Помогите настроить полнотекстовый поиск

[mepsy]

Пытаюсь настроить полнотекстовый локальный поиск в wheesy xfce. Чтобы не устанавливать Tracker и KDE, загрузил и установил Sphinx 2.2.4.  Не могу его никак настроить. К настоящему моменту есть две проблемы, с которыми никак не могу справиться: 1. Редактирование файла /etc/sphinxsearch/sphinx.conf: устанваливаю разные значения - 100, 1000 - max_matches, но sphinx всё время предупреждает о том, что "key 'max_matches' was permanently removed from Sphinx configuration" и отсылает к неким документам (Refer to documentation for details).
Вывод команды sudo indexer --all

Открыть содержимое (спойлер)

Sphinx 2.2.4-id64-release (r4806)
Copyright (c) 2001-2014, Andrew Aksyonoff
Copyright (c) 2008-2014, Sphinx Technologies Inc (http://sphinxsearch.com)
using config file '/etc/sphinxsearch/sphinx.conf'...
WARNING: key 'max_matches' was permanently removed from Sphinx configuration. Refer to documentation for details.
indexing index 'test1'...
ERROR: index 'test1': sql_connect: \0414\043E\0441\0442\0443\043F \0437\0430\043A\0440\044B\0442 \0434\043B\044F \043F\043E\043B\044C\0437\043E\0432\0430\0442\0435\043B\044F 'test'@'localhost' (\0431\044B\043B \0438\0441\043F\043E\043B\044C\0437\043E\0432\0430\043D \043F\0430\0440\043E\043B\044C: \041D\0415\0422) (DSN=mysql://test:***@localhost:3306/test).
total 0 docs, 0 bytes
total 0.002 sec, 0 bytes/sec, 0.00 docs/sec
indexing index 'test1stemmed'...
ERROR: index 'test1stemmed': sql_connect: \0414\043E\0441\0442\0443\043F \0437\0430\043A\0440\044B\0442 \0434\043B\044F \043F\043E\043B\044C\0437\043E\0432\0430\0442\0435\043B\044F 'test'@'localhost' (\0431\044B\043B \0438\0441\043F\043E\043B\044C\0437\043E\0432\0430\043D \043F\0430\0440\043E\043B\044C: \041D\0415\0422) (DSN=mysql://test:***@localhost:3306/test).
total 0 docs, 0 bytes
total 0.000 sec, 0 bytes/sec, 0.00 docs/sec
skipping non-plain index 'dist1'...
skipping non-plain index 'rt'...
total 0 reads, 0.000 sec, 0.0 kb/call avg, 0.0 msec/call avg
total 0 writes, 0.000 sec, 0.0 kb/call avg, 0.0 msec/call avg

[свернуть]

Вывод команды sudo service sphinxsearch start

Открыть содержимое (спойлер)

using config file '/etc/sphinxsearch/sphinx.conf'...
WARNING: key 'max_matches' was permanently removed from Sphinx configuration. Refer to documentation for details.
FATAL: failed to create pid file '/var/run/sphinxsearch/searchd.pid': No such file or directory
ERROR.

[свернуть]

В /etc/init.d/sphinxsearch всё закомментировано

Открыть содержимое (спойлер)

#! /bin/sh
#
#       Written by Miquel van Smoorenburg <miquels@cistron.nl>.
#       Modified for Debian
#       by Ian Murdock <imurdock@gnu.ai.mit.edu>.
#       Further changes by Javier Fernandez-Sanguino <jfs@debian.org>
#       Modified for sphinx by Radu Spineanu <radu@debian.org>
#   Modified by Anton Tsitlionok <deogar@sphinxsearch.com>, 2013
### BEGIN INIT INFO
# Provides:          sphinxsearch
# Required-Start:    $local_fs $remote_fs $syslog $network $time
# Required-Stop:     $local_fs $remote_fs $syslog $network
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: Fast standalone full-text SQL search engine
### END INIT INFO

[свернуть]

Раскомментирование отдельных строк в sphinx.conf, который пытаюсь слепить методом тыка из взятого в Sphinx примера, к видимому успеху не приводит. На всякий случай, привожу весь кривой конфиг

Открыть содержимое (спойлер)

#
# Sphinx configuration file
#############################################################################
## data source definition
#############################################################################

source src1
{
   # data source type. mandatory, no default value
   # known types are mysql, pgsql, mssql, xmlpipe, xmlpipe2, odbc
   type         = mysql

   #####################################################################
   ## SQL settings (for 'mysql' and 'pgsql' types)
   #####################################################################

   # some straightforward parameters for SQL source types
   sql_host      = localhost
   sql_user      = test
   sql_pass      =
   sql_db         = test
   sql_port      = 3306   # optional, default is 3306

