.org/~joe/foo/bar.html to /home/joe/www/foo/bar.html, try the example below.
The fallback to HTTP can be disabled by setting the local_urls_only attribute to true. As of version 3.1.5, you can provide multiple mappings of a given URL to different directories, and htdig will use the first mapping that works. Special characters can be embedded in these names using %xx hex encoding. For example, you can use %3D to embed an "=" sign in an URL pattern. C en_US Set this to whatever locale you want your search database cover. It affects the way international characters are dealt with. On most systems a list of legal locales can be found in /usr/lib/locale. Also check the setlocale(3C) man page. Note that depending the locale you choose, and whether your system's locale implementation affects floating point input, you may need to specify the decimal point as a comma rather than a period. This will affect settings of search_algorithm and any of the scoring factors. false true This sets whether htsearch should use the syslog() to log search requests. If set, this will log requests with a default level of LOG_INFO and a facility of LOG_LOCAL5. For details on redirecting the log into a separate file or other actions, see the syslog.conf(5) man page. To set the level and facility used in logging, change LOG_LEVEL and LOG_FACILITY in the include/htconfig.h file before compiling.
Each line logged by htsearch contains the following:
REMOTE_ADDR [config] (match_method) [words] [logicalWords] (matches/matches_per_page) - page, HTTP_REFERER
where any of the above are null or empty, it either puts in '-' or 'default' (for config). bogus@unconfigured.htdig.user ben.dover@uptight.com This should be the email address of the person in charge of the digging operation. This string is added to the user-agent: field when the digger sends a request to a server. and boolean This is the default method for matching that htsearch uses. The valid choices are:
  • or
  • and
  • boolean
This attribute will only be used if the HTML form that calls htsearch didn't have the method value set. 10 999 If this is set to a relatively small number, the matches will be shown in pages instead of all at once. This attribute will only be used if the HTML form that calls htsearch didn't have the matchesperpage value set. -1 100 This attribute tells htdig to limit the number of requests it will send to a server using a single, persistent HTTP connection. This only applies when the persistent_connections attribute is set. You may set the limit as high as you want, but it must be at least 1. A value of -1 specifies no limit. Requests in the queue for a server will be combined until either the limit is reached, or the queue is empty. 60 40 While gathering descriptions of URLs, htdig will only record those descriptions which are shorter than this length. This is used mostly to deal with broken HTML. (If a hyperlink is not terminated with a </a> the description will go on until the end of the document.) 5 15 While gathering descriptions of URLs, htdig will only record up to this number of descriptions, in the order in which it encounters them. This is used to prevent the database entry for a document from growing out of control if the document has a huge number of links to it. 100000 5000000 This is the upper limit to the amount of data retrieved for documents. This is mainly used to prevent unreasonable memory consumption since each document will be read into memory by htdig. 1 10 This value determines the maximum number of excerpts that can be displayed for one matching document in the search results. 512 50000 For each document retrieved, the top of the document is stored. This attribute determines the size of this block. The text that will be stored is only the text; no markup is stored.
We found that storing 50,000 bytes will store about 95% of all the documents completely. This really depends on how much storage is available and how much you want to show. 999999 4 Instead of limiting the indexing process by URL pattern, it can also be limited by the number of hops or clicks a document is removed from the starting URL.
