New project structure

This commit is contained in:
Michael RICOIS 2013-05-13 11:46:57 +00:00
parent 3cd79992d6
commit d6a5260603
31 changed files with 369 additions and 291 deletions

0
README Normal file
View File

View File

@ -1,50 +0,0 @@
source ent
{
type = mysql
sql_host = 192.168.78.230
sql_user = sphinx
sql_pass = indexer
sql_db = jo
sql_query_pre = SET NAMES utf8
sql_query = SELECT id, siren, nic, siege, CONCAT(raisonSociale,\
' "" ',enseigne,' "" ',sigle,' "" ',identite_pre,' "" ',\
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(sigle,' ',''),\
'@','A'), '&',' ET '),'/',''),',',''),'-',''),'.',''),' ',\
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(sigle,\
'@','A'), '&',' ET '),'/',''),',',''),'-',''),'.','')) AS nom, adr_dep, \
actif, adr_num, adr_typeVoie, CONCAT(adr_libVoie,' ',adr_comp) AS adresse, adr_cp, \
adr_ville AS ville, cj, ape_etab, (siren>200) AS sirenValide, rang \
FROM etablissements;
sql_attr_str2ordinal = adr_typeVoie
sql_attr_uint = siren
sql_attr_uint = nic
sql_attr_uint = siege
sql_attr_uint = adr_num
sql_attr_uint = adr_cp
sql_attr_uint = adr_dep
sql_attr_uint = actif
sql_attr_uint = cj
sql_attr_uint = sirenValide
sql_attr_uint = rang
}
index ent
{
source = ent
path = /dbs/sphinx/ent
docinfo = extern
charset_type = utf-8
charset_table = 0..9, A..Z->a..z, a..z, \
U+23, U+25, U+26, U+2B, U+3D, U+40, \
U+C0->a, U+C1->a, U+C2->a, U+C3->a, U+C4->a, U+C5->a, U+C6->a, U+C7->c, \
U+C8->e, U+C8->e, U+C9->e, U+CA->e, U+CB->e, U+CC->i, U+CD->i, U+CE->i, \
U+CF->i, U+D0->d, U+D1->n, U+D2->o, U+D3->o, U+D4->o, U+D5->o, U+D6->o, \
U+D8->o, U+D9->u, U+DA->u, U+DB->u, U+DC->u, U+DD->y, U+E0->a, U+E1->a, \
U+E2->a, U+E3->a, U+E4->a, U+E5->a, U+E7->c, U+E8->e, U+E9->e, U+EA->e, \
U+EB->e, U+EC->i, U+ED->i, U+EE->i, U+EF->i, U+F1->n, U+F2->o, U+F3->o, \
U+F4->o, U+F5->o, U+F6->o, U+F8->o, U+F9->u, U+FA->u,U+FB->u, U+FC->u, \
U+FD->y, U+FF->y, U+0152->U+0153, U+0153
wordforms = /usr/local/sphinx/etc/wordforms-ent.txt
}

