llamaparserb 0.2.2 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a1761244b0d8ac8c6ee13eb10b50d3eadcb182d8e7bad24828ba70a290bc7507
-  data.tar.gz: cf7908683f630f17ef39b4c7e6393cb4637e350ea89c674c78a03a68b1bdf1c5
+  metadata.gz: 9c49dc1624b1b955cad8032696308fc81357533113cf429f53fe063d5db0cab2
+  data.tar.gz: 2eee40a9054fe05d02a094828bb47a6cc3a5b65ca6fcb5864b9648cff5f0b418
 SHA512:
-  metadata.gz: 7bc3f4c44814c1cf63ad480882ee7f2b647af3ee8b34b46d95da7ae90c372b656c0f69fd224ba1c03d26cfa5185ee89a356f05e080169375945fc0f8d3548d8d
-  data.tar.gz: 5e8d6c3d234e298836f2f373d515321337d1f1ab1904a38e9dfb8a1e6076b6307e69e84d93b5fd27fd02f3f0642e19f8adc38b9de20e5d0ac87b47600b13de90
+  metadata.gz: 50ed45e0d813d776b79fce2e87593cdc88063158045c62ba76e1bc0e353fb1b27b2b03b87dc4253443b4b2a355fec179708d04e1920dc0f47730c271e3ff81cc
+  data.tar.gz: 9f86c50b3f5c4bd987c9bab987c5b139ce1dd998577d04ea3fba53d64b430d3cdcc0af3d880d73ebea1914341bc185fb78469f78d9521098f06510c7024b8746

data/CHANGELOG.md

@@ -6,10 +6,22 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.3.0] - 2024-11-28
+### Added
+- Add support for parsing files from URLs
+
+## [0.2.3] - 2024-11-28
+### Added
+- Add support for all supported optional llamaparsse parameters to `parse_file`
+
+[0.2.3]: https://github.com/horizing/llamaparserb/releases/tag/v0.2.3...v0.2.2
+
 ## [0.2.2] - 2024-11-28
 ### Fixed
 - Fix issue with handling file path
 
+[0.2.2]: https://github.com/horizing/llamaparserb/releases/tag/v0.2.2
+
 ## [0.2.1] - 2024-11-28
 ### Fixed
 - Fix parse_file to handle files that are not on the local filesystem
@@ -25,8 +37,4 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [0.1.0] - 2024-11-27
 ### Added
-- Initial release
-
-[Unreleased]: https://github.com/horizing/llamaparserb/compare/v0.1.0...HEAD
-[0.1.1]: https://github.com/horizing/llamaparserb/releases/tag/v0.1.1...v0.1.0
-[0.1.0]: https://github.com/horizing/llamaparserb/releases/tag/v0.1.0
+- Initial release

data/README.md

@@ -43,11 +43,14 @@ text = client.parse_file(file_content, 'pdf')
 # Parse a file to markdown
 client = Llamaparserb::Client.new(ENV['LLAMA_CLOUD_API_KEY'], result_type: "markdown")
 markdown = client.parse_file('path/to/document.pdf')
+
+# Parse a file from a URL
+markdown = client.parse_file('https://example.com/document.pdf')
 ```
 
 ### File Input Options
 
-The `parse_file` method accepts two types of inputs:
+The `parse_file` method accepts three types of inputs:
 
 1. File path (String):
 ```ruby
@@ -69,25 +72,123 @@ temp_file = Tempfile.new(['document', '.pdf'])
 client.parse_file(temp_file, 'pdf')
 ```
 
+3. URL (String):
+```ruby
+client.parse_file('https://example.com/document.pdf')
+```
+
 ### Advanced Options
 
