Add project files:

- Add database initialization scripts - Add configuration files - Add documentation - Add public assets - Add source code structure - Update README
2025-07-11 19:18:40 +00:00 · 2025-01-04 17:22:46 -07:00
parent 372943801d
commit fde5b5e318
39 changed files with 10099 additions and 187 deletions
--- a/db/init.sql
+++ b/db/init.sql
@ -0,0 +1,171 @@
+-- Enable required extensions
+create extension if not exists "uuid-ossp";      -- For UUID generation
+create extension if not exists pg_cron;          -- For scheduled jobs
+
+-- Create the search_cache table
+create table public.search_cache (
+  id uuid default uuid_generate_v4() primary key,
+  query text not null,
+  results jsonb not null,
+  location text not null,
+  category text not null,
+  created_at timestamp with time zone default timezone('utc'::text, now()) not null,
+  updated_at timestamp with time zone default timezone('utc'::text, now()) not null,
+  expires_at timestamp with time zone default timezone('utc'::text, now() + interval '7 days') not null
+);
+
+-- Create indexes
+create index search_cache_query_idx on public.search_cache (query);
+create index search_cache_location_category_idx on public.search_cache (location, category);
+create index search_cache_expires_at_idx on public.search_cache (expires_at);
+
+-- Enable RLS
+alter table public.search_cache enable row level security;
+
+-- Create policies
+create policy "Allow public read access"
+  on public.search_cache for select
+  using (true);
+
+create policy "Allow service write access"
+  on public.search_cache for insert
+  with check (true);
+
+create policy "Allow service update access"
+  on public.search_cache for update
+  using (true)
+  with check (true);
+
+create policy "Allow delete expired records"
+  on public.search_cache for delete
+  using (expires_at < now());
+
+-- Create function to clean up expired records
+create or replace function cleanup_expired_cache()
+returns void
+language plpgsql
+security definer
+as $$
+begin
+  delete from public.search_cache
+  where expires_at < now();
+end;
+$$;
+
+-- Create a manual cleanup function since pg_cron might not be available
+create or replace function manual_cleanup()
+returns void
+language plpgsql
+security definer
+as $$
+begin
+  delete from public.search_cache
+  where expires_at < now();
+end;
+$$;
+
+-- Create a view for cache statistics
+create or replace view cache_stats as
+select
+  count(*) as total_entries,
+  count(*) filter (where expires_at < now()) as expired_entries,
+  count(*) filter (where expires_at >= now()) as valid_entries,
+  min(created_at) as oldest_entry,
+  max(created_at) as newest_entry,
+  count(distinct category) as unique_categories,
+  count(distinct location) as unique_locations
+from public.search_cache;
+
+-- Grant permissions to access the view
+grant select on cache_stats to postgres;
+
+-- Create table if not exists businesses
+create table if not exists businesses (
+  id text primary key,
+  name text not null,
+  phone text,
+  email text,
+  address text,
+  rating numeric,
+  website text,
+  logo text,
+  source text,
+  description text,
+  latitude numeric,
+  longitude numeric,
+  last_updated timestamp with time zone default timezone('utc'::text, now()),
+  search_count integer default 1,
+  created_at timestamp with time zone default timezone('utc'::text, now())
+);
+
+-- Create indexes for common queries
+create index if not exists businesses_name_idx on businesses (name);
+create index if not exists businesses_rating_idx on businesses (rating desc);
+create index if not exists businesses_search_count_idx on businesses (search_count desc);
+create index if not exists businesses_last_updated_idx on businesses (last_updated desc);
+
+-- Create tables if they don't exist
+CREATE TABLE IF NOT EXISTS businesses (
+    id TEXT PRIMARY KEY,
+    name TEXT NOT NULL,
+    phone TEXT,
+    email TEXT,
+    address TEXT,
+    rating INTEGER,
+    website TEXT,
+    logo TEXT,
+    source TEXT,
+    description TEXT,
+    location JSONB,