The starting page or pages will have hop count 0. -1 10 This attribute can be used to limit the number of keywords per document that htdig will accept from meta keywords tags. A value of -1 or less means no limit. This can help combat meta keyword spamming, by limiting the amount of keywords that will be indexed, but it will not completely prevent irrelevant matches in a search if the first few keywords in an offending document are not relevant to its contents. 512 1000 While gathering descriptions from meta description tags, htdig will only store up to this much of the text for each document. 1000 100 The Prefix fuzzy algorithm could potentially match a very large number of words. This value limits the number of words each prefix can match. Note that this does not limit the number of documents that are matched in any way. 3 6 This option set the maximum number of retries when retrieving a document fails (mainly for reasons of connection). 4 6 When stars are used to display the score of a match, this value determines the maximum number of stars that can be displayed. ${maximum_pages} 20 This value limits the number of page links that will be included in the page list at the bottom of the search results page. By default, it takes on the value of the maximum_pages attribute, but you can set it to something lower to allow more pages than buttons. In this case, pages above this number will have no corresponding button. 10 20 This value limits the number of page links that will be included in the page list at the bottom of the search results page. As of version 3.1.4, this will limit the total number of matching documents that are shown. You can make the number of page buttons smaller than the number of allowed pages by setting the maximum_page_buttons attribute. 32 15 This sets the maximum length of words that will be indexed. Words longer than this value will be silently truncated when put into the index, or searched in the index. ${database_base}.md5hash.db ${database_base}.md5.db This file holds a database of md5 and date hashes of pages to catch and eliminate duplicates of pages. See also the check_unique_md5 and check_unique_date attributes. 50 20 This is a factor which will be used to multiply the weight of words in any META description tags in a document. The number may be a floating point number. See also the heading_factor attribute and the description_factor attribute. ${database_base}.metaphone.db ${database_base}.mp.db The database file used for the fuzzy "metaphone" search algorithm. This database is created by htfuzzy and used by htsearch. and All or Any boolean Boolean or Or and And These values are used to create the method menu. It consists of pairs. The first element of each pair is one of the known methods, the second element is the text that will be shown in the menu for that method. This text needs to be quoted if it contains spaces. See the select list documentation for more information on how this attribute is used. ${config_dir}/mime.types /etc/mime.types This file is used by htdig for local file access and resolving file:// URLs to ensure the files are parsable. If you are running a webserver with its own MIME file, you should set this attribute to point to that file. 1 2 This sets the minimum length of prefix matches used by the "prefix" fuzzy matching algorithm. Words shorter than this will not be used in prefix matching. 5 3 This sets the minimum length of words used by the "speling" fuzzy matching algorithm. Words shorter than this will not be used in this fuzzy matching. 3 2 This sets the minimum length of words that will be indexed. Words shorter than this value will be silently ignored but still put into the excerpt.
Note that by making this value less than 3, a lot more words that are very frequent will be indexed. It might be advisable to add some of these to the bad_word_list. 1 1000 This factor gives higher rankings to documents that have more than one matching search word when the or match_method is used. In version 3.1.6, the matching words' combined scores were multiplied by this factor for each additional matching word. Currently, this multiplier is applied at most once. [next] <img src="/htdig/buttonr.gif"> The text displayed in the hyperlink to go to the next page of matches. false yes If no excerpt is available, this option will act the same as excerpt_show_top, that is, it will show the top of the document. <em>(None of the search words were found in the top of this document.)</em> This text will be displayed in place of the excerpt if there is no excerpt available. If this attribute is set to nothing (blank), the excerpt label will not be displayed in this case. [next] The text displayed where there would normally be a hyperlink to go to the next page of matches. <hr noshade size=2>All results on this page.<br> This text will be used as the value of the PAGEHEADER variable, for use in templates or the search_results_footer file, when all search results fit on a single page. <strong>1</strong> <strong>2</strong> \ <strong>3</strong> <strong>4</strong> \ <strong>5</strong> <strong>6</strong> \ <strong>7</strong> <strong>8</strong> \ <strong>9</strong> <strong>10</strong> The text strings in this list will be used when putting together the PAGELIST variable, for use in templates or the search_results_footer file, when search results fit on more than page. The PAGELIST is the list of links at the bottom of the search results page. There should be as many strings in the list as there are pages allowed by the maximum_page_buttons attribute. If there are not enough, or the list is empty, the page numbers alone will be used as the text for the links. An entry from this list is used for the current page, as the current page is shown in the page list without a hypertext link, while entries from the page_number_text list are used for the links to other pages. The text strings can contain HTML tags to highlight page numbers or embed images. The strings need to be quoted if they contain spaces. [prev] The text displayed where there would normally be a hyperlink to go to the previous page of matches. filename "No Title Found" This specifies the text to use in search results when no title is found in the document itself. If it is set to filename, htsearch will use the name of the file itself, enclosed in brackets (e.g. [index.html]). <!--/htdig_noindex--> </SCRIPT> This string marks the end of a section of an HTML file that should be completely ignored when indexing. It works together with noindex_start. As in the defaults, this can be SGML comment declarations that can be inserted anywhere in the documents to exclude different sections from being indexed. However, existing tags can also be used; this is especially useful to exclude some sections from being indexed where the files to be indexed can not be edited. The example shows how SCRIPT sections in 'uneditable' documents can be skipped. Note that the match for this string is case insensitive. <!--htdig_noindex--> <SCRIPT This string marks the start of a section of an HTML file that should be completely ignored when indexing. It works together with noindex_end. As in the defaults, this can be SGML comment declarations that can be inserted anywhere in the documents to exclude different sections from being indexed. However, existing tags can also be used; this is especially useful to exclude some sections from being indexed where the files to be indexed can not be edited. The example shows how SCRIPT sections in 'uneditable' documents can be skipped; note how noindex_start does not contain an ending >: this allows for all SCRIPT tags to be matched regardless of attributes defined (different types or languages). Note that the match for this string is case insensitive. ${common_dir}/nomatch.html /www/searching/nothing.html This specifies the file which contains the HTML text to display when no matches were found. The file should contain a complete HTML document.