View File

@ -1,206 +0,0 @@
& > et
un > 1
deux > 2
trois > 3
quatre > 4
cinq > 5
six > 6
sept > 7
huit > 8
neuf > 9
dix > 10
onze > 11
douze > 12
treize > 13
quatorze > 14
quinze > 15
seize > 16
vingt > 20
vingts > 20
trente > 30
quarante > 40
cinquante > 50
soixante > 60
quatrevingt > 80
cent > 100
cents > 100
mille > 1000
gal > general
abe > abbaye
agl > agglomeration
all > allee
ach > ancien
art > ancienne
anse > anse
arc > arcade
aut > autoroute
av > avenue
bre > barriere
bch > bas
bstd > bastide
bast > baston
begi > beguinage
ber > berge
bois > bois
bcle > boucle
bd > boulevard
brg > bourg
but > butte
cale > cale
camp > camp
cgne > campagne
cpg > camping
carr > carre
cau > carreau
car > carrefour
care > carriere
cst > castel
cav > cavee
central > ctre
centre > ctre
chl > chalet
chp > chapelle
chi > charmille
cht > chateau
chs > chaussee
che > chemin
chv > chemin
chem > cheminement
cloi > cloitre
clos > clos
coli > colline
ctr > contour
cor > corniche
cott > cottage
crs > cours
dars > darse
deg > degre
dsg > descente
dsc > descente
dig > digue
dom > domaine
dr > docteur
eca > ecart
ecl > ecluse
egl > eglise
en > enceinte
env > enclave
enc > enclos
esc > escalier
espa > espace
esp > esplanade
eting > etang
fg > faubourg
frm > ferme
fon > fontaine
fort > fort
form > forum
fos > fosse
foyr > foyer
gal > galerie
gare > gare
garn > garenne
gbd > grand
gden > grande
gr > grande
gri > grille
grim > grimpette
gpe > groupe
gpt > groupement
hle > halle
ham > hameau
hch > haut
hip > hippodrome
imm > immeuble
imp > impasse
jard > jardin
jte > jetee
leve > levee
lieu > ld
lieudit > ld
lot > lotissement
mf > maison
man > manoir
mar > marche
mas > mas
met > metro
mte > montee
mln > moulin
mus > musee
nte > nouvelle
pal > palais
parc > parc
pkg > parking
prv > parvis
pas > passage
pn > passage
pass > passe
ple > passerelle
pat > patio
pav > pavillon
peri > peripherique
psty > peristyle
pta > petites
pae > petite
pim > petite
prt > petite
ptr > petite
pl > place
plci > placis
plag > plage
pln > plaine
plan > plan
plt > plateau
pnt > pointe
pont > pont
pch > porche
port > port
pte > porte
porq > portique
pot > poterne
pour > pourtour
prq > presqu'ile
prom > promenade
qu > quai
quai > quai
qua > quartier
rac > raccourci
raid > raidillon
rpe > rampe
rem > rempart
res > residence
roc > rocade
rpt > rond
roqt > roquet
rtd > rotonde
rte > route
r > rue
rle > ruelle
sente > sen
sentier > sen
sq > square
st > saint
stde > stade
ste > sainte
sta > station
tpl > terre
trn > terrain
tsse > terrasse
trt > tertre
tra > traverse
vallon > val
vallee > val
ven > venelle
via > via
vte > vieille
vche > vieux
vla > villa
vge > village
vlge > village
voi > voie
zac > zone
zad > zone
za > zone
zi > zone
zup > zone

View File

@ -90,7 +90,7 @@ case "$1" in
echo -n "Starting $DESC: "
# Check if we have the configuration file
if [ ! -f /usr/local/sphinx/etc/sphinx.conf ]; then
if [ ! -f /etc/sphinxsearch/sphinx.conf ]; then
echo "Please create an /usr/local/sphinx/sphinx.conf configuration file."
echo "Templates are in the directory."
exit 0

View File

@ -0,0 +1,10 @@
#
# Settings for the sphinxsearch searchd daemon
# Please read /usr/share/doc/sphinxsearch/README.Debian for details.
#
# Should sphinxsearch run automatically on startup? (default: no)
# Before doing this you might want to modify /etc/sphinxsearch/sphinx.conf
# so that it works for you.
START=yes

View File

@ -0,0 +1 @@
/lib/init/upstart-job

View File

@ -0,0 +1,34 @@
# SphinxSearch Service
description "SphinxSearch Daemon"
author "Andrey Aksyonoff <shodan@sphinxsearch.com>"
start on (net-device-up
and local-filesystems
and runlevel [2345])
stop on runlevel [016]
respawn
respawn limit 10 35
# The default of 5 seconds is too low if we have rt indices and have to flush them
kill timeout 30
env DEFAULTFILE="/etc/default/sphinxsearch"
pre-start script
if [ -f "$DEFAULTFILE" ]; then
. "$DEFAULTFILE"
fi
if [ "z$START" != "zyes" ]; then
stop
logger "To enable sphinxsearch, edit /etc/default/sphinxsearch and set START=yes"
exit 0
fi
if [ ! -f /etc/sphinxsearch/sphinx.conf ]; then
logger "Please create an /etc/sphinxsearch/sphinx.conf configuration file."
logger "Templates are in the /etc/sphinxsearch/ directory."
exit 0
fi
end script
exec /usr/local/sphinx/bin/searchd --nodetach