   # UNIX socket name
   # optional, default is empty (reuse client library defaults)
   # usually '/var/lib/mysql/mysql.sock' on Linux
   # usually '/tmp/mysql.sock' on FreeBSD
   #
   # sql_sock      = /tmp/mysql.sock


   # MySQL specific client connection flags
   # optional, default is 0
   #
   # mysql_connect_flags   = 32 # enable compression

   # MySQL specific SSL certificate settings
   # optional, defaults are empty
   #
   # mysql_ssl_cert      = /etc/ssl/client-cert.pem
   # mysql_ssl_key      = /etc/ssl/client-key.pem
   # mysql_ssl_ca      = /etc/ssl/cacert.pem

   # MS SQL specific Windows authentication mode flag
   # MUST be in sync with charset_type index-level setting
   # optional, default is 0
   #
   # mssql_winauth      = 1 # use currently logged on user credentials


   # ODBC specific DSN (data source name)
   # mandatory for odbc source type, no default value
   #
   # odbc_dsn      = DBQ=C:\data;DefaultDir=C:\data;Driver={Microsoft Text Driver (*.txt; *.csv)};
   # sql_query      = SELECT id, data FROM documents.csv


   # ODBC and MS SQL specific, per-column buffer sizes
   # optional, default is auto-detect
   #
   # sql_column_buffers   = content=12M, comments=1M


   # pre-query, executed before the main fetch query
   # multi-value, optional, default is empty list of queries
   #
   # sql_query_pre      = SET NAMES utf8
   # sql_query_pre      = SET SESSION query_cache_type=OFF


   # main document fetch query
   # mandatory, integer document ID field MUST be the first selected column
   sql_query      = \
      SELECT id, group_id, UNIX_TIMESTAMP(date_added) AS date_added, title, content \
      FROM documents


   # joined/payload field fetch query
   # joined fields let you avoid (slow) JOIN and GROUP_CONCAT
   # payload fields let you attach custom per-keyword values (eg. for ranking)
   #
   # syntax is FIELD-NAME 'from'  ( 'query' | 'payload-query' ); QUERY
   # joined field QUERY should return 2 columns (docid, text)
   # payload field QUERY should return 3 columns (docid, keyword, weight)
   #
   # REQUIRES that query results are in ascending document ID order!
   # multi-value, optional, default is empty list of queries
   #
   # sql_joined_field   = tags from query; SELECT docid, CONCAT('tag',tagid) FROM tags ORDER BY docid ASC
   # sql_joined_field   = wtags from payload-query; SELECT docid, tag, tagweight FROM tags ORDER BY docid ASC


   # file based field declaration
   #
   # content of this field is treated as a file name
   # and the file gets loaded and indexed in place of a field
   #
   # max file size is limited by max_file_field_buffer indexer setting
   # file IO errors are non-fatal and get reported as warnings
   #
   # sql_file_field      = content_file_path


   # range query setup, query that must return min and max ID values
   # optional, default is empty
   #
   # sql_query will need to reference $start and $end boundaries
   # if using ranged query:
   #
   # sql_query      = \
   #   SELECT doc.id, doc.id AS group, doc.title, doc.data \
   #   FROM documents doc \
   #   WHERE id>=$start AND id<=$end
   #
   # sql_query_range      = SELECT MIN(id),MAX(id) FROM documents


   # range query step
   # optional, default is 1024
   #
   # sql_range_step      = 1000


   # unsigned integer attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # optional bit size can be specified, default is 32
   #
   # sql_attr_uint      = author_id
   # sql_attr_uint      = forum_id:9 # 9 bits for forum_id
   sql_attr_uint      = group_id

   # boolean attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # equivalent to sql_attr_uint with 1-bit size
   #
   # sql_attr_bool      = is_deleted


   # bigint attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # declares a signed (unlike uint!) 64-bit attribute
   #
   # sql_attr_bigint      = my_bigint_id


   # UNIX timestamp attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # similar to integer, but can also be used in date functions
   #
   # sql_attr_timestamp   = posted_ts
   # sql_attr_timestamp   = last_edited_ts
   sql_attr_timestamp   = date_added


   # floating point attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # values are stored in single precision, 32-bit IEEE 754 format
   #
   # sql_attr_float      = lat_radians
   # sql_attr_float      = long_radians


