Subversion Repositories Applications.papyrus

Rev

Rev 2137 | Details | Compare with Previous | Last modification | View Log | RSS feed

Rev Author Line No. Line
2103 drzraf 1 
#
2 
# Minimal Sphinx configuration sample (clean, simple, functional)
3 
#
4 
# http://www.tela-botanica.org/page:recherche
5 
# mysql -h0 -P 9306
6 
# SELECT * FROM i_spip WHERE MATCH('@texte mousse');
7 
 
8 
# sudo sphinx-indexer i_spip -c /etc/sphinx/sphinx-min.conf
9 
# sudo sphinx-searchd -c /etc/sphinx/sphinx-min.conf
10 
source common
11 
{
12 
	type			= mysql
13 
 
14 
	sql_host		= localhost
2149 mathias 15 
	sql_user		= root
16 
	sql_pass		= mat87cho
2137 drzraf 17 
	sql_db			= tela_prod_v4
2149 mathias 18 
	#sql_sock		= /run/mysqld/mysqld.sock
2137 drzraf 19 
 
2115 drzraf 20 
    sql_query_pre   = SET NAMES utf8
21 
    sql_query_pre   = SET CHARACTER_SET_RESULTS=utf8
22 
 
23 
    sql_query       = SELECT 1
2103 drzraf 24 
}
25 
 
26 
index i_common
27 
{
2115 drzraf 28 
    # ces deux directive permettent à la source abstraite d'être valide
29 
    # lors de l'utilisation de sphinx-search sans précision d'index
30 
	source			= common
2137 drzraf 31 
	path			= /var/lib/sphinxsearch/void
2103 drzraf 32 
	docinfo			= extern
33 
	stopwords		= /etc/sphinx/stopwords.txt
34 
 
35 
    # http://sphinxsearch.com/docs/2.1.1/conf-morphology.html
36 
	morphology		= libstemmer_french
37 
	enable_star		= 1
38 
    min_word_len    = 3
39 
    # min_prefix_len  = 3
40 
    min_infix_len   = 3
41 
 
