DEPRECATION WARNING

This documentation is not using the current rendering mechanism and is probably outdated. The extension maintainer should switch to the new system. Details on how to use the rendering mechanism can be found here.

Indexed Search Reference¶

Author:	Kasper Skårhøj
Created:	2002-11-01T00:32:00
Changed by:	Michael Stucki
Changed:	2008-12-27T10:04:45
Classification:	doc_indexed_search
Keywords:	indexed search, reference, forDevelopers, forAdvanced
Author:	TYPO3 Core Development Team
Email:	info@typo3.org
Info 4:
Language:	en

Indexed Search Reference - doc_indexed_search

Indexed Search Reference¶

Extension Key: doc_indexed_search

Language: en

Keywords: indexed search, reference, forDevelopers, forAdvanced

This document is published under the Open Content License

available from http://www.opencontent.org/opl.shtml

The content of this document is related to TYPO3

- a GNU/GPL CMS/Framework available from www.typo3.org

Table of Contents¶

Indexed Search Reference 1

`Introduction 3 <#1.1.Introduction|outline>`_

`User manual 5 <#1.2.User%20manual|outline>`_

`Adminstration 7 <#1.3.Adminstration|outline>`_

`Indexing configurations 10 <#1.4.Indexing%20configurations|outline>`_

`Configuration 21 <#1.5.Configuration|outline>`_

`Technical details 23 <#1.6.Technical%20details|outline>`_

`Analysing the indexed data 25 <#1.7.Analysing%20the%20indexed%20data|outline>`_

`Database Tables 30 <#1.8.Database%20Tables|outline>`_

`Known problems 33 <#1.9.Known%20problems|outline>`_

Introduction¶

What does it do?¶

The Indexed Search Engine provides two major elements to TYPO3:

Indexing: An indexing engine which indexes TYPO3 pages on-the-fly as they are rendered by TYPO3's frontend. Indexing a page means that all words from the page (or specifically defined areas on the page) are registered, counted, weighted and finally inserted into a database table of words. Then another table will be filled with relation records between the word table and the page. This is the basic idea.

Searching: A plugin you can insert on your website which allows website users to search for information on your website. By searching the plugin first looks in the word-table if the word exist and if it does all pages which has a relation to that word will be considered for the search result display. The search results are ordered based on factors like where on the page the word was found or the frequency of the word on the page.

This is an example of how the search interface on a website looks:

Features of the indexer¶

The indexing engine has several features:

HTML data priority: 1) <title>-data 2) <meta-keywords>, 3) <meta- description>, 4) <body>
Indexing external files: Text formats like html and txt and doc, pdf by external programs (catdoc / pdftotext)
Wordcounting and frequency used to rate results
Exact, partially or metaphone search
Searching freely for sentences (non-indexed).
NOT case-sensitive in any ways though.

Features of the search frontend (the plugin)¶

The search interface has several options for advanced searching. Any of those can be disabled and/or preset with default values:

Searching whole word, part of word, sounds like, sentence
Logical AND and OR search including syntactical recognition of AND, OR and NOT as logical keywords. Furthermore sentences encapsulated in quotes will be recognized.
Searching can be targeted at specific media, for instance searching only indexed PDF files, HTML-files, Word-files, TYPO3-pages or everything
The engine is language sensitive based on the multiple-language feature of TYPO3's CMS frontend.
Searching can be performed in specific sections of the website.
Results can be sorted descending or ascending and ordered by word frequency, weight, location relative to page top, page modification date, page title, etc.
The display of search results can be intelligently divided into sections based on the internal page hierarchy. Thus results are primarily grouped by relation, then by hit-relevance.

This shows the full range of default options for “advanced search”:

Warning¶

The search frontend plugin is optimized for features, not speed. Especially it will be slow on a website with many pages in the page tree because it traverses the whole tree each time to build a list of accessible pages. However you can circumvent this by modifications to the search plugin so it does not check page access based on the id- list. But then you loose that feature of course. Can't have both.

In any case; The indexing of pages and searching the indexed information are two different processes and therefore you can easily use another frontend plugin for making searches in the same data for whatever reason you might have for discarding the default search plugin.