   # multi-valued attribute (MVA) attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # MVA values are variable length lists of unsigned 32-bit integers
   #
   # syntax is ATTR-TYPE ATTR-NAME 'from' SOURCE-TYPE [;QUERY] [;RANGE-QUERY]
   # ATTR-TYPE is 'uint' or 'timestamp'
   # SOURCE-TYPE is 'field', 'query', or 'ranged-query'
   # QUERY is SQL query used to fetch all ( docid, attrvalue ) pairs
   # RANGE-QUERY is SQL query used to fetch min and max ID values, similar to 'sql_query_range'
   #
   # sql_attr_multi      = uint tag from query; SELECT docid, tagid FROM tags
   # sql_attr_multi      = uint tag from ranged-query; \
   #   SELECT docid, tagid FROM tags WHERE id>=$start AND id<=$end; \
   #   SELECT MIN(docid), MAX(docid) FROM tags


   # string attribute declaration
   # multi-value (an arbitrary number of these is allowed), optional
   # lets you store and retrieve strings
   #
   # sql_attr_string      = stitle


   # JSON attribute declaration
   # multi-value (an arbitrary number of these is allowed), optional
   # lets you store a JSON document as an (in-memory) attribute for later use
   #
   # sql_attr_json      = properties


   # combined field plus attribute declaration (from a single column)
   # stores column as an attribute, but also indexes it as a full-text field
   #
   # sql_field_string   = author

   
   # post-query, executed on sql_query completion
   # optional, default is empty
   #
   # sql_query_post      =

   
   # post-index-query, executed on successful indexing completion
   # optional, default is empty
   # $maxid expands to max document ID actually fetched from DB
   #
   # sql_query_post_index   = REPLACE INTO counters ( id, val ) \
   #   VALUES ( 'max_indexed_id', $maxid )


   # ranged query throttling, in milliseconds
   # optional, default is 0 which means no delay
   # enforces given delay before each query step
   sql_ranged_throttle   = 0


   # kill-list query, fetches the document IDs for kill-list
   # k-list will suppress matches from preceding indexes in the same query
   # optional, default is empty
   #
   # sql_query_killlist   = SELECT id FROM documents WHERE edited>=@last_reindex


   # columns to unpack on indexer side when indexing
   # multi-value, optional, default is empty list
   #
   # unpack_zlib      = zlib_column
   # unpack_mysqlcompress   = compressed_column
   # unpack_mysqlcompress   = compressed_column_2


   # maximum unpacked length allowed in MySQL COMPRESS() unpacker
   # optional, default is 16M
   #
   # unpack_mysqlcompress_maxsize   = 16M


   # hook command to run when SQL connection succeeds
   # optional, default value is empty (do nothing)
   #
   # hook_connect         = bash sql_connect.sh


   # hook command to run after (any) SQL range query
   # it may print out "minid maxid" (w/o quotes) to override the range
   # optional, default value is empty (do nothing)
   #
   # hook_query_range      = bash sql_query_range.sh


   # hook command to run on successful indexing completion
   # $maxid expands to max document ID actually fetched from DB
   # optional, default value is empty (do nothing)
   #
   # hook_post_index      = bash sql_post_index.sh $maxid

   #####################################################################
   ## xmlpipe2 settings
   #####################################################################

   # type         = xmlpipe

   # shell command to invoke xmlpipe stream producer
   # mandatory
   #
   # xmlpipe_command      = cat /var/lib/sphinxsearch/test.xml

   # xmlpipe2 field declaration
   # multi-value, optional, default is empty
   #
   # xmlpipe_field      = subject
   # xmlpipe_field      = content


   # xmlpipe2 attribute declaration
   # multi-value, optional, default is empty
   # all xmlpipe_attr_XXX options are fully similar to sql_attr_XXX
   # examples:
   #
   # xmlpipe_attr_timestamp   = published
   # xmlpipe_attr_uint   = author_id
   # xmlpipe_attr_bool   = is_enabled
   # xmlpipe_attr_float   = latitude
   # xmlpipe_attr_bigint   = guid
   # xmlpipe_attr_multi   = tags
   # xmlpipe_attr_multi_64   = tags64
   # xmlpipe_attr_string   = title
   # xmlpipe_attr_json   = extra_data
   # xmlpipe_field_string   = content


   # perform UTF-8 validation, and filter out incorrect codes
   # avoids XML parser choking on non-UTF-8 documents
   # optional, default is 0
   #
   # xmlpipe_fixup_utf8   = 1
}


# inherited source example
#
# all the parameters are copied from the parent source,
# and may then be overridden in this source definition
source src1throttled : src1
{
   sql_ranged_throttle   = 100
}

#############################################################################
## index definition
#############################################################################

# local index example
#
# this is an index which is stored locally in the filesystem
#
# all indexing-time options (such as morphology and charsets)
# are configured per local index
index test1
{
   # index type
   # optional, default is 'plain'
   # known values are 'plain', 'distributed', and 'rt' (see samples below)
   # type   = plain

   # document source(s) to index
   # multi-value, mandatory
   # document IDs must be globally unique across all sources
   source   = src1