42 
    # prefix_fields   =
43 
    # infix_fields    =
44 
 
45 
    phrase_boundary = ., ?, !, U+2026
46 
 
47 
    # http://sphinxsearch.com/wiki/doku.php?id=charset_tables
2137 drzraf 48 
    # mysql tela_prod_v4<<<"SELECT p_description FROM projet WHERE p_id = 121"|xxd
2103 drzraf 49 
    # sphinx-search -q -c sphinx-min.conf -i i_projet "homoge*|homogé"|grep -w hits
50 
    # SELECT * FROM i_projet WHERE MATCH('homoge*|homogé'); SHOW META;
2115 drzraf 51 
	charset_type	= utf-8
52 
    # valeur par défaut, cf http://sphinxsearch.com/docs/archives/1.10/conf-charset-table.html
53 
    # pour écraser i_common
54 
    charset_table = 0..9, A..Z->a..z, _, a..z, U+410..U+42F->U+430..U+44F, U+430..U+44F, \
55 
	    U+041..U+05a, U+0c6, U+0d0, U+0d8, U+110, U+126, U+132, U+13f, U+141, U+14a, \
56 
		U+152, U+166, U+181, U+186, U+187, U+18a, U+18e..U+190, U+193, U+198, U+19f, U+1a2, \
57 
		U+1b3, U+1c4, U+1c7, U+1ca, U+1e4, U+21c, U+061->U+041, U+0c0->U+041, U+0c1->U+041, \
58 
		U+0c2->U+041, U+0c3->U+041, U+0c4->U+041, U+0c5->U+041, U+0e0->U+041, U+0e1->U+041, \
59 
		U+0e2->U+041, U+0e3->U+041, U+0e4->U+041, U+0e5->U+041, U+100->U+041, U+101->U+041, \
60 
		U+102->U+041, U+103->U+041, U+104->U+041, U+105->U+041, U+1cd->U+041, U+1ce->U+041, \
61 
		U+1de->U+041, U+1df->U+041, U+1e0->U+041, U+1e1->U+041, U+1fa->U+041, U+1fb->U+041, \
62 
		U+200->U+041, U+201->U+041, U+202->U+041, U+203->U+041, U+226->U+041, U+227->U+041, \
63 
		U+1e01->U+041, U+1ea0->U+041, U+1ea1->U+041, U+1ea2->U+041, U+1ea3->U+041, \
64 
		U+1ea4->U+041, U+1ea5->U+041, U+1ea6->U+041, U+1ea7->U+041, U+1ea8->U+041, \
65 
		U+1ea9->U+041, U+1eaa->U+041, U+1eab->U+041, U+1eac->U+041, U+1ead->U+041, \
66 
		U+1eae->U+041, U+1eaf->U+041, U+1eb0->U+041, U+1eb1->U+041, U+1eb2->U+041, \
67 
		U+1eb3->U+041, U+1eb4->U+041, U+1eb5->U+041, U+1eb6->U+041, U+1eb7->U+041, \
68 
		U+062->U+042, U+1e02->U+042, U+1e03->U+042, U+1e04->U+042, U+1e05->U+042, \
69 
		U+1e06->U+042, U+1e07->U+042, U+063->U+043, U+0c7->U+043, U+0e7->U+043, \
70 
		U+106->U+043, U+107->U+043, U+108->U+043, U+109->U+043, U+10a->U+043, U+10b->U+043, \
71 
		U+10c->U+043, U+10d->U+043, U+1e08->U+043, U+1e09->U+043, U+064->U+044, \
72 
		U+10e->U+044, U+10f->U+044, U+1e0a->U+044, U+1e0b->U+044, U+1e0c->U+044, \
73 
		U+1e0d->U+044, U+1e0e->U+044, U+1e0f->U+044, U+1e10->U+044, U+1e11->U+044, \
74 
		U+1e12->U+044, U+1e13->U+044, U+065->U+045, U+0c8->U+045, U+0c9->U+045, \
75 
		U+0ca->U+045, U+0cb->U+045, U+0e8->U+045, U+0e9->U+045, U+0ea->U+045, U+0eb->U+045, \
76 
		U+112->U+045, U+113->U+045, U+114->U+045, U+115->U+045, U+116->U+045, U+117->U+045, \
77 
		U+118->U+045, U+119->U+045, U+11a->U+045, U+11b->U+045, U+204->U+045, U+205->U+045, \
78 
		U+206->U+045, U+207->U+045, U+228->U+045, U+229->U+045, U+1e14->U+045, \
79 
		U+1e15->U+045, U+1e16->U+045, U+1e17->U+045, U+1e18->U+045, U+1e19->U+045, \
80 
		U+1e1a->U+045, U+1e1b->U+045, U+1e1c->U+045, U+1e1d->U+045, U+1eb8->U+045, \
81 
		U+1eb9->U+045, U+1eba->U+045, U+1ebb->U+045, U+1ebc->U+045, U+1ebd->U+045, \
82 
		U+1ebe->U+045, U+1ebf->U+045, U+1ec0->U+045, U+1ec1->U+045, U+1ec2->U+045, \
83 
		U+1ec3->U+045, U+1ec4->U+045, U+1ec5->U+045, U+1ec6->U+045, U+1ec7->U+045, \
84 
		U+066->U+046, U+1e1e->U+046, U+1e1f->U+046, U+067->U+047, U+11c->U+047, \
85 
		U+11d->U+047, U+11e->U+047, U+11f->U+047, U+120->U+047, U+121->U+047, U+122->U+047, \
86 
		U+123->U+047, U+1e6->U+047, U+1e7->U+047, U+1f4->U+047, U+1f5->U+047, \
87 
		U+1e20->U+047, U+1e21->U+047, U+068->U+048, U+124->U+048, U+125->U+048, \
88 
		U+21e->U+048, U+21f->U+048, U+1e22->U+048, U+1e23->U+048, U+1e24->U+048, \
89 
		U+1e25->U+048, U+1e26->U+048, U+1e27->U+048, U+1e28->U+048, U+1e29->U+048, \
90 
		U+1e2a->U+048, U+1e2b->U+048, U+1e96->U+048, U+069->U+049, U+0cc->U+049, \
91 
		U+0cd->U+049, U+0ce->U+049, U+0cf->U+049, U+0ec->U+049, U+0ed->U+049, U+0ee->U+049, \
92 
		U+0ef->U+049, U+128->U+049, U+129->U+049, U+12a->U+049, U+12b->U+049, U+12c->U+049, \
93 
		U+12d->U+049, U+12e->U+049, U+12f->U+049, U+130->U+049, U+131->U+049, U+1cf->U+049, \
94 
		U+1d0->U+049, U+208->U+049, U+209->U+049, U+20a->U+049, U+20b->U+049, \
95 
		U+1e2c->U+049, U+1e2d->U+049, U+1e2e->U+049, U+1e2f->U+049, U+1ec8->U+049, \
96 
		U+1ec9->U+049, U+1eca->U+049, U+1ecb->U+049, U+06a->U+04a, U+134->U+04a, \
97 
		U+135->U+04a, U+1f0->U+04a, U+06b->U+04b, U+136->U+04b, U+137->U+04b, U+1e8->U+04b, \
98 
		U+1e9->U+04b, U+1e30->U+04b, U+1e31->U+04b, U+1e32->U+04b, U+1e33->U+04b, \
99 
		U+1e34->U+04b, U+1e35->U+04b, U+06c->U+04c, U+139->U+04c, U+13a->U+04c, \
100 
		U+13b->U+04c, U+13c->U+04c, U+13d->U+04c, U+13e->U+04c, U+1e36->U+04c, \
101 
		U+1e37->U+04c, U+1e38->U+04c, U+1e39->U+04c, U+1e3a->U+04c, U+1e3b->U+04c, \
102 
		U+1e3c->U+04c, U+1e3d->U+04c, U+06d->U+04d, U+1e3e->U+04d, U+1e3f->U+04d, \
103 
		U+1e40->U+04d, U+1e41->U+04d, U+1e42->U+04d, U+1e43->U+04d, U+06e->U+04e, \
104 
		U+0d1->U+04e, U+0f1->U+04e, U+143->U+04e, U+144->U+04e, U+145->U+04e, U+146->U+04e, \