User manual¶

Adding the search plugin to a page¶

That is really easy:

Create a page called “Search” or something like that. This is where the search box will appear.

Then create a new content element on that page. From the Web>Page module you can do it like this:

Then select some plugin-type if you can. It doesn't matter if it's a guestbook or forum. Or if no plugins are available, just select a “Regular text element” as in the top of the page.

Then make sure “Insert plugin” is selected (if not, select it and save the element, then you'll see the form below), enter a title and select the “Plugin” type to be “Indexed search”:

Then select the root page of your website as the “Starting point” of the plugin content element:

And that's it. Your frontend should now look like this:

The styles are most likely different from this, but that is controlled by the developer having administration access to the system.

Adminstration¶

Monitoring indexed content¶

The Indexed Search extension adds two backend modules, one as a global database-wide statistics module and a page specific analysis module.

In the Web>Info module you can see an overview of how many instances are indexed per TYPO3-page. Look at this image:

As you can see most pages here are indexed only one time. However a few are indexed twice. This can happen for several reasons and here the reason is most likely due to a user login or something related.

The most interesting occurence is the page “References” which has more than 20 indexed instances available. The reason is that this page holds multiple cached views due to some parameters which are used by a plugin on that page. Each instance will be searchable as a unique search result.

Now imagine that you want to clear out all those instances of the “References” page to let them be re-indexed when viewed again; Simply click the page “References” in the page tree to the left. Then you see this:

You can either click the red garbage bin (1) in order to clear all listed instances or alternatively pick out single instances by clicking the local garbage bin (2).

Monitoring the global picture of indexed pages¶

By the Tools>Indexing module you can get statistics about the indexing engine. Currently they are sparse and very roughly presented. This view needs some more work to be friendly and really useful.

“General statistics”¶

This shows that 217 pages are indexed, comprising 7000+ words and using 40.000 records in the relation table to glue things together.

“List: TYPO3 Pages”¶

This view shows a list of indexed pages with all the technical details:

Indexing configurations¶

Setting up the “crawler” extension¶

Before you can work with “Indexing configurations” you must make sure you have set up the “crawler” extension and have a cron-job running that will process the crawler queue as we fill it! For this, please refer to the documentation of the “crawler” extension!

Generally about indexing configurations¶

Indexing configuration sets up indexing jobs that are performed by a cron-script independently of frontend requests. The “crawler” extension is used as a service to perform the execution of queue entries that controls the indexing.

The Indexing configuration contains two parts

Definition of execution time and periodicality.

Definition of indexing type and settings.

Below you see what all Indexing Configurations have in common:

These settings are described in the context sensitive help so please refer to that for more information.

The “Session ID” requires a show introduction: When an indexing job is started it will set this value to a unique number which is used as ID for that process and all indexed entries are tagged with it. When the processing of an indexing configuration is done it will be reset to zero again.

Periodic indexing of the website (“Page tree”)¶

You can have the whole page tree indexed overnight using this indexing configuration of type “Page tree”:

This defines that the page tree is to be crawled to a depth of 3 levels from the root point “Testsite”. For each page a combination of parameters is calculated based on the “crawler” configurations for the “Re-index” processing instruction (See “crawler” extension for more information) and those URLs are committed to the crawler log plus entries for all subpages to the processed page (so that each of those pages are indexed as well.)

This is what the crawler log may look like after processing:

Here you can notice that the visited URLs have additional parameters added - those are combined based on the “crawler” extensions configuration in Page TSconfig.

Also notice the special crawler log entries found in the “Storage folder”. These are the “meta-entries” which calls an indexed search hook which in turn generates the URL entries and pushed them to the queue.

On the far right in this view you can see that noted as well, including the “set_id”:

Finally, in the Web>Info, “Indexed search” you will see that these visited URLs were re-indexed:

Location: Indexing configurations for indexing of the page tree should be placed in a SysFolder since their location in the page tree is not relevant to their function.

Periodic indexing of records (“Database Records”)¶

You can also use the Indexing Configuration to index single records.

