534 lines
20 KiB
Plaintext
534 lines
20 KiB
Plaintext
create or replace package body pkg_26ai_TRAFFIC_LOAD
|
|
as
|
|
|
|
/*
|
|
|
|
|
|
Criado por: fernando.leal@oracle.com
|
|
Data: Abril/2026
|
|
Objetivo: Identificar informacoes de imagens como placas, modelo do veiculo e infracoes (uso de celular)
|
|
|
|
v1 - Tarffic Load - leal
|
|
|
|
*/
|
|
|
|
--
|
|
-- chamada a API do Vision. Definir modo de extracao com p_feature_type
|
|
--
|
|
function fnc_26ai_traffic_vision( p_base64_image in clob,
|
|
p_feature_type in varchar2)
|
|
return clob
|
|
--
|
|
-- Visio: https://docs.oracle.com/pt-br/solutions/ai-vision-extract-data/index.html#GUID-A4FD65D0-BF62-472B-B4C7-0ADF5425A566
|
|
--
|
|
as
|
|
/*
|
|
p_feature_type: https://docs.oracle.com/en-us/iaas/api/#/en/vision/20220125/datatypes/ImageFeature
|
|
IMAGE_CLASSIFICATION: Label the image.
|
|
OBJECT_DETECTION: Identify objects in the image with bounding boxes.
|
|
TEXT_DETECTION: Recognize text at the word and line level.
|
|
FACE_DETECTION: Identify faces in the image with bounding boxes and face landmarks.
|
|
*/
|
|
v_endpoint varchar2(500) := 'https://vision.aiservice.us-chicago-1.oci.oraclecloud.com/20220125/actions/analyzeImage';
|
|
request_json CLOB;
|
|
l_response_body clob;
|
|
begin
|
|
|
|
request_json := to_clob('{
|
|
"compartmentId": "' || g_comp_id || '",
|
|
"image": {
|
|
"source":"INLINE",
|
|
"data":"' || p_base64_image || '"
|
|
},
|
|
"features":[
|
|
{
|
|
"featureType":"' || p_feature_type || '",
|
|
"maxResults": 1
|
|
}
|
|
]
|
|
}' );
|
|
|
|
-- Definir os cabeçalhos da requisição
|
|
APEX_WEB_SERVICE.G_REQUEST_HEADERS(1).NAME := 'Content-Type';
|
|
APEX_WEB_SERVICE.G_REQUEST_HEADERS(1).VALUE := 'application/json';
|
|
APEX_WEB_SERVICE.G_REQUEST_HEADERS(2).NAME := 'Accept';
|
|
APEX_WEB_SERVICE.G_REQUEST_HEADERS(2).VALUE := 'application/json';
|
|
|
|
-- Faça a chamada POST usando APEX_WEB_SERVICE e a credencial OCI
|
|
l_response_body := APEX_WEB_SERVICE.make_rest_request(
|
|
p_url => v_endpoint,
|
|
p_http_method => 'POST',
|
|
p_body => request_json,
|
|
p_credential_static_id => 'apex_cred'
|
|
);
|
|
|
|
return l_response_body;
|
|
|
|
end;
|
|
|
|
--
|
|
-- funcao de rag para extracao de informacoes da imagem
|
|
--
|
|
function fnc_26ai_traffic_rag (p_base64_image IN clob,
|
|
p_oci_cred IN VARCHAR2 )
|
|
return clob
|
|
as
|
|
|
|
-- modelos: https://docs.oracle.com/en-us/iaas/Content/generative-ai/pretrained-models.htm
|
|
-- cuidado com pre requisito (1)
|
|
gen_ai_endpoint varchar2(500) := 'https://inference.generativeai.us-chicago-1.oci.oraclecloud.com';
|
|
gen_ai_model varchar2(500) := 'meta.llama-4-maverick-17b-128e-instruct-fp8';
|
|
|
|
chat_resp dbms_cloud_types.RESP;
|
|
|
|
request_json_part1 CLOB;
|
|
request_json_part2 CLOB;
|
|
request_body BLOB;
|
|
BEGIN
|
|
-- create temp blobs
|
|
dbms_lob.createtemporary(request_body, FALSE);
|
|
|
|
request_json_part1 := to_clob(
|
|
'{
|
|
"compartmentId": "' || g_comp_id || '",
