mirror of
https://github.com/gosom/google-maps-scraper.git
synced 2026-09-19 07:27:12 +08:00
Georgios Komninos
af155eecb7
This patch adds functionality to automatically provisiomn a webserver that exposes: - a REST API to submit keywords for scraping - a WebUI to automatically workers
469 lines
17 KiB
Go
469 lines
17 KiB
Go
// Package docs Code generated by swaggo/swag. DO NOT EDIT
|
|
package docs
|
|
|
|
import "github.com/swaggo/swag"
|
|
|
|
const docTemplate = `{
|
|
"schemes": {{ marshal .Schemes }},
|
|
"swagger": "2.0",
|
|
"info": {
|
|
"description": "{{escape .Description}}",
|
|
"title": "{{.Title}}",
|
|
"contact": {
|
|
"name": "API Support",
|
|
"url": "https://github.com/gosom/google-maps-scraper"
|
|
},
|
|
"license": {
|
|
"name": "MIT",
|
|
"url": "https://opensource.org/licenses/MIT"
|
|
},
|
|
"version": "{{.Version}}"
|
|
},
|
|
"host": "{{.Host}}",
|
|
"basePath": "{{.BasePath}}",
|
|
"paths": {
|
|
"/api/v1/health": {
|
|
"get": {
|
|
"description": "Check if the API service is healthy",
|
|
"produces": [
|
|
"application/json"
|
|
],
|
|
"tags": [
|
|
"health"
|
|
],
|
|
"summary": "Health check",
|
|
"responses": {
|
|
"200": {
|
|
"description": "OK",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.HealthCheckResponse"
|
|
}
|
|
}
|
|
}
|
|
}
|
|
},
|
|
"/api/v1/jobs": {
|
|
"get": {
|
|
"description": "List all scrape jobs with pagination and optional state filtering",
|
|
"produces": [
|
|
"application/json"
|
|
],
|
|
"tags": [
|
|
"scrape"
|
|
],
|
|
"summary": "List jobs",
|
|
"parameters": [
|
|
{
|
|
"type": "string",
|
|
"description": "Filter by state (available, running, completed, cancelled, discarded, pending, retryable, scheduled)",
|
|
"name": "state",
|
|
"in": "query"
|
|
},
|
|
{
|
|
"type": "integer",
|
|
"description": "Number of jobs to return (default: 20, max: 100)",
|
|
"name": "limit",
|
|
"in": "query"
|
|
},
|
|
{
|
|
"type": "string",
|
|
"description": "Cursor for pagination",
|
|
"name": "cursor",
|
|
"in": "query"
|
|
}
|
|
],
|
|
"responses": {
|
|
"200": {
|
|
"description": "OK",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ListJobsResponse"
|
|
}
|
|
},
|
|
"400": {
|
|
"description": "Bad Request",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
},
|
|
"500": {
|
|
"description": "Internal Server Error",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
}
|
|
}
|
|
}
|
|
},
|
|
"/api/v1/jobs/{job_id}": {
|
|
"get": {
|
|
"description": "Get the status and results of a scrape job",
|
|
"produces": [
|
|
"application/json"
|
|
],
|
|
"tags": [
|
|
"scrape"
|
|
],
|
|
"summary": "Get job status",
|
|
"parameters": [
|
|
{
|
|
"type": "string",
|
|
"description": "Job ID",
|
|
"name": "job_id",
|
|
"in": "path",
|
|
"required": true
|
|
}
|
|
],
|
|
"responses": {
|
|
"200": {
|
|
"description": "OK",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.JobStatusResponse"
|
|
}
|
|
},
|
|
"400": {
|
|
"description": "Bad Request",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
},
|
|
"404": {
|
|
"description": "Not Found",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
}
|
|
}
|
|
},
|
|
"delete": {
|
|
"description": "Queues deletion of a job and its results. Running jobs are cancelled first.",
|
|
"tags": [
|
|
"scrape"
|
|
],
|
|
"summary": "Delete a job",
|
|
"parameters": [
|
|
{
|
|
"type": "string",
|
|
"description": "Job ID",
|
|
"name": "job_id",
|
|
"in": "path",
|
|
"required": true
|
|
}
|
|
],
|
|
"responses": {
|
|
"202": {
|
|
"description": "Accepted",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.DeleteJobResponse"
|
|
}
|
|
},
|
|
"400": {
|
|
"description": "Invalid job ID",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
},
|