   # index files path and file name, without extension
   # mandatory, path must be writable, extensions will be auto-appended
   path   = /var/lib/sphinxsearch/data/test1

   # document attribute values (docinfo) storage mode
   # optional, default is 'extern'
   # known values are 'none', 'extern' and 'inline'
   docinfo   = extern

   # dictionary type, 'crc' or 'keywords'
   # crc is faster to index when no substring/wildcards searches are needed
   # crc with substrings might be faster to search but is much slower to index
   # (because all substrings are pre-extracted as individual keywords)
   # keywords is much faster to index with substrings, and index is much (3-10x) smaller
   # keywords supports wildcards, crc does not, and never will
   # optional, default is 'keywords'
   dict   = keywords

   # memory locking for cached data (.spa and .spi), to prevent swapping
   # optional, default is 0 (do not mlock)
   # requires searchd to be run from root
   mlock   = 0

   # a list of morphology preprocessors to apply
   # optional, default is empty
   #
   # builtin preprocessors are 'none', 'stem_en', 'stem_ru', 'stem_enru',
   # 'soundex', and 'metaphone'; additional preprocessors available from
   # libstemmer are 'libstemmer_XXX', where XXX is algorithm code
   # (see libstemmer_c/libstemmer/modules.txt)
   #
   # morphology      = stem_en, stem_ru, soundex
   # morphology      = libstemmer_german
   # morphology      = libstemmer_sv
   morphology      = none

   # minimum word length at which to enable stemming
   # optional, default is 1 (stem everything)
   #
   # min_stemming_len   = 1


   # stopword files list (space separated)
   # optional, default is empty
   # contents are plain text, charset_table and stemming are both applied
   #
   # stopwords      = /var/lib/sphinxsearch/data/stopwords.txt


   # wordforms file, in "mapfrom > mapto" plain text format
   # optional, default is empty
   #
   # wordforms      = /var/lib/sphinxsearch/data/wordforms.txt


   # tokenizing exceptions file
   # optional, default is empty
   #
   # plain text, case sensitive, space insensitive in map-from part
   # one "Map Several Words => ToASingleOne" entry per line
   #
   # exceptions      = /var/lib/sphinxsearch/data/exceptions.txt


   # embedded file size limit
   # optional, default is 16K
   #
   # exceptions, wordforms, and stopwords files smaller than this limit
   # are stored in the index; otherwise, their paths and sizes are stored
   #
   # embedded_limit      = 16K

   # minimum indexed word length
   # default is 1 (index everything)
   min_word_len      = 1


   # ignored characters list
   # optional, default value is empty
   #
   # ignore_chars      = U+00AD


   # minimum word prefix length to index
   # optional, default is 0 (do not index prefixes)
   #
   # min_prefix_len      = 0


   # minimum word infix length to index
   # optional, default is 0 (do not index infixes)
   #
   # min_infix_len      = 0


   # maximum substring (prefix or infix) length to index
   # optional, default is 0 (do not limit substring length)
   #
   # max_substring_len   = 8


   # list of fields to limit prefix/infix indexing to
   # optional, default value is empty (index all fields in prefix/infix mode)
   #
   # prefix_fields      = filename
   # infix_fields      = url, domain


   # expand keywords with exact forms and/or stars when searching fit indexes
   # search-time only, does not affect indexing, can be 0 or 1
   # optional, default is 0 (do not expand keywords)
   #
   # expand_keywords      = 1

   
   # n-gram length to index, for CJK indexing
   # only supports 0 and 1 for now, other lengths to be implemented
   # optional, default is 0 (disable n-grams)
   #
   # ngram_len      = 1


   # n-gram characters list, for CJK indexing
   # optional, default is empty
   #
   # ngram_chars      = U+3000..U+2FA1F


   # phrase boundary characters list
   # optional, default is empty
   #
   # phrase_boundary      = ., ?, !, U+2026 # horizontal ellipsis


   # phrase boundary word position increment
   # optional, default is 0
   #
   # phrase_boundary_step   = 100


   # blended characters list
   # blended chars are indexed both as separators and valid characters
   # for instance, AT&T will results in 3 tokens ("at", "t", and "at&t")
   # optional, default is empty
   #
   # blend_chars      = +, &, U+23


   # blended token indexing mode
   # a comma separated list of blended token indexing variants
   # known variants are trim_none, trim_head, trim_tail, trim_both, skip_pure
   # optional, default is trim_none
   #
   # blend_mode      = trim_tail, skip_pure


   # whether to strip HTML tags from incoming documents
   # known values are 0 (do not strip) and 1 (do strip)
   # optional, default is 0
   html_strip      = 0