Location: You must place the indexing configuration on the page where you want the search results to be displayed - typically on the page where a plugin exists that can process the parameters pointing to the record. In the case below the Indexing Configuration is placed on the same page as the frontend plugin (“Morbi diam enim...”) that can display the search results:

The configuration record looks like this:

If the records you want to index is not located on the page where the indexing configuration and fronend plugin is, then you can point to the location. Notice how the field with “GET parameters” is used to define how the search results are shown - this must correspond with what the plugin takes of parameters.

A fancy option is the “Index Records immediately when saved” - which will index records as they are saved through “TCEmain”!

In the crawler log you will see the entries for record indexing like this:

After processing the Web>Info, “Indexed search” view will show this view:

Notice how the GET parameters are nicely added and how the “CfgUid” column contains the UID of the indexing configuration / the “set_id” of the processing.

In fact, if a record is removed its indexing entry will also be removed upon next indexing - simply because the “set_id” is used to finally clear out old entries after a re-index!

Indexing External websites (“External URL”)¶

You can index external websites using Indexing Configurations. They can actually crawl an external URL! Configuration looks like this:

It pretty much explains itself how it works. The Context Sensitive Help will provide enough information to complete configuration.

Location: You should place the Indexing Configuration on a “Not- in-menu” page in the root of the site for instance. The page must be “searchable” since the external URL results are bound to a page in the page tree, namely the page where the configuration is found.

This is how the crawler log looks immediately after the crawling has begun:

The initial entry is “http://typo3.org/” which is already processed. When this process was executed it added entries for all found subpages to the queue as well. When their execution time comes the crawler will request those URLs as well and if subpages are found on them, entries for those subpages are added until the configured depth is reached.

After a few minutes you see more entries processed like this:

In Web>Info, “Indexed search” the indexed entries looks like this:

Indexing directories of files (“Filepath on server”)¶

You can also have directories of files on your server indexed periodically, using the type “Filepath on server”.

Again, the options are either easy to understand or your can read more about them in the Context Sensitive Help.

Location: The Indexed Search configuration should be located on a not- in-menu page, just like the “External URL” type required. Same reasons; results are bound to a page in the page tree.

The process of indexing a directory of files is the same as for the external URL: For each directory a) all files are indexed and b) all sub-directories added to the crawler queue for later processing. This is shown in the crawler log:

When processing is done the result is shown in the Web>Info, “Indexed search”:

Showing the search results¶

By default the search results are shown with no distinction between those from local TYPO3 pages, records indexed, the file path and external URLs. Only division follows that of the page on which the result is found:

However, you can configure to have a division of the search results into categories following the indexing configurations:

To obtain this categorization you must set TypoScript configuration in the Setup field like this:

plugin.tx_indexedsearch.search.defaultFreeIndexUidList = 0,6,7,8
plugin.tx_indexedsearch.blind.freeIndexUid = 0

The “defaultFreeIndexUidList” is uid numbers of indexing configurations to show in the categorization! The order determines which are shown in top. Changing it could bring results from TYPO3.org and TYPO3.com in top:

The categorization happens when the “Category” selector in the “Advanced” search form is set like this:

(Notice, you can preset this value from TypoScript as well!)

Searching a specific category from URL

If you want search forms on the site to make look up directly in results belonging to one or more indexing configurations you can use a set or GET variables like these, here using UID values 7 and 8 since they look up in TYPO3.org and TYPO3.com results:

index.php?id=78&tx_indexedsearch[sword]=level&tx_indexedsearch[_freeIndexUid]=7,8

Grouping more indexing configurations in one search category¶

You might find that you want to group the results from multiple indexing configurations in the same category. For instance, I have an indexing configuration for both “TYPO3.org” and “TYPO3.com” but I want all search results to appear under the category “External URLs”. This can be done by creating a special type of indexing configuration which only points to other indexing configurations:

This indexing configuration is not used during indexing but during searching. So a reconfiguration of the TypoScript to use uid 9 instead of 7,8 will yield this result:

TypoScript:

plugin.tx_indexedsearch.search.defaultFreeIndexUidList = 9,6,0

Disable frontend initiated indexing¶

If you choose to index your site using Indexing Configurations you can disable indexing through the user requests in the frontend. This is easily done via the configuration of the Indexed Search extension in the Extension Manager:

Indexing files on pages separately¶

If enabled, links to local files found on pages will initiate indexing of those external files. However, this often has the unpleasant effect that too many files are indexed during the same page request. Using the crawler extension you can configure the indexer to add a queue entry instead of immediate indexing of external files. Thus the indexing will happen outside the frontend user request, using the cronscript!