 ```ruby
 client = Llamaparserb::Client.new(
   ENV['LLAMA_CLOUD_API_KEY'],
   {
-    result_type: "markdown",  # Output format: "text" or "markdown"
-    num_workers: 4,           # Number of workers for concurrent processing
-    check_interval: 1,        # How often to check job status (seconds)
-    max_timeout: 2000,        # Maximum time to wait for parsing (seconds)
-    verbose: true,            # Enable detailed logging
-    language: :en,            # Target language
-    parsing_instruction: "",  # Custom parsing instructions
-    premium_mode: false,      # Enable premium parsing features
-    split_by_page: true       # Split result by pages
+    # Basic Configuration
+    result_type: "markdown",    # Output format: "text" or "markdown"
+    num_workers: 4,             # Number of workers for concurrent processing
+    check_interval: 1,          # How often to check job status (seconds)
+    max_timeout: 2000,          # Maximum time to wait for parsing (seconds)
+    verbose: true,              # Enable detailed logging
+    show_progress: true,        # Show progress during parsing
+    ignore_errors: true,        # Return nil instead of raising errors
+    
+    # Language and Parsing Options
+    language: :en,              # Target language for parsing
+    parsing_instruction: "",    # Custom parsing instructions
+    skip_diagonal_text: false,  # Skip diagonal text in documents
+    invalidate_cache: false,    # Force reprocessing of cached documents
+    do_not_cache: false,        # Disable caching of results
+    
+    # Processing Modes
+    fast_mode: false,          # Enable faster processing (may reduce quality)
+    premium_mode: false,       # Enable premium parsing features
+    continuous_mode: false,    # Process document as continuous text
+    do_not_unroll_columns: false, # Keep columnar text structure
+    
+    # Page Handling
+    split_by_page: true,       # Split result by pages
+    page_separator: "\n\n",    # Custom page separator
+    page_prefix: "Page ",      # Text to prepend to each page
+    page_suffix: "\n",         # Text to append to each page
+    target_pages: [1,2,3],     # Array of specific pages to process
+    bounding_box: {            # Specify area to parse (coordinates in pixels)
+      x1: 0, y1: 0,           # Top-left corner
+      x2: 612, y2: 792        # Bottom-right corner
+    },
+    
+    # OCR and Image Processing
+    disable_ocr: false,        # Disable Optical Character Recognition
+    take_screenshot: false,    # Capture screenshot of document
+    
+    # Advanced Processing Features
+    gpt4o_mode: false,         # Enable GPT-4 Optimization mode
+    gpt4o_api_key: "key",      # API key for GPT-4 Optimization
+    guess_xlsx_sheet_names: false, # Attempt to guess Excel sheet names
+    is_formatting_instruction: false, # Use formatting instructions
+    annotate_links: false,     # Include link annotations in output
+    
+    # Multimodal Processing
+    vendor_multimodal_api_key: "key",      # API key for multimodal processing
+    use_vendor_multimodal_model: false,     # Enable multimodal model
+    vendor_multimodal_model_name: "model",  # Specify multimodal model
+    
+    # Integration Options
+    webhook_url: "https://...", # URL for webhook notifications
+    http_proxy: "http://...",   # HTTP proxy configuration
+    
+    # Azure OpenAI Configuration
+    azure_openai_deployment_name: "deployment", # Azure OpenAI deployment name
+    azure_openai_endpoint: "endpoint",         # Azure OpenAI endpoint
+    azure_openai_api_version: "2023-05-15",    # Azure OpenAI API version
+    azure_openai_key: "key"                    # Azure OpenAI API key
   }
 )
 ```
 
+### Feature-Specific Options
+
+#### Page Processing
+- `split_by_page`: Split the document into separate pages
+- `page_separator`: Custom text to insert between pages
+- `page_prefix`/`page_suffix`: Add custom text before/after each page
+- `target_pages`: Process only specific pages
+- `bounding_box`: Parse only a specific area of the document
+
+#### OCR and Image Processing
+- `disable_ocr`: Turn off Optical Character Recognition
+- `take_screenshot`: Generate document screenshots
+- `skip_diagonal_text`: Ignore text at diagonal angles
+
+#### Advanced Processing
+- `continuous_mode`: Process text as a continuous stream
+- `do_not_unroll_columns`: Preserve column structure
+- `guess_xlsx_sheet_names`: Auto-detect Excel sheet names
+- `annotate_links`: Include document hyperlinks in output
+- `is_formatting_instruction`: Use special formatting instructions
+
+#### Performance Options
+- `fast_mode`: Faster processing with potential quality trade-offs
+- `premium_mode`: Access to premium features
+- `invalidate_cache`/`do_not_cache`: Control result caching
+- `num_workers`: Configure concurrent processing
+
+#### Integration Features
+- `webhook_url`: Receive processing notifications
+- `http_proxy`: Configure proxy settings
+
+#### Azure OpenAI Integration
+Configure Azure OpenAI services with:
+- `azure_openai_deployment_name`
+- `azure_openai_endpoint`
+- `azure_openai_api_version`
+- `azure_openai_key`
+
+#### Multimodal Processing
+Enable advanced multimodal processing with:
+- `vendor_multimodal_api_key`
+- `use_vendor_multimodal_model`
+- `vendor_multimodal_model_name`
