diff --git a/README b/README new file mode 100644 index 0000000..e69de29 diff --git a/Sphinx.sh b/Sphinx.sh deleted file mode 100755 index 3433974..0000000 --- a/Sphinx.sh +++ /dev/null @@ -1,137 +0,0 @@ -#!/bin/bash - -allVersion=("2.0.1" "0.9.9" "2.0.2") -declare -A ERREURS -declare -A ASTUCES - -declare -A ASTUCESDEV - -ROUGE="\\033[1;31m" -NORMAL="\\033[0;39m" -CYAN="\\033[1;36m" -VERT="\\033[1;32m" -BOLD="\\033[1m" -UNDERLINE="\\033[4m" -TEXTNONE="\\033[0m" - -#Var -dir="Exist" -log="Exist" -dbs="Exist" -sphinx="Exist" -sphinxlog="Exist" - -ERREURS=( - ["CHMOD"]="Execute the script with SUDO or in ROOT" - ["DIRECTORY"]="Directory not create" - ["CMD"]="This command not exist" - ["NOCMD"]="Please enter options") - -ASTUCES=( - ["CREATE"]="Sphinx --create (prepare l'environement)"); - -function loadIndication() { - while [ -d "/proc/$1" ] - do #Current PID in /proc - echo -en "Loading... /\033[1G"; sleep .07 - echo -en "Loading... -\033[1G"; sleep .07 - echo -en "Loading... / \033[1G"; sleep .07 - echo -en "Loading... |\033[1G"; sleep .07 - done -} - -if [ `id -u` = 0 ] #Need Root or Sudo -then - commande=('--install' '--all') - for i; do - case $i in - '--help') - echo -e $ROUGE"*************************************************************************************************************************************"$NORMAL - echo -e "--install version : Vous permet d'installer configurer et crée les indexe et la libstemmer compilation, installation et configuration"; - echo -e "--delete : Vous permet de restaurer les parametres de la machines en supprimer la configuration et l'installation" - echo -e "--version : Vous retournes les dernieres version disponnibles en téléchargement sur Sphinx." - echo -e $ROUGE"**************************************************************************************************************************************"$NORMAL - exit; - ;; - '--delete') - `rm -r indexeConf tmp sphinx-* wget* libste* *~ 2> /dev/null` - echo -e "- Clean directory" - exit - ;; - '--version') - echo -e "1. Last version : 2.0.1" - echo -e "2. Past version : 0.9.9\n" - ;; - '--install') - if [ ! -e /dbs ];then echo -e "/dbs not exist !"; exit; fi - if [ ! -e tmp ];then `mkdir tmp`; fi - if [ ! -e /dbs/sphinx ];then `mkdir /dbs/sphinx`; fi - if [ ! -e /usr/local/sphinx ];then `mkdir /dbs/sphinx`; fi - if [ ! -e /dbs/sphinxlog ];then `mkdir /dbs/sphinxlog`; fi - if [ ! -e /init.d/sphinxsearch ];then `cp initscript/sphinxsearch /etc/init.d/ > /dev/null` ;`chmod +x /etc/init.d/sphinxsearch > /dev/null`; `update-rc.d sphinxsearch defaults > /dev/null`; fi - if [ ! -e config/${HOSTNAME} ];then `mkdir config/${HOSTNAME}`;fi - if [ ! -e sphinx-2.0.1.tar.gz ] || [ ! -e sphinx-0.9.9.tar.gz ] || [ ! -e sphinx-2.0.2.tar.gz ] - then - if [ ! -z $2 ]; then - if [ $2 == "2.0.1" ]; then - version="$2-beta" - elif [ $2 == "0.9.9" ]; then - version="$2-beta" - elif [ $2 == "2.0.2" ]; then - version="$2" - else - echo -e $ROUGE"[Erreur]: This version not exist !"$NORMAL - exit - fi - echo -e $CYAN"Download Sphinx-$version..."$NORMAL - `wget -b http://sphinxsearch.com/files/sphinx-$version.tar.gz >/dev/null` - loadIndication `(ps -C wget -o pid=)` - `rm wget-log*` - else - echo -e $ROUGE"[Erreur]: Please specifie version, exemple 2.0.1 or 0.9.9"$NORMAL - exit - fi - fi - `tar -z -xf sphinx-$version.tar.gz -C tmp ` - `rm sphinx-$version.tar.gz*` - echo -e " - Extract files..." - echo -e " - Complited" - #If libstemmer not installed in tmp/ - if [ -e libstemmer_c.tgz ] - then - echo -e $CYAN"Extract file from libstemmer_c..."$NORMAL - `mv libstemmer_c.tgz tmp/sphinx-$version/libstemmer_c; cd tmp/sphinx-$version/libstemmer_c/ ; tar -zxf libstemmer_c.tgz` - echo -e " - Complited" - else - echo -e $CYAN"Download and Configure libstemmer..."$NORMAL - `cd tmp/sphinx-$version ; wget -b http://snowball.tartarus.org/dist/libstemmer_c.tgz > /dev/null` - loadIndication `(ps -C wget -o pid=)` - `cd tmp/sphinx-$version ; tar -zxf libstemmer_c.tgz` - echo -e "- Complited" - fi - echo -e "- Extract files..." - libstemmer="100%" - echo "" - echo -e $CYAN"Configuration progress..."$ROUGE - `cd tmp/sphinx-$version ; sh configure --with-libstemmer --prefix=/usr/local/sphinx 1> /dev/null ` - echo -e $NORMAL"- Complited" - echo -e $NORMAL - echo -e $CYAN"Compilation progress..."$ROUGE - if [ $version != "2.0.1-beta" ];then `cd tmp/sphinx-$version ; make; make install ` ; - else - `cd tmp/sphinx-$version/ ; make; make install`;fi - echo -e $NORMAL" - Complited" - echo -e $VERT"The installation as been finish ! great ! thank's damien :)"$NORMAL - exit - ;; - *) - #Default - echo -e $ROUGE"[Erreur]" ${ERREURS["CMD"]} $NORMAL - exit - ;; - esac - done - echo -e $ROUGE"[Erreur]" ${ERREURS["NOCMD"]} $NORMAL -else - echo -e $ROUGE"[Erreur]" ${ERREURS["CHMOD"]} $NORMAL -fi \ No newline at end of file diff --git a/config/DOC b/config/DOC deleted file mode 100644 index 918f4b1..0000000 --- a/config/DOC +++ /dev/null @@ -1,156 +0,0 @@ -La base de données doit communiqué en UTF-8 - -sql_query_pre = SET NAMES utf8 -charset_type = utf-8 - -# 'utf-8' defaults for English and Russian -charset_table = 0..9, A..Z->a..z, _, a..z, \ - U+410..U+42F->U+430..U+44F, U+430..U+44F - -# For french -charset_table = 0..9, A..Z->a..z, a..z, \ - U+00C0..U+00D6->U+00E0..U+00F6, U+00E0..U+00F6, \ - U+00D8..U+00DE->U+00F8..U+00FE, U+00F8..U+00FE, \ - U+0152->U+0153, U+0153 \ - -Morphology -========== -# 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 = libstemmer_french - -Taille minimum des mots ou l'on applique la morphology -min_stemming_len = 4 - - -According to libstimmer.c/libstimmer/modules.txt, the french module can be refered to by either - french, fr, fre, fra... french UTF_8,ISO_8859_1 french,fr,fre,fra - - -Utilisation des stopwords -========================= -une, le, la, les, de, du, dans, l', d', @ - -Utilisation des wordforms -========================= - -& > ET - -Les chiffres - -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 - -Les chiffres romain - - -Index DIR -========= - -source dir -{ - type = mysql - sql_host = 192.168.3.30 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = - sql_query = SELECT id, siren, actif, \ - IF(civilite='M',1,IF(civilite='MME' OR civilite='MLLE',2,0)) AS genre, \ - CONCAT_WS(' ',dirRS,nom,naissance_nom) AS nom, prenom, \ - YEAR(naissance_date) AS naiss_annee, \ - MONTH(naissance_date) AS naiss_mois, \ - DAY(naissance_date) AS naiss_jour, \ - naissance_lieu, adr_dep \ - FROM rncs_dirigeants; - sql_attr_uint = naiss_annee - sql_attr_uint = naiss_mois - sql_attr_uint = naiss_jour - sql_attr_uint = siren - sql_attr_uint = actif - sql_attr_uint = genre - sql_attr_uint = adr_dep -} - -index dir -{ - source = dir - path = /dbs/sphinx/dir - docinfo = extern - morphology = none - charset_type = sbcs - charset_table = 0..9, A..Z->a..z, a..z, \ - U+23, U+25, U+26, U+2B, U+3D, U+40, \ - U+C0..U+DE->U+E0..U+FE, U+DF, U+E0..U+FF -} - -Index ENT -========= -source ent -{ - type = mysql - sql_host = 192.168.3.30 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = - sql_query = SELECT id, siren, nic, siege, raisonSociale, enseigne, sigle, identite_pre, \ - CONCAT_WS(' ', REPLACE(raisonSociale,' ',''), REPLACE(enseigne,' ',''), \ - REPLACE(sigle,' ',''),REPLACE(identite_pre,' ','')) 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, tel, fax, rang \ - FROM etablissements_tmp; - - 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 = source_ent - path = /dbs/sphinx/ent - docinfo = extern - morphology = libstemmer_french - wordforms = /usr/local/sphinx/etc/wordforms.txt - charset_type = sbcs - charset_table = 0..9, A..Z->a..z, a..z, \ - U+23, U+25, U+26, U+2B, U+3D, U+40, \ - U+C0..U+DE->U+E0..U+FE, U+DF, U+E0..U+FF -} - diff --git a/config/Dev/ent.conf b/config/Dev/ent.conf deleted file mode 100644 index 3460965..0000000 --- a/config/Dev/ent.conf +++ /dev/null @@ -1,40 +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 = sbcs - charset_table = 0..9, A..Z->a..z, a..z, \ - U+23, U+25, U+26, U+2B, U+3D, U+40, \ - U+C0..U+DE->U+E0..U+FE, U+DF, U+E0..U+FF -} diff --git a/config/Dev/enttmp.conf b/config/Dev/enttmp.conf deleted file mode 100644 index 609cf1d..0000000 --- a/config/Dev/enttmp.conf +++ /dev/null @@ -1,41 +0,0 @@ -source ent -{ - type = mysql - sql_host = 192.168.78.230 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = - 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_tmp; - 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 = sbcs - charset_table = 0..9, A..Z->a..z, a..z, \ - U+23, U+25, U+26, U+2B, U+3D, U+40, \ - U+C0..U+DE->U+E0..U+FE, U+DF, U+E0..U+FF -} diff --git a/config/Sphinx/MysqlServer/ent.conf b/config/Sphinx/MysqlServer/ent.conf deleted file mode 100644 index 0017ecf..0000000 --- a/config/Sphinx/MysqlServer/ent.conf +++ /dev/null @@ -1,40 +0,0 @@ -source ent -{ - type = mysql - sql_host = 192.168.3.30 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = - 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 = sbcs - charset_table = 0..9, A..Z->a..z, a..z, \ - U+23, U+25, U+26, U+2B, U+3D, U+40, \ - U+C0..U+DE->U+E0..U+FE, U+DF, U+E0..U+FF -} diff --git a/config/Sphinx/MysqlServer/enttmp.conf b/config/Sphinx/MysqlServer/enttmp.conf deleted file mode 100644 index 5ca1118..0000000 --- a/config/Sphinx/MysqlServer/enttmp.conf +++ /dev/null @@ -1,41 +0,0 @@ -source ent -{ - type = mysql - sql_host = 192.168.3.30 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = - 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_tmp; - 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 = sbcs - charset_table = 0..9, A..Z->a..z, a..z, \ - U+23, U+25, U+26, U+2B, U+3D, U+40, \ - U+C0..U+DE->U+E0..U+FE, U+DF, U+E0..U+FF -} diff --git a/config/SphinxHisto/MysqlServer/dir-rescue.conf b/config/SphinxHisto/MysqlServer/dir-rescue.conf deleted file mode 100644 index 1947c59..0000000 --- a/config/SphinxHisto/MysqlServer/dir-rescue.conf +++ /dev/null @@ -1,64 +0,0 @@ - -source dir -{ - type = mysql - sql_host = 192.168.3.30 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = SET NAMES utf8 - sql_query = SELECT id, siren, adr_dep, typeDir, dirSiren, civilite, CONCAT(nom,' ',naissance_nom, ' ', dirRS) AS nom, prenom, \ - YEAR(naissance_date) AS naiss_annee, \ - MONTH(naissance_date) AS naiss_mois, \ - DAY(naissance_date) AS naiss_jour, \ - naissance_lieu, fonction_code, actif \ - FROM dirigeants; - sql_attr_str2ordinal= civilite - sql_attr_str2ordinal= typeDir - sql_attr_str2ordinal= fonction_code - sql_attr_uint = dirSiren - sql_attr_uint = naiss_annee - sql_attr_uint = naiss_mois - sql_attr_uint = naiss_jour - sql_attr_uint = actif -# sql_attr_uint = genre - sql_attr_uint = adr_dep -} - -index dir -{ - source = dir - path = /dbs/sphinx/dir - docinfo = extern - morphology = none - 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 -} - -index dir_phx -{ - source = dir - path = /dbs/sphinx/dir_phx - docinfo = extern - morphology = libstemmer_french - 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 -} \ No newline at end of file diff --git a/config/SphinxHisto/MysqlServer/dir.conf b/config/SphinxHisto/MysqlServer/dir.conf deleted file mode 100644 index f4186a6..0000000 --- a/config/SphinxHisto/MysqlServer/dir.conf +++ /dev/null @@ -1,64 +0,0 @@ - -source dir -{ - type = mysql - sql_host = 192.168.3.30 - sql_user = sphinx - sql_pass = indexer - sql_db = jo - sql_query_pre = SET NAMES utf8 - sql_query = SELECT id, siren, adr_dep, typeDir, dirSiren, civilite, CONCAT(nom,' ',naissance_nom, ' ', dirRS) AS nom, prenom, \ - YEAR(naissance_date) AS naiss_annee, \ - MONTH(naissance_date) AS naiss_mois, \ - DAY(naissance_date) AS naiss_jour, \ - naissance_lieu, fonction_code, actif \ - FROM dirigeants_tmp; - sql_attr_str2ordinal= civilite - sql_attr_str2ordinal= typeDir - sql_attr_str2ordinal= fonction_code - sql_attr_uint = dirSiren - sql_attr_uint = naiss_annee - sql_attr_uint = naiss_mois - sql_attr_uint = naiss_jour - sql_attr_uint = actif -# sql_attr_uint = genre - sql_attr_uint = adr_dep -} - -index dir -{ - source = dir - path = /dbs/sphinx/dir - docinfo = extern - morphology = none - 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 -} - -index dir_phx -{ - source = dir - path = /dbs/sphinx/dir_phx - docinfo = extern - morphology = libstemmer_french - 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 -} \ No newline at end of file diff --git a/configure.php b/configure.php deleted file mode 100644 index 1428d04..0000000 --- a/configure.php +++ /dev/null @@ -1,83 +0,0 @@ -#!/usr/bin/php - -Utilisation : [actions] - -Installation : - - install list - - install version -Configuration : - - config initscript - - config logrotate - - config list - - config server *.conf - 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'), -); - -if ($argv[1]=='install') -{ - //Version - if (!array_key_exists($argv[2], $SPHINX_VERSION)){ - exit; - } - $version = $argv[2]; - - //Compilation - $cmd = array(); - $cmd[] = "tar xzvf ".$SPHINX_VERSION[$version]['archive']; - $cmd[] = "cp libstemmer_c.tgz ".$SPHINX_VERSION[$version]['dir']."/"; - $cmd[] = "cd ".$SPHINX_VERSION[$version]['dir']."/"; - $cmd[] = "tar xzvf libstemmer_c.tgz"; - $cmd[] = "./configure --with-libstemmer --prefix=/usr/local/sphinx"; - $cmd[] = "make"; - $cmd[] = "make install"; - foreach($cmd as $c){ - passthru($c); - } -} - -if ($argv[1]=='config') -{ - //Initscript - if ($argv[2]=='initscript') { - exec("cp -v initscript/sphinxsearch /etc/init.d/"); - exit; - } - //Logrotate - if ($argv[2]=='logrotate') { - echo exec("which logrotate"); - exec("cp -v logrotate/searchd /etc/logrotate.d/"); - exit; - } - - //Liste des configuration - if ($argv[2]=='list') { - echo passthru("tree config/".$hostname."/"); - exit; - } - - //Configuration - if (!file_exist("config/".$hostname."/".