Note that this attribute could also be defined in terms of database_base to make is specific to the current search database. false true This attribute determines whether htsearch sends out full HTTP headers as required for an NPH (non-parsed header) CGI. Some servers assume CGIs will act in this fashion, for example MS IIS. If your server does not send out full HTTP headers, you should set this to true. <hr noshade size=2>Pages:<br> This text will be used as the value of the PAGEHEADER variable, for use in templates or the search_results_footer file, when all search results fit on more than one page. " " "</td> <td>" The text strings in this list will be used when putting together the PAGELIST variable, for use in templates or the search_results_footer file, when search results fit on more than page. The PAGELIST is the list of links at the bottom of the search results page. The strings in the list will be used in rotation, and will separate individual entries taken from page_number_text and no_page_number_text. There can be as many or as few strings in the list as you like. If there are not enough for the number of pages listed, it goes back to the start of the list. If the list is empty, a space is used. The text strings can contain HTML tags. The strings need to be quoted if they contain spaces, or to specify an empty string. <em>1</em> <em>2</em> \ <em>3</em> <em>4</em> \ <em>5</em> <em>6</em> \ <em>7</em> <em>8</em> \ <em>9</em> <em>10</em> The text strings in this list will be used when putting together the PAGELIST variable, for use in templates or the search_results_footer file, when search results fit on more than page. The PAGELIST is the list of links at the bottom of the search results page. There should be as many strings in the list as there are pages allowed by the maximum_page_buttons attribute. If there are not enough, or the list is empty, the page numbers alone will be used as the text for the links. Entries from this list are used for the links to other pages, while an entry from the no_page_number_text list is used for the current page, as the current page is shown in the page list without a hypertext link. The text strings can contain HTML tags to highlight page numbers or embed images. The strings need to be quoted if they contain spaces. true false If set to true, when servers make it possible, htdig can take advantage of persistent connections, as defined by HTTP/1.1 (RFC2616). This permits to reduce the number of open/close operations of connections, when retrieving a document with HTTP. s en Specifies the value of the PLURAL_MATCHES template variable used in the header, footer and template files. This can be used for localization for non-English languages where 's' is not the appropriate suffix. * ing A null prefix character means that prefix matching should be applied to every search word. Otherwise a match is returned only if the word does not end in the characters specified. [prev] <img src="/htdig/buttonl.gif"> The text displayed in the hyperlink to go to the previous page of matches. 25 10 The "regex" fuzzy algorithm could potentially match a very large number of words. This value limits the number of words each regular expression can match. Note that this does not limit the number of documents that are matched in any way. true true If TRUE, htpurge will remove any URLs which were marked as unreachable by htdig from the database. If FALSE, it will not do this. When htdig is run in initial mode, documents which were referred to but could not be accessed should probably be removed, and hence this option should then be set to TRUE, however, if htdig is run to update the database, this may cause documents on a server which is temporarily unavailable to be removed. This is probably NOT what was intended, so hence this option should be set to FALSE in that case. index.html default.html default.htm index.html index.htm Set this to the default documents in a directory used by the servers you are indexing. These document names will be stripped off of URLs when they are normalized, if one of these names appears after the final slash, to translate URLs like http://foo.com/index.html into http://foo.com/