+
 ### Supported File Types
 
 The client supports a wide range of file formats including:

data/lib/llamaparserb/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Llamaparserb
-  VERSION = "0.2.2"
+  VERSION = "0.3.0"
 end

data/lib/llamaparserb.rb

@@ -51,6 +51,9 @@ module Llamaparserb
         elsif File.exist?(file_input)
           job_id = create_job_from_path(file_input)
           log "Started parsing file under job_id #{job_id}", :info
+        elsif URI::DEFAULT_PARSER.make_regexp.match?(file_input)
+          job_id = create_job_from_url(file_input)
+          log "Started parsing URL under job_id #{job_id}", :info
         else
           raise Error, "file_type parameter is required for binary string input"
         end
@@ -100,7 +103,20 @@ module Llamaparserb
         bounding_box: nil,
         target_pages: nil,
         ignore_errors: true,
-        split_by_page: true
+        split_by_page: true,
+        vendor_multimodal_api_key: nil,
+        use_vendor_multimodal_model: false,
+        vendor_multimodal_model_name: nil,
+        take_screenshot: false,
+        disable_ocr: false,
+        is_formatting_instruction: false,
+        annotate_links: false,
+        webhook_url: nil,
+        azure_openai_deployment_name: nil,
+        azure_openai_endpoint: nil,
+        azure_openai_api_version: nil,
+        azure_openai_key: nil,
+        http_proxy: nil
       }
     end
 
@@ -184,7 +200,7 @@ module Llamaparserb
     def build_connection
       Faraday.new(url: base_url) do |f|
         f.request :multipart
-        f.request :json
+        f.request :url_encoded
         f.response :json
         f.response :raise_error
         f.adapter Faraday.default_adapter
@@ -221,7 +237,13 @@ module Llamaparserb
         temp_file,
         detect_content_type(temp_file.path)
       )
-      create_job(file)
+
+      response = @connection.post("upload") do |req|
+        req.headers["Authorization"] = "Bearer #{api_key}"
+        req.body = {file: file}
+      end
+
+      response.body["id"]
     ensure
       temp_file&.close
       temp_file&.unlink
@@ -236,9 +258,8 @@ module Llamaparserb
       response.body["id"]
     end
 
-    def upload_params(file)
-      {
-        file: file,
+    def upload_params(file = nil, url = nil)
+      params = {
         language: @options[:language].to_s,
         parsing_instruction: @options[:parsing_instruction],
         invalidate_cache: @options[:invalidate_cache],
@@ -250,8 +271,36 @@ module Llamaparserb
         do_not_unroll_columns: @options[:do_not_unroll_columns],
         gpt4o_mode: @options[:gpt4o_mode],
         gpt4o_api_key: @options[:gpt4o_api_key],
+        vendor_multimodal_api_key: @options[:vendor_multimodal_api_key],
+        use_vendor_multimodal_model: @options[:use_vendor_multimodal_model],
+        vendor_multimodal_model_name: @options[:vendor_multimodal_model_name],
+        take_screenshot: @options[:take_screenshot],
+        disable_ocr: @options[:disable_ocr],
+        guess_xlsx_sheet_names: @options[:guess_xlsx_sheet_names],
+        is_formatting_instruction: @options[:is_formatting_instruction],
+        annotate_links: @options[:annotate_links],
         from_ruby_package: true
-      }.compact
+      }
+
+      params[:page_separator] = @options[:page_separator] if @options[:page_separator]
+      params[:page_prefix] = @options[:page_prefix] if @options[:page_prefix]
+      params[:page_suffix] = @options[:page_suffix] if @options[:page_suffix]
+      params[:bounding_box] = @options[:bounding_box] if @options[:bounding_box]
+      params[:target_pages] = @options[:target_pages] if @options[:target_pages]
+      params[:webhook_url] = @options[:webhook_url] if @options[:webhook_url]
+      params[:azure_openai_deployment_name] = @options[:azure_openai_deployment_name] if @options[:azure_openai_deployment_name]
+      params[:azure_openai_endpoint] = @options[:azure_openai_endpoint] if @options[:azure_openai_endpoint]
+      params[:azure_openai_api_version] = @options[:azure_openai_api_version] if @options[:azure_openai_api_version]
+      params[:azure_openai_key] = @options[:azure_openai_key] if @options[:azure_openai_key]
+      params[:http_proxy] = @options[:http_proxy] if @options[:http_proxy]
+
+      if url
+        params[:input_url] = url.to_s
+      elsif file
+        params[:file] = file
+      end
+
+      params.compact
     end
 
     def get_job_status(job_id)
@@ -300,5 +349,20 @@ module Llamaparserb
         raise Error, "Unsupported file type: #{extension}. Supported types: #{SUPPORTED_FILE_TYPES.join(", ")}"
       end
     end
+
+    def create_job_from_url(url)
+      log "Creating job from URL: #{url}", :debug
+
+      response = @connection.post("upload") do |req|
+        req.headers["Authorization"] = "Bearer #{api_key}"
+        req.headers["Accept"] = "application/json"
+        # Create a simple form data request
+        req.options.timeout = 30  # Optional: add timeout
+        req.body = {"input_url" => url.to_s}
+      end
+
+      log "Response: #{response.body.inspect}", :debug
+      response.body["id"]
+    end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: llamaparserb
 version: !ruby/object:Gem::Version
-  version: 0.2.2
+  version: 0.3.0
 platform: ruby
 authors:
 - Heidar Bernhardsson