|
"404": {
|
|
"description": "Job not found",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
},
|
|
"500": {
|
|
"description": "Internal server error",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
}
|
|
}
|
|
}
|
|
},
|
|
"/api/v1/scrape": {
|
|
"post": {
|
|
"description": "Submit a new Google Maps scrape job for processing",
|
|
"consumes": [
|
|
"application/json"
|
|
],
|
|
"produces": [
|
|
"application/json"
|
|
],
|
|
"tags": [
|
|
"scrape"
|
|
],
|
|
"summary": "Submit a scrape job",
|
|
"parameters": [
|
|
{
|
|
"description": "Scrape request parameters",
|
|
"name": "request",
|
|
"in": "body",
|
|
"required": true,
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ScrapeRequest"
|
|
}
|
|
}
|
|
],
|
|
"responses": {
|
|
"202": {
|
|
"description": "Accepted",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ScrapeResponse"
|
|
}
|
|
},
|
|
"400": {
|
|
"description": "Bad Request",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
},
|
|
"500": {
|
|
"description": "Internal Server Error",
|
|
"schema": {
|
|
"$ref": "#/definitions/api.ErrorResponse"
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
},
|
|
"definitions": {
|
|
"api.DeleteJobResponse": {
|
|
"description": "Response returned after successfully queueing a job for deletion",
|
|
"type": "object",
|
|
"properties": {
|
|
"message": {
|
|
"description": "Status message",
|
|
"type": "string",
|
|
"example": "deletion queued"
|
|
}
|
|
}
|
|
},
|
|
"api.ErrorResponse": {
|
|
"description": "Error response returned when a request fails",
|
|
"type": "object",
|
|
"properties": {
|
|
"message": {
|
|
"type": "string",
|
|
"example": "invalid request body"
|
|
}
|
|
}
|
|
},
|
|
"api.HealthCheckResponse": {
|
|
"description": "Health check response indicating service status",
|
|
"type": "object",
|
|
"properties": {
|
|
"status": {
|
|
"type": "string",
|
|
"example": "ok"
|
|
}
|
|
}
|
|
},
|
|
"api.JobStatusResponse": {
|
|
"description": "Detailed status of a scrape job including results when completed",
|
|
"type": "object",
|
|
"properties": {
|
|
"completed_at": {
|
|
"description": "When the job completed",
|
|
"type": "string",
|
|
"example": "2024-01-15T10:35:00Z"
|
|
},
|
|
"created_at": {
|
|
"description": "When the job was created",
|
|
"type": "string",
|
|
"example": "2024-01-15T10:30:00Z"
|
|
},
|
|
"error": {
|
|
"description": "Error message if job failed",
|
|
"type": "string",
|
|
"example": ""
|
|
},
|
|
"job_id": {
|
|
"description": "Unique job identifier",
|
|
"type": "string",
|
|
"example": "kYzR8xLmNpQvWjX3"
|
|
},
|
|
"keyword": {
|
|
"description": "Search keyword used for this job",
|
|
"type": "string",
|
|
"example": "restaurants in New York"
|
|
},
|
|
"result_count": {
|
|
"description": "Number of results found",
|
|
"type": "integer",
|
|
"example": 25
|
|
},
|
|
"results": {
|
|
"description": "Scrape results (array of place data)"
|
|
},
|
|
"started_at": {
|
|
"description": "When the job started processing",
|
|
"type": "string",
|
|
"example": "2024-01-15T10:30:05Z"
|
|
},
|
|
"status": {
|
|
"description": "Current job status (pending, running, completed, failed)",
|
|
"type": "string",
|
|
"example": "completed"
|
|
}
|
|
}
|
|
},
|
|
"api.JobSummary": {
|
|
"description": "Summary of a job without the full result data",
|
|
"type": "object",
|
|
"properties": {
|
|
"completed_at": {
|
|
"description": "When the job completed",
|
|
"type": "string",
|
|
"example": "2024-01-15T10:35:00Z"
|
|
},
|
|
"created_at": {
|
|
"description": "When the job was created",
|
|
"type": "string",
|
|
"example": "2024-01-15T10:30:00Z"
|
|
},
|
|
"error": {
|
|
"description": "Error message if job failed",
|
|
"type": "string",