View File

@ -0,0 +1,12 @@
/dbs/sphinxlog/query.log /dbs/sphinxlog/searchd.log {
missingok
notifempty
weekly
rotate 5
compress
delaycompress
create 640 root root
postrotate
killall -SIGUSR1 searchd
endscript
}

View File

@ -32,7 +32,7 @@ According to libstimmer.c/libstimmer/modules.txt, the french module can be refer
Utilisation des stopwords
=========================
une, le, la, les, de, du, dans, l', d', @
un, une, le, la, les, de, des, du, dans, l', d', @

View File

@ -1,23 +1,25 @@
source act
{
type = mysql
sql_host = 192.168.78.230
sql_user = sphinx
sql_pass = indexer
sql_db = jo
sql_query_pre = SET NAMES utf8
sql_query = \
SELECT id, Siren1, RaisonSociale, Siren2, actif, dateLien, IF(dateUpdate>dateInsert,dateUpdate,dateInsert) as dateMaj, \
source, Pmin*1 AS PminNum, MajMin, PpPm, IF(Pays='','FRA',Pays) AS Pays \
FROM liens \
WHERE ActionPart=1 AND Siren1>1000 AND Siren1 IN (SELECT DISTINCT SIREN FROM etablissements);
sql_attr_uint = actif
sql_attr_uint = source
sql_attr_uint = PminNum
sql_attr_str2ordinal= MajMin
sql_attr_str2ordinal= PpPm
sql_attr_str2ordinal= Pays
type = mysql
sql_host = 192.168.78.230
sql_user = sphinx
sql_pass = indexer
sql_db = jo
sql_query_pre = SET NAMES utf8
sql_query = \
SELECT l.id, l.idPar, l.idAct, r.siren, r.actif, r.PpPm, r.RS, r.adresse_cp, r.adresse_ville, p.libPays AS pays, l.PDetention \
FROM liens2 l, liensRef r, tabPays p \
WHERE l.dateSuppr = '0000-00-00 00:00:00' \
AND r.id = l.idAct \
AND ( r.siren>1000 OR (r.siren=0 AND r.adresse_pays!='FRA') OR (r.siren=0 AND r.PpPm='PP') ) \
AND p.codPays3 = r.adresse_pays;
sql_attr_uint = idAct
sql_attr_uint = actif
sql_attr_float= PDetention
sql_attr_str2ordinal = pays
}
index act

