Search

Prerequisite for a search in the cache archives with the WebAssistant is an indexation.

The search is available:

with the URL http://127.0.0.1:8080/search/
with the Command ls.
In the search bar of the browser (Firefox 2, Internet Explorer 7)
Add for this at the search page the search plug-in to your browser.

At first you select a cache archive. Make sure that only indexed archives are available. You can search for words, domains and URLs. Several search criterias are combined with the operation AND.

Search Terms

Search for words

Examples:
Input	Search for pages which … the word.
archiv	Exactly contain: archiv
webpage*	Start with: webpage
*website	End with: website
mirror	Contain: mirror

Search for a domain

You in addition indicate the keyword: site

Examples:
Input	Search for pages which … the characters in the domain name.
site:mm3tools	Exactly contain: mm3tools
site:proxy*	Start with: proxy
site:*browser	End with: browser
site:offline	Contain: offline

Search in a part of URL

You in addition indicate the keyword: url

Examples
Input	Search for pages which … the characters in the URL.
url:download	Contain: download

Output of a Search

The result of a search is displayed as a hitlist. The files (pages) are listed with their URL, size, date of archiving as well as 200 characters.
Text files are marked by TXT in addition.
The title and the description are reported to HTML files in addition.
The sequence of the files corresponds to the alphabetical sort of URL. Several files from the same domain are reported intentionally. Files with a red archiving date were actualized after construction of their index. About the link Marker the page is displayed with highlighted Search-Words. Marking isn't possible for all files.

Information about the Index

Word Histogram

The histogram displays a sorting of the words and the number of the files in which the corresponding word occurs.

For an alphabetical sort you use keyword: wordAlphabetical

Examples:
Input	Histogram about words which … the characters.
wordAlphabetical:archiv	Exactly contain: archiv
wordAlphabetical:webpage*	Start with: webpage
wordAlphabetical:*website	End with: website
wordAlphabetical:mirror	Contain: mirror
wordAlphabetical:*	Contain arbitrary characters (all words)

For a sorting after frequency you use the keyword: wordFrequency
For a sorting after word length you use the keyword: wordLength

Domain Histogram

The histogram displays an alphabetical sort of the domains and the number of the files which are included in the domain.

Therefore you use the keyword: siteAlphabetical

Examples:
Input	Histogram about domain which domain names … the characters.
siteAlphabetical:mm3tools	Exactly contain: mm3tools
siteAlphabetical:proxy*	Start with: proxy
siteAlphabetical:*browser	End with: browser
siteAlphabetical:offline	Contain: offline
siteAlphabetical:*	Contain arbitrary characters (all domain names)

For a sorting after frequency you use the keyword: siteFrequency

Indexing

The search in the cache archives with the WebAssistant presupposes an indexation. It becomes indexed text and HTML files (pages). The algorithm of the Indexer, works essentially language independent. At this the corresponding lower case characters are always used for capital characters. The Latin and Russian alphabet as well as some special signs of European languages are supported.
Please, inform Tools, if you need another language.

Script file

You start the update with one of the following script files in the folder: MM3-WebAssistantProfessional/script/

Script	Operation System
MM3-Utility.bat	Windows of Microsoft
MM3-Utility.sh	Linux and UNIX
MM3-Utility.command	Mac OS X of Apple

In the first dialog all utilities are displayed.

Select: Create an index for the search about an cache archive
With Next you get the configuration dialog: Indexer

Configuration of the Indexer

For the indexing you can set the following configuration:

Select the cache archive to be indexed
Specification of the minimal word length.
Only words which have a minimal word length are included into the indexing. Simplified, this word length consists of the characters of a word.
Display of the positive and negative word list
- Negative word list
  These words aren't included into the index.
  Stop words for English, Russian and German are existing.
  If you have created additional stop words, please inform us about these.
- Positive word list
  These words are taken despite fall below the minimal word length.
The corresponding files are in the files positive.*.txt and negative.*.txt of the folder MM3-WebAssistantProfessional/config/search/. You can adapt the word lists to your need. The characters * stands for a language specific word list, e.g. en for the English and de for the German language. All files with a name structured correspondingly are used. We recommend for the identification of the language to use the abbreviations to ISO LanguageCode (ISO-639).

You start the indexation after you have done the settings. The needed duration is dependent on the size of the archives. The indexation can take up some time. Please, close the WebAssistant before indexation.

Log output

You can take from the output of the Indexer:

Indexed cache archives
Number of the file still to be indexed
At the moment indexed domain
Time needed till now
Progress bar
Summarizing statistics about the indexation

Starting with a command line

You also can start the Indexer with the following command line:

java -jar MM3-WebAssistant.jar Indexer cacheActive=D:\CacheArchiv\ minWordLength=2 withNumber=yes start

Out of Memory

The needed memory is dependent on the size of the archives and the chosen minimal word length. You can increase the available memory for the Indexer in the script file, if the indexation needs more memory. You can alternatively subdivide the cache archive into several archives or increase the minimal word length.