{ "openapi": "3.0.1", "info": { "title": "Website Media Extractor & Scraper", "description": "The Advanced Website Media Extractor & Scraper is a comprehensive media extraction tool that supports images, videos, audio, documents, archives, e-books, fonts, apps, and contact information from websites. Features advanced filtering, proxy support, and detailed analytics.", "version": "2.0", "x-build-id": "kPLX0eK2gz54BTBM4" }, "servers": [ { "url": "https://api.apify.com/v2" } ], "paths": { "/acts/hlymrk~html-web-media-scraper/run-sync-get-dataset-items": { "post": { "operationId": "run-sync-get-dataset-items-hlymrk-html-web-media-scraper", "x-openai-isConsequential": false, "summary": "Executes an Actor, waits for its completion, and returns Actor's dataset items in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK" } } } }, "/acts/hlymrk~html-web-media-scraper/runs": { "post": { "operationId": "runs-sync-hlymrk-html-web-media-scraper", "x-openai-isConsequential": false, "summary": "Executes an Actor and returns information about the initiated run in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/runsResponseSchema" } } } } } } }, "/acts/hlymrk~html-web-media-scraper/run-sync": { "post": { "operationId": "run-sync-hlymrk-html-web-media-scraper", "x-openai-isConsequential": false, "summary": "Executes an Actor, waits for completion, and returns the OUTPUT from Key-value store in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK" } } } } }, "components": { "schemas": { "inputSchema": { "type": "object", "required": [ "startUrls" ], "properties": { "startUrls": { "title": "Start URLs", "type": "array", "description": "URLs to start with.", "items": { "type": "object", "required": [ "url" ], "properties": { "url": { "type": "string", "title": "URL of a web page", "format": "uri" } } } }, "maxRequestsPerCrawl": { "title": "Max Requests per Crawl", "minimum": 1, "maximum": 10000, "type": "integer", "description": "Maximum number of requests that can be made by this crawler.", "default": 100 }, "proxyConfiguration": { "title": "Proxy Configuration", "type": "object", "description": "Proxy settings for the crawler", "default": { "useApifyProxy": true } }, "mediaTypes": { "title": "Media Types to Extract", "type": "array", "description": "Select which media types to extract", "items": { "type": "string", "enum": [ "images", "videos", "audios", "documents", "archives", "ebooks", "fonts", "apps", "contacts", "svg", "embed", "objects" ] }, "default": [ "images", "videos", "audios", "documents" ] }, "imageOptions": { "title": "Image Extraction Options", "type": "object", "description": "Configure image extraction settings", "properties": { "includeBackgroundImages": { "title": "Include Background Images", "type": "boolean", "description": "Extract images from CSS background-image properties", "editor": "checkbox", "default": true }, "minImageSize": { "title": "Minimum Image Size (pixels)", "type": "integer", "description": "Skip images smaller than this size (width or height)", "editor": "number", "default": 50, "minimum": 1 }, "includeDataUrls": { "title": "Include Data URLs", "type": "boolean", "description": "Include base64 encoded images (data: URLs)", "editor": "checkbox", "default": false } }, "default": { "includeBackgroundImages": true, "minImageSize": 50, "includeDataUrls": false } }, "fileFilters": { "title": "File Filtering Options", "type": "object", "description": "Filter files by size, type, and other criteria", "properties": { "maxFileSize": { "title": "Maximum File Size (MB)", "type": "number", "description": "Skip files larger than this size", "editor": "number", "default": 100, "minimum": 0.1, "maximum": 1000 }, "allowedExtensions": { "title": "Allowed File Extensions", "type": "array", "description": "Only extract files with these extensions (leave empty for all)", "editor": "requestListSources", "items": { "type": "object" }, "default": [] }, "blockedExtensions": { "title": "Blocked File Extensions", "type": "array", "description": "Skip files with these extensions", "editor": "requestListSources", "items": { "type": "object" }, "default": [] } }, "default": { "maxFileSize": 100, "allowedExtensions": [], "blockedExtensions": [] } }, "contactExtraction": { "title": "Contact Information Extraction", "type": "object", "description": "Configure contact information extraction", "properties": { "extractContacts": { "title": "Extract Contact Information", "type": "boolean", "description": "Extract emails, phone numbers, and