+    place_id TEXT,
+    photos TEXT[],
+    opening_hours TEXT[],
+    distance JSONB,
+    last_updated TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP,
+    search_count INTEGER DEFAULT 0
+);
+
+CREATE TABLE IF NOT EXISTS searches (
+    id SERIAL PRIMARY KEY,
+    query TEXT NOT NULL,
+    location TEXT NOT NULL,
+    timestamp TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP,
+    results_count INTEGER
+);
+
+CREATE TABLE IF NOT EXISTS cache (
+    key TEXT PRIMARY KEY,
+    value JSONB NOT NULL,
+    created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP,
+    expires_at TIMESTAMP WITH TIME ZONE NOT NULL
+);
+
+-- Create indexes
+CREATE INDEX IF NOT EXISTS idx_businesses_location ON businesses USING GIN (location);
+CREATE INDEX IF NOT EXISTS idx_businesses_search ON businesses USING GIN (to_tsvector('english', name || ' ' || COALESCE(description, '')));
+CREATE INDEX IF NOT EXISTS idx_cache_expires ON cache (expires_at);
+
+-- Set up RLS (Row Level Security)
+ALTER TABLE businesses ENABLE ROW LEVEL SECURITY;
+ALTER TABLE searches ENABLE ROW LEVEL SECURITY;
+ALTER TABLE cache ENABLE ROW LEVEL SECURITY;
+
+-- Create policies
+CREATE POLICY "Allow anonymous select" ON businesses FOR SELECT USING (true);
+CREATE POLICY "Allow service role insert" ON businesses FOR INSERT WITH CHECK (true);
+CREATE POLICY "Allow service role update" ON businesses FOR UPDATE USING (true);
+
+CREATE POLICY "Allow anonymous select" ON searches FOR SELECT USING (true);
+CREATE POLICY "Allow service role insert" ON searches FOR INSERT WITH CHECK (true);
+
+CREATE POLICY "Allow anonymous select" ON cache FOR SELECT USING (true);
+CREATE POLICY "Allow service role all" ON cache USING (true);
+
+-- Add place_id column to businesses table if it doesn't exist
+ALTER TABLE businesses ADD COLUMN IF NOT EXISTS place_id TEXT;
+CREATE INDEX IF NOT EXISTS idx_businesses_place_id ON businesses(place_id);
+
+-- Create a unique constraint on place_id (excluding nulls)
+CREATE UNIQUE INDEX IF NOT EXISTS idx_businesses_place_id_unique 
+ON businesses(place_id) 
+WHERE place_id IS NOT NULL; 
--- a/db/schema.sql
+++ b/db/schema.sql
@ -0,0 +1,44 @@
+-- Create the businesses table
+create table businesses (
+  id uuid primary key,
+  name text not null,
+  phone text,
+  address text,
+  city text,
+  state text,
+  zip text,
+  category text[],
+  rating numeric,
+  review_count integer,
+  license text,
+  services text[],
+  hours jsonb,
+  website text,
+  email text,
+  verified boolean default false,
+  last_updated timestamp with time zone,
+  search_query text,
+  search_location text,
+  search_timestamp timestamp with time zone,
+  reliability_score integer,
+  
+  -- Create a composite index for deduplication
+  constraint unique_business unique (phone, address)
+);
+
+-- Create indexes for common queries
+create index idx_business_location on businesses (city, state);
+create index idx_business_category on businesses using gin (category);
+create index idx_search_query on businesses using gin (search_query gin_trgm_ops);
+create index idx_search_location on businesses using gin (search_location gin_trgm_ops);
+create index idx_reliability on businesses (reliability_score);
+
+-- Enable full text search
+alter table businesses add column search_vector tsvector 
+  generated always as (
+    setweight(to_tsvector('english', coalesce(name, '')), 'A') ||
+    setweight(to_tsvector('english', coalesce(search_query, '')), 'B') ||
+    setweight(to_tsvector('english', coalesce(search_location, '')), 'C')
+  ) stored;
+
+create index idx_business_search on businesses using gin(search_vector); 
--- a/db/verify.sql
+++ b/db/verify.sql
@ -0,0 +1,15 @@
+-- Check if table exists
+SELECT EXISTS (
+   SELECT FROM information_schema.tables 
+   WHERE table_schema = 'public'
+   AND table_name = 'businesses'
+);
+
+-- Check table structure
+SELECT column_name, data_type, is_nullable
+FROM information_schema.columns
+WHERE table_schema = 'public'
+AND table_name = 'businesses';
+
+-- Check row count
+SELECT count(*) FROM businesses;