This behaviour is configured in the extension managers configuration for “Indexed search”:

Configuration¶

General¶

The most basic requirement for the search engine to work is that pages are getting indexed. That will not happen by just installing the plugin! You will have to set up in TypoScript that a certain page should be indexed. That is needed for several good reasons. First of all not all sites in a TYPO3 database might need indexing. So therefore we disable it on a per-site basis. Secondly a single site may have frames and in that case we need only index the page-object which actually shows the page content.

Lets say that you have a PAGE object called "page" (that is pretty typical), then you will have to set this config-option:

page.config.index_enable = 1

When this option is set you should begin to see your pages being indexed when they are shown next time. Remember that only cached pages are indexed!

This is documented in TSref in the CONFIG section. Please look there for further options. For instance indexing of external media can also be enabled there.

Languages¶

The plugin supports all system languages in TYPO3. Translation is done using the typo3.org tools.

If you want to use eg. danish language that will automatically be used if this option is set in your template (the value is the internal language key):

config.language = dk

TypoScript¶

[Still missing the major parts here. Just use the object browser for now since that includes all options]

Property

search.rootPidList

Data type

list of int

Description

A list of integer which should be root-pages to search from. Thus you can search multiple branches of the page tree by setting this property to a list of page id numbers.

If this value is set to less than zero (eg. -1) searching will happen in ALL of the page tree with no regard to branches at all.

Notice that by “root-page” we mean a website root defined by

a TypoScript Template! If you just want to search in branches of your site, use the possibility of searching in levels.

Default

The current root-page id

search.detect_sys_domain_records¶

Property

search.detect_sys_domain_records

Data type

boolean

Description

If set, then the search results are linked to the proper domains where they are found.

Default

search.detect_sys_domain_records.target¶

Property

search.detect_sys_domain_records.target

Data type

string

Description

Target for external URLs.

Default

search.mediaList¶

Property

search.mediaList

Data type

string

Description

Restrict the file type list when searching for files.

Property

specConfs.[pid]

Data type

Description

“specConfs” is an array of objects with properties that can customize certain behaviours of the display of a result row depending on it's position in the rootline. For instance you can define that all results which links to pages in a branch from page id 123 should have another page icon displayed. Of you can add a suffix to the class names so you can style that section differently.

Examples:

If a page “Contact” is found in a search for “address” and that “Contact” page is in the rootline “Frontpage [ID=23] > About us [ID=45] > Contact [ID=77]” then you should set the pid value to either “77” or “45”. If “45” then all subpages including the “About us” page will have similar configuration.

If the pid value is set to 0 (zero) it will apply to all pages.

Please see the options below.

Default

specConfs.[pid].pageIcon¶

Property

specConfs.[pid].pageIcon

Data type

->IMAGE cObject

Description

Alternative page icon.

Default

specConfs.[pid].CSSsuffix¶

Property

specConfs.[pid].CSSsuffix

Data type

string

Description

A string that will be appended to the class-names of all the class- attributes used within the result row presentation. The prefix will be like this:

Example:

If “...CSSsuffix = doc” then eg. the class name “tx-indexedsearch- title” will be “tx-indexedsearch-title-doc”

Default

whatis_stdWrap¶

Property

whatis_stdWrap

Data type

->stdWrap

Description

Parse input through the stdWrap function

Default

[tsref:plugin.tx_indexedsearch]

Technical details¶

HTML content¶

HTML content is weighted by the indexing engine in this order:

<title>-data

<meta-keywords>

<meta-description>

<body>

In addition you can insert markers as HTML comments which define which part of the body-text to include or exclude in the indexing:

The marker is  or 

Rules:

If there is no marker at all, everything is included.

If the first found marker is an “end” marker, the previous content until that point is included and the preceeding code until next “begin” marker is excluded.

If the first found marker is a “begin” marker, the previous content until that point is excluded and preceeding content until next “end” marker is included.