105 
		U+147->U+04e, U+148->U+04e, U+1f8->U+04e, U+1f9->U+04e, U+1e44->U+04e, \
106 
		U+1e45->U+04e, U+1e46->U+04e, U+1e47->U+04e, U+1e48->U+04e, U+1e49->U+04e, \
107 
		U+1e4a->U+04e, U+1e4b->U+04e, U+06f->U+04f, U+0d2->U+04f, U+0d3->U+04f, \
108 
		U+0d4->U+04f, U+0d5->U+04f, U+0d6->U+04f, U+0f2->U+04f, U+0f3->U+04f, U+0f4->U+04f, \
109 
		U+0f5->U+04f, U+0f6->U+04f, U+14c->U+04f, U+14d->U+04f, U+14e->U+04f, U+14f->U+04f, \
110 
		U+150->U+04f, U+151->U+04f, U+1a0->U+04f, U+1a1->U+04f, U+1d1->U+04f, U+1d2->U+04f, \
111 
		U+1ea->U+04f, U+1eb->U+04f, U+1ec->U+04f, U+1ed->U+04f, U+20c->U+04f, U+20d->U+04f, \
112 
		U+20e->U+04f, U+20f->U+04f, U+22a->U+04f, U+22b->U+04f, U+22c->U+04f, U+22d->U+04f, \
113 
		U+22e->U+04f, U+22f->U+04f, U+230->U+04f, U+231->U+04f, U+1e4c->U+04f, \
114 
		U+1e4d->U+04f, U+1e4e->U+04f, U+1e4f->U+04f, U+1e50->U+04f, U+1e51->U+04f, \
115 
		U+1e52->U+04f, U+1e53->U+04f, U+1ecc->U+04f, U+1ecd->U+04f, U+1ece->U+04f, \
116 
		U+1ecf->U+04f, U+1ed0->U+04f, U+1ed1->U+04f, U+1ed2->U+04f, U+1ed3->U+04f, \
117 
		U+1ed4->U+04f, U+1ed5->U+04f, U+1ed6->U+04f, U+1ed7->U+04f, U+1ed8->U+04f, \
118 
		U+1ed9->U+04f, U+1eda->U+04f, U+1edb->U+04f, U+1edc->U+04f, U+1edd->U+04f, \
119 
		U+1ede->U+04f, U+1edf->U+04f, U+1ee0->U+04f, U+1ee1->U+04f, U+1ee2->U+04f, \
120 
		U+1ee3->U+04f, U+070->U+050, U+1e54->U+050, U+1e55->U+050, U+1e56->U+050, \
121 
		U+1e57->U+050, U+071->U+051, U+072->U+052, U+154->U+052, U+155->U+052, \
122 
		U+156->U+052, U+157->U+052, U+158->U+052, U+159->U+052, U+210->U+052, U+211->U+052, \
123 
		U+212->U+052, U+213->U+052, U+1e58->U+052, U+1e59->U+052, U+1e5a->U+052, \
124 
		U+1e5b->U+052, U+1e5c->U+052, U+1e5d->U+052, U+1e5e->U+052, U+1e5f->U+052, \
125 
		U+073->U+053, U+0df->U+053, U+15a->U+053, U+15b->U+053, U+15c->U+053, U+15d->U+053, \
126 
		U+15e->U+053, U+15f->U+053, U+160->U+053, U+161->U+053, U+17f->U+053, U+218->U+053, \
127 
		U+219->U+053, U+1e60->U+053, U+1e61->U+053, U+1e62->U+053, U+1e63->U+053, \
128 
		U+1e64->U+053, U+1e65->U+053, U+1e66->U+053, U+1e67->U+053, U+1e68->U+053, \
129 
		U+1e69->U+053, U+1e9b->U+053, U+074->U+054, U+162->U+054, U+163->U+054, \
130 
		U+164->U+054, U+165->U+054, U+21a->U+054, U+21b->U+054, U+1e6a->U+054, \
131 
		U+1e6b->U+054, U+1e6c->U+054, U+1e6d->U+054, U+1e6e->U+054, U+1e6f->U+054, \
132 
		U+1e70->U+054, U+1e71->U+054, U+1e97->U+054, U+075->U+055, U+0d9->U+055, \
133 
		U+0da->U+055, U+0db->U+055, U+0dc->U+055, U+0f9->U+055, U+0fa->U+055, U+0fb->U+055, \
134 
		U+0fc->U+055, U+168->U+055, U+169->U+055, U+16a->U+055, U+16b->U+055, U+16c->U+055, \
135 
		U+16d->U+055, U+16e->U+055, U+16f->U+055, U+170->U+055, U+171->U+055, U+172->U+055, \
136 
		U+173->U+055, U+1af->U+055, U+1b0->U+055, U+1d3->U+055, U+1d4->U+055, U+1d5->U+055, \
137 
		U+1d6->U+055, U+1d7->U+055, U+1d8->U+055, U+1d9->U+055, U+1da->U+055, U+1db->U+055, \
138 
		U+1dc->U+055, U+214->U+055, U+215->U+055, U+216->U+055, U+217->U+055, \
139 
		U+1e72->U+055, U+1e73->U+055, U+1e74->U+055, U+1e75->U+055, U+1e76->U+055, \
140 
		U+1e77->U+055, U+1e78->U+055, U+1e79->U+055, U+1e7a->U+055, U+1e7b->U+055, \
141 
		U+1ee4->U+055, U+1ee5->U+055, U+1ee6->U+055, U+1ee7->U+055, U+1ee8->U+055, \
142 
		U+1ee9->U+055, U+1eea->U+055, U+1eeb->U+055, U+1eec->U+055, U+1eed->U+055, \
143 
		U+1eee->U+055, U+1eef->U+055, U+1ef0->U+055, U+1ef1->U+055, U+076->U+056, \
144 
		U+1e7c->U+056, U+1e7d->U+056, U+1e7e->U+056, U+1e7f->U+056, U+077->U+057, \
145 
		U+174->U+057, U+175->U+057, U+1e80->U+057, U+1e81->U+057, U+1e82->U+057, \
146 
		U+1e83->U+057, U+1e84->U+057, U+1e85->U+057, U+1e86->U+057, U+1e87->U+057, \
147 
		U+1e88->U+057, U+1e89->U+057, U+1e98->U+057, U+078->U+058, U+1e8a->U+058, \
148 
		U+1e8b->U+058, U+1e8c->U+058, U+1e8d->U+058, U+079->U+059, U+0dd->U+059, \
149 
		U+0fd->U+059, U+0ff->U+059, U+176->U+059, U+177->U+059, U+178->U+059, U+232->U+059, \
150 
		U+233->U+059, U+1e8e->U+059, U+1e8f->U+059, U+1e99->U+059, U+1ef2->U+059, \
151 
		U+1ef3->U+059, U+1ef4->U+059, U+1ef5->U+059, U+1ef6->U+059, U+1ef7->U+059, \
152 
		U+1ef8->U+059, U+1ef9->U+059, U+07a->U+05a, U+179->U+05a, U+17a->U+05a, \
153 
		U+17b->U+05a, U+17c->U+05a, U+17d->U+05a, U+17e->U+05a, U+1e90->U+05a, \
154 
		U+1e91->U+05a, U+1e92->U+05a, U+1e93->U+05a, U+1e94->U+05a, U+1e95->U+05a, \
155 
		U+0e6->U+0c6, U+1e2->U+0c6, U+1e3->U+0c6, U+1fc->U+0c6, U+1fd->U+0c6, U+0f0->U+0d0, \
156 
		U+0f8->U+0d8, U+1fe->U+0d8, U+1ff->U+0d8, U+111->U+110, U+127->U+126, U+133->U+132, \
157 
		U+140->U+13f, U+142->U+141, U+14b->U+14a, U+153->U+152, U+167->U+166, U+253->U+181, \
158 
		U+254->U+186, U+188->U+187, U+257->U+18a, U+1dd->U+18e, U+259->U+18f, U+25b->U+190, \
159 
		U+260->U+193, U+199->U+198, U+275->U+19f, U+1a3->U+1a2, U+1b4->U+1b3, U+1c5->U+1c4, \
160 
		U+1c6->U+1c4, U+1c8->U+1c7, U+1c9->U+1c7, U+1cb->U+1ca, U+1cc->U+1ca, U+1e5->U+1e4, \
161 
		U+1f2->U+1f1, U+1f3->U+1f1, U+21d->U+21c
2103 drzraf 162 
 