social media profiles", "editor": "checkbox", "default": true }, "includeEmails": { "title": "Include Email Addresses", "type": "boolean", "description": "Extract email addresses from page content", "editor": "checkbox", "default": true }, "includePhones": { "title": "Include Phone Numbers", "type": "boolean", "description": "Extract phone numbers from page content", "editor": "checkbox", "default": true }, "includeSocialMedia": { "title": "Include Social Media Profiles", "type": "boolean", "description": "Extract social media profile links", "editor": "checkbox", "default": true } }, "default": { "extractContacts": true, "includeEmails": true, "includePhones": true, "includeSocialMedia": true } }, "outputOptions": { "title": "Output Configuration", "type": "object", "description": "Configure output format and content", "properties": { "includeSummary": { "title": "Include Summary Statistics", "type": "boolean", "description": "Add summary statistics to each result", "editor": "checkbox", "default": true }, "includeMetadata": { "title": "Include File Metadata", "type": "boolean", "description": "Include additional file metadata (size, type, etc.)", "editor": "checkbox", "default": true }, "groupByType": { "title": "Group Results by Media Type", "type": "boolean", "description": "Organize results by media type categories", "editor": "checkbox", "default": true } }, "default": { "includeSummary": true, "includeMetadata": true, "groupByType": true } }, "crawlingOptions": { "title": "Advanced Crawling Options", "type": "object", "description": "Advanced settings for crawling behavior", "properties": { "respectRobotsTxt": { "title": "Respect robots.txt", "type": "boolean", "description": "Follow robots.txt directives", "editor": "checkbox", "default": true }, "userAgent": { "title": "Custom User Agent", "type": "string", "description": "Custom user agent string (leave empty for default)", "editor": "textfield", "default": "" }, "maxRetries": { "title": "Maximum Retries", "type": "integer", "description": "Maximum number of retries for failed requests", "editor": "number", "default": 3, "minimum": 0, "maximum": 10 } }, "default": { "respectRobotsTxt": true, "userAgent": "", "maxRetries": 3 } }, "conversionOptions": { "title": "Media Conversion Options", "type": "object", "description": "Convert media formats for better compatibility", "properties": { "convertSvgToImage": { "title": "Convert SVG to Images", "type": "boolean", "editor": "checkbox", "description": "Convert SVG elements to raster images", "default": false }, "convertCanvasToImage": { "title": "Convert Canvas to Images", "type": "boolean", "editor": "checkbox", "description": "Convert canvas elements to images", "default": false }, "imageFormat": { "title": "Conversion Image Format", "type": "string", "description": "Format for converted images", "editor": "select", "enum": [ "png", "jpeg", "webp" ], "default": "png" }, "imageQuality": { "title": "Image Quality (1-100)", "type": "integer", "description": "Quality for JPEG/WebP conversion", "editor": "number", "minimum": 1, "maximum": 100, "default": 90 }, "maxConversionWidth": { "title": "Max Conversion Width", "type": "integer", "description": "Maximum width for converted images", "editor": "number", "default": 2048, "minimum": 100 }, "maxConversionHeight": { "title": "Max Conversion Height", "type": "integer", "description": "Maximum height for converted images", "editor": "number", "default": 2048, "minimum": 100 } }, "default": { "convertSvgToImage": false, "convertCanvasToImage": false, "imageFormat": "png", "imageQuality": 90, "maxConversionWidth": 2048, "maxConversionHeight": 2048 } }, "duplicateDetection": { "title": "Duplicate Detection", "type": "object", "description": "Remove duplicate media items", "properties": { "enableDuplicateDetection": { "title": "Enable Duplicate Detection", "type": "boolean", "editor": "checkbox", "description": "Remove duplicate media items", "default": true }, "compareBy": { "title": "Compare By", "type": "array", "description": "Criteria for duplicate detection", "editor": "select", "items": { "type": "string", "enum": [ "src", "alt", "fileSize", "dimensions" ] }, "default": [ "src", "dimensions" ] }, "similarityThreshold": { "title": "Similarity Threshold (0-1)", "type": "number", "description": "Threshold for considering items as duplicates", "editor": "number", "minimum": 0, "maximum": 1, "default": 0.8 } }, "default": { "enableDuplicateDetection": true, "compareBy": [ "src", "dimensions" ], "similarityThreshold": 0.8 } }, "validationOptions": { "title": "Media Validation", "type": "object", "description": "Validate media files and URLs", "properties": { "enableValidation": { "title": "Enable Media Validation", "type": "boolean", "editor": "checkbox", "description": "Validate media URLs and file properties", "default": true }, "checkUrlAccessibility": { "title": "Check URL Accessibility", "type": "boolean", "editor": "checkbox", "description": "Verify that media URLs are accessible", "default": true }, "validateFileHeaders": { "title": "Validate File Headers", "type": "boolean", "editor": "checkbox", "description": "Check MIME types and file headers", "default": true }, "validationTimeout": { "title": "Validation Timeout (ms)", "type": "integer", "description": "Timeout for URL validation requests", "editor": "number", "default": 10000, "minimum": 1000, "maximum": 30000 } }, "default": { "enableValidation": true, "checkUrlAccessibility": true, "validateFileHeaders": true, "validationTimeout": 10000 } }, "customSelectors": { "title": "Custom CSS Selectors", "type": "array", "description": "Define custom CSS selectors for media extraction", "items": { "type": "object", "properties": { "name": { "title": "Selector Name", "type": "string", "description": "Unique name for this selector", "editor": "textfield" }, "selector": { "title": "CSS Selector", "type": "string", "description": "CSS selector to find elements", "editor": "textarea" }, "mediaType": { "title": "Media Type", "type": "string", "description": "Type of media this selector finds", "editor": "textfield" }, "srcAttribute": { "title": "Source Attribute", "type": "string", "description": "Attribute containing the media URL", "editor": "textfield", "default": "src" }, "altAttribute": { "title": "Alt Text Attribute", "type": "string", "description": "Attribute containing description text", "editor": "textfield", "default": "alt" } }, "required": [ "name", "selector", "mediaType" ] }, "default": [] }, "batchProcessing": { "title": "Batch Processing Options", "type": "object", "description": "Configure batch processing for large URL lists", "properties": { "enableBatchProcessing": { "title": "Enable Batch Processing", "type": "boolean", "editor": "checkbox", "description": "Process URLs in batches for better performance and reliability", "default": true }, "batchSize": { "title": "Batch Size", "type": "integer", "description": "Number of URLs to process in each batch", "editor": "number", "default": 10, "minimum": 1, "maximum": 100 }, "concurrency": { "title": "Concurrent Requests", "type": "integer", "description": "Number of URLs to process simultaneously within each batch", "editor": "number", "default": 3, "minimum": 1, "maximum": 10 }, "delayBetweenBatches": { "title": "Delay Between Batches (ms)", "type": "integer", "description": "Delay between processing batches in milliseconds", "editor": "number", "default": 1000, "minimum": 0, "maximum": 10000 }, "maxRetries": { "title": "Maximum Retries", "type": "integer", "description": "Maximum number of retries for failed URLs", "editor": "number", "default": 3, "minimum": 0, "maximum": 10 }, "failureThreshold": { "title": "Failure Threshold (0-1)", "type": "number", "description": "Stop processing if failure rate exceeds this threshold", "editor": "number", "default": 0.5, "minimum": 0, "maximum": 1 }, "enableProgressTracking": { "title": "Enable Progress Tracking", "type": "boolean", "editor": "checkbox", "description": "Track and save processing progress for resumption", "default": true }, "resumeFromLastBatch": { "title": "Resume from Last Batch", "type": "boolean", "editor": "checkbox", "description": "Resume processing from the last completed batch if interrupted", "default": true } }, "default": { "enableBatchProcessing": true, "batchSize": 10, "concurrency": 3, "delayBetweenBatches": 1000, "maxRetries": 3, "failureThreshold": 0.5, "enableProgressTracking": true, "resumeFromLastBatch": true } }, "urlListManagement": { "title": "URL List Management", "type": "object", "description": "Configure URL list processing and filtering", "properties": { "enableDeduplication": { "title": "Remove Duplicate URLs", "type": "boolean", "editor": "checkbox", "description": "Remove duplicate URLs from the input list", "default": true }, "enableValidation": { "title": "Validate URLs", "type": "boolean", "editor": "checkbox", "description": "Validate URL format and accessibility", "default": true }, "maxUrlsPerBatch": { "title": "Maximum