Use of hashes¶

The hashes used are md5 hashes where the first 7 chars are converted into an integer which is used as the hash in the database. This is done in order to save space in the database, thus using only 4 bytes and not a varchar of 32 bytes. It's estimated that a hash of 7 chars (32) is sufficient (originally 8, but at some point PHP changed behavior with hexdec-function so that where originally a 32 bit value was input half the values would be negative, they were suddenly positive all of them. That would require a similar change of the fields in the database. To cut it simple, the length was reduced to 7, all being positive then).

How pages are indexed¶

First of all a page must be cachable. For pages where the cache is disabled, no indexing will occur.

The “phash” is a unique identification of a “page” with regard to the indexer. So an entry in the index_phash table equals 1 resultrow in the search-results (called a phash-row).

A phash is a combination of the page-id, type, sys_language id, gr_list, MP and the cHash parameters of the page (function setT3Hashes()). If the phash is made for EXTERNAL media (item_type > 0) then it's a combination of the absolute filename hashes with any “subpage” indication, for instance if a PDF-document is splitted into subsections.

So for external media there is one phash-row for each file (except PDF-files where there may be more). But for TYPO3-pages there can be more phash-rows matching one single page. Obviously the type-parameter would normally always be only one, namely the type-number of the content page. And the cHash may be of importance for the result as well with regard to plugins using that. For instance a message board may make pages cachable by using the cHash params. If so, each cached page will also be indexed. Thus many phash-rows for a single page-id.

But the most tricky reason for having multiple phash-rows for a single TYPO3-page id is if the gr_list is set! This works like this: If a page has exactly the same content both with and without logins, then it's stored only once! If the page-content differs whether a user is logged in or not - it may even do so based on the fe_groups! - then it's indexed as many times as the content differs. The phash is of course different, but the phash_grouping value is the same.

The table index_grlist will always hold one record per phash-row (of item_type=0, that is TYPO3 pages). But it may also hold many more records. These point to the phash-row in question in the case of other gr_list combinations which actually had the SAME content - and thus refers to the same phash-row.

External media¶

External media (pdf, doc, html, txt) is tricky. External media is always detected as links to local files in the content of a TYPO3 page which is being indexed. But external media can the linked to from more than one page. So the index_section table may hold many entries for a single external phash-record, one for each position it's found. Also it's important to notice that external media is only indexed or updated if a “parent” TYPO3 page is re-indexed. Only then will the links to the external files be found. In a searching operation external media will be listed only once (grouping by phash), but say two TYPO3 pages are linking to the document, then only one of them will be shown as the path where the link can be found. However if both TYPO3 pages are not available, then the document will not be shown.

Access restricted pages¶

A TYPO3 page will always be available in the search result only if there is access to the page. This is secured in the final result query. Whether extendToSubpages is taken into account depends on the join_pages-flag (see above). But the page will only be listed if the user has access.

However a page may be indexed more than once if the content differs from usergroup to usergroup or just without login. Still the result display will display only one occurrence, because similar pages (determined based on phash_grouping) will be detected.

The tricky scenario:

Say that a page has a content element with some secret information visible for only one usergroup. The page as a whole will be visible for all users. The page will be indexed twice - both without login and with login because page content differs. The problem is that if a search is conducted and matching one of the secret words in the access restricted section, then the page will be in the search result even if the user is not logged in!

The best solution to this problem is to allow the result to be listed anyway, but then HIDE the resume if the index_grlist table cannot confirm positively that the combination of usergroups of the user has access to the result. So the result is there, but no resume shown (The resume might contain hidden text).

External media

Equally for external media they are linked from a TYPO3 page. When an external media is selected we can be sure that the page linking to it can be selected. But we cannot be sure that the link was in a section accessible for the user. Similarly we should make a lookup in the index_grlist table selecting the phash/gr_list by the phash_t3-value of the section record for the search-result. If this is not available we should not display a link to the document and not show resume, but rather link to the page, from which the user can see the real link to the document.

Note:

These tricky scenarios exist only if the content on a page differs based on login. It does not affect situations with access restriction to the page as a whole. A general lesson from this is to reduce the number of hidden content elements! Instead use hidden pages. Better, more reliable.