2115 drzraf 163 
 
164 
 
165 
 
2103 drzraf 166 
    # later (2.1.1)
167 
    # lemmatizer_base = /usr/share/sphinx/dicts/
168 
}
169 
 
170 
 
171 
# SPIP
172 
source spip : common
173 
{
2118 drzraf 174 
	# sql_db = tela_prod_spip_actu
2103 drzraf 175 
	# bibliotheque/more_recherche_spip_article.class.php
176 
	# premier champ = index (obligatoire)
177 
 
178 
	# surtitre = '' partout
179 
	# descriptif = '' partout
180 
	# chapo = '' partout
181 
	# soustitre = '' partout sauf id = 19
182 
	# ps = '' partout sauf id = 19
183 
	sql_query		= \
2116 drzraf 184 
		SELECT id_article, titre, texte, UNIX_TIMESTAMP(date) AS date, lang, "spip" AS group_id, CONCAT("spip-", id_article) as main_id \
2103 drzraf 185 
              FROM spip_articles WHERE statut = "publie"
186 
 
187 
#	sql_attr_uint		= id_article
188 
	sql_attr_timestamp	= date
189 
	sql_attr_string		= lang
2105 drzraf 190 
	sql_field_string	= group_id      # sql field pour permettre du filtrage @group_id != spip par exemple
2103 drzraf 191 
	sql_attr_string		= main_id
192 
 
193 
	sql_field_string	= titre
194 
 
2115 drzraf 195 
	sql_query_info		= SELECT * FROM spip_articles WHERE id_article=$id
2103 drzraf 196 
}
197 
 
198 
index i_spip : i_common {
199 
	source			= spip
2137 drzraf 200 
	path			= /var/lib/sphinxsearch/spip
2103 drzraf 201 
 