|
|
"servingMode":
|
|
{
|
|
"modelId": "' || gen_ai_model || '",
|
|
"servingType": "ON_DEMAND"
|
|
}
|
|
,
|
|
"chatRequest": {
|
|
"apiFormat": "GENERIC",
|
|
"messages": [
|
|
{
|
|
"role": "USER",
|
|
"content": [
|
|
{
|
|
"type": "TEXT",
|
|
"text": "' || 'Gere um JSON com a placa do veiculo, modelo do veiculo e se motorista estiver visivel com uso celular na direção, aponte a infração. Exemplo de saida: {placa:XXXX,modelo:XXXXXXXXXXXXXXXXX,infracao:XXXXXXXXXXXXXX}. Retorne apenas o JSON, sem nenhuma mensagem de introducao nem de explicacao.' || '"
|
|
},
|
|
{
|
|
"type": "IMAGE",
|
|
"imageUrl": {
|
|
"url": "data:image/' || 'png' || ';base64,');
|
|
|
|
request_json_part2 := to_clob('"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"temperature": 0.4,
|
|
"numGenerations": 5,
|
|
"topK": 1
|
|
}
|
|
}');
|
|
|
|
-- append part1 json to request blob
|
|
dbms_lob.append(request_body, apex_util.clob_to_blob(p_clob => request_json_part1,p_charset => 'AL32UTF8'));
|
|
|
|
-- append base64 image to request blob
|
|
dbms_lob.append(request_body, apex_util.clob_to_blob(p_clob => p_base64_image,p_charset => 'AL32UTF8'));
|
|
|
|
-- append part2 json to request blob
|
|
dbms_lob.append(request_body, apex_util.clob_to_blob(p_clob => request_json_part2,p_charset => 'AL32UTF8'));
|
|
|
|
-- call Gen AI Chat
|
|
chat_resp := dbms_cloud.send_request(
|
|
credential_name => p_oci_cred,
|
|
uri => gen_ai_endpoint || '/20231130/actions/chat',
|
|
method => dbms_cloud.METHOD_POST,
|
|
body => request_body
|
|
);
|
|
|
|
-- clear temp blobs
|
|
dbms_lob.freetemporary(request_body);
|
|
|
|
RETURN json_value( dbms_cloud.get_response_text(chat_resp),'$.chatResponse.choices[0].message.content[0].text') ;
|
|
END;
|
|
|
|
|
|
--
|
|
-- Rotina que define a lista de dados a serem vetorizados
|
|
-- Define-se o tipo de embedding para que outros jobs ja existentes possam continuar execucao sem impacto de novas cargas, e assim, testar novos embeddings
|
|
--
|
|
PROCEDURE prc_refresh_files(p_oci_cred IN VARCHAR2 default 'OCI_CRED',
|
|
p_bucket in varchar2 default null,
|
|
p_mimetype in varchar2 default 'PNG',
|
|
p_embedding_name in varchar2 default 'COHERE')
|
|
as
|
|
-- ao levar para ambiente de cliente, definir parametro para p_bucket
|
|
-- default: associado ao bucket de testes - Leal 17-04-26
|
|
v_bucket varchar2(600) := 'https://objectstorage.sa-saopaulo-1.oraclecloud.com/p/cCeVS9davcdjSieWS1H4JOkJs51Ae_-roo4Cr9DGMCE0A7tmx3cHs60ex75D-BX7/n/idi1o0a010nx/b/bucket-public-sector/o/'; -- := p_bucket
|
|
begin
|
|
|
|
-- apenas processar arquivos nao existentes no log de controle tb_26ai_TRAFFIC
|
|
-- tabela de controle dos arquivos que devem ser processados
|
|
-- (1) os nomes de arquivos sao unicos para processamento, por isso ha uma clausula not in que nao insere nomes de arquivos da fila
|
|
-- (2) os arquivos com status "NULL" ainda nao foram vetorizados pela rotina proc_process_files
|
|