   # what HTML attributes to index if stripping HTML
   # optional, default is empty (do not index anything)
   #
   # html_index_attrs   = img=alt,title; a=title;


   # what HTML elements contents to strip
   # optional, default is empty (do not strip element contents)
   #
   # html_remove_elements   = style, script


   # whether to preopen index data files on startup
   # optional, default is 0 (do not preopen), searchd-only
   #
   # preopen         = 1


   # whether to enable in-place inversion (2x less disk, 90-95% speed)
   # optional, default is 0 (use separate temporary files), indexer-only
   #
   # inplace_enable      = 1


   # in-place fine-tuning options
   # optional, defaults are listed below
   #
   # inplace_hit_gap      = 0 # preallocated hitlist gap size
   # inplace_docinfo_gap   = 0 # preallocated docinfo gap size
   # inplace_reloc_factor   = 0.1 # relocation buffer size within arena
   # inplace_write_factor   = 0.1 # write buffer size within arena


   # whether to index original keywords along with stemmed versions
   # enables "=exactform" operator to work
   # optional, default is 0
   #
   # index_exact_words   = 1


   # position increment on overshort (less that min_word_len) words
   # optional, allowed values are 0 and 1, default is 1
   #
   # overshort_step      = 1


   # position increment on stopword
   # optional, allowed values are 0 and 1, default is 1
   #
   # stopword_step      = 1


   # hitless words list
   # positions for these keywords will not be stored in the index
   # optional, allowed values are 'all', or a list file name
   #
   # hitless_words      = all
   # hitless_words      = hitless.txt


   # detect and index sentence and paragraph boundaries
   # required for the SENTENCE and PARAGRAPH operators to work
   # optional, allowed values are 0 and 1, default is 0
   #
   # index_sp         = 1


   # index zones, delimited by HTML/XML tags
   # a comma separated list of tags and wildcards
   # required for the ZONE operator to work
   # optional, default is empty string (do not index zones)
   #
   # index_zones      = title, h*, th


   # index per-document and average per-index field lengths, in tokens
   # required for the BM25A(), BM25F() in expression ranker
   # optional, default is 0 (do not index field lenghts)
   #
   # index_field_lengths   = 1


   # regular expressions (regexps) to filter the fields and queries with
   # gets applied to data source fields when indexing
   # gets applied to search queries when searching
   # multi-value, optional, default is empty list of regexps
   #
   # regexp_filter      = \b(\d+)\" => \1inch
   # regexp_filter      = (blue|red) => color


   # list of the words considered frequent with respect to bigram indexing
   # optional, default is empty
   #
   # bigram_freq_words   = the, a, i, you, my


   # bigram indexing mode
   # known values are none, all, first_freq, both_freq
   # option, default is none (do not index bigrams)
   #
   # bigram_index      = both_freq


   # snippet document file name prefix
   # preprended to file names when generating snippets using load_files option
   # WARNING, this is a prefix (not a path), trailing slash matters!
   # optional, default is empty
   #
   # snippets_file_prefix   = /mnt/mydocs/server1


   # whether to apply stopwords before or after stemming
   # optional, default is 0 (apply stopwords after stemming)
   #
   # stopwords_unstemmed   = 0


   # path to a global (cluster-wide) keyword IDFs file
   # optional, default is empty (use local IDFs)
   #
   # global_idf      = /usr/local/sphinx/var/global.idf
}


# inherited index example
#
# all the parameters are copied from the parent index,
# and may then be overridden in this index definition
index test1stemmed : test1
{
   path         = /var/lib/sphinxsearch/data/test1stemmed
   morphology      = stem_en
}


# distributed index example
#
# this is a virtual index which can NOT be directly indexed,
# and only contains references to other local and/or remote indexes
index dist1
{
   # 'distributed' index type MUST be specified
   type         = distributed

   # local index to be searched
   # there can be many local indexes configured
   local         = test1
   local         = test1stemmed

   # remote agent
   # multiple remote agents may be specified
   # syntax for TCP connections is 'hostname:port:index1,[index2[,...]]'
   # syntax for local UNIX connections is '/path/to/socket:index1,[index2[,...]]'
   agent         = localhost:9313:remote1
   agent         = localhost:9314:remote2,remote3
   # agent         = /var/run/searchd.sock:remote4

   # remote agent mirrors groups, aka mirrors, aka HA agents
   # defines 2 or more interchangeable mirrors for a given index part
   #
   # agent         = server3:9312 | server4:9312 :indexchunk2
   # agent         = server3:9312:chunk2server3 | server4:9312:chunk2server4
   # agent         = server3:chunk2server3 | server4:chunk2server4
   # agent         = server21|server22|server23:chunk2


   # blackhole remote agent, for debugging/testing
   # network errors and search results will be ignored
   #
   # agent_blackhole      = testbox:9312:testindex1,testindex2