Note that you can disable stripping of these names during normalization by setting the list to an empty string. The list should only contain names that all servers you index recognize as default documents for directory URLs, as defined by the DirectoryIndex setting in Apache's srm.conf, for example. This only applies to http:// and https:// URLS. false true If TRUE, htpurge will remove any URLs which were discovered and included as stubs in the database but not yet retrieved. If FALSE, it will not do this. When htdig is run in initial mode with no restrictions on hopcount or maximum documents, these should probably be removed and set to true. However, if you are hoping to index a small set of documents and eventually get to the rest, you should probably leave this as false. //www.acme.com/widgets/ This specifies a set of patterns that all URLs have to match against in order for them to be included in the search results. Any number of strings can be specified, separated by spaces. If multiple patterns are given, at least one of the patterns has to match the URL. The list can be specified from within the configuration file, and can be overridden with the "restrict" input parameter in the search form. Note that the restrict list does not take precedence over the exclude list - if a URL matches patterns in both lists it is still excluded from the search results. htdig myhtdig Sets the name that htdig will look for when parsing robots.txt files. This can be used to make htdig appear as a different spider than ht://Dig. Useful to distinguish between a private and a global index. /search/results.shtml Overrides the value of the SCRIPT_NAME environment attribute. This is useful if htsearch is not being called directly as a CGI program, but indirectly from within a dynamic .shtml page using SSI directives. Previously, you needed a wrapper script to do this, but this configuration attribute makes wrapper scripts obsolete for SSI and possibly for other server scripting languages, as well. (You still need a wrapper script when using PHP, though.)
Check out the contrib/scriptname directory for a small example. Note that this attribute also affects the value of the CGI variable used in htsearch templates. exact:1 0.3 Specifies the search algorithms and their weight to use when searching. Each entry in the list consists of the algorithm name, followed by a colon (:) followed by a weight multiplier. The multiplier is a floating point number between 0 and 1. Note that depending on your locale setting, and whether your system's locale implementation affects floating point input, you may need to specify the decimal point as a comma rather than a period.
Note:If the exact method is not listed, the search may not work since the original terms will not be used.
Current algorithms supported are:
exact
The default exact word matching algorithm. This will find only exactly matched words.
soundex
Uses a slightly modified soundex algorithm to match words. This requires that the soundex database be present. It is generated with the htfuzzy program.
metaphone
Uses the metaphone algorithm for matching words. This algorithm is more specific to the english language than soundex. It requires the metaphone database, which is generated with the htfuzzy program.
accents
Uses the accents algorithm for matching words. This algorithm will treat all accented letters as equivalent to their unaccented counterparts. It requires the accents database, which is generated with the htfuzzy program.
endings
This algorithm uses language specific word endings to find matches. Each word is first reduced to its word root and then all known legal endings are used for the matching. This algorithm uses two databases which are generated with htfuzzy.
synonyms
Performs a dictionary lookup on all the words. This algorithm uses a database generated with the htfuzzy program.
substring
Matches all words containing the queries as substrings. Since this requires checking every word in the database, this can really slow down searches considerably.
prefix
Matches all words beginning with the query strings. Uses the option prefix_match_character to decide whether a query requires prefix matching. For example "abc*" would perform prefix matching on "abc" since * is the default prefix_match_character.
regex
Matches all words that match the patterns given as regular expressions. Since this requires checking every word in the database, this can really slow down searches considerably.
speling
A simple fuzzy algorithm that tries to find one-off spelling mistakes, such as transposition of two letters or an extra character. Since this usually generates just a few possibilities, it is relatively quick.
${common_dir}/footer.html /usr/local/etc/ht/end-stuff.html This specifies a filename to be output at the end of search results. While outputting the footer, some variables will be expanded. Variables use the same syntax as the Bourne shell. If there is a variable VAR, the following will all be recognized:
  • $VAR
  • $(VAR)
  • ${VAR}
The following variables are available. See hts_template.html for a complete list.
MATCHES
The number of documents that were matched.
PLURAL_MATCHES
If MATCHES is not 1, this will be the string "s", else it is an empty string. This can be used to say something like "$(MATCHES) document$(PLURAL_MATCHES) were found"
MAX_STARS
The value of the max_stars attribute.