|
|
"example": ""
|
|
},
|
|
"job_id": {
|
|
"description": "Unique job identifier (encoded)",
|
|
"type": "string",
|
|
"example": "kYzR8xLmNpQvWjX3"
|
|
},
|
|
"keyword": {
|
|
"description": "Search keyword used for this job",
|
|
"type": "string",
|
|
"example": "restaurants in New York"
|
|
},
|
|
"result_count": {
|
|
"description": "Number of results found",
|
|
"type": "integer",
|
|
"example": 25
|
|
},
|
|
"started_at": {
|
|
"description": "When the job started processing",
|
|
"type": "string",
|
|
"example": "2024-01-15T10:30:05Z"
|
|
},
|
|
"status": {
|
|
"description": "Current job status",
|
|
"type": "string",
|
|
"example": "completed"
|
|
}
|
|
}
|
|
},
|
|
"api.ListJobsResponse": {
|
|
"description": "Paginated list of jobs with cursor for next page",
|
|
"type": "object",
|
|
"properties": {
|
|
"has_more": {
|
|
"description": "Whether there are more results",
|
|
"type": "boolean",
|
|
"example": true
|
|
},
|
|
"jobs": {
|
|
"description": "List of jobs",
|
|
"type": "array",
|
|
"items": {
|
|
"$ref": "#/definitions/api.JobSummary"
|
|
}
|
|
},
|
|
"next_cursor": {
|
|
"description": "Cursor for next page (empty if no more results)",
|
|
"type": "string",
|
|
"example": "eyJpZCI6MTIzfQ=="
|
|
}
|
|
}
|
|
},
|
|
"api.ScrapeRequest": {
|
|
"description": "Request body for submitting a scrape job",
|
|
"type": "object",
|
|
"properties": {
|
|
"email": {
|
|
"description": "Whether to extract email addresses from websites",
|
|
"type": "boolean",
|
|
"example": false
|
|
},
|
|
"extra_reviews": {
|
|
"description": "Extract additional reviews",
|
|
"type": "boolean",
|
|
"example": false
|
|
},
|
|
"fast_mode": {
|
|
"description": "Use fast mode (stealth HTTP instead of browser)",
|
|
"type": "boolean",
|
|
"example": false
|
|
},
|
|
"geo_coordinates": {
|
|
"description": "Geographic coordinates in \"lat,lon\" format",
|
|
"type": "string",
|
|
"example": "40.7128,-74.0060"
|
|
},
|
|
"keyword": {
|
|
"description": "Search keyword (e.g., \"restaurants in New York\")",
|
|
"type": "string",
|
|
"example": "restaurants in New York"
|
|
},
|
|
"lang": {
|
|
"description": "Language code for results (default: \"en\")",
|
|
"type": "string",
|
|
"example": "en"
|
|
},
|
|
"max_depth": {
|
|
"description": "Maximum depth for pagination (default: 1, max: 100)",
|
|
"type": "integer",
|
|
"example": 1
|
|
},
|
|
"radius": {
|
|
"description": "Search radius in kilometers",
|
|
"type": "number",
|
|
"example": 5
|
|
},
|
|
"timeout": {
|
|
"description": "Job timeout in seconds (1-300, default: 300)",
|
|
"type": "integer",
|
|
"example": 300
|
|
},
|
|
"zoom": {
|
|
"description": "Zoom level for map search (1-21)",
|
|
"type": "integer",
|
|
"example": 14
|
|
}
|
|
}
|
|
},
|
|
"api.ScrapeResponse": {
|
|
"description": "Response returned after successfully submitting a scrape job",
|
|
"type": "object",
|
|
"properties": {
|
|
"job_id": {
|
|
"description": "Unique job identifier",
|
|
"type": "string",
|
|
"example": "kYzR8xLmNpQvWjX3"
|
|
},
|
|
"status": {
|
|
"description": "Current job status",
|
|
"type": "string",
|
|
"example": "pending"
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}`
|
|
|
|
// SwaggerInfo holds exported Swagger Info so clients can modify it
|
|
var SwaggerInfo = &swag.Spec{
|
|
Version: "1.0",
|
|
Host: "localhost:8080",
|
|
BasePath: "/",
|
|
Schemes: []string{"http", "https"},
|
|
Title: "Google Maps Scraper Pro API",
|
|
Description: "API for scraping Google Maps data asynchronously",
|
|
InfoInstanceName: "swagger",
|
|
SwaggerTemplate: docTemplate,
|
|
LeftDelim: "{{",
|
|
RightDelim: "}}",
|
|
}
|
|
|
|
func init() {
|
|
swag.Register(SwaggerInfo.InstanceName(), SwaggerInfo)
|
|
}
|