URLs per Batch", "type": "integer", "description": "Maximum number of URLs to process in a single run", "editor": "number", "default": 1000, "minimum": 1, "maximum": 50000 }, "blockedDomains": { "title": "Blocked Domains", "type": "array", "description": "Domains to exclude from processing", "editor": "requestListSources", "items": { "type": "object" }, "default": [] }, "allowedDomains": { "title": "Allowed Domains", "type": "array", "description": "Only process URLs from these domains (leave empty for all)", "editor": "requestListSources", "items": { "type": "object" }, "default": [] }, "urlPatterns": { "title": "URL Pattern Filters", "type": "object", "editor": "json", "description": "Include/exclude URLs based on patterns", "properties": { "includePatterns": { "title": "Include Patterns", "type": "array", "description": "Only process URLs matching these regex patterns", "editor": "json", "items": { "type": "string" }, "default": [] }, "excludePatterns": { "title": "Exclude Patterns", "type": "array", "description": "Skip URLs matching these regex patterns", "editor": "json", "items": { "type": "string" }, "default": [] } }, "default": { "includePatterns": [], "excludePatterns": [] } } }, "default": { "enableDeduplication": true, "enableValidation": true, "maxUrlsPerBatch": 1000, "blockedDomains": [], "allowedDomains": [], "urlPatterns": { "includePatterns": [], "excludePatterns": [] } } } } }, "runsResponseSchema": { "type": "object", "properties": { "data": { "type": "object", "properties": { "id": { "type": "string" }, "actId": { "type": "string" }, "userId": { "type": "string" }, "startedAt": { "type": "string", "format": "date-time", "example": "2025-01-08T00:00:00.000Z" }, "finishedAt": { "type": "string", "format": "date-time", "example": "2025-01-08T00:00:00.000Z" }, "status": { "type": "string", "example": "READY" }, "meta": { "type": "object", "properties": { "origin": { "type": "string", "example": "API" }, "userAgent": { "type": "string" } } }, "stats": { "type": "object", "properties": { "inputBodyLen": { "type": "integer", "example": 2000 }, "rebootCount": { "type": "integer", "example": 0 }, "restartCount": { "type": "integer", "example": 0 }, "resurrectCount": { "type": "integer", "example": 0 }, "computeUnits": { "type": "integer", "example": 0 } } }, "options": { "type": "object", "properties": { "build": { "type": "string", "example": "latest" }, "timeoutSecs": { "type": "integer", "example": 300 }, "memoryMbytes": { "type": "integer", "example": 1024 }, "diskMbytes": { "type": "integer", "example": 2048 } } }, "buildId": { "type": "string" }, "defaultKeyValueStoreId": { "type": "string" }, "defaultDatasetId": { "type": "string" }, "defaultRequestQueueId": { "type": "string" }, "buildNumber": { "type": "string", "example": "1.0.0" }, "containerUrl": { "type": "string" }, "usage": { "type": "object", "properties": { "ACTOR_COMPUTE_UNITS": { "type": "integer", "example": 0 }, "DATASET_READS": { "type": "integer", "example": 0 }, "DATASET_WRITES": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_READS": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_WRITES": { "type": "integer", "example": 1 }, "KEY_VALUE_STORE_LISTS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_READS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_WRITES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_INTERNAL_GBYTES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_EXTERNAL_GBYTES": { "type": "integer", "example": 0 }, "PROXY_RESIDENTIAL_TRANSFER_GBYTES": { "type": "integer", "example": 0 }, "PROXY_SERPS": { "type": "integer", "example": 0 } } }, "usageTotalUsd": { "type": "number", "example": 0.00005 }, "usageUsd": { "type": "object", "properties": { "ACTOR_COMPUTE_UNITS": { "type": "integer", "example": 0 }, "DATASET_READS": { "type": "integer", "example": 0 }, "DATASET_WRITES": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_READS": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_WRITES": { "type": "number", "example": 0.00005 }, "KEY_VALUE_STORE_LISTS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_READS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_WRITES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_INTERNAL_GBYTES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_EXTERNAL_GBYTES": { "type": "integer", "example": 0 }, "PROXY_RESIDENTIAL_TRANSFER_GBYTES": { "type": "integer", "example": 0 }, "PROXY_SERPS": { "type": "integer", "example": 0 } } } } } } } } } }