LOGICAL_WORDS
A string of the search words with either "and" or "or" between the words, depending on the type of search.
WORDS
A string of the search words with spaces in between.
PAGEHEADER
This expands to either the value of the page_list_header or no_page_list_header attribute depending on how many pages there are.
Note that this file will NOT be output if no matches were found. In this case the nothing_found_file attribute is used instead. Also, this file will not be output if it is overridden by defining the search_results_wrapper attribute. ${common_dir}/header.html /usr/local/etc/ht/start-stuff.html This specifies a filename to be output at the start of search results. While outputting the header, some variables will be expanded. Variables use the same syntax as the Bourne shell. If there is a variable VAR, the following will all be recognized:
  • $VAR
  • $(VAR)
  • ${VAR}
The following variables are available. See hts_template.html for a complete list.
MATCHES
The number of documents that were matched.
PLURAL_MATCHES
If MATCHES is not 1, this will be the string "s", else it is an empty string. This can be used to say something like "$(MATCHES) document$(PLURAL_MATCHES) were found"
MAX_STARS
The value of the max_stars attribute.
LOGICAL_WORDS
A string of the search words with either "and" or "or" between the words, depending on the type of search.
WORDS
A string of the search words with spaces in between.
Note that this file will NOT be output if no matches were found. In this case the nothing_found_file attribute is used instead. Also, this file will not be output if it is overridden by defining the search_results_wrapper attribute. /docs/|faq.html * /maillist/ /testresults/ This specifies a list of patterns for URLs in search results. Results will be displayed in the specified order, with the search algorithm result as the second order. Remaining areas, that do not match any of the specified patterns, can be placed by using * as the pattern. If no * is specified, one will be implicitly placed at the end of the list.
See also url_seed_score. ${common_dir}/wrapper.html This specifies a filename to be output at the start and end of search results. This file replaces the search_results_header and search_results_footer files, with the contents of both in one file, and uses the pseudo-variable $(HTSEARCH_RESULTS) as a separator for the header and footer sections. If the filename is not specified, the file is unreadable, or the pseudo-variable above is not found, htsearch reverts to the separate header and footer files instead. While outputting the wrapper, some variables will be expanded, just as for the search_results_header and search_results_footer files.
Note that this file will NOT be output if no matches were found. In this case the nothing_found_file attribute is used instead. http://(.*)\\.mydomain\\.org/([^/]*) http://\\2.\\1.com \ http://www\\.myschool\\.edu/myorgs/([^/]*) http://\\1.org This is a list of pairs, regex replacement, used to rewrite URLs in the search results. The left hand string is a regular expression; the right hand string is a literal string with embedded placeholders for fragments that matched inside brackets in the regular expression. \0 is the whole matched string, \1 to \9 are bracketted substrings. The backslash must be doubled-up in the attribute setting to get past the variable expansion parsing. Rewrite rules are applied sequentially to each URL before it is displayed or checked against the restrict or exclude lists. Rewriting does not stop once a match has been made, so multiple rules may affect a given URL. See also url_part_aliases which allows URLs to be of one form during indexing and translated for results, and url_rewrite_rules which allows URLs to be rewritten while indexing. foo.mydomain.com:80=www.mydomain.com:80 \ bar.mydomain.com:80=www.mydomain.com:80 This attribute tells the indexer that servers have several DNS aliases, which all point to the same machine and are NOT virtual hosts. This allows you to ensure pages are indexed only once on a given machine, despite the alias used in a URL. As shown in the example, the mapping goes from left to right, so the server name on the right hand side is the one that is used. As of version 3.1.3, the port number is optional, and is assumed to be 80 if omitted. There is no easy way to map all ports from one alias to another without listing them all. -1 50 This attribute tells htdig to limit the dig to retrieve a maximum number of documents from each server. This can cause unusual behavior on update digs since the old URLs are stored alphabetically. Therefore, update digs will add additional URLs in pseudo-alphabetical order, up to the limit of the attribute. However, it is most useful to partially index a server as the URLs of additional documents are entered into the database, marked as never retrieved.