Analysing the indexed data¶

The indexer is constructed to work with TYPO3's page structure. Opposite to a crawler which simply indexes all the pages it can find, the TYPO3 indexer MUST take the following into account:

Only cached pages can be indexed.Pages with dynamic content - such as search pages etc - should supply their own search engine for lookup in specific tables. Another option is to selectively allow certain of those “dynamic” pages to be cached anyways (see the cHashParams concept used by some plugins)
Pages in more than one language must be indexed separately as “different pages”.
Pages with messageboards may have multiple indexed versions based on what is displayed on the page: The overview or a single messageboard item? This is determined by the cHashParams value.
Pages with access restricted to must be observed!
Because pages can contain different content whether a user is logged in or not and even based on which groups he is a member of, a single page (identified by the combination of id/type/language/cHashParams) may even be available in more than one indexed version based on the user-groups. But while the same page may have different content based on the user-groups (and so must be indexed once for each) such pages may just as well present the SAME content regardless of usergroups! This is the very most tricky thing.

Understanding these complex scenarios...¶

The best thing to do is to grab an example. Please refer to the picture below while reading the bulletlist here:

The overview in general shows one line per “phash-row” (a single row from the index_phash table). Such a row represents a single hit in a searching session. In other words, each line with grayish background in the overview may be a search-hit. The columns of these rows are:

Title: The search-result title.
[icon]: Click here to remove the indexed information for this entry (will be re-indexed on the next hit).
pHash: The “id” of the search row. The hash is calculated based on id/type/language/MP/cHashParams/gr_list of the page when indexed. For external media this is based on filepath/page-interval (for PDF's only)
cHash: Calculated based on the actual content which was indexed.
rl-012: This is the rootline ids for level 0,1,2. Used when searching in certain sections. For instance a search-operation may select all pages with rl1=123 which will result in a search within pages which exist ONLY in the branch of the website where the level1-page has uid=123.
pid.t.l: This is the page-id / type-number / sys_language uid
Size: How many bytes the indexed page consumed
grlist: This is the gr_list of the user which initiated the indexing operation.
cHashParams: Additional parameters which are identifying the page in addition to the id/type number which usually does that.

The page “Content elements” has one indexed version. The page-id of the root-page is “1” and the page on level-1 in the rootline had the uid “2”. Notice how all subpages to “Content elements” has the exact same rl0 and rl1 value. Where the page “Content elements” does NOT have a value for rl2 so does all the subpages (because they ARE the level 2 themselves!). Furthermore the page has the page-id “2”, a type-value of “0” and is indexed with the default language “0”. The size was 10.6 KB and the user who initiated the indexing operation was a member of the groups 0,-2,1 (which is effectively fe_group “1”, because 0 and -2 is pseudogroups).

On the page “Special content” there must have been a link to a local PDF and Word file, since those two are indexed in relation to this page. The PDF-file is located in the path “uploads/media/tsref_onepage.pdf” relative to the website. Notice that the PDF file is actually indexed three times, one time per page. This is of course configurable. Each indexed section of the PDF-file has the potential to show up as a search-result row of course (because the phash is different per indexed part). The whole point with this is that a large PDF file might contain so much information that it might match all too many search-queries. So breaking a PDF-file down into smaller parts makes it possible for us to indicate exactly WHERE in the PDF-file the search word was found!

Looking at the word file (and the PDF-file as well) we see that they are found on BOTH the page “Special content” and on the page “ISEARCH example”. But looking at the phash values (for the word-file it is “268192666”) it is the SAME value in both cases. So this means, that the Word and PDF file is indexed only once - when it is first discovered! Later when another page is indexed and a link to the same document appears, then the document is not indexed as another document, but rather an entry in the index_section table is made indicating that this result row is also found available (linked to) from another page/section.Say you are doing a search in the section from “Content elements” and outwards in the page tree. The word- document is matched in the search, but it will appear only once in the search result. Now, if one of the two pages where the Word document was either hidden or access restricted the word-document would still be matched (because one of the pages is accessible for the user). But if BOTH pages with the link to the word document is not accessible for the user doing the search, then the word document will not be included in the search result.