   # persistenly connected remote agent
   # reduces connect() pressure, requires that workers IS threads
   #
   # agent_persistent      = testbox:9312:testindex1,testindex2


   # remote agent connection timeout, milliseconds
   # optional, default is 1000 ms, ie. 1 sec
   agent_connect_timeout   = 1000

   # remote agent query timeout, milliseconds
   # optional, default is 3000 ms, ie. 3 sec
   agent_query_timeout      = 3000

   # HA mirror agent strategy
   # optional, defaults to (random mirror)
   # know values are nodeads, noerrors, roundrobin, nodeadstm, noerrorstm
   #
   # ha_strategy            = nodeads

   # path to RLP context file
   # optional, defaut is empty
   #
   # rlp_context = /usr/local/share/sphinx/rlp/rlp-context.xml
}


# realtime index example
#
# you can run INSERT, REPLACE, and DELETE on this index on the fly
# using MySQL protocol (see 'listen' directive below)
index rt
{
   # 'rt' index type must be specified to use RT index
   type         = rt

   # index files path and file name, without extension
   # mandatory, path must be writable, extensions will be auto-appended
   path         = /var/lib/sphinxsearch/data/rt

   # RAM chunk size limit
   # RT index will keep at most this much data in RAM, then flush to disk
   # optional, default is 128M
   #
   # rt_mem_limit      = 512M

   # full-text field declaration
   # multi-value, mandatory
   rt_field      = title
   rt_field      = content

   # unsigned integer attribute declaration
   # multi-value (an arbitrary number of attributes is allowed), optional
   # declares an unsigned 32-bit attribute
   rt_attr_uint      = gid

   # RT indexes currently support the following attribute types:
   # uint, bigint, float, timestamp, string, mva, mva64, json
   #
   # rt_attr_bigint      = guid
   # rt_attr_float      = gpa
   # rt_attr_timestamp   = ts_added
   # rt_attr_string      = author
   # rt_attr_multi      = tags
   # rt_attr_multi_64   = tags64
   # rt_attr_json      = extra_data
}

#############################################################################
## indexer settings
#############################################################################

indexer
{
   # memory limit, in bytes, kiloytes (16384K) or megabytes (256M)
   # optional, default is 128M, max is 2047M, recommended is 256M to 1024M
   mem_limit      = 128M

   # maximum IO calls per second (for I/O throttling)
   # optional, default is 0 (unlimited)
   #
   # max_iops      = 40


   # maximum IO call size, bytes (for I/O throttling)
   # optional, default is 0 (unlimited)
   #
   # max_iosize      = 1048576


   # maximum xmlpipe2 field length, bytes
   # optional, default is 2M
   #
   # max_xmlpipe2_field   = 4M


   # write buffer size, bytes
   # several (currently up to 4) buffers will be allocated
   # write buffers are allocated in addition to mem_limit
   # optional, default is 1M
   #
   # write_buffer      = 1M


   # maximum file field adaptive buffer size
   # optional, default is 8M, minimum is 1M
   #
   # max_file_field_buffer   = 32M


   # how to handle IO errors in file fields
   # known values are 'ignore_field', 'skip_document', and 'fail_index'
   # optional, default is 'ignore_field'
   #
   # on_file_field_error = skip_document


   # how to handle syntax errors in JSON attributes
   # known values are 'ignore_attr' and 'fail_index'
   # optional, default is 'ignore_attr'
   #
   # on_json_attr_error = fail_index


   # whether to auto-convert numeric values from strings in JSON attributes
   # with auto-conversion, string value with actually numeric data
   # (as in {"key":"12345"}) gets stored as a number, rather than string
   # optional, allowed values are 0 and 1, default is 0 (do not convert)
   #
   # json_autoconv_numbers = 1


   # whether and how to auto-convert key names in JSON attributes
   # known value is 'lowercase'
   # optional, default is unspecified (do nothing)
   #
   # json_autoconv_keynames = lowercase


   # lemmatizer cache size
   # improves the indexing time when the lemmatization is enabled
   # optional, default is 256K
   #
   # lemmatizer_cache = 512M
}

#############################################################################
## searchd settings
#############################################################################

searchd
{
   # [hostname:]port[:protocol], or /unix/socket/path to listen on
   # known protocols are 'sphinx' (SphinxAPI) and 'mysql41' (SphinxQL)
   #
   # multi-value, multiple listen points are allowed
   # optional, defaults are 9312:sphinx and 9306:mysql41, as below
   #
   # listen         = 127.0.0.1
   # listen         = 192.168.0.1:9312
   # listen         = 9312
   # listen         = /var/run/searchd.sock
   listen         = 9312
   listen         = 9306:mysql41