A value of -1 specifies no limit. 0 20 This attribute tells htdig to ensure a server has had a delay (in seconds) from the beginning of the last connection. This can be used to prevent "server abuse" by digging without delay. It's recommended to set this to 10-30 (seconds) when indexing servers that you don't monitor yourself. Additionally, this attribute can slow down local indexing if set, which may or may not be what you intended. score revtime This is the default sorting method that htsearch uses to determine the order in which matches are displayed. The valid choices are:
  • score
  • time
  • title
  • revscore
  • revtime
  • revtitle
This attribute will only be used if the HTML form that calls htsearch didn't have the sort value set. The words date and revdate can be used instead of time and revtime, as both will sort by the time that the document was last modified, if this information is given by the server. The default is to sort by the score, which ranks documents by best match. The sort methods that begin with "rev" simply reverse the order of the sort. Note that setting this to something other than "score" will incur a slowdown in searches. score Score time Time title Title revscore 'Reverse Score' revtime 'Reverse Time' revtitle 'Reverse Title' score 'Best Match' time Newest title A-Z \ revscore 'Worst Match' revtime Oldest revtitle Z-A These values are used to create the sort menu. It consists of pairs. The first element of each pair is one of the known sort methods, the second element is the text that will be shown in the menu for that sort method. This text needs to be quoted if it contains spaces. See the select list documentation for more information on how this attribute is used. ${database_base}.soundex.db ${database_base}.snd.db The database file used for the fuzzy "soundex" search algorithm. This database is created by htfuzzy and used by htsearch. ${image_url_prefix}/star_blank.gif //www.somewhere.org/icons/noelephant.gif This specifies the URL to use to display a blank of the same size as the star defined in the star_image attribute or in the star_patterns attribute. ${image_url_prefix}/star.gif //www.somewhere.org/icons/elephant.gif This specifies the URL to use to display a star. This allows you to use some other icon instead of a star. (We like the star...)
The display of stars can be turned on or off with the use_star_image attribute and the maximum number of stars that can be displayed is determined by the max_stars attribute.
Even though the image can be changed, the ALT value for the image will always be a '*'. http://www.sdsu.edu /sdsu.gif \ http://www.ucsd.edu /ucsd.gif This attribute allows the star image to be changed depending on the URL or the match it is used for. This is mainly to make a visual distinction between matches on different web sites. The star image could be replaced with the logo of the company the match refers to.
It is advisable to keep all the images the same size in order to line things up properly in a short result listing.
The format is simple. It is a list of pairs. The first element of each pair is a pattern, the second element is a URL to the image for that pattern. 1 Day component of first date allowed as last-modified date of returned docutments. This is most usefully specified as a GCI argument. See also startyear. <strong><code>... </code></strong> ... When excerpts are displayed in the search output, this string will be prepended to the excerpt if there is text before the text displayed. This is just a visual reminder to the user that the excerpt is only part of the complete document. <strong> <font color="#FF0000"> When excerpts are displayed in the search output, matched words will be highlighted using this string and end_highlight. You should ensure that highlighting tags are balanced, that is, any formatting tags that this string opens should be closed by end_highlight. 1 Month component of first date allowed as last-modified date of returned docutments. This is most usefully specified as a GCI argument. See also startyear. http://www.htdig.org/ //www.somewhere.org/alldata/index.html This is the list of URLs that will be used to start a dig when there was no existing database. Note that multiple URLs can be given here.
Note also that the value of start_url will be the default value for limit_urls_to, so if you set start_url to the URLs for specific files, rather than a site or subdirectory URL, you may need to set limit_urls_to to something less restrictive so htdig doesn't reject links in the documents. 1970 2001 This specifies the year of the cutoff start date for search results. If the start or end date are specified, only results with a last modified date within this range are shown. See also startday, startmonth, endday, endmonth, endyear. These are most usefully specified as a GCI argument.
For each component, if a negative number is given, it is taken as relative to the current date. Relative days can span several months or even years if desired, and relative months can span several years. A startday of -90 will select matching documents modified within the last 90 days. 25 100 The Substring fuzzy algorithm could potentially match a very large number of words. This value limits the number of words each substring pattern can match. Note that this does not limit the number of documents that are matched in any way. ${common_dir}/synonyms.db ${database_base}.syn.db Points to the database that htfuzzy creates when the synonyms algorithm is used.
htsearch uses this to perform synonym dictionary lookups. ${common_dir}/synonyms /usr/dict/synonyms This points to a text file containing the synonym dictionary used for the synonyms search algorithm.