-- (3) os arquivos com status "P" foram vetorizados pela rotina proc_process_files
|
|
insert into tb_26ai_traffic(dt_ref,
|
|
object_name,
|
|
status,
|
|
worker_id,
|
|
bytes,
|
|
created,
|
|
last_modified,
|
|
EMBEDDING_NAME,
|
|
EMBEDDING_MIMETYPE,
|
|
IMAGE_BASE64)
|
|
SELECT sysdate,
|
|
object_name , -- cuidado: caracteres especiais tem tratamento de acesso para object storage
|
|
null,
|
|
null,
|
|
BYTES,
|
|
CREATED,
|
|
LAST_MODIFIED,
|
|
p_embedding_name,
|
|
p_mimetype,
|
|
APEX_WEB_SERVICE.BLOB2CLOBBASE64(
|
|
DBMS_CLOUD.GET_OBJECT(
|
|
credential_name => p_oci_cred,
|
|
object_uri => v_bucket || object_name
|
|
),'N','N' )
|
|
FROM table(dbms_cloud.list_objects(credential_name => p_oci_cred,
|
|
location_uri => v_bucket)) mod
|
|
WHERE UTL_URL.ESCAPE( object_name ,TRUE,'AL32UTF8') NOT IN (
|
|
SELECT UTL_URL.ESCAPE( object_name ,TRUE,'AL32UTF8')
|
|
FROM tb_26ai_TRAFFIC)
|
|
AND ( UTL_URL.ESCAPE( lower(object_name) ,TRUE,'AL32UTF8') like '%jpg'
|
|
or UTL_URL.ESCAPE( lower(object_name) ,TRUE,'AL32UTF8') like '%jpeg'
|
|
or UTL_URL.ESCAPE( lower(object_name) ,TRUE,'AL32UTF8') like '%png')
|
|
order by CREATED;
|
|
|
|
commit;
|
|
|
|
end;
|
|
|
|
|
|
--
|
|
-- Isola de modo unico, por worker id (job), o batch de dados que deverao ser processados. Status definido em R (Reservado)
|
|
-- Nao é chamado diretamente, mas faz parte da rotina de embedding
|
|
--
|
|
PROCEDURE prc_reserve_files (
|
|
p_limit IN NUMBER,
|
|
p_worker_id IN VARCHAR2, -- sera setado via job para definir que este bloco de dados sera usado pelo job N
|
|
p_oci_cred IN VARCHAR2 default 'OCI_CRED',
|
|
p_bucket in varchar2 ,
|
|
p_docs OUT SYS.ODCIVARCHAR2LIST,
|
|
p_mimetype in varchar2 default 'PNG',
|
|
p_embedding_name in varchar2 default 'COHERE'
|
|
)
|
|
IS
|
|
BEGIN
|
|
|
|
-- inicialização obrigatória
|
|
p_docs := SYS.ODCIVARCHAR2LIST();
|
|
|
|
FOR r1 IN (
|
|
SELECT object_name
|
|
FROM tb_26ai_TRAFFIC
|
|
WHERE status IS NULL
|
|
and ROWNUM <= p_limit -- CUIDADO: requer ajustes de acordo com tamanho dos dados e servidor
|
|
and EMBEDDING_NAME = p_embedding_name
|
|
and EMBEDDING_MIMETYPE = p_mimetype
|
|
FOR UPDATE SKIP LOCKED
|
|
) LOOP
|
|
|
|
UPDATE tb_26ai_TRAFFIC
|
|
SET status = 'R', -- reserved
|
|
worker_id = p_worker_id -- sera setado via job para definir que este bloco de dados sera usado pelo job N
|
|
WHERE object_name = r1.object_name
|
|
and EMBEDDING_NAME = p_embedding_name
|
|
and EMBEDDING_MIMETYPE = p_mimetype
|
|
and status IS NULL;
|
|
|
|
p_docs.EXTEND;
|
|
p_docs(p_docs.COUNT) := r1.object_name;
|
|
|
|
END LOOP;
|
|
|
|
commit;
|
|
|
|
END;
|
|
|
|
|
|
--
|
|
-- Rotina principal de Embedding
|
|
-- Status de Reserva (R) torna se Started (S)
|
|
-- Se concluir com sucesso, Started (S) torna se Processado (P)
|
|
-- Senao, torna-se Error (E)
|
|
--
|
|
-- Para usar ONNX, importe previamente o ONNX ao banco com comando abaixo. Em seguida, ajuste string de uso, com nome do modelo, nas linhas de codigo do inicio desta rotina.