202 
    # prefix_fields   = titre, texte
203 
    infix_fields    = titre, texte
204 
}
205 
 
206 
 
207 
# PAPYRUS
208 
source papyrus : common
209 
{
210 
    #gm_nom
211 
	#gm_titre
212 
	#gm_titre_alternatif
213 
    #gm_mots_cles                       # TODO
214 
    #gm_description_libre               # TODO
215 
	#gm_description_resume              # TODO
216 
	#gm_description_table_matieres      partout '' sauf gm_id_menu = 1
217 
    #gm_source                          partout '' sauf gm_id_menu = 587
218 
	#gm_auteur                          # TODO
219 
	#gm_contributeur                    # TODO
220 
	#gm_editeur                         # TODO
221 
	#gm_categorie                       # TODO
222 
    #gm_public                          partout ''
223 
    #gmc_contenu
224 
    sql_query = \
225 
        SELECT m.gm_id_menu, \
226 
               IF(gm_nom != '', gm_nom, IF(gm_titre != '', gm_titre, gm_titre_alternatif)) as titre, \
227 
               gmc_contenu AS texte, \
228 
               gm_mots_cles,gm_source, gm_auteur, gm_contributeur, gm_editeur, gm_categorie, \
229 
               \
2116 drzraf 230 
               UNIX_TIMESTAMP(gm_date_creation) AS date,        \
2103 drzraf 231 
               "papyrus" AS group_id,    \
232 
               CONCAT("papyrus-", m.gm_id_menu) as main_id \
233 
               \
234 
               FROM gen_menu m \
235 
               LEFT JOIN gen_menu_contenu mc  ON mc.gmc_ce_menu      = m.gm_id_menu \
236 
                                              AND mc.gmc_bool_dernier = 1 \
237 
               WHERE m.gm_date_fin_validite = '0000-00-00 00:00:00' OR LOWER(m.gm_date_fin_validite) < now()
238 
 
2116 drzraf 239 
	sql_attr_timestamp	= date
2105 drzraf 240 
	sql_field_string	= group_id
2103 drzraf 241 
	sql_attr_string		= main_id
2115 drzraf 242 
 
243 
	sql_query_info		= SELECT * FROM gen_menu WHERE gm_id_menu=$id
2103 drzraf 244 
}
245 
 
246 
index i_papyrus : i_common {
247 
	source			= papyrus
2137 drzraf 248 
	path			= /var/lib/sphinxsearch/papyrus
2103 drzraf 249 
 
250 
    # prefix_fields   = titre, texte
251 
    infix_fields    = titre, texte
252 
}
253 
 
254 
 
255 
 
256 
# PROJET
257 
source projet : common {
258 
    # p_id
259 
    # p_titre
260 
    # p_resume
261 
    # p_description
262 
    # pd_nom            # erreur dans more_recherche_projet.class.php
263 
    # pd_description    # erreur dans more_recherche_projet.class.php
264 
	sql_query		= \
265 
        SELECT p_id, p_titre AS titre, p_description AS texte, \
266 
               p_resume, \
2116 drzraf 267 
               UNIX_TIMESTAMP(p_date_creation) AS date, \
2103 drzraf 268 
               "projet" AS group_id,    \
269 
               CONCAT("projet-", p_id) as main_id \
270 
        FROM projet
271 
 
2116 drzraf 272 
	sql_attr_timestamp	= date
2105 drzraf 273 
	sql_field_string	= group_id
2103 drzraf 274 
	sql_attr_string		= main_id
2115 drzraf 275 
 
276 
	sql_query_info		= SELECT * FROM projet WHERE p_id=$id
2103 drzraf 277 
}
278 
 
279 
index i_projet : i_common {
280 
	source			= projet
2137 drzraf 281 
	path			= /var/lib/sphinxsearch/projet
2103 drzraf 282 
 
283 
    # prefix_fields   = titre, texte
284 
    infix_fields    = titre, texte
285 
}
286 
 
287 
 
288 
 
289 
# BAZAR
290 
source bazar : common {
291 
	sql_query		= \
2116 drzraf 292 
        SELECT bf_id_fiche, bf_description AS texte, bf_titre AS titre, UNIX_TIMESTAMP(bf_date_debut_evenement) AS date, \
293 
        "bazar" AS group_id, CONCAT("bazar-", bf_id_fiche) as main_id \
2103 drzraf 294 
        FROM bazar_fiche
295 
 
2116 drzraf 296 
	sql_attr_timestamp	= date
2105 drzraf 297 
	sql_field_string	= group_id
2103 drzraf 298 
	sql_attr_string		= main_id
2115 drzraf 299 
 
300 
	sql_query_info		= SELECT * FROM bazar_fiche WHERE bf_id_fiche=$id
2103 drzraf 301 
}
302 
 
303 
index i_bazar : i_common {
304 
	source			= bazar
2137 drzraf 305 
	path			= /var/lib/sphinxsearch/bazar
2103 drzraf 306 
 
307 
    # prefix_fields   = titre, texte
308 
    infix_fields    = titre, texte
309 
}
310 
 