View File

@ -0,0 +1,75 @@
source ent
{
type = mysql
sql_host = 192.168.78.230
sql_user = sphinx
sql_pass = indexer
sql_db = jo
sql_query_pre = SET NAMES utf8
sql_query = SELECT id, siren, nic, siege, \
CONCAT_WS(" ", \
raisonSociale, \
enseigne, \
sigle, \
identite_pre\
) AS nom,
REPLACE(REPLACE(adr_dep, '2B', '202'), '2A', '201') AS adr_dep, \
actif, adr_num, CONCAT_WS(" ", adr_typeVoie, adr_libVoie ,adr_comp) AS adresse, adr_cp, \
adr_ville AS ville, cj, ape_etab, (siren>200) AS sirenValide, rang \
FROM etablissements;
sql_attr_uint = siren
sql_attr_uint = nic
sql_attr_uint = siege
sql_attr_uint = adr_dep
sql_attr_uint = actif
sql_attr_uint = adr_num
sql_attr_uint = adr_cp
sql_attr_uint = cj
sql_attr_uint = sirenValide
sql_attr_uint = rang
}
index ent
{
source = ent
path = /dbs/sphinx/ent
docinfo = extern
charset_type = utf-8
charset_table = 0..9, A..Z->a..z, a..z, \
U+23, U+25, U+26, U+2B, U+3D, U+40, \
U+C0->a, U+C1->a, U+C2->a, U+C3->a, U+C4->a, U+C5->a, U+C6->a, U+C7->c, \
U+C8->e, U+C8->e, U+C9->e, U+CA->e, U+CB->e, U+CC->i, U+CD->i, U+CE->i, \
U+CF->i, U+D0->d, U+D1->n, U+D2->o, U+D3->o, U+D4->o, U+D5->o, U+D6->o, \
U+D8->o, U+D9->u, U+DA->u, U+DB->u, U+DC->u, U+DD->y, U+E0->a, U+E1->a, \
U+E2->a, U+E3->a, U+E4->a, U+E5->a, U+E7->c, U+E8->e, U+E9->e, U+EA->e, \
U+EB->e, U+EC->i, U+ED->i, U+EE->i, U+EF->i, U+F1->n, U+F2->o, U+F3->o, \
U+F4->o, U+F5->o, U+F6->o, U+F8->o, U+F9->u, U+FA->u,U+FB->u, U+FC->u, \
U+FD->y, U+FF->y, U+0152->U+0153, U+0153
wordforms = /usr/local/sphinx/etc/wordforms-ent.txt
enable_star = 1
}
index ent_phx
{
source = ent
path = /dbs/sphinx/ent_phx
docinfo = extern
charset_type = utf-8
charset_table = 0..9, A..Z->a..z, a..z, \
U+23, U+25, U+26, U+2B, U+3D, U+40, \
U+C0->a, U+C1->a, U+C2->a, U+C3->a, U+C4->a, U+C5->a, U+C6->a, U+C7->c, \
U+C8->e, U+C8->e, U+C9->e, U+CA->e, U+CB->e, U+CC->i, U+CD->i, U+CE->i, \
U+CF->i, U+D0->d, U+D1->n, U+D2->o, U+D3->o, U+D4->o, U+D5->o, U+D6->o, \
U+D8->o, U+D9->u, U+DA->u, U+DB->u, U+DC->u, U+DD->y, U+E0->a, U+E1->a, \
U+E2->a, U+E3->a, U+E4->a, U+E5->a, U+E7->c, U+E8->e, U+E9->e, U+EA->e, \
U+EB->e, U+EC->i, U+ED->i, U+EE->i, U+EF->i, U+F1->n, U+F2->o, U+F3->o, \
U+F4->o, U+F5->o, U+F6->o, U+F8->o, U+F9->u, U+FA->u,U+FB->u, U+FC->u, \
U+FD->y, U+FF->y, U+0152->U+0153, U+0153
wordforms = /usr/local/sphinx/etc/wordforms-ent.txt
enable_star = 1
morphology = libstemmer_fr
min_stemming_len = 4
}

View File

@ -0,0 +1,209 @@
& > et
un > 1
deux > 2
trois > 3
quatre > 4
cinq > 5
six > 6
sept > 7
huit > 8
neuf > 9
dix > 10
onze > 11
douze > 12
treize > 13
quatorze > 14
quinze > 15
seize > 16
vingt > 20
vingts > 20
trente > 30
quarante > 40
cinquante > 50
soixante > 60
quatrevingt > 80
cent > 100
cents > 100
mille > 1000
zac > zone
zad > zone
za > zone
zi > zone
zup > zone
general > gal
abbaye > abe
agglomeration > agl
aglo > agl
allee > all
ancien > ach
ancienne > art
anse > anse
arcade > arc
autoroute > aut
avenue > av
barriere > bre
bas > bch
bastide > bstd
baston > bast
beguinage > begi
berge > ber
bois > bois
boite postal > bp
boucle > bcle
boulevard > bd
bourg > brg
butte > but
campagne > cgne
camping > cpg
carre > carr
carreau > cau
carrefour > car
carriere > care
castel > cst
cavee > cav
central > ctre
centre > ctre
chalet > chl
chapelle > chp
charmille > chi
chateau > cht
chaussee > chs
che > chemin
chv > chemin
cheminement > chem
cloitre > cloi
colline > coli
contour > ctr
corniche > cor
cottage > cott
cours > crs
darse > dars
degre > deg
dsg > descente
dsc > descente
digue > dig
domaine > dom
docteur > dr
ecart > eca
ecluse > ecl
eglise > egl
enceinte > en
enclave > env
enclos > enc
escalier > esc
espace > espa
esplanade > esp
etang > eting
faubourg > fg
ferme > frm
fontaine > fon
fort > fort
forum > form
fosse > fos
foyer > foyr
galerie > gal
garenne > garn
grand > gbd
gden > grande
gr > grande
grille > gri
grimpette > grim
groupe > gpe
groupement > gpt
halle > hle
hameau > ham
haut > hch
hippodrome > hip
immeuble > imm
impasse > imp
jardin > jard
jetee > jte
levee > leve
lieu > ld
lieudit > ld
lotissement > lot
maison > mf
manoir > man
marche > mar
metro > met
montee > mte
moulin > mln
musee > mus
nouvelle > nte
palais > pal
parc > parc
parking > pkg
parvis > prv
pas > passage
pn > passage
passe > pass
passerelle > ple
patio > pat
pavillon > pav
peripherique > peri
peristyle > psty
petites > pta
pae > petite
pim > petite
prt > petite
ptr > petite
place > pl
placis > plci
plage > plag
plaine > pln
plan > plan
plateau > plt
pointe > pnt
porche > pch
porte > pte
portique > porq
poterne > pot
pourtour > pour
presqu ile > prq
promenade > prom
quai > qu
quartier > qua
raccourci > rac
raidillon > raid
rampe > rpe
rempart > rem
residence > res
rocade > roc
rond > rpt
roquet > roqt
rotonde > rtd
route > rte
rue > r
ruelle > rle
sente > sen
sentier > sen
square > sq
st > saint
ste > saint
sainte > saint
stade > stde
station > sta
terre > tpl
terrain > trn
terrasse > tsse
tertre > trt
traverse > tra
vallon > val
vallee > val
venelle > ven
vieille > vte
villa > vla
vge > village
vlge > village
voie > voi
centre cial > ccal
centre com > ccal
centre comm > ccal
centre commercial > ccal
ville > mairie
commune > mairie
conseil regional > region
conseil régional > region
conseil general > departement
conseil général > departement
companie > cie