|
|
--
|
|
/* -- importacao do modelo ao banco
|
|
BEGIN
|
|
DBMS_VECTOR.LOAD_ONNX_MODEL(directory=>'DATA_PUMP_DIR',
|
|
file_name=>'clip-vit-large-patch14_img.onnx',
|
|
model_name=>'OPENAI_CLIP_MULTI_IMG',
|
|
metadata=>JSON('{"function" : "embedding", "embeddingOutput":"embedding", "input": {"input": ["DATA"]}}') );
|
|
|
|
END;
|
|
*/
|
|
--
|
|
-- Pre requisito: create sequence seq_26ai_TRAFFIC MINVALUE 1 INCREMENT BY 1 START WITH 1 CACHE 20 NOORDER NOCYCLE NOKEEP NOSCALE GLOBAL ;
|
|
--
|
|
--
|
|
PROCEDURE proc_process_files(p_limit in number default 10,
|
|
p_worker_id in number,
|
|
p_stop_process_list in varchar2 default 'N',
|
|
p_mimetype in varchar2 default 'PNG',
|
|
p_embedding_name in varchar2 default 'COHERE')
|
|
AS
|
|
v_error CLOB;
|
|
v_session_id NUMBER;
|
|
v_dt_start TIMESTAMP;
|
|
v_oci_cred VARCHAR2(20) := 'OCI_CRED';
|
|
v_bucket VARCHAR2(600) := 'https://objectstorage.sa-saopaulo-1.oraclecloud.com/p/cCeVS9davcdjSieWS1H4JOkJs51Ae_-roo4Cr9DGMCE0A7tmx3cHs60ex75D-BX7/n/idi1o0a010nx/b/bucket-public-sector/o/';
|
|
l_docs SYS.ODCIVARCHAR2LIST;
|
|
|
|
-- CREATE OR REPLACE TYPE t_audio_id_list AS TABLE OF VARCHAR2(600);
|
|
v_all_ids SYS.ODCIVARCHAR2LIST;
|
|
v_json_embedding varchar2(2000);
|
|
|
|
--
|
|
-- CUIDADI: geracao de vetores ira onerar tempo de processamento das imagens
|
|
--
|
|
v_call_vector_embedding varchar2(1) := 'N';
|
|
v_call_rag varchar2(1) := 'Y';
|
|
|
|
BEGIN
|
|
|
|
IF p_stop_process_list != 'N' THEN
|
|
RETURN;
|
|
END IF;
|
|
|
|
if p_embedding_name = 'COHERE' then
|
|
v_json_embedding := '{"provider": "OCIGenAI","credential_name": "OCI_CRED","url": "https://inference.generativeai.us-chicago-1.oci.oraclecloud.com/20231130/actions/embedText","batch_size": 50,"model": "cohere.embed-v4.0"}';
|
|
|
|
elsif p_embedding_name = 'VLLM' then
|
|
v_json_embedding := '{"provider": "openai","url": "https://hub-gpus.DOMINIO.com.br/embed/v1/embeddings","host":"local","batch_size": 100,"model": "Qwen/Qwen3-Embedding-4B"}';
|
|
|
|
elsif p_embedding_name = 'OPENAI' then
|
|
v_json_embedding := '{"provider" : "openai","credential_name" : "CRED_OPENAI", "url":"https://api.openai.com/v1/chat/completions", "model" : "gpt-4.1-mini" }';
|
|
|
|
elsif p_embedding_name = 'ONNX-E5' then
|
|
v_json_embedding := '{"provider":"database", "model":"MULTILINGUAL_E5_BASE"}';
|
|
|
|
elsif p_embedding_name = 'ONNX-VIT' then
|
|
v_json_embedding := '{"provider":"database", "model":"VIT_BASE_PATCH16_224"}';
|
|
|
|
end if;
|
|
|
|
--
|
|
-- rotina que reserva arquivos de modo exclusivo, permitindo uso de scheduler paralelos no banco
|
|
-- objetivo: embedding em sessoes paralelas do banco para diminui tempo de carga
|
|
-- (1) deve ser definido um valor adequado de arquivos por job, definido no limite de linhas (limit)
|
|
-- (2) cada job tem seu worker id definido pelo proprio scheduler