2112 drzraf 311 
 
312 
 
313 
# COSTE
314 
source coste : common {
2115 drzraf 315 
	sql_db			= tb_eflore
316 
 
2112 drzraf 317 
	sql_query		= \
2117 drzraf 318 
        SELECT c.flore_bdtfx_nn AS id, c.nom_sci AS titre, dsc.body AS texte, cle.body AS determination, \
319 
        "coste" AS group_id, CONCAT("coste-", c.flore_bdtfx_nn) as main_id	\
2112 drzraf 320 
        FROM tb_eflore.coste_v2_00 c \
321 
        LEFT JOIN tela_prod_wikini.florecoste_pages dsc ON c.page_wiki_dsc = dsc.tag AND dsc.latest = 'Y' \
2117 drzraf 322 
        LEFT JOIN tela_prod_wikini.florecoste_pages cle ON c.page_wiki_cle = cle.tag AND cle.latest = 'Y' \
323 
        WHERE c.flore_bdtfx_nn != ''
2112 drzraf 324 
 
325 
	sql_field_string	= group_id
2117 drzraf 326 
	sql_attr_string		= main_id
2113 drzraf 327 
 
2117 drzraf 328 
	sql_query_info	= SELECT * FROM coste_v2_00  WHERE flore_bdtfx_nn=$id
2112 drzraf 329 
}
330 
 
331 
index i_coste : i_common {
332 
	source			= coste
2137 drzraf 333 
	path			= /var/lib/sphinxsearch/coste
2112 drzraf 334 
 
2117 drzraf 335 
    infix_fields    = texte, titre
2112 drzraf 336 
}
337 
 
2123 drzraf 338 
 
339 
 
340 
# NVJFL
341 
source nvjfl : common {
342 
	sql_db			= tb_eflore
343 
 
344 
    # note: 66 taxons nvjfl n'ont pas de numéro nomenclatural
345 
	sql_query		= \
2129 drzraf 346 
        SELECT b.num_nom, num_taxon, nom_sci AS titre, GROUP_CONCAT(DISTINCT nom_vernaculaire) AS texte, \
2123 drzraf 347 
        "nvjfl" AS group_id, CONCAT("nvjfl-", b.num_nom) as main_id	\
348 
        FROM nvjfl_v2007 n \
2125 drzraf 349 
        LEFT JOIN bdtfx_v1_01 b ON n.num_taxon = b.num_taxonomique \
2123 drzraf 350 
        WHERE b.num_nom IS NOT NULL \
351 
        GROUP by num_taxon
352 
 
2129 drzraf 353 
	sql_field_string	= titre
354 
	sql_field_string	= texte
355 
 
2123 drzraf 356 
	sql_field_string	= group_id
357 
	sql_attr_string		= main_id
358 
    sql_attr_uint       = num_taxon
359 
 
360 
	sql_query_info	= SELECT * FROM nvjfl_v2007 WHERE num_taxon=$id
361 
}
362 
 
363 
index i_nvjfl : i_common {
364 
	source			= nvjfl
2137 drzraf 365 
	path			= /var/lib/sphinxsearch/nvjfl
2123 drzraf 366 
 
367 
    infix_fields    = texte, titre
368 
}
369 
 
370 
 
371 
 
2128 drzraf 372 
# noms scientifiques / noms d'auteur BDTFX
373 
source bdtfx : common {
374 
	sql_db			= tb_eflore
375 
 
376 
	sql_query		= \
377 
        SELECT b.num_nom, b.nom_sci AS nom_sci, b.auteur AS auteur, \
2134 drzraf 378 
        "bdtfx" AS group_id, CONCAT("bdtfx-", b.num_nom) as main_id, b.num_nom, \
379 
        CONCAT(bret.nom_sci, ' ', bret.auteur) AS nom_ret \
380 
        FROM bdtfx_v1_01 b LEFT JOIN bdtfx_v1_01 bret ON b.num_nom_retenu = bret.num_nom \
2128 drzraf 381 
        UNION ALL \
382 
        SELECT a.num_nom + 500000, a.nom_sci, a.auteur, \
2134 drzraf 383 
        "bdtxa" AS group_id, CONCAT("bdtxa-", a.num_nom) AS main_id, a.num_nom,	\
384 
        CONCAT(aret.nom_sci, ' ', aret.auteur) AS nom_ret \
2137 drzraf 385 
        FROM bdtxa_v1_01 a LEFT JOIN bdtxa_v1_01 aret ON a.num_nom_retenu = aret.num_nom \
2128 drzraf 386 
        UNION ALL \
387 
        SELECT i.num_nom + 1000000, i.nom_sci, i.auteur, \
2134 drzraf 388 
        "isfan" AS group_id, CONCAT("isfan-", i.num_nom) AS main_id, i.num_nom, \
389 
        CONCAT(iret.nom_sci, ' ', iret.auteur) AS nom_ret \
390 
        FROM isfan_v2013 i LEFT JOIN isfan_v2013 iret ON i.num_nom_retenu = iret.num_nom
2128 drzraf 391 
 