Each line of this file has at least two words. The first word is the word to replace, the rest of the words are synonyms for that word. ${common_dir}/syntax.html ${common_dir}/synerror.html This points to the file which will be displayed if a boolean expression syntax error was found. 1 6 This option set the maximum number of attempts when a connection timeouts. After all these retries, the connection attempt results <timed out>. 5 10 This attribute sets the wait time (in seconds) after a connection fails and the timeout is raised. Long builtin-long builtin-long Short builtin-short builtin-short Short short ${common_dir}/short.html \ Normal normal builtin-long \ Detailed detail ${common_dir}/detail.html This maps match template names to internal names and template file names. It is a list of triplets. The first element in each triplet is the name that will be displayed in the FORMAT menu. The second element is the name used internally and the third element is a filename of the template to use.
There are two predefined templates, namely builtin-long and builtin-short. If the filename is one of those, they will be used instead.
More information about templates can be found in the htsearch documentation. The particular template is selecterd by the format cgi argument, and the default is given by template_name in the config file. builtin-long long Specifies the default template if no format field is given by the search form. This needs to map to the template_map. http://www.sdsu.edu ${common_dir}/sdsu.html \ http://www.ucsd.edu ${common_dir}/ucsd.html This attribute allows the results template to be changed depending on the URL or the match it is used for. This is mainly to make a visual distinction between matches on different web sites. The results for each site could thus be shown in a style matching that site.
The format is simply a list of pairs. The first element of each pair is a pattern, the second element is the name of the template file for that pattern.
More information about templates can be found in the htsearch documentation.
Normally, when using this template selection method, you would disable user selection of templates via the format input parameter in search forms, as the two methods were not really designed to interact. Templates selected by URL patterns would override any user selection made in the form. If you want to use the two methods together, see the notes on combining them for an example of how to do this. 1 0 This is a factor which will be used to multiply the weight of words that are not in any special part of a document. Setting a factor to 0 will cause normal words to be ignored. The number may be a floating point number. See also the heading_factor attribute. 30 42 Specifies the time the digger will wait to complete a network read. This is just a safeguard against unforeseen things like the all too common transformation from a network to a notwork.
The timeout is specified in seconds. 100 12 This is a factor which will be used to multiply the weight of words in the title of a document. Setting a factor to 0 will cause words in the title to be ignored. The number may be a floating point number. See also the heading_factor attribute. ${database_base}.urls /tmp/urls This file is only created if create_url_list is set to true. It will contain a list of all URLs that were seen. ${database_base}.log /tmp/htdig.progress If htdig is run with the -l option and interrupted, it will write out its progress to this file. Note that if it has a large number of URLs to write, it may take some time to exit. This can especially happen when running update digs and the run is interrupted soon after beginning. http://search.example.com/~htdig *site \ http://www.htdig.org/this/ *1 \ .html *2 http://www.htdig.org/ *site \ http://www.htdig.org/that/ *1 \ .htm *2 A list of translations pairs from and to, used when accessing the database. If a part of an URL matches with the from-string of each pair, it will be translated into the to-string just before writing the URL to the database, and translated back just after reading it from the database.
This is primarily used to provide an easy way to rename parts of URLs for e.g. changing www.example.com/~htdig to www.htdig.org. Two different configuration files for digging and searching are then used, with url_part_aliases having different from strings, but identical to-strings.
See also common_url_parts.
Strings that are normally incorrect in URLs or very seldom used, should be used as to-strings, since extra storage will be used each time one is found as normal part of a URL. Translations will be performed with priority for the leftmost longest match. Each to-string must be unique and not be a part of any other to-string.
Note that when this attribute is changed, the database should be rebuilt, unless the effect of "moving" the affected URLs in the database is wanted, as described above.