|
|
--
|
|
prc_reserve_files(
|
|
p_limit => p_limit,
|
|
p_worker_id=>p_worker_id,
|
|
p_oci_cred=> v_oci_cred,
|
|
p_bucket=> v_bucket,
|
|
p_docs=>l_docs,
|
|
p_mimetype => p_mimetype,
|
|
p_embedding_name => p_embedding_name ) ;
|
|
|
|
v_dt_start := CURRENT_TIMESTAMP;
|
|
-- create sequence seq_26ai_TRAFFIC MINVALUE 1 INCREMENT BY 1 START WITH 1 CACHE 20 NOORDER NOCYCLE NOKEEP NOSCALE GLOBAL ;
|
|
v_session_id := seq_26ai_TRAFFIC.nextval;
|
|
|
|
SELECT object_name
|
|
BULK COLLECT INTO v_all_ids
|
|
FROM (
|
|
SELECT UTL_URL.ESCAPE( object_name ,TRUE,'AL32UTF8') object_name
|
|
FROM tb_26ai_TRAFFIC src
|
|
WHERE worker_id = p_worker_id
|
|
and status = 'R'
|
|
and EMBEDDING_NAME = p_embedding_name
|
|
and EMBEDDING_MIMETYPE = p_mimetype
|
|
);
|
|
|
|
-- Marca batch de processamento como Iniciado
|
|
UPDATE tb_26ai_TRAFFIC
|
|
SET status = 'S', -- started
|
|
dt_start_process = SYSDATE
|
|
WHERE worker_id = p_worker_id
|
|
and status = 'R'
|
|
and EMBEDDING_NAME = p_embedding_name
|
|
and EMBEDDING_MIMETYPE = p_mimetype;
|
|
|
|
COMMIT;
|
|
|
|
-- embedding
|
|
BEGIN
|
|
|
|
if v_call_vector_embedding = 'Y' then
|
|
|
|
FORALL i IN 1 .. v_all_ids.COUNT
|
|
INSERT INTO tb_26ai_TRAFFIC_vector (
|
|
ID, FILE_NAME, CREATED_DATE, CREATED_BY,
|
|
EMBED_ID, EMBED_DATA, EMBED_VECTOR, EMBED_MODE, MIMETYPE
|
|
)
|
|
SELECT
|
|
v_session_id,
|
|
lower(replace( v_all_ids(i) ,' ','_')),
|
|
CURRENT_TIMESTAMP,
|
|
'admin',
|
|
rownum embed_id,
|
|
null text_chunk,
|
|
t.vec,
|
|
p_embedding_name,
|
|
p_mimetype
|
|
FROM (
|
|
select DBMS_VECTOR.UTL_TO_EMBEDDING(
|
|
DBMS_CLOUD.GET_OBJECT(
|
|
credential_name => v_oci_cred,
|
|
object_uri => v_bucket || v_all_ids(i) ),
|
|
'image',
|
|
JSON('{
|
|
"provider": "OCIGenAI",
|
|
"credential_name": "OCI_CRED",
|
|
"url": "https://inference.generativeai.us-chicago-1.oci.oraclecloud.com/20231130/actions/embedText",
|
|
"model": "cohere.embed-v4.0"}')
|
|
) vec) t;
|
|
|
|
end if; -- v_call_vector_embedding
|
|
|
|
-- Marca batch omo processado e gera analise da imagem com RAG
|
|
if v_call_rag = 'Y' then
|
|
|
|
UPDATE tb_26ai_TRAFFIC
|
|
SET status = 'P', -- processado
|
|
dt_end_process = SYSDATE,
|
|
JSON_DATA = fnc_26ai_traffic_rag (p_base64_image => IMAGE_BASE64, p_oci_cred => v_oci_cred )
|
|
WHERE worker_id = p_worker_id
|
|
and status = 'S';
|
|
|
|
else -- Marca batch omo processado
|
|
|
|
UPDATE tb_26ai_TRAFFIC
|
|
SET status = 'P', -- processado
|
|
dt_end_process = SYSDATE
|
|
WHERE worker_id = p_worker_id
|
|
and status = 'S';
|
|
|
|
end if;
|
|
|
|
|
|
COMMIT;
|
|
EXCEPTION
|
|
WHEN OTHERS THEN
|
|
|
|
v_error := SQLERRM;
|
|
|
|
INSERT INTO tb_26ai_TRAFFIC_debug
|
|
VALUES ('JOB_TRAFFIC_WORKER_' || p_worker_id , v_error, null, SYSDATE);
|
|
|
|
-- Marca batch de PDF como erro
|
|
UPDATE tb_26ai_TRAFFIC
|
|