   # log file, searchd run info is logged here
   # optional, default is 'searchd.log'
   log         = /var/log/sphinxsearch/searchd.log

   # query log file, all search queries are logged here
   # optional, default is empty (do not log queries)
   query_log      = /var/log/sphinxsearch/query.log

   # client read timeout, seconds
   # optional, default is 5
   read_timeout      = 5

   # request timeout, seconds
   # optional, default is 5 minutes
   client_timeout      = 300

   # maximum amount of children to fork (concurrent searches to run)
   # optional, default is 0 (unlimited)
   max_children      = 30

   # maximum amount of persistent connections from this master to each agent host
   # optional, but necessary if you use agent_persistent. It is reasonable to set the value
   # as max_children, or less on the agent's hosts.
   persistent_connections_limit   = 30

   # PID file, searchd process ID file name
   # mandatory
   pid_file   = /var/run/sphinxsearch/searchd.pid

   # max amount of matches the daemon ever keeps in RAM, per-index
   # WARNING, THERE'S ALSO PER-QUERY LIMIT, SEE SetLimits() API CALL
   # default is 1000 (just like Google)
   max_matches   = 1000

   # seamless rotate, prevents rotate stalls if precaching huge datasets
   # optional, default is 1
   seamless_rotate   = 1

   # whether to forcibly preopen all indexes on startup
   # optional, default is 1 (preopen everything)
   preopen_indexes      = 1

   # whether to unlink .old index copies on succesful rotation.
   # optional, default is 1 (do unlink)
   unlink_old      = 1

   # attribute updates periodic flush timeout, seconds
   # updates will be automatically dumped to disk this frequently
   # optional, default is 0 (disable periodic flush)
   #
   # attr_flush_period   = 900


   # MVA updates pool size
   # shared between all instances of searchd, disables attr flushes!
   # optional, default size is 1M
   mva_updates_pool   = 1M

   # max allowed network packet size
   # limits both query packets from clients, and responses from agents
   # optional, default size is 8M
   max_packet_size      = 8M

   # max allowed per-query filter count
   # optional, default is 256
   max_filters      = 256

   # max allowed per-filter values count
   # optional, default is 4096
   max_filter_values   = 4096


   # socket listen queue length
   # optional, default is 5
   #
   # listen_backlog      = 5


   # per-keyword read buffer size
   # optional, default is 256K
   #
   # read_buffer      = 256K


   # unhinted read size (currently used when reading hits)
   # optional, default is 32K
   #
   # read_unhinted      = 32K


   # max allowed per-batch query count (aka multi-query count)
   # optional, default is 32
   max_batch_queries   = 32


   # max common subtree document cache size, per-query
   # optional, default is 0 (disable subtree optimization)
   #
   # subtree_docs_cache   = 4M


   # max common subtree hit cache size, per-query
   # optional, default is 0 (disable subtree optimization)
   #
   # subtree_hits_cache   = 8M


   # multi-processing mode (MPM)
   # known values are none, fork, prefork, and threads
   # threads is required for RT backend to work
   # optional, default is threads
   workers         = threads # for RT to work


   # max threads to create for searching local parts of a distributed index
   # optional, default is 0, which means disable multi-threaded searching
   # should work with all MPMs (ie. does NOT require workers=threads)
   #
   # dist_threads      = 4


   # binlog files path; use empty string to disable binlog
   # optional, default is build-time configured data directory
   #
   # binlog_path      = # disable logging
   # binlog_path      = /var/lib/sphinxsearch/data # binlog.001 etc will be created there


   # binlog flush/sync mode
   # 0 means flush and sync every second
   # 1 means flush and sync every transaction
   # 2 means flush every transaction, sync every second
   # optional, default is 2
   #
   # binlog_flush      = 2


   # binlog per-file size limit
   # optional, default is 128M, 0 means no limit
   #
   # binlog_max_log_size   = 256M


   # per-thread stack size, only affects workers=threads mode
   # optional, default is 64K
   #
   # thread_stack         = 128K


   # per-keyword expansion limit (for dict=keywords prefix searches)
   # optional, default is 0 (no limit)
   #
   # expansion_limit      = 1000


   # RT RAM chunks flush period
   # optional, default is 0 (no periodic flush)
   #
   # rt_flush_period      = 900


   # query log file format
   # optional, known values are plain and sphinxql, default is plain
   #
   # query_log_format      = sphinxql


   # version string returned to MySQL network protocol clients
   # optional, default is empty (use Sphinx version)
   #
   # mysql_version_string   = 5.0.37


   # trusted plugin directory
   # optional, default is empty (disable UDFs)
   #
   # plugin_dir         = /usr/local/sphinx/lib