Please note: Don't just copy the example below into a single configuration file. There are two separate settings of url_part_aliases below; the first one is for the configuration file to be used by htdig, htmerge, and htnotify, and the second one is for the configuration file to be used by htsearch. (.*)\\?JServSessionIdroot=.* \\1 \ (.*)\\&JServSessionIdroot=.* \\1 \ (.*)&context=.* \\1 This is a list of pairs, regex replacement used to permanently rewrite URLs as they are indexed. The left hand string is a regex; the right hand string is a literal string with embedded placeholders for fragments that matched inside brackets in the regex. \0 is the whole matched string, \1 to \9 are bracketted substrings. Rewrite rules are applied sequentially to each incoming URL before normalization occurs. Rewriting does not stop once a match has been made, so multiple rules may affect a given URL. See also url_part_aliases which allows URLs to be of one form during indexing and translated for results. /mailinglist/ *.5-1e6 /docs/|/news/ *1.5 /testresults/ "*.7 -200" /faq-area/ *2+10000 This is a list of pairs, pattern formula, used to weigh the score of hits, depending on the URL of the document.
The pattern part is a substring to match against the URL. Pipe ('|') characters can be used in the pattern to concatenate substrings for web-areas that have the same formula.
The formula describes a factor and a constant, by which the hit score is weighed. The factor part is multiplied to the original score, then the constant part is added.
The format of the formula is the factor part: "*N" optionally followed by comma and spaces, followed by the constant part : "+M", where the plus sign may be emitted for negative numbers. Either part is optional, but must come in this order.
The numbers N and M are floating point constants.
More straightforward is to think of the format as "newscore = oldscore*N+M", but with the "newscore = oldscore" part left out. 1 1 TO BE COMPLETED
See also heading_factor. false true If set to true, htdig will use META date tags in documents, overriding the modification date returned by the server. Any documents that do not have META date tags will retain the last modified date returned by the server or found on the local file system. false true If set to true, any META description tags will be used as excerpts by htsearch. Any documents that do not have META descriptions will retain their normal excerpts. true no If set to true, the star_image attribute is used to display upto max_stars images for each match. htdig htdig-digger This allows customization of the user_agent: field sent when the digger requests a file from a server. .html .htm .shtml This is a list of extensions on URLs which are the only ones considered acceptable. This list is used to supplement the MIME-types that the HTTP server provides with documents. Some HTTP servers do not have a correct list of MIME-types and so can advertise certain documents as text while they are some binary format. If the list is empty, then all extensions are acceptable, provided they pass other criteria for acceptance or rejection. If the list is not empty, only documents with one of the extensions in the list are parsed. See also bad_extensions. .-_/!#$%^&' -' This is the set of characters which will be deleted from the document before determining what a word is. This means that if a document contains something like Andrew's the digger will see this as Andrews.
The same transformation is performed on the keywords the search engine gets.
See also the extra_word_characters attribute. VERSION 3.2.0 This specifies the value of the VERSION variable which can be used in search templates. The default value of this attribute is determined at compile time, and will not normally be set in configuration files. ${database_base}.words.db ${database_base}.allwords.db This is the main word database. It is an index of all the words to a list of documents that contain the words. This database can grow large pretty quickly. ${database_base}.worddump /tmp/words.txt This file is basically a text version of the file specified in word_db. Its only use is to have a human readable database of all words. The file is easy to parse with tools like perl or tcl. 10000000 40000000 Size of memory cache used by Berkeley DB (DB used by the indexer) IMPORTANT: It makes a huge difference. The rule is that the cache size should be at least 2% of the expected index size. The Berkeley DB file has 1% of internal pages that *must* be cached for good performances. Giving an additional 1% leaves room for caching leaf pages. true true Enables or disables the default compression system for the indexer. This currently compresses the index by a factor of 8. If the Zlib library is not found on the system, the default is false. true true Enables or disables the zlib compression system for the indexer. wordlist_compress must be true to use this option!` false true This enables monitoring of what's happening in the indexer. It can help to detect performance/configuration problems. 0 .1 Sets the number of seconds between each monitor output. myfile Print monitoring output on file instead of the default stderr. 0 8192 Size of pages used by Berkeley DB (DB used by the indexer) true wordlist_verbose 1 walk logic
wordlist_verbose 2 walk logic details
wordlist_verbose 2 walk logic lots of details
 Word/DocID 32/Flags 8/Location 16 DATA