392 
	sql_field_string	= group_id
393 
	sql_field_string	= nom_sci
394 
	sql_field_string	= auteur
395 
	sql_attr_string		= main_id
396 
    sql_attr_uint       = num_nom
2134 drzraf 397 
    sql_attr_string     = nom_ret
2128 drzraf 398 
 
2137 drzraf 399 
	sql_query_info	= SELECT num_nom, num_nom_retenu, num_tax_sup, rang, nom_sci, genre, auteur, annee, homonyme, nom_francais, num_taxonomique, nom_complet, famille FROM bdtfx_v1_01 \
400 
                          WHERE num_nom=$id UNION \
401 
                          SELECT num_nom, num_nom_retenu, num_tax_sup, rang, nom_sci, genre, auteur, annee, homonyme, nom_francais, num_tax, nom_complet, famille FROM bdtxa_v1_01 \
402 
                          WHERE num_nom=$id UNION \
403 
                          SELECT num_nom, num_nom_retenu, num_tax_sup, rang, nom_sci, genre, auteur, annee, NULL, NULL, num_taxonomique, nom_complet, famille FROM isfan_v2013 \
404 
                          WHERE num_nom=$id
2128 drzraf 405 
}
406 
 
407 
index i_bdtfx : i_common {
408 
	source			= bdtfx
2137 drzraf 409 
	path			= /var/lib/sphinxsearch/bdtfx
2128 drzraf 410 
 
2132 drzraf 411 
	stopwords		=
412 
	morphology		=
2128 drzraf 413 
    infix_fields    = b.nom_sci, b.auteur
2132 drzraf 414 
	enable_star		= 0
415 
    min_word_len    = 1
416 
    min_prefix_len  = 0
417 
    min_infix_len   = 3
418 
 
2128 drzraf 419 
}
420 
 
421 
 
2134 drzraf 422 
# noms scientifiques / noms d'auteur BDTFX2
423 
source bdtfx2 : common {
424 
	sql_db			= tb_eflore
2128 drzraf 425 
 
2134 drzraf 426 
	sql_query		= \
427 
        SELECT b.num_nom, b.nom_sci AS nom_sci, b.auteur AS auteur, \
428 
        "bdtfx2" AS group_id, CONCAT("bdtfx2-", b.num_nom) as main_id, b.num_nom, \
429 
        CONCAT(bret.nom_sci, ' ', bret.auteur) AS nom_ret \
430 
        FROM bdtfx_v2_00 b LEFT JOIN bdtfx_v2_00 bret ON b.num_nom_retenu = bret.num_nom \
431 
        UNION ALL \
432 
        SELECT a.num_nom + 500000, a.nom_sci, a.auteur, \
433 
        "bdtxa" AS group_id, CONCAT("bdtxa-", a.num_nom) AS main_id, a.num_nom,	\
434 
        CONCAT(aret.nom_sci, ' ', aret.auteur) AS nom_ret \
2137 drzraf 435 
        FROM bdtxa_v1_01 a LEFT JOIN bdtxa_v1_01 aret ON a.num_nom_retenu = aret.num_nom \
2134 drzraf 436 
        UNION ALL \
437 
        SELECT i.num_nom + 1000000, i.nom_sci, i.auteur, \
438 
        "isfan" AS group_id, CONCAT("isfan-", i.num_nom) AS main_id, i.num_nom, \
439 
        CONCAT(iret.nom_sci, ' ', iret.auteur) AS nom_ret \
440 
        FROM isfan_v2013 i LEFT JOIN isfan_v2013 iret ON i.num_nom_retenu = iret.num_nom
441 
 
442 
	sql_field_string	= group_id
443 
	sql_field_string	= nom_sci
444 
	sql_field_string	= auteur
445 
	sql_attr_string		= main_id
446 
    sql_attr_uint       = num_nom
447 
    sql_attr_string     = nom_ret
448 
 
2137 drzraf 449 
	sql_query_info	= SELECT num_nom, num_nom_retenu, num_tax_sup, rang, nom_sci, genre, auteur, annee, homonyme, nom_francais, num_taxonomique, nom_complet, famille FROM bdtfx_v2_00 \
450 
                          WHERE num_nom=$id UNION \
451 
                          SELECT num_nom, num_nom_retenu, num_tax_sup, rang, nom_sci, genre, auteur, annee, homonyme, nom_francais, num_tax, nom_complet, famille FROM bdtxa_v1_01 \
452 
                          WHERE num_nom=$id UNION \
453 
                          SELECT num_nom, num_nom_retenu, num_tax_sup, rang, nom_sci, genre, auteur, annee, NULL, NULL, num_taxonomique, nom_complet, famille FROM isfan_v2013 \
454 
                          WHERE num_nom=$id
2134 drzraf 455 
}
456 
 
457 
index i_bdtfx2 : i_common {
458 
	source			= bdtfx2
2137 drzraf 459 
	path			= /var/lib/sphinxsearch/bdtfx2
2134 drzraf 460 
 
461 
	stopwords		=
462 
	morphology		=
463 
    infix_fields    = b.nom_sci, b.auteur
464 
	enable_star		= 0
465 
    min_word_len    = 1
466 
    min_prefix_len  = 0
467 
    min_infix_len   = 3
468 
 