$argv[2])){ - echo "Erreur\n"; exit; - } - $server = $argv[2]; - - $concat = ''; - for($i=3;$i /usr/local/sphinx/etc/sphinx.conf"); -} - diff --git a/indexer/indexerAct.sh b/indexer/indexerAct.sh index 4e88909..95fa4c4 100644 --- a/indexer/indexerAct.sh +++ b/indexer/indexerAct.sh @@ -2,8 +2,6 @@ PATH_BIN=/usr/local/sphinx/bin PATH_LOG=/dbs/sphinxlog -DATE=`date '+%Y-%m-%d %H-%M-%S'` - -echo "DEBUT de l'indexation ==== $(date +%H:%M:%S)" >> $PATH_LOG/indexerAct-$(date +%F).log -$PATH_BIN/indexer --rotate act act_phx >> $PATH_LOG/indexerAct-$(date +%F).log -echo "FIN ===== $(date +%H:%M:%S)" >> $PATH_LOG/indexerAct-$(date +%F).log +echo "$(date '+%Y-%m-%d %H:%M:%S') === DEBUT de l'indexation" >> $PATH_LOG/indexer.log +$PATH_BIN/indexer --rotate act >> $PATH_LOG/indexer.log +echo "$(date '+%Y-%m-%d %H:%M:%S') === FIN" >> $PATH_LOG/indexer.log diff --git a/indexer/indexerDiri.sh b/indexer/indexerDiri.sh index c4511d9..247bf80 100755 --- a/indexer/indexerDiri.sh +++ b/indexer/indexerDiri.sh @@ -1,13 +1,12 @@ #!/bin/bash PATH_BIN=/usr/local/sphinx/bin PATH_LOG=/dbs/sphinxlog +PATH_SQL=/home/scripts/indexer/sql -DATE=`date '+%Y-%m-%d %H-%M-%S'` - -echo "DEBUT de l'indexation ==== $(date +%H:%M:%S)" >> $PATH_LOG/indexerDiri-$(date +%F).log +echo "$(date '+%Y-%m-%d %H:%M:%S') === DEBUT de l'indexation" >> $PATH_LOG/indexer.log # nombre de lignes dans la table etab -output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < sql/countDiri.sql) +output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < $PATH_SQL/countDiri.sql) for line in "$output"; do etab="$line" done @@ -15,7 +14,7 @@ done etab=$(echo $etab|sed -e "s/^[etab ]*//g"||sed -e "s/[ ]*$//g") # nombre de lignes dans la table diri_tmp -output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < sql/countDiri_tmp.sql) +output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < $PATH_SQL/countDiri_tmp.sql) for line in "$output"; do etab_tmp="$line" done @@ -23,15 +22,18 @@ done etab_tmp=$(echo $etab_tmp|sed -e "s/^[etab_tmp ]*//g"||sed -e "s/[ ]*$//g") # Comparaison du nombre de lignes entre etab (ancien) et etab_tmp (nouveau) -if test $etab_tmp -gt $etab; then - echo "Il y a $etab_tmp lignes dans la nouvelle table dirigeants ($etab lignes dans l'ancienne)" >> $PATH_LOG/indexerDiri-$(date +%F).log - $PATH_BIN/indexer --rotate dir dir_phx >> $PATH_LOG/indexerDiri-$(date +%F).log - echo "Sphinx - termine : $(date +%H:%M:%S)" >> $PATH_LOG/indexerDiri-$(date +%F).log - mysql -f -usphinx -h192.168.3.30 -pindexer jo < sql/finIndexationDiri.sql >> $PATH_LOG/indexerDiri-$(date +%F).log +if [ -z "$etab" ]; then + echo "etab = null" +elif [ -z "$etab_tmp" ]; then + echo "etab_tmp = null" +elif [ $etab_tmp -gt $etab ]; then + echo "Il y a $etab_tmp lignes dans la nouvelle table dirigeants ($etab lignes dans l'ancienne)" >> $PATH_LOG/indexer.log + $PATH_BIN/indexer --config /etc/sphinxsearch/sphinx.conf --rotate dir dir_phx >> $PATH_LOG/indexer.log + echo "Sphinx - termine : $(date +%H:%M:%S)" >> $PATH_LOG/indexer.log + mysql -f -usphinx -h192.168.3.30 -pindexer jo < $PATH_SQL/finIndexationDiri.sql >> $PATH_LOG/indexer.log else # Il y a plus de lignes dans l'ancienne table dirigeants ==> On ne fait rien - echo "Il n'y a pas assez de lignes dans la nouvelle table dirigeants ($etab_tmp / $etab) !" >> $PATH_LOG/indexerDiri-$(date +%F).log - exit + echo "Il n'y a pas assez de lignes dans la nouvelle table dirigeants ($etab_tmp / $etab) !" >> $PATH_LOG/indexer.log fi -echo "FIN ===== $(date +%H:%M:%S)" >> $PATH_LOG/indexerDiri-$(date +%F).log +echo "$(date '+%Y-%m-%d %H:%M:%S') === FIN" >> $PATH_LOG/indexer.log diff --git a/indexer/indexerEtab.sh b/indexer/indexerEtab.sh index 5ef9e32..cdeb932 100755 --- a/indexer/indexerEtab.sh +++ b/indexer/indexerEtab.sh @@ -1,13 +1,12 @@ #!/bin/bash PATH_BIN=/usr/local/sphinx/bin PATH_LOG=/dbs/sphinxlog +PATH_SQL=/home/scripts/indexer/sql -DATE=`date '+%Y-%m-%d %H-%M-%S'` - -echo "DEBUT de l'indexation ==== $(date +%H:%M:%S)" >> $PATH_LOG/indexerEtab-$(date +%F).log +echo "$(date '+%Y-%m-%d %H:%M:%S') === DEBUT de l'indexation" >> $PATH_LOG/indexer.log # nombre de lignes dans la table etab -output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < sql/countEtab.sql) +output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < $PATH_SQL/countEtab.sql) for line in "$output"; do etab="$line" done @@ -17,7 +16,7 @@ etab=$(echo $etab|sed -e "s/^[etab ]*//g"||sed -e "s/[ ]*$//g") #etab=$(echo $etab| gawk '{print + 0}') # nombre de lignes dans la table etab_tmp -output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < sql/countEtab_tmp.sql) +output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < $PATH_SQL/countEtab_tmp.sql) for line in "$output"; do etab_tmp="$line" done @@ -27,16 +26,18 @@ etab_tmp=$(echo $etab_tmp|sed -e "s/^[etab_tmp ]*//g"||sed -e "s/[ ]*$//g") #etab_tmp=$(echo $etab_tmp| gawk '{print + 0}') # Comparaison du nombre de lignes entre etab (ancien) et etab_tmp (nouveau) -if test $etab_tmp -gt $etab; then - echo "Il y a $etab_tmp lignes dans la nouvelle table etablissement ($etab lignes dans l'ancienne)" >> $PATH_LOG/indexerEtab-$(date +%F).log -# $PATH_BIN/indexer --rotate --all >> $PATH_LOG/indexer-$(date +%F).log - $PATH_BIN/indexer --rotate ent ent_mns ent_phx >> $PATH_LOG/indexerEtab-$(date +%F).log - echo "Sphinx - terminé : $(date +%H:%M:%S)" >> $PATH_LOG/indexerEtab-$(date +%F).log - mysql -f -usphinx -h192.168.3.30 -pindexer jo < sql/finIndexationEtab.sql >> $PATH_LOG/indexerEtab-$(date +%F).log +if [ -z "$etab" ]; then + echo "etab = null" +elif [ -z "$etab_tmp" ]; then + echo "etab_tmp = null" +elif [ $etab_tmp -gt $etab ]; then + echo "Il y a $etab_tmp lignes dans la nouvelle table etablissement ($etab lignes dans l'ancienne)" >> $PATH_LOG/indexer.log + $PATH_BIN/indexer --config /etc/sphinxsearch/sphinx.conf --rotate ent ent_phx >> $PATH_LOG/indexer.log + echo "Sphinx - terminé : $(date +%H:%M:%S)" >> $PATH_LOG/indexer.log + mysql -f -usphinx -h192.168.3.30 -pindexer jo < $PATH_SQL/finIndexationEtab.sql >> $PATH_LOG/indexer.log else # Il y a plus de lignes dans l'ancienne table etablissements ==> On ne fait rien - echo "Il n'y a pas assez de lignes dans la nouvelle table etablissement ($etab_tmp / $etab) !" >> $PATH_LOG/indexerEtab-$(date +%F).log - exit + echo "Il n'y a pas assez de lignes dans la nouvelle table etablissement ($etab_tmp / $etab) !" >> $PATH_LOG/indexer.log fi -echo "FIN ===== $(date +%H:%M:%S)" >> $PATH_LOG/indexerEtab-$(date +%F).log +echo "$(date '+%Y-%m-%d %H:%M:%S') === FIN" >> $PATH_LOG/indexer.log diff --git a/indexer/sql/countDiri.sql b/indexer/sql/countDiri.sql index 609a33a..7605dc6 100644 --- a/indexer/sql/countDiri.sql +++ b/indexer/sql/countDiri.sql @@ -1 +1 @@ -SELECT count(*) AS etab_tmp from dirigeants; \ No newline at end of file +SELECT count(*) AS etab from dirigeants; \ No newline at end of file diff --git a/indexer/test.sh b/indexer/test.sh new file mode 100644 index 0000000..3fd2dfa --- /dev/null +++ b/indexer/test.sh @@ -0,0 +1,36 @@ +#!