View File

@ -7,10 +7,9 @@ Utilisation : <?php echo $argv[0]; ?> [actions]
Installation :
- install list
- install version
- install [version]
Configuration :
- config initscript
- config logrotate
- config etc [version]
- config list
- config server *.conf
<?php
@ -21,9 +20,7 @@ $hostname = exec('echo $HOSTNAME');
$SPHINX_VERSION = array(
'0.9.9' => array('archive'=>'sphinx-0.9.9.tar.gz', 'dir'=>'sphinx-0.9.9'),
'1.10' => array('archive'=>'sphinx-1.10-beta.tar.gz', 'dir'=>'sphinx-1.10-beta'),
'2.0.2' => array('archive'=>'sphinx-2.0.2-beta.tar.gz', 'dir'=>'sphinx-2.0.2-beta'),
'2.0.3' => array('archive'=>'sphinx-2.0.3-release.tar.gz', 'dir'=>'sphinx-2.0.3-release'),
'2.0.4' => array('archive'=>'sphinx-2.0.4-release.tar.gz', 'dir'=>'sphinx-2.0.4-release'),
'2.0.8' => array('archive'=>'sphinx-2.0.8-release.tar.gz', 'dir'=>'sphinx-2.0.8-release'),
);
/**
@ -36,7 +33,6 @@ $SPHINX_VERSION = array(
*
*/
if ($argv[1]=='install')
{
//Version
@ -72,14 +68,9 @@ if ($argv[1]=='install')
if ($argv[1]=='config')
{
//Initscript
if ($argv[2]=='initscript') {
exec("cp -v initscript/sphinxsearch /etc/init.d/");
exit;
}
//Logrotate
if ($argv[2]=='logrotate') {
echo exec("which logrotate");
exec("cp -v logrotate/searchd /etc/logrotate.d/");
if ($argv[2]=='etc') {
$version = $argv[3];
exec("cp -Ri scripts/build/$version/etc/* /etc/");
exit;
}
@ -99,6 +90,6 @@ if ($argv[1]=='config')
for($i=3;$i<count($argv);$i++){
$concat.= " config/".$hostname."/".$server."/".$argv[$i];
}
exec("cat config/".$hostname."/sphinx.conf".$concat." > /usr/local/sphinx/etc/sphinx.conf");
exec("cat config/".$hostname."/sphinx.conf".$concat." > /etc/sphinxsearch/sphinx.conf");
}