Here we can see that the pages “Special content”, “Advanced” and “Menu/Sitemap” is indexed twice each. The reason is that those three pages has had different content depending on whether or not a user was logged in!In the case of the page “Special content” the reason is that the page contained a content element which was visible for users which was a member of group number #1. Therefore the page was different in the two cases.The page “Advanced” has a user-login form and that form looks different whether a user is logged in or not.Finally the page “Menu/Sitemap” apparently changed. There reason was that this page includes a sitemap and that sitemap displayed some extra pages when the logged in users hit the page and so the content was not the same as without login.Another thing which is interesting is that two different users must have visited those pages. We can see that because the page “Special content” was apparently indexed with the usergroup combination “1,2”. Later another user hit the page but only a member of group “1”. However the page content was the SAME. And because those two users saw the very same page, it was not indexed a third time, but it was instead noted down that a user with membership of only group “1” did also see this same page. That comparison was based on the cHash (contentHash) which is a hash-value based on the actual content being indexed. So when the user with group “1” only came to the page, the indexer engine realize that the page as it looked has already been indexed because another phash-row with that content hash was already available.

These pages does not contain any tricks it appears. According to the grlist's both users with membership of group “1,2” and group “1” only as well as surfers who did not at all login (“0,-1” is the pseudo- group for no login) as visited the page. And because only one indexed version exist the page must have had the same content to present all users regardless of their login-status.The reason why the page “Your own scripts” does not contain a grlist value “0,-2,1,2” as the others do is simply because no user with that combination of usergroups has ever visited the page!

txt and html documents can also be indexed as external media. In the case of HTML-documents the documents <title> is detected and used.

On the image below we are looking at another scenario. In this case the cHashParams is obviously used by the plugin “tt_board”. The plugin has been constructed so intelligently that it links to the messages in the message board without disabling the normal page-cache but rather sending the tt_board_uid parameter along with a so called “cHash”. If this is combined correctly the caching engine allows the page to be cached. Not only does this mean a quicker display of pages in the message board - it also means we can index the page!

As you see the main board page showing the list of messages/threads (“Sourcream and Oni...”) is indexed without any values for the parameter tt_board_uid (the cHashParams field is blank). Then it has also been indexed one time for each display of a message. In a search result any of these five rows may appear as an independent result row - after all they are to be regarded as a single page with unique content, despite sharing the same page-id!

Another interesting thing is that while the main page has inherited the page title for the search-result (“Sourcream and ...”) each of the indexed pages with a message has got another title - namely the subject line of the message shown! Thus a search matching three of these five pages will not shown three similar page-titles but a unique page title relative to the actual content on the page. It is the tt_board plugin that sets the page-title itself by an API-call.

The only glitch here is that the tt_board plugin has falsely allowed the main page to be cached twice. See the first and last phash-row. The last row has got the parameter “&tt_board_uid= “ sent and the tt_board plugin should not have allowed that! Because looking at the content hash of the first and last we realize that it's the SAME hash (84186444) and therefore the SAME content! However being two separate result rows they will both be displayed in the search result as separate hits. The responsibility for this lies with the plugin. However such occurrences can be automatically filtered out during the search result display. But it's better to avoid this kind of stuff.

The last example below has three main issues to discuss:

The page “Other languages” is apparently available in three languages. Which ones are not possible to determine unless we know the value from the sys_languages table. In this case the default language (zero - 0) is english and the language with id 1 and id 2 is danish and german versions of the page.When a search is conducted each page may turn up as a result page but with a little flag telling if the page was found in another language than the main language on the website (see second illustration hereafter)

If there is no phash-rows found for a page this can mean three things:

Either the page is not cached. In this case both the tt_products and tt_news plugins apparently disables the caching of the page thereby disabling any indexing of the pages. Searching in news and products must be done with a searching function looking up directly in the news and products tables.

In the case with other pages the reason may be that the pages has never been visited and therefore not indexed yet! Indexing of pages in TYPO3 happens during the rendering of the page - there is currently no “crawler” to assist this job.

Finally the reason for a page not being indexed can be the combination of 1 and 2: That the page has never been visited. And if it was visited, the cache would have been disabled.