/bin/bash +PATH_BIN=/usr/local/sphinx/bin +PATH_LOG=/dbs/sphinxlog +PATH_SQL=/home/scripts/indexer/sql + +# nombre de lignes dans la table etab +output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < $PATH_SQL/countEtab.sql) +for line in "$output"; do + etab="$line" +done +# Suppression fin de ligne +etab=$(echo $etab|sed -e "s/^[etab ]*//g"||sed -e "s/[ ]*$//g") +# Cast de la variable au cas ou +#etab=$(echo $etab| gawk '{print + 0}') + +# nombre de lignes dans la table etab_tmp +output=$(mysql -h192.168.3.30 -usphinx -pindexer jo < $PATH_SQL/countEtab_tmp.sql) +for line in "$output"; do + etab_tmp="$line" +done +# Suppression fin de ligne +etab_tmp=$(echo $etab_tmp|sed -e "s/^[etab_tmp ]*//g"||sed -e "s/[ ]*$//g") +# Cast de la variable au cas ou +#etab_tmp=$(echo $etab_tmp| gawk '{print + 0}') + +# Comparaison du nombre de lignes entre etab (ancien) et etab_tmp (nouveau) +if [ -z "$etab" ]; then + echo "etab = null" +elif [ -z "$etab_tmp" ]; then + echo "etab_tmp = null" +elif [ $etab_tmp -gt $etab ]; then + echo "Il y a $etab_tmp lignes dans la nouvelle table etablissement ($etab lignes dans l'ancienne)" +else + # Il y a plus de lignes dans l'ancienne table etablissements ==> On ne fait rien + echo "Il n'y a pas assez de lignes dans la nouvelle table etablissement ($etab_tmp / $etab) !" +fi \ No newline at end of file diff --git a/initscript/sphinxsearch b/initscript/sphinxsearch deleted file mode 100644 index db33880..0000000 --- a/initscript/sphinxsearch +++ /dev/null @@ -1,147 +0,0 @@ -#! /bin/sh -# -# Written by Miquel van Smoorenburg . -# Modified for Debian -# by Ian Murdock . -# Further changes by Javier Fernandez-Sanguino -# Modified for sphinx by Radu Spineanu -# -# - -### BEGIN INIT INFO -# Provides: sphinxsearch -# Required-Start: $local_fs $remote_fs $syslog $network $time -# Required-Stop: $local_fs $remote_fs $syslog $network -# Default-Start: 2 3 4 5 -# Default-Stop: 0 1 6 -# Short-Description: Fast standalone full-text SQL search engine -### END INIT INFO - -PREFIX=/usr/local/sphinx -DATA=/dbs/sphinx -PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin -DAEMON=$PREFIX/bin/searchd -NAME=sphinx -DESC=sphinx - -test -x $DAEMON || exit 0 - -PIDFILE=/var/log/searchd.pid -DODTIME=1 # Time to wait for the server to die, in seconds - # If this value is set too low you might not - # let some servers to die gracefully and - # 'restart' will not work - -# Include sphinx defaults if available -if [ -f /etc/default/sphinx ] ; then - . /etc/default/sphinx -fi - -set -e - -running_pid() -{ - # Check if a given process pid's cmdline matches a given name - pid=$1 - name=$2 - [ -z "$pid" ] && return 1 - [ ! -d /proc/$pid ] && return 1 - cmd=`cat /proc/$pid/cmdline | tr "\000" "\n"|head -n 1 |cut -d : -f 1` - # Is this the expected child? - [ "$cmd" != "$name" ] && return 1 - return 0 -} - -running() -{ -# Check if the process is running looking at /proc -# (works for all users) - - # No pidfile, probably no daemon present - [ ! -f "$PIDFILE" ] && return 1 - # Obtain the pid and check it against the binary name - pid=`cat $PIDFILE` - running_pid $pid $DAEMON || return 1 - return 0 -} - -force_stop() { -# Forcefully kill the process - [ ! -f "$PIDFILE" ] && return - if running ; then - kill -15 $pid - # Is it really dead? - [ -n "$DODTIME" ] && sleep "$DODTIME"s - if running ; then - kill -9 $pid - [ -n "$DODTIME" ] && sleep "$DODTIME"s - if running ; then - echo "Cannot kill $LABEL (pid=$pid)!" - exit 1 - fi - fi - fi - rm -f $PIDFILE - return 0 -} - -case "$1" in - start) - echo -n "Starting $DESC: " - - # Check if we have the configuration file - if [ ! -f /usr/local/sphinx/etc/sphinx.conf ]; then - echo "Please create an /usr/local/sphinx/sphinx.conf configuration file." - echo "Templates are in the directory." - exit 0 - fi - - start-stop-daemon --start --exec ${DAEMON} - if running ; then - echo "$NAME." - else - echo " ERROR." - fi - ;; - stop) - echo -n "Stopping $DESC: " - start-stop-daemon --stop --quiet --oknodo --pidfile $PIDFILE \ - --exec $DAEMON - echo "$NAME." - ;; - force-stop) - echo -n "Forcefully stopping $DESC: " - force_stop - if ! running ; then - echo "$NAME." - else - echo " ERROR." - fi - ;; - restart) - echo -n "Restarting $DESC: " - start-stop-daemon --stop --quiet --oknodo --pidfile $PIDFILE \ - --exec $DAEMON - [ -n "$DODTIME" ] && sleep $DODTIME - start-stop-daemon --start --exec ${DAEMON} - echo "$NAME." - ;; - - status) - echo -n "$LABEL is " - if running ; then - echo "running" - else - echo " not running." - exit 1 - fi - ;; - *) - N=/etc/init.d/$NAME - # echo "Usage: $N {start|stop|restart|reload|force-reload}" >&2 - echo "Usage: $N {start|stop|restart|force-reload|status|force-stop}" >&2 - exit 1 - ;; -esac - -exit 0 diff --git a/scripts/build/config/Odea/MySqlServer/ciblage.conf b/scripts/build/config/Odea/MySqlServer/ciblage.conf new file mode 100644 index 0000000..4867f7e --- /dev/null +++ b/scripts/build/config/Odea/MySqlServer/ciblage.conf @@ -0,0 +1,157 @@ + +source ciblage +{ + type = mysql + sql_host = 192.168.3.30 + sql_user = sphinx + sql_pass = indexer + sql_db = jo + sql_query_pre = SET NAMES utf8 + sql_query = \ + SELECT id, LPAD(siren, 9, '000000000') AS siren, LPAD(nic, 5, '00000') AS nic, \ + siege, \ + adr_cp, \ + REPLACE(REPLACE(adr_dep, '2B', '202'), '2A', '201') AS adr_dep, \ + IF(tel>0,1,0) AS tel, \ + IF(fax>0,1,0) AS fax, \ + cj, \ + capital, \ + CONCAT('EX ', ape_etab) AS ape_etab, \ + CONCAT('EX ', ape_entrep) AS ape_entrep, \ + age_entrep, \ + age_etab, \ + tca, \ + tcaexp, \ + IF(teff_entrep IS NULL,99,teff_entrep) AS teff_entrep, \ + IF(teff_etab IS NULL,99,teff_etab) AS teff_etab, \ + IF(web='',0,1) AS web, \ + IF(mail='',0,1) AS mail, \ + IF(adrDom>0,1,0) AS adrDom, \ + actifEco, \ + presentRcs, \ + procolHisto, \ + tvaIntraValide, \ + dateCrea_etab,\ + dateCrea_ent, \ + dateImmat, \ + eff_entrep, \ + eff_etab, \ + IF(dirNom='',0,1) AS dirNom, \ + nbEtab, \ + IF(nbMPubli>0,1,0) AS nbMPubli, \ + IF(CAST(sirenGrp AS UNSIGNED)>100,1,0) AS sirenGrp, \ + nbActio, \ + IF(nbActio>0,1,0) AS actio, \ + nbPart, \ + IF(nbPart>0,1,0) AS part, \ + CASE bilType WHEN 'I' THEN 1 WHEN 'R' THEN 2 WHEN 'E' THEN 3 ELSE 0 END as bilType, \ + bilAnnee, \ + bilCloture, \ + bilDuree, \ + bilTca, \ + bilEE, \ + bilFL, \ + bilFK, \ + bilFR, \ + bilGF, \ + bilGP, \ + bilGU, \ + bilGW, \ + bilHD, \ + bilHH, \ + bilHL, \ + bilHM, \ + bilHN, \ + bilYP, \ + CAST(codeCommune AS UNSIGNED) AS codeCommune, \ + CASE zus WHEN '' THEN 0 WHEN 'HORSZONE' THEN 0 WHEN 'NSP' THEN 0 WHEN 'X' THEN 0 ELSE 1 END as zus, \ + CASE zfu WHEN '' THEN 0 WHEN 'HORSZONE' THEN 0 WHEN 'NSP' THEN 0 WHEN 'X' THEN 0 ELSE 1 END as zfu, \ + CASE zru WHEN '' THEN 0 WHEN 'NSP' THEN 0 ELSE 1 END as zru, \ + CASE cucs WHEN '' THEN 0 WHEN 'NSP' THEN 0 ELSE 1 END as cucs, \ + IF(zrr=1,1,0) as zrr, \ + IF(zafr=1,1,0) as zafr, \ + CASE avisCs WHEN 0 THEN 1 WHEN 10 THEN 2 WHEN 15 THEN 3 WHEN 23 THEN 4 WHEN 29 THEN 4 WHEN 39 THEN 4 WHEN 43 THEN 4 WHEN 21 THEN 5 WHEN 26 THEN 5 WHEN 28 THEN 5 WHEN 31 THEN 6 WHEN 50 THEN 6 WHEN 24 THEN 7 ELSE 0 END as avisCs \ + FROM etablissements_act WHERE siren>100; + + sql_field_string = siren + sql_field_string = nic + sql_attr_uint = siege + sql_attr_uint = adr_cp + sql_attr_uint = adr_dep + sql_attr_uint = tel + sql_attr_uint = fax + sql_attr_uint = cj + sql_attr_uint = capital + sql_attr_uint = age_entrep + sql_attr_uint = age_etab + sql_attr_uint = tca + sql_attr_uint = tcaexp + sql_attr_uint = teff_entrep + sql_attr_uint = teff_etab + sql_attr_uint = web + sql_attr_uint = mail + sql_attr_uint = adrDom + sql_attr_uint = actifEco + sql_attr_uint = presentRcs + sql_attr_uint = procolHisto + sql_attr_uint = tvaIntraValide + sql_attr_uint = dateCrea_etab + sql_attr_uint = dateCrea_ent + sql_attr_uint = dateImmat + sql_attr_uint = eff_entrep + sql_attr_uint = eff_etab + sql_attr_uint = dirNom + sql_attr_uint = nbEtab + sql_attr_uint = nbMPubli + sql_attr_uint = sirenGrp + sql_attr_uint = nbActio + sql_attr_uint = actio + sql_attr_uint = nbPart + sql_attr_uint = part + sql_attr_uint = bilType + sql_attr_uint = bilAnnee + sql_attr_uint = bilCloture + sql_attr_uint = bilDuree + sql_attr_uint = bilTca + sql_attr_uint = bilEE + sql_attr_uint = bilFL + sql_attr_uint = bilFK + sql_attr_uint = bilFR + sql_attr_uint = bilGF + sql_attr_uint = bilGP + sql_attr_uint = bilGU + sql_attr_uint = bilGW + sql_attr_uint = bilHD + sql_attr_uint = bilHH + sql_attr_uint = bilHL + sql_attr_uint = bilHM + sql_attr_uint = bilHN + sql_attr_uint = bilYP + sql_attr_uint = codeCommune + sql_attr_uint = zus + sql_attr_uint = zru + sql_attr_uint = zfu + sql_attr_uint = cucs + sql_attr_uint = zrr + sql_attr_uint = zafr + sql_attr_uint = avisCs +} + +index ciblage +{ + source = ciblage + path = /dbs/sphinx/ciblage + docinfo = extern + morphology = none + 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 +} diff --git a/scripts/build/config/Odea/sphinx.conf b/scripts/build/config/Odea/sphinx.conf new file mode 100644 index 0000000..eb7f627 --- /dev/null +++ b/scripts/build/config/Odea/sphinx.conf @@ -0,0 +1,79 @@ + +############################################################################# +## indexer settings +############################################################################# + +indexer +{ + # memory limit, in bytes, kiloytes (16384K) or megabytes (256M) + # optional, default is 32M, max is 2047M, recommended is 256M to 1024M + mem_limit = 256M + + # 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 +} + +############################################################################# +## searchd settings +############################################################################# + +searchd +{ + # IP address to bind on + # optional, default is 0.0.0.0 (ie. listen on all interfaces) + # + # address = 127.0.0.1 + # address = 192.168.0.1 + + + # searchd TCP port number + # mandatory, default is 3312 + listen = 3312 + + # log file, searchd run info is logged here + # optional, default is 'searchd.log' + log = /dbs/sphinxlog/searchd.log + + # query log file, all search queries are logged here + # optional, default is empty (do not log queries) + query_log = /dbs/sphinxlog/query.log + + # client read timeout, seconds + # optional, default is 5 + read_timeout = 5 + + # maximum amount of children to fork (concurrent searches to run) + # optional, default is 0 (unlimited) + max_children = 30 + + # PID file, searchd process ID file name + # mandatory + pid_file = /var/log/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 0 (do not preopen) + preopen_indexes = 1 + + # whether to unlink .old index copies on succesful rotation. + # optional, default is 1 (do unlink) + unlink_old = 1 + + compat_sphinxql_magics=0 +} diff --git a/scripts/build/config/Sphinx/DOC b/scripts/build/config/Sphinx/DOC new file mode 100644 index 0000000..9ee4c35 --- /dev/null +++ b/scripts/build/config/Sphinx/DOC @@ -0,0 +1,90 @@ +La base de données doit communiqué en UTF-8 + +sql_query_pre = SET NAMES utf8 +charset_type = utf-8 + +# 'utf-8' defaults for English and Russian +charset_table = 0..9, A..Z->a..z, _, a..z, \ + U+410..U+42F->U+430..U+44F, U+430..U+44F + +# For french +charset_table = 0..9, A..Z->a..z, a..z, \ + U+00C0..U+00D6->U+00E0..U+00F6, U+00E0..U+00F6, \ + U+00D8..U+00DE->U+00F8..U+00FE, U+00F8..U+00FE, \ + U+0152->U+0153, U+0153 \ + +Morphology +========== +# 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 = libstemmer_french + +Taille minimum des mots ou l'on applique la morphology +min_stemming_len = 4 + + +According to libstimmer.c/libstimmer/modules.txt, the french module can be refered to by either + french, fr, fre, fra... french UTF_8,ISO_8859_1 french,fr,fre,fra + + +Utilisation des stopwords +========================= +un, une, le, la, les, de, des, du, dans, l', d', @ + + + + + +Utilisation des wordforms +========================= +Appliquer après les règles de charset_table +Stemming n'est pas appliquer à ces mots , ils deviennent des exceptions + +Les mots définis sont utilisé pour normalisé les mots durant l'indexation et la recherche, +il est donc nécessaire de réindexé et redémarrer la recherche pour appliquer les changements + +Divers +------ + +& > ET + +Les chiffres +------------ + +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 + +Les chiffres romain +------------------- + + +Les éléments de voies et abbréviations +-------------------------------------- +Voir le wordforms diff --git a/config/SphinxHisto/MysqlServer/act.conf b/scripts/build/config/Sphinx/Dev/act.conf similarity index 50% rename from config/SphinxHisto/MysqlServer/act.conf rename to scripts/build/config/Sphinx/Dev/act.conf index 257e50c..0ba5e99 100644 --- a/config/SphinxHisto/MysqlServer/act.conf +++ b/scripts/build/config/Sphinx/Dev/act.conf @@ -2,7 +2,7 @@ source act { type = mysql - sql_host = 192.168.3.30 + sql_host = 192.168.78.230 sql_user = sphinx sql_pass = indexer sql_db = jo @@ -19,7 +19,7 @@ source act sql_attr_uint = idAct sql_attr_uint = actif sql_attr_float= PDetention - sql_attr_str2ordinal = pays + sql_attr_string = pays } index act @@ -28,15 +28,15 @@ index act path = /dbs/sphinx/act docinfo = extern morphology = none - charset_type = sbcs + 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 -} \ No newline at end of file + 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 +} diff --git a/scripts/build/config/Sphinx/Dev/ciblage.conf b/scripts/build/config/Sphinx/Dev/ciblage.conf new file mode 100644 index 0000000..088bbd8 --- /dev/null +++ b/scripts/build/config/Sphinx/Dev/ciblage.conf @@ -0,0 +1,157 @@ + +source ciblage +{ + 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, LPAD(siren, 9, '000000000') AS siren, LPAD(nic, 5, '00000') AS nic, \ + siege, \ + adr_cp, \ + REPLACE(REPLACE(adr_dep, '2B', '202'), '2A', '201') AS adr_dep, \ + IF(tel>0,1,0) AS tel, \ + IF(fax>0,1,0) AS fax, \ + cj, \ + capital, \ + CONCAT('EX ', ape_etab) AS ape_etab, \ + CONCAT('EX ', ape_entrep) AS ape_entrep, \ + age_entrep, \ + age_etab, \ + tca, \ + tcaexp, \ + IF(teff_entrep IS NULL,99,teff_entrep) AS teff_entrep, \ + IF(teff_etab IS NULL,99,teff_etab) AS teff_etab, \ + IF(web='',0,1) AS web, \ + IF(mail='',0,1) AS mail, \ + IF(adrDom>0,1,0) AS adrDom, \ + actifEco, \ + presentRcs, \ + procolHisto, \ + tvaIntraValide, \ + dateCrea_etab,\ + dateCrea_ent, \ + dateImmat, \ + eff_entrep, \ + eff_etab, \ + IF(dirNom='',0,1) AS dirNom, \ + nbEtab, \ + IF(nbMPubli>0,1,0) AS nbMPubli, \ + IF(CAST(sirenGrp AS UNSIGNED)>100,1,0) AS sirenGrp, \ + nbActio, \ + IF(nbActio>0,1,0) AS actio, \ + nbPart, \ + IF(nbPart>0,1,0) AS part, \ + CASE bilType WHEN 'I' THEN 1 WHEN 'R' THEN 2 WHEN 'E' THEN 3 ELSE 0 END as bilType, \ + bilAnnee, \ + bilCloture, \ + bilDuree, \ + bilTca, \ + bilEE, \ + bilFL, \ + bilFK, \ + bilFR, \ + bilGF, \ + bilGP, \ + bilGU, \ + bilGW, \ + bilHD, \ + bilHH, \ + bilHL, \ + bilHM, \ + bilHN, \ + bilYP, \ + CAST(codeCommune AS UNSIGNED) AS codeCommune, \ + CASE zus WHEN '' THEN 0 WHEN 'HORSZONE' THEN 0 WHEN 'NSP' THEN 0 WHEN 'X' THEN 0 ELSE 1 END as zus, \ + CASE zfu WHEN '' THEN 0 WHEN 'HORSZONE' THEN 0 WHEN 'NSP' THEN 0 WHEN 'X' THEN 0 ELSE 1 END as zfu, \ + CASE zru WHEN '' THEN 0 WHEN 'NSP' THEN 0 ELSE 1 END as zru, \ + CASE cucs WHEN '' THEN 0 WHEN 'NSP' THEN 0 ELSE 1 END as cucs, \ + IF(zrr=1,1,0) as zrr, \ + IF(zafr=1,1,0) as zafr, \ + CASE avisCs WHEN 0 THEN 1 WHEN 10 THEN 2 WHEN 15 THEN 3 WHEN 23 THEN 4 WHEN 29 THEN 4 WHEN 39 THEN 4 WHEN 43 THEN 4 WHEN 21 THEN 5 WHEN 26 THEN 5 WHEN 28 THEN 5 WHEN 31 THEN 6 WHEN 50 THEN 6 WHEN 24 THEN 7 ELSE 0 END as avisCs \ + FROM etablissements_act WHERE siren>100; + + sql_field_string = siren + sql_field_string = nic + sql_attr_uint = siege + sql_attr_uint = adr_cp + sql_attr_uint = adr_dep + sql_attr_uint = tel + sql_attr_uint = fax + sql_attr_uint = cj + sql_attr_uint = capital + sql_attr_uint = age_entrep + sql_attr_uint = age_etab + sql_attr_uint = tca + sql_attr_uint = tcaexp + sql_attr_uint = teff_entrep + sql_attr_uint = teff_etab + sql_attr_uint = web + sql_attr_uint = mail + sql_attr_uint = adrDom + sql_attr_uint = actifEco + sql_attr_uint = presentRcs + sql_attr_uint = procolHisto + sql_attr_uint = tvaIntraValide + sql_attr_uint = dateCrea_etab + sql_attr_uint = dateCrea_ent + sql_attr_uint = dateImmat + sql_attr_uint = eff_entrep + sql_attr_uint = eff_etab + sql_attr_uint = dirNom + sql_attr_uint = nbEtab + sql_attr_uint = nbMPubli + sql_attr_uint = sirenGrp + sql_attr_uint = nbActio + sql_attr_uint = actio + sql_attr_uint = nbPart + sql_attr_uint = part + sql_attr_uint = bilType + sql_attr_uint = bilAnnee + sql_attr_uint = bilCloture + sql_attr_uint = bilDuree + sql_attr_uint = bilTca + sql_attr_uint = bilEE + sql_attr_uint = bilFL + sql_attr_uint = bilFK + sql_attr_uint = bilFR + sql_attr_uint = bilGF + sql_attr_uint = bilGP + sql_attr_uint = bilGU + sql_attr_uint = bilGW + sql_attr_uint = bilHD + sql_attr_uint = bilHH + sql_attr_uint = bilHL + sql_attr_uint = bilHM + sql_attr_uint = bilHN + sql_attr_uint = bilYP + sql_attr_uint = codeCommune + sql_attr_uint = zus + sql_attr_uint = zru + sql_attr_uint = zfu + sql_attr_uint = cucs + sql_attr_uint = zrr + sql_attr_uint = zafr + sql_attr_uint = avisCs +} + +index ciblage +{ + source = ciblage + path = /dbs/sphinx/ciblage + docinfo = extern + morphology = none + 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 +} diff --git a/scripts/build/config/Sphinx/Dev/ciblage.txt b/scripts/build/config/Sphinx/Dev/ciblage.txt new file mode 100644 index 0000000..5cede41 --- /dev/null +++ b/scripts/build/config/Sphinx/Dev/ciblage.txt @@ -0,0 +1,110 @@ +############################################################################# +# Documentation +# ============= +# +# Présence d'éléments +# IF(element=='',0,1) AS pElement +# +# Liste des champs dans la BDD +# ----------------------------- +# id +# source +# source_id +# triCode +# autre_id +# siren => LPAD(siren, 9, '000000000') +# nic => LPAD(nic, 5, '00000') +# actif => 0 ou 1 +# siege => 0 ou 1 +# raisonSociale +# enseigne +# sigle +# identite_pre +# marques => IF(marques=='',0,1) AS pMarques +# adr_num +# adr_btq +# adr_typeVoie +# adr_libVoie +# adr_comp +# adr_cp => Code postal +# adr_ville => Ville +# adr_dep => Departement (Corse 2A,2B => 201, 202) +# adr_com +# tel => IF(tel>0,1,0) AS pTel +# fax => IF(fax>0,1,0) AS pFax +# cj => Texte +# capital => Float +# capitalDev +# capitalSrc +# ape_etab => Texte => Code APE de l'etablissement +# ape_entrep => Texte => Code APE de l'entreprise +# age_entrep => Entier => Age de l'entreprise +# age_etab => Entier => Age de l'etablissement +# tca => Entier => Tranche de chiffre d'affaire +# tcaexp => Entier => Tranche de chiffre d'affaire à l'export +# teff_entrep => Tranche Effectif de l'entreprise +# teff_etab => Tranche Effectif de l'etablissement +# rang +# web => IF(web=='',0,1) AS pWeb +# mail => IF(mail=='',0,1) AS pMail +# adrDom => 0,1,2 +# lieuAct +# actifEco => 0,1 +# presentRcs => 0,1 +# procolHisto => +# tvaIntraCle +# tvaIntraValide => 0,1 +# ape4_etab +# ape4_entrep +# NaceEtab +# NaceEntrep +# dateCrea_etab => Date +# dateCrea_ent => Date +# dateImmat => Date +# eff_entrep => Entier => Effectif de l'entreprise +# eff_etab => Entier => Effectir de l'etablissement +# distSP +# achPost +# rivoli +# dirCiv +# dirNom => IF(pDirNom=='',0,1) AS pDirNom +# dirPrenom +# dirDateNaiss +# dirFct +# nbEtab => Nombre d'établissement +# nbMPubli +# sirenGrp => IF(pSirenGrp>0,1,0) AS pSirenGrp +# nbActio => Entier => Nombre d'actionnaires (actio,bool) +# nbPart => Entier => Nombre de participations (part, bool) +# bilType => +# bilAnnee => +# bilCloture => +# bilDuree => +# bilTca => +# bilEE => +# bilFL => +# bilFK => +# bilFR => +# bilGF => +# bilGP => +# bilGU => +# bilGW => +# bilHD => +# bilHH => +# bilHL => +# bilHM => +# bilHN => +# bilYP => +# avisCs +# codeCommune => +# l93_x +# l93_y +# alt +# precis +# zus => +# zru => +# zfu => +# cucs => +# zrr => +# zafr => +############################################################################# \ No newline at end of file diff --git a/scripts/build/config/Sphinx/Dev/dir.conf b/scripts/build/config/Sphinx/Dev/dir.conf new file mode 100644 index 0000000..c390aa5 --- /dev/null +++ b/scripts/build/config/Sphinx/Dev/dir.conf @@ -0,0 +1,64 @@ + +source dir +{ + 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, adr_dep, typeDir, dirSiren, civilite, CONCAT(nom,' ',naissance_nom, ' ', dirRS) AS nom, prenom, \ + YEAR(naissance_date) AS naiss_annee, \ + MONTH(naissance_date) AS naiss_mois, \ + DAY(naissance_date) AS naiss_jour, \ + naissance_lieu, fonction_code, actif \ + FROM dirigeants; + sql_attr_string = civilite + sql_attr_string = typeDir + sql_attr_string = fonction_code + sql_attr_uint = dirSiren + sql_attr_uint = naiss_annee + sql_attr_uint = naiss_mois + sql_attr_uint = naiss_jour + sql_attr_uint = actif + sql_attr_uint = adr_dep +} + +index dir +{ + source = dir + path = /dbs/sphinx/dir + docinfo = extern + morphology = none + 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 +} + +index dir_phx +{ + source = dir + path = /dbs/sphinx/dir_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 + min_stemming_len = 4 +} \ No newline at end of file diff --git a/scripts/build/config/Sphinx/Dev/dirtmp.conf b/scripts/build/config/Sphinx/Dev/dirtmp.conf new file mode 100644 index 0000000..87c3399 --- /dev/null +++ b/scripts/build/config/Sphinx/Dev/dirtmp.conf @@ -0,0 +1,63 @@ +source dir +{ + type = mysql + sql_host = 192.168.78.230 + sql_user = sphinx + sql_pass = indexer + sql_db = jo + sql_query_pre = + sql_query = \ + SELECT id, siren, adr_dep, typeDir, dirSiren, civilite, CONCAT(nom,' ',naissance_nom, ' ', dirRS) AS nom, prenom, \ + YEAR(naissance_date) AS naiss_annee, \ + MONTH(naissance_date) AS naiss_mois, \ + DAY(naissance_date) AS naiss_jour, \ + naissance_lieu, fonction_code, actif \ + FROM dirigeants_tmp; + sql_attr_string = civilite + sql_attr_string = typeDir + sql_attr_string = fonction_code + sql_attr_uint = dirSiren + sql_attr_uint = naiss_annee + sql_attr_uint = naiss_mois + sql_attr_uint = naiss_jour + sql_attr_uint = actif + sql_attr_uint = adr_dep +} + +index dir +{ + source = dir + path = /dbs/sphinx/dir + docinfo = extern + morphology = none + 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 +} + +index dir_phx +{ + source = dir + path = /dbs/sphinx/dir_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 + min_stemming_len = 4 +} \ No newline at end of file diff --git a/scripts/build/config/Sphinx/Dev/ent.conf b/scripts/build/config/Sphinx/Dev/ent.conf new file mode 100644 index 0000000..58b4b9b --- /dev/null +++ b/scripts/build/config/Sphinx/Dev/ent.conf @@ -0,0 +1,74 @@ + +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 = actif + sql_attr_uint = adr_num + sql_attr_uint = adr_cp + sql_attr_uint = adr_dep + 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 = /etc/sphinxsearch/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 +} diff --git a/scripts/build/config/Sphinx/Dev/enttmp.conf b/scripts/build/config/Sphinx/Dev/enttmp.conf new file mode 100644 index 0000000..6f7ed4a --- /dev/null +++ b/scripts/build/config/Sphinx/Dev/enttmp.conf @@ -0,0 +1,74 @@ + +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_tmp; + + sql_attr_uint = siren + sql_attr_uint = nic + sql_attr_uint = siege + sql_attr_uint = actif + sql_attr_uint = adr_num + sql_attr_uint = adr_cp + sql_attr_uint = adr_dep + 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 = /etc/sphinxsearch/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 +} diff --git a/config/Dev/histo.conf b/scripts/build/config/Sphinx/Dev/histo.conf similarity index 78% rename from config/Dev/histo.conf rename to scripts/build/config/Sphinx/Dev/histo.conf index 063282c..35bdff2 100644 --- a/config/Dev/histo.conf +++ b/scripts/build/config/Sphinx/Dev/histo.conf @@ -1,9 +1,10 @@ + source histo { type = mysql sql_host = 192.168.78.230 - sql_user = sphinx_histo - sql_pass = sphinx + sql_user = sphinx + sql_pass = indexer sql_db = histobodacc sql_port = 3306 # optional, default is 3306 sql_query = SELECT id, nomFichier, annee1, dateBod, texte FROM bodacc_ocr; @@ -14,10 +15,10 @@ source histo index histo { source = histo - path = /home/sphinx/histo + path = /dbs/sphinx/histo docinfo = extern morphology = none - min_word_len = 1 + min_word_len = 2 charset_type = sbcs html_strip = 1 -} +} \ No newline at end of file diff --git a/scripts/build/config/Sphinx/MysqlServer/ent.conf b/scripts/build/config/Sphinx/MysqlServer/ent.conf new file mode 100644 index 0000000..fe84d10 --- /dev/null +++ b/scripts/build/config/Sphinx/MysqlServer/ent.conf @@ -0,0 +1,73 @@ +source ent +{ + type = mysql + sql_host = 192.168.3.30 + 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 = actif + sql_attr_uint = adr_num + sql_attr_uint = adr_cp + sql_attr_uint = adr_dep + 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 = /etc/sphinxsearch/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 +} diff --git a/scripts/build/config/Sphinx/MysqlServer/enttmp.conf b/scripts/build/config/Sphinx/MysqlServer/enttmp.conf new file mode 100644 index 0000000..103d122 --- /dev/null +++ b/scripts/build/config/Sphinx/MysqlServer/enttmp.conf @@ -0,0 +1,73 @@ +source ent +{ + type = mysql + sql_host = 192.168.3.30 + 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_tmp; + + sql_attr_uint = siren + sql_attr_uint = nic + sql_attr_uint = siege + sql_attr_uint = actif + sql_attr_uint = adr_num + sql_attr_uint = adr_cp + sql_attr_uint = adr_dep + 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 = /etc/sphinxsearch/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 +} diff --git a/scripts/build/config/Sphinx/sphinx.conf b/scripts/build/config/Sphinx/sphinx.conf new file mode 100644 index 0000000..eb7f627 --- /dev/null +++ b/scripts/build/config/Sphinx/sphinx.conf @@ -0,0 +1,79 @@ + +############################################################################# +## indexer settings +############################################################################# + +indexer +{ + # memory limit, in bytes, kiloytes (16384K) or megabytes (256M) + # optional, default is 32M, max is 2047M, recommended is 256M to 1024M + mem_limit = 256M + + # 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 +} + +############################################################################# +## searchd settings +############################################################################# + +searchd +{ + # IP address to bind on + # optional, default is 0.0.0.0 (ie. listen on all interfaces) + # + # address = 127.0.0.1 + # address = 192.168.0.1 + + + # searchd TCP port number + # mandatory, default is 3312 + listen = 3312 + + # log file, searchd run info is logged here + # optional, default is 'searchd.log' + log = /dbs/sphinxlog/searchd.log + + # query log file, all search queries are logged here + # optional, default is empty (do not log queries) + query_log = /dbs/sphinxlog/query.log + + # client read timeout, seconds + # optional, default is 5 + read_timeout = 5 + + # maximum amount of children to fork (concurrent searches to run) + # optional, default is 0 (unlimited) + max_children = 30 + + # PID file, searchd process ID file name + # mandatory + pid_file = /var/log/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 0 (do not preopen) + preopen_indexes = 1 + + # whether to unlink .old index copies on succesful rotation. + # optional, default is 1 (do unlink) + unlink_old = 1 + + compat_sphinxql_magics=0 +} diff --git a/scripts/build/config/Sphinx/wordforms-ent.txt b/scripts/build/config/Sphinx/wordforms-ent.txt new file mode 100644 index 0000000..929912d --- /dev/null +++ b/scripts/build/config/Sphinx/wordforms-ent.txt @@ -0,0 +1,207 @@ +& > 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 general > departement +companie > cie diff --git a/config/Dev/act.conf b/scripts/build/config/SphinxHisto/MysqlServer/act.conf similarity index 52% rename from config/Dev/act.conf rename to scripts/build/config/SphinxHisto/MysqlServer/act.conf index 18f3370..6613a28 100644 --- a/config/Dev/act.conf +++ b/scripts/build/config/SphinxHisto/MysqlServer/act.conf @@ -1,22 +1,24 @@ + 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_old); - 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 = 192.168.3.30 + 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_string = pays } index act diff --git a/config/Dev/dir.conf b/scripts/build/config/SphinxHisto/MysqlServer/dirtmp.conf similarity index 75% rename from config/Dev/dir.conf rename to scripts/build/config/SphinxHisto/MysqlServer/dirtmp.conf index ca23d07..cf94333 100644 --- a/config/Dev/dir.conf +++ b/scripts/build/config/SphinxHisto/MysqlServer/dirtmp.conf @@ -1,7 +1,7 @@ source dir { type = mysql - sql_host = 192.168.78.230 + sql_host = 192.168.3.30 sql_user = sphinx sql_pass = indexer sql_db = jo @@ -12,16 +12,16 @@ source dir MONTH(naissance_date) AS naiss_mois, \ DAY(naissance_date) AS naiss_jour, \ naissance_lieu, fonction_code, actif \ - FROM dirigeants; - sql_attr_str2ordinal = civilite - sql_attr_str2ordinal = typeDir - sql_attr_str2ordinal = fonction_code - sql_attr_uint = dirSiren - sql_attr_uint = naiss_annee - sql_attr_uint = naiss_mois - sql_attr_uint = naiss_jour - sql_attr_uint = actif - sql_attr_uint = adr_dep + FROM dirigeants_tmp; + sql_attr_string = civilite + sql_attr_string = typeDir + sql_attr_string = fonction_code + sql_attr_uint = dirSiren + sql_attr_uint = naiss_annee + sql_attr_uint = naiss_mois + sql_attr_uint = naiss_jour + sql_attr_uint = actif + sql_attr_uint = adr_dep } index dir @@ -30,7 +30,7 @@ index dir path = /dbs/sphinx/dir docinfo = extern morphology = none - charset_type = utf8 + 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, \ @@ -41,4 +41,4 @@ index dir 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 -} +} \ No newline at end of file diff --git a/config/SphinxHisto/MysqlServer/histo.conf b/scripts/build/config/SphinxHisto/MysqlServer/histo.conf similarity index 100% rename from config/SphinxHisto/MysqlServer/histo.conf rename to scripts/build/config/SphinxHisto/MysqlServer/histo.conf diff --git a/config/SphinxHisto/sphinx.conf b/scripts/build/config/SphinxHisto/sphinx.conf similarity index 100% rename from config/SphinxHisto/sphinx.conf rename to scripts/build/config/SphinxHisto/sphinx.conf diff --git a/config/Sphinx/sphinx.conf b/scripts/build/config/srvws02/sphinx.conf similarity index 98% rename from config/Sphinx/sphinx.conf rename to scripts/build/config/srvws02/sphinx.conf index af10d95..b518273 100644 --- a/config/Sphinx/sphinx.conf +++ b/scripts/build/config/srvws02/sphinx.conf @@ -12,7 +12,7 @@ indexer # maximum IO calls per second (for I/O throttling) # optional, default is 0 (unlimited) # - # max_iops = 40 + max_iops = 40 # maximum IO call size, bytes (for I/O throttling) @@ -36,7 +36,7 @@ searchd # searchd TCP port number # mandatory, default is 3312 - listen = 3312 + listen = 9312 # log file, searchd run info is logged here # optional, default is 'searchd.log' diff --git a/config/Dev/dirtmp.conf b/scripts/build/config/srvws02/srvws02/dir.conf similarity index 88% rename from config/Dev/dirtmp.conf rename to scripts/build/config/srvws02/srvws02/dir.conf index 7b056a5..129a5aa 100644 --- a/config/Dev/dirtmp.conf +++ b/scripts/build/config/srvws02/srvws02/dir.conf @@ -1,7 +1,7 @@ source dir { type = mysql - sql_host = 192.168.78.230 + sql_host = 192.168.3.30 sql_user = sphinx sql_pass = indexer sql_db = jo @@ -12,7 +12,7 @@ source dir MONTH(naissance_date) AS naiss_mois, \ DAY(naissance_date) AS naiss_jour, \ naissance_lieu, fonction_code, actif \ - FROM dirigeants_tmp; + FROM dirigeants; sql_attr_str2ordinal = civilite sql_attr_str2ordinal = typeDir sql_attr_str2ordinal = fonction_code @@ -21,12 +21,12 @@ source dir sql_attr_uint = naiss_mois sql_attr_uint = naiss_jour sql_attr_uint = actif - sql_attr_uint = adr_dep + sql_attr_uint = adr_dep } index dir { - source = dir + source = source_dir path = /dbs/sphinx/dir docinfo = extern morphology = none diff --git a/scripts/build/configure.php b/scripts/build/configure.php new file mode 100644 index 0000000..ac69154 --- /dev/null +++ b/scripts/build/configure.php @@ -0,0 +1,97 @@ +#!/usr/bin/php + +Utilisation : [actions] + +Installation : + - install list + - install [version] +Configuration : + - config etc [version] + - config list + - config server *.conf + 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'), + '2.0.8' => array('archive'=>'sphinx-2.0.8-release.tar.gz', 'dir'=>'sphinx-2.0.8-release'), + '2.1.2' => array('archive'=>'sphinx-2.1.2-release.tar.gz', 'dir'=>'sphinx-2.1.2-release'), + '2.1.7' => array('archive'=>'sphinx-2.1.7-release.tar.gz', 'dir'=>'sphinx-2.1.7-release'), +); + +/** + * Nécessaire : + * ============ + * GCC : gcc g++ make + * MySQL : libmysqld-dev + * + * tree + * + */ + +if ($argv[1]=='install') +{ + //Version + if (!array_key_exists($argv[2], $SPHINX_VERSION)){ + echo "Erreur.\n"; exit; + } + $version = $argv[2]; + + //Vérification nécessaires + + //Compilation + $cmd = array(); + $cmd[] = "tar xzvf ".$SPHINX_VERSION[$version]['archive']; + $cmd[] = "cp libstemmer_c.tgz ".$SPHINX_VERSION[$version]['dir']."/"; + + foreach($cmd as $c){ + passthru($c); + } + + //Changement du répertoire pour la compilation + chdir($SPHINX_VERSION[$version]['dir']); + + $cmd = array(); + $cmd[] = "tar xzvf libstemmer_c.tgz"; + $cmd[] = "./configure --with-libstemmer --prefix=/usr/local/sphinx"; + $cmd[] = "make"; + $cmd[] = "make install"; + foreach($cmd as $c){ + passthru($c); + } +} + +if ($argv[1]=='config') +{ + //Initscript + if ($argv[2]=='etc') { + $version = $argv[3]; + exec("cp -vRi scripts/build/$version/etc/* /etc/"); + exit; + } + + //Liste des configurations + if ($argv[2]=='list') { + echo passthru("tree config/".$hostname."/"); + exit; + } + + //Configuration + $server = $argv[2]; + if (!file_exists("config/".$hostname."/".$server)){ + echo "Erreur\n"; exit; + } + + $concat = ''; + for($i=3;$i /etc/sphinxsearch/sphinx.conf"); +} + diff --git a/scripts/build/ubuntu-12.04/etc/default/sphinxsearch b/scripts/build/ubuntu-12.04/etc/default/sphinxsearch new file mode 100644 index 0000000..8d9bf66 --- /dev/null +++ b/scripts/build/ubuntu-12.04/etc/default/sphinxsearch @@ -0,0 +1,10 @@ +# +# Settings for the sphinxsearch searchd daemon +# Please read /usr/share/doc/sphinxsearch/README.Debian for details. +# + +# Should sphinxsearch run automatically on startup? (default: no) +# Before doing this you might want to modify /etc/sphinxsearch/sphinx.conf +# so that it works for you. +START=yes + diff --git a/scripts/build/ubuntu-12.04/etc/init.d/sphinxsearch b/scripts/build/ubuntu-12.04/etc/init.d/sphinxsearch new file mode 100644 index 0000000..0138977 --- /dev/null +++ b/scripts/build/ubuntu-12.04/etc/init.d/sphinxsearch @@ -0,0 +1 @@ +/lib/init/upstart-job \ No newline at end of file diff --git a/scripts/build/ubuntu-12.04/etc/init/sphinxsearch.conf b/scripts/build/ubuntu-12.04/etc/init/sphinxsearch.conf new file mode 100644 index 0000000..c2da12e --- /dev/null +++ b/scripts/build/ubuntu-12.04/etc/init/sphinxsearch.conf @@ -0,0 +1,34 @@ +# SphinxSearch Service + +description "SphinxSearch Daemon" +author "Andrey Aksyonoff " + +start on (net-device-up + and local-filesystems + and runlevel [2345]) +stop on runlevel [016] + +respawn +respawn limit 10 35 + +# The default of 5 seconds is too low if we have rt indices and have to flush them +kill timeout 30 + +env DEFAULTFILE="/etc/default/sphinxsearch" +pre-start script + if [ -f "$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 --config /etc/sphinxsearch/sphinx.conf diff --git a/scripts/build/ubuntu-12.04/etc/logrotate.d/indexer b/scripts/build/ubuntu-12.04/etc/logrotate.d/indexer new file mode 100644 index 0000000..222a5dc --- /dev/null +++ b/scripts/build/ubuntu-12.04/etc/logrotate.d/indexer @@ -0,0 +1,8 @@ +/dbs/sphinxlog/indexer.log { + missingok + notifempty + weekly + rotate 5 + compress + delaycompress +} \ No newline at end of file diff --git a/logrotate/searchd b/scripts/build/ubuntu-12.04/etc/logrotate.d/searchd similarity index 100% rename from logrotate/searchd rename to scripts/build/ubuntu-12.04/etc/logrotate.d/searchd