469 
}
470 
 
471 
 
472 
 
2123 drzraf 473 
# COSTE / NVJFL
474 
# source costenvjfl : common {
475 
# 	sql_db			= tb_eflore
476 
 
477 
# 	sql_query		= \
478 
#         SELECT c.flore_bdtfx_nn AS id, c.nom_sci AS titre, dsc.body AS texte, cle.body AS determination, \
479 
#         GROUP_CONCAT(n.nom_vernaculaire) AS nvjfl, n.num_taxon AS num_taxonomique, \
480 
#         "costenvjfl" AS group_id, CONCAT("costenvjfl-", c.flore_bdtfx_nn) as main_id	\
481 
#         FROM tb_eflore.coste_v2_00 c \
482 
#         LEFT JOIN tela_prod_wikini.florecoste_pages dsc ON c.page_wiki_dsc = dsc.tag AND dsc.latest = 'Y' \
483 
#         LEFT JOIN tela_prod_wikini.florecoste_pages cle ON c.page_wiki_cle = cle.tag AND cle.latest = 'Y' \
2125 drzraf 484 
#         LEFT JOIN bdtfx_v1_01 b ON c.flore_bdtfx_nn = b.num_nom \
2123 drzraf 485 
#         LEFT JOIN nvjfl_v2007 n ON b.num_taxonomique = n.num_taxon \
486 
#         WHERE c.flore_bdtfx_nn != ''
487 
 
488 
# 	sql_field_string	= group_id
489 
# 	sql_attr_string		= main_id
490 
#     sql_attr_uint       = num_taxonomique
491 
 
492 
# 	sql_query_info	= SELECT * FROM coste_v2_00  WHERE flore_bdtfx_nn=$id
493 
# }
494 
 
495 
# index i_costenvjfl : i_common {
496 
# 	source			= costenvjfl
2137 drzraf 497 
# 	path			= /var/lib/sphinxsearch/costenvjfl
2123 drzraf 498 
 
499 
#     infix_fields    = texte, titre
500 
# }
501 
 
2117 drzraf 502 
# notes:
503 
# une nouvelle source d'indexation *DOIT*, pour être utilisable par le moteur de recherche papyrus,
504 
# définir les attributs *group_id* et *main_id*.
505 
# cf: sphinx_search dans moteur_recherche_sphinx.php
2112 drzraf 506 
 
2117 drzraf 507 
 
508 
 
2112 drzraf 509 
# misc...
2116 drzraf 510 
# index testrt
511 
# {
512 
# 	type			= rt
513 
# 	rt_mem_limit		= 32M
2103 drzraf 514 
 
2137 drzraf 515 
# 	path			= /var/lib/sphinxsearch/testrt
2116 drzraf 516 
# 	charset_type		= utf-8
2103 drzraf 517 
 
2116 drzraf 518 
# 	rt_field		= title
519 
# 	rt_field		= content
520 
# 	rt_attr_uint		= gid
521 
# }
2103 drzraf 522 
 
523 
 
524 
indexer
525 
{
526 
	mem_limit		= 32M
527 
}
528 
 
529 
 
530 
searchd
531 
{
532 
	listen			= 127.0.0.1:9312
533 
	listen			= 9306:mysql41
2108 drzraf 534 
	log             = /var/log/sphinx/sphinx-searchd.log
2103 drzraf 535 
	query_log		= /var/log/sphinx/sphinx-query.log
2108 drzraf 536 
	read_timeout	= 5
537 
	max_children	= 30
2103 drzraf 538 
	pid_file		= /run/sphinx/sphinx-searchd.pid
539 
	max_matches		= 1000
2108 drzraf 540 
	seamless_rotate	= 1
541 
	preopen_indexes	= 1
2103 drzraf 542 
	unlink_old		= 1
543 
	workers			= threads # for RT to work
544 
	binlog_path		= /var/lib/sphinx
545 
 
2113 drzraf 546 
    # note: les collations n'affectent pas l'indexation mais seulement
547 
    # les attributs:
548 
    # cf: http://sphinxsearch.com/bugs/view.php?id=1597
549 
    # collation_server = utf8_general_ci
2112 drzraf 550 
    # collation_libc_locale = fr_FR.UTF-8
2103 drzraf 551 
    # collation_libc_locale = fr_FR@euro
552 
}
553 
 
554 
 
2112 drzraf 555 
# sudo sphinx-searchd --stopwait -c sphinx-min.conf; sudo sphinx-indexer -c sphinx-min.conf i_projet; sphinx-search -q -c sphinx-min.conf -i i_projet "@texte homogé*"
2103 drzraf 556 
# sudo sphinx-searchd -c sphinx-min.conf
2112 drzraf 557 
 
558 
 
2113 drzraf 559 
# problème de charsets et génération de charset_table pour i_coste:
560 
# + https://raw.github.com/melo/scripts/master/bin/x-sphinx-charset-generator [ moins fiable ]
2112 drzraf 561 
# $ wget https://raw.github.com/tom--/Collation-to-Charset-Table/master/collation_2_charset_table-{1,2}.php
562 
# $ php collation_2_charset_table-1.php|sed -n -e '/^[A-Z]/p'|php collation_2_charset_table-2.php