   # default server-wide collation
   # optional, default is libc_ci
   #
   # collation_server      = utf8_general_ci


   # server-wide locale for libc based collations
   # optional, default is C
   #
   # collation_libc_locale   = ru_RU.UTF-8


   # threaded server watchdog (only used in workers=threads mode)
   # optional, values are 0 and 1, default is 1 (watchdog on)
   #
   # watchdog            = 1

   
   # costs for max_predicted_time model, in (imaginary) nanoseconds
   # optional, default is "doc=64, hit=48, skip=2048, match=64"
   #
   # predicted_time_costs   = doc=64, hit=48, skip=2048, match=64


   # current SphinxQL state (uservars etc) serialization path
   # optional, default is none (do not serialize SphinxQL state)
   #
   # sphinxql_state         = sphinxvars.sql


   # maximum RT merge thread IO calls per second, and per-call IO size
   # useful for throttling (the background) OPTIMIZE INDEX impact
   # optional, default is 0 (unlimited)
   #
   # rt_merge_iops         = 40
   # rt_merge_maxiosize      = 1M


   # interval between agent mirror pings, in milliseconds
   # 0 means disable pings
   # optional, default is 1000
   #
   # ha_ping_interval      = 0


   # agent mirror statistics window size, in seconds
   # stats older than the window size (karma) are retired
   # that is, they will not affect master choice of agents in any way
   # optional, default is 60 seconds
   #
   # ha_period_karma         = 60


   # delay between preforked children restarts on rotation, in milliseconds
   # optional, default is 0 (no delay)
   #
   # prefork_rotation_throttle   = 100


   # a prefix to prepend to the local file names when creating snippets
   # with load_files and/or load_files_scatter options
   # optional, default is empty
   #
   # snippets_file_prefix      = /mnt/common/server1/
}

#############################################################################
## common settings
#############################################################################

common
{

   # lemmatizer dictionaries base path
   # optional, defaut is /usr/local/share (see ./configure --datadir)
   #
   # lemmatizer_base = /usr/local/share/sphinx/dicts

   # path to RLP root directory
   # optional, defaut is /usr/local/share (see ./configure --datadir)
   #
   # rlp_root = /usr/local/share/sphinx/rlp

   # path to RLP environment file
   # optional, defaut is /usr/local/share/rlp-environment.xml (see ./configure --datadir)
   #
   # rlp_environment = /usr/local/share/sphinx/rlp/rlp/etc/rlp-environment.xml
}

# --eof--

[свернуть]

Вводил команды для создания базы, пользователя и передачи ему прав, но дальше не продвинулся. Замена значений sql_user, sql_pass, sql_db в предложенном конфиге на те, которые создавались в MySQL, не привела к успеху, после чего я решил обратиться за помощью. 
Глядя на английские слова, интуитивно понимаю, что всё дело в MySQL, которую Nepomuk+Strigi настраивали в KDE без меня и которую мне настроить, видимо, так и не удалось, хотя вход есть.   
Вывод sudo mysql -p

Открыть содержимое (спойлер)

Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 42
Server version: 5.5.38-0+wheezy1 (Debian)
Copyright (c) 2000, 2014, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
mysql>

[свернуть]

Вывод mysqladmin --help

Открыть содержимое (спойлер)

mysqladmin  Ver 8.42 Distrib 5.5.38, for debian-linux-gnu on x86_64
Variables (--variable-name=value)
and boolean options {FALSE|TRUE}  Value (after reading options)
--------------------------------- ----------------------------------------
count                             0
debug-check                       FALSE
debug-info                        FALSE
force                             FALSE
compress                          FALSE
character-sets-dir                /usr/share/mysql/charsets
default-character-set             utf8
host                              (No default value)
no-beep                           FALSE
port                              3306
relative                          FALSE
socket                            /var/run/mysqld/mysqld.sock
sleep                             0
ssl                               FALSE
ssl-ca                            (No default value)
ssl-capath                        (No default value)
ssl-cert                          (No default value)
ssl-cipher                        (No default value)
ssl-key                           (No default value)
ssl-verify-server-cert            FALSE
user                              (No default value)
verbose                           FALSE
vertical                          FALSE
connect-timeout                   43200
shutdown-timeout                  3600
plugin-dir                        (No default value)
default-auth                      (No default value)
enable-cleartext-plugin           FALSE
Default options are read from the following files in the given order:
/etc/my.cnf /etc/mysql/my.cnf /usr/etc/my.cnf ~/.my.cnf

[свернуть]

Подскажите, что делать дальше. Может я вообще не тем занимаюсь или есть что-нибудь лучше Sphinx. Может легче установить nepomuk+strigi, но чувствую, что тут ведь будет снова не как в KDE,  а всё сложнее.