SET status = 'E', -- erro
|
|
dt_end_process = SYSDATE
|
|
WHERE worker_id = p_worker_id
|
|
and status = 'S';
|
|
COMMIT;
|
|
END;
|
|
|
|
|
|
|
|
COMMIT;
|
|
|
|
END;
|
|
|
|
|
|
PROCEDURE proc_remove_jobs
|
|
AS
|
|
BEGIN
|
|
-- nao pode fazer pelo numero de jobs existentes pois senao a eliminacao seria falha:
|
|
-- cada job tem um padrao de ome associado ao worker id, e nao a contabilizacao que pode ter gaps
|
|
-- ajustar de acordo com maximo permitido pelo item de definicao de novos jobs
|
|
FOR i IN 1..300 LOOP
|
|
begin
|
|
DBMS_SCHEDULER.STOP_JOB( job_name => 'JOB_TRAFFIC_WORKER_' || i );
|
|
exception
|
|
when others then null;
|
|
end;
|
|
|
|
begin
|
|
DBMS_SCHEDULER.DROP_JOB( job_name => 'JOB_TRAFFIC_WORKER_' || i );
|
|
exception
|
|
when others then null;
|
|
end;
|
|
|
|
|
|
update tb_26ai_TRAFFIC
|
|
set status = null, worker_id = null
|
|
where status = 'R'; -- estava reservado, mas com remocao do job volta pra status null sem worker id definido
|
|
|
|
commit;
|
|
|
|
END LOOP;
|
|
END;
|
|
|
|
|
|
PROCEDURE proc_add_jobs(p_limit in number,
|
|
p_total_jobs in number,
|
|
p_mimetype in varchar2,
|
|
p_embedding_name in varchar2 default 'COHERE')
|
|
AS
|
|
v_job_count number;
|
|
v_has number;
|
|
BEGIN
|
|
select count(1)
|
|
into v_job_count
|
|
from user_scheduler_jobs
|
|
where JOB_NAME like 'JOB_TRAFFIC_WORKER_%';
|
|
|
|
-- if p_total_jobs <= v_job_count or p_total_jobs is null then
|
|
|
|
--raise_application_error(-20002,'The number of scheduler jobs must be greater than what already exists');
|
|
|
|
-- else
|
|
|
|
|
|
FOR i IN 1..p_total_jobs LOOP
|
|
|
|
select count(1)
|
|
into v_has
|
|
from user_scheduler_jobs
|
|
where JOB_NAME = 'JOB_TRAFFIC_WORKER_' || i;
|
|
|
|
if v_has = 0 then -- nao existe com worker id "i"
|
|
|
|
DBMS_SCHEDULER.CREATE_JOB(
|
|
job_name => 'JOB_TRAFFIC_WORKER_' || i,
|
|
job_type => 'PLSQL_BLOCK',
|
|
job_action => 'BEGIN pkg_26ai_TRAFFIC_LOAD.proc_process_files(p_limit=>' || p_limit || ',p_worker_id=>' || i || ',p_mimetype=>''' || p_mimetype || ''',p_embedding_name=>''' || p_embedding_name || '''); END;',
|
|
start_date => SYSTIMESTAMP,
|
|
repeat_interval => 'FREQ=SECONDLY; INTERVAL=2;',
|
|
enabled => TRUE
|
|
);
|
|
|
|
else -- cria-se um novo job alem do ultimo
|
|
|
|
DBMS_SCHEDULER.CREATE_JOB(
|
|
job_name => 'JOB_TRAFFIC_WORKER_' || to_char(i+v_job_count),
|
|
job_type => 'PLSQL_BLOCK',
|
|
job_action => 'BEGIN pkg_26ai_TRAFFIC_LOAD.proc_process_files(p_limit=>' || p_limit || ',p_worker_id=>' || to_char(i+v_job_count) || ',p_mimetype=>''' || p_mimetype || ''',p_embedding_name=>''' || p_embedding_name || '''); END;',
|
|
start_date => SYSTIMESTAMP,
|
|
repeat_interval => 'FREQ=SECONDLY; INTERVAL=2;',
|
|
enabled => TRUE
|
|
);
|
|
|
|
end if;
|
|
|
|
END LOOP;
|
|
|
|
|
|
-- end if;
|
|
|
|
END;
|
|
|
|
end;
|
|
/ |