These numbers just tells us that:

the page “Lists” was indexed once by a user with membership of group 1 and 2.
the page “Addresses” was also indexed by a user with membership of group 1 and 2 but has since been visited by a user without login. Both instances yielded a similar page and it was therefore not indexed twice.This raises the question about the page “Lists”: Is that access- restricted for users without login or has a user without login just never visited that page since no “0,-1” grlist has been detected? Both could be the answer. On pages which has access-restriction (or a whole section in an intranet) such pages would obviously not have been indexed by no-login users. However in this case nothing indicates that the page should be hidden for non-login users and so we must conclude that the page has simply not yet been visited by a no-login user - otherwise it would look like the page “Addresses” having also the “0,-1” list detected.
The “Guestbook” page was indexed by a user without login only.

Illustration 1A seach result showing how localized versions of a page are displayed.

Database Tables¶

index_phash¶

This table contains references to TYPO3 pages or external documents. The fields are like this:

phash¶

a

phash

b

7md5/int hash. It's an integer based on a 7-char md5-hash.

This is a unique representation of the 'page' indexed.

For TYPO3 pages this is a serialization of id,type,gr_list (see later), MP and cHashParams (which enables 'subcaching' with extra parameters). This concept is also used for TYPO3 caching (although the caching hash includes the all-array and thus takes the template into account, which this hash does not! It's expected that template changes through conditions would not seriously alter the page content)

For external media this is a serialization of 1) unique filename id, 2) any subpage indication (parallel to cHashParams). gr_list is NOT taken into consideration here!

phash_grouping¶

a

phash_grouping

b

7md5/int hash.

This is a non-unique hash exactly like phash, but WITHOUT the gr_list and (in addition) for external media without subpage indication. Thus this field will indicate a 'unique' page (or file) while this page may exist twice or more due to gr_list. Use this field to GROUP BY the search so you get only one hit per page when selecting with gr_list in mind.

Currently a seach result does not either group or limit by this, but rather the result display may group the result into logical units.

item_mtime¶

a

item_mtime

b

Modification time:

For TYPO3 pages: the SYS_LASTCHANGED value

For external media: The filemtime() value.

Depending on config, if mtime hasn't changed compared to this value the file/page is not indexed again.

tstamp¶

a

tstamp

b

time stamp of the indexing operation. You can configure min/max ages which are checked with this timestamp.

A min-age defines how long an indexed page must be indexed before it's reconsidered to index it again.

A max-age defines an absolute point at which re-indexing will occur (unless the content has not changed according to an md5-hash)

cHashParams¶

a

cHashParams

b

The cHashParams.

For TYPO3 pages: These are used to re-generate the actual url of the TYPO3 page in question

For files this is an empty array. Not used.

item_type¶

a

item_type

b

An integer indicating the content type,

0 is TYPO3 pages

1- external files like pdf (2), doc (3), html (1), txt (4) and so on. See the class.indexer.php file

item_title¶

a

item_title

a

gr_list

b

Contains the gr_list of the user initiating the indexing of the document.

index_section¶

Points out the section where an entry in index_phash belongs.

phash¶

a

phash

b

The phash of the indexed document.

phash_t3¶

a

phash_t3

b

For free text searching, eg with a sentence, in all content: title, description, keywords, body

phash¶

a

phash

b

The phash of the indexed document.

fulltextdata¶

a

fulltextdata

b

The total content stripped for any HTML codes.

Currently the MySQL FULLTEXT search is not used (something with MATCH ... AGAINST), but this will be added in the future.

index_grlist¶

flags

b

Bits, which describes the weight of the words:

8th bit (128) = word found in title,

7th bit (64) = word found in keywords,

6th bit (32) = word found in description,

Last 5 bits are not used yet, but if used they will enter the weight hierarchy. The result rows are ordered by this value if the “Weight/Frequency” sorting is selected. Thus results with a hit in the title, keywords or description are ranked higher in the result list.

Known problems¶

Currently the extension is under observation because instances of heavy server load/unstability has been reported. It is not yet clear if THIS extension has anything to do with. So it's only under suspicion at this point until further data has been collected. But for now it is adviced to be careful with the application of the extension for mission critical, high-load environments.
It's still uncertain how performance is under heavy load conditions and when MANY pages are indexed. Currently benchmarks has been done only up to 2000 pages indexed/approx. 400.000 relation records. It is probably that some parts has to be optimized for such scenarios.

33