Nomad changes

docs/snippets/r/advanced/batch_extraction.md

@@ -0,0 +1,13 @@
+```r
+library(kreuzberg)
+
+# Batch extract from multiple files
+files <- c("report.pdf", "slides.pptx", "data.xlsx")
+results <- batch_extract_files_sync(files)
+
+for (i in seq_along(results)) {
+  cat(sprintf("File: %s\n", files[i]))
+  cat(sprintf("  MIME: %s\n", results[[i]]$mime_type))
+  cat(sprintf("  Length: %d chars\n\n", nchar(results[[i]]$content)))
+}
+```

docs/snippets/r/advanced/chunk_page_mapping.md

@@ -0,0 +1,27 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  chunking = list(max_characters = 500L, overlap = 50L),
+  pages = list(extract_pages = TRUE)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+for (i in seq_along(result$chunks)) {
+  chunk <- result$chunks[[i]]
+  metadata <- result$chunk_metadata[[i]]
+
+  if (!is.null(metadata$first_page) && !is.null(metadata$last_page)) {
+    page_range <- if (metadata$first_page == metadata$last_page) {
+      sprintf("Page %d", metadata$first_page)
+    } else {
+      sprintf("Pages %d-%d", metadata$first_page, metadata$last_page)
+    }
+
+    preview <- substr(chunk, 1L, min(50L, nchar(chunk)))
+    cat(sprintf("Chunk: %s... (%s)\n", preview, page_range))
+  }
+}
+```

docs/snippets/r/advanced/chunking_config.md

@@ -0,0 +1,37 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  chunking = list(max_characters = 1000L, overlap = 200L)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Chunks produced: %d\n", length(result$chunks)))
+for (i in seq_len(min(3L, length(result$chunks)))) {
+  cat(sprintf("Chunk %d length: %d characters\n", i, nchar(result$chunks[[i]])))
+}
+```
+
+```r title="R - Prepend Heading Context"
+library(kreuzberg)
+
+config <- list(
+  chunking = list(
+    max_characters = 500L,
+    overlap = 50L,
+    chunker_type = "markdown",
+    prepend_heading_context = TRUE
+  )
+)
+
+json <- extract_file_sync("document.md", "text/markdown", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+for (i in seq_len(min(3L, length(result$chunks)))) {
+  chunk <- result$chunks[[i]]
+  preview <- substr(chunk, 1L, min(100L, nchar(chunk)))
+  cat(sprintf("%s\n", preview))
+}
+```

docs/snippets/r/advanced/chunking_rag.md

@@ -0,0 +1,18 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  chunking = list(max_characters = 800L, overlap = 150L)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Total chunks: %d\n", length(result$chunks)))
+cat("Processing chunks for RAG pipeline:\n")
+
+for (i in seq_len(min(3L, length(result$chunks)))) {
+  chunk <- result$chunks[[i]]
+  cat(sprintf("Chunk %d: %d characters\n", i, nchar(chunk)))
+}
+```

docs/snippets/r/advanced/embedding_with_chunking.md

@@ -0,0 +1,23 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  chunking = list(max_characters = 1000L, overlap = 200L)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Preparing %d chunks for embedding:\n", length(result$chunks)))
+
+embeddings_data <- list()
+for (i in seq_along(result$chunks)) {
+  embeddings_data[[i]] <- list(
+    chunk_id = i,
+    text = result$chunks[[i]],
+    length = nchar(result$chunks[[i]])
+  )
+}
+
+cat(sprintf("Ready to embed %d chunks\n", length(embeddings_data)))
+```

docs/snippets/r/advanced/keyword_extraction_config.md

@@ -0,0 +1,18 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  keywords = list(
+    algorithm = "yake",
+    max_keywords = 10L,
+    min_score = 0.3,
+    ngram_range = c(1L, 3L),
+    language = "en"
+  )
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Keywords extracted: %d\n", length(result$keywords)))
+```

docs/snippets/r/advanced/keyword_extraction_example.md

@@ -0,0 +1,19 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  keywords = list(enabled = TRUE)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Keywords extracted: %d\n", length(result$keywords)))
+
+if (length(result$keywords) > 0) {
+  cat("Top keywords:\n")
+  for (i in seq_len(min(10L, length(result$keywords)))) {
+    cat(sprintf("  %d. %s\n", i, result$keywords[[i]]))
+  }
+}
+```

docs/snippets/r/advanced/language_detection_config.md

@@ -0,0 +1,22 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  language_detection = list(
+    enabled = TRUE,
+    min_confidence = 0.8,
+    detect_multiple = FALSE
+  )
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+if (length(result$detected_languages) > 0) {
+  cat(sprintf("Detected language: %s\n", result$detected_languages[[1]]))
+} else {
+  cat("No language detected\n")
+}
+
+cat(sprintf("Content length: %d characters\n", nchar(result$content)))
+```

docs/snippets/r/advanced/language_detection_multilingual.md

@@ -0,0 +1,13 @@
+```r title="R"
+library(kreuzberg)
+
+files <- c("english.pdf", "spanish.pdf", "french.pdf")
+config <- list(language_detection = list(enabled = TRUE))
+
+for (file in files) {
+  json <- extract_file_sync(file, "application/pdf", config)
+  result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+  cat(sprintf("%s: detected language = %s\n",
+              file, result$detected_language))
+}
+```

docs/snippets/r/advanced/quality_processing_config.md

@@ -0,0 +1,10 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(enable_quality_processing = TRUE)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Quality score: %.2f\n", result$quality_score))
+```

docs/snippets/r/advanced/quality_processing_example.md

@@ -0,0 +1,13 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(enable_quality_processing = TRUE)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat("Quality Metrics:\n")
+cat(sprintf("Quality Score: %.2f\n", result$quality_score))
+cat(sprintf("Content Length: %d characters\n", nchar(result$content)))
+cat(sprintf("Pages: %d\n", length(result$pages)))
+```

docs/snippets/r/advanced/token_reduction_config.md

@@ -0,0 +1,17 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  token_reduction = list(
+    mode = "moderate",
+    preserve_markdown = TRUE,
+    preserve_code = TRUE,
+    language_hint = "eng"
+  )
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat(sprintf("Reduced content length: %d characters\n", nchar(result$content)))
+```

docs/snippets/r/advanced/token_reduction_example.md

@@ -0,0 +1,14 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  token_reduction = list(enabled = TRUE)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+cat("Token-reduced content:\n")
+cat(sprintf("Length: %d characters\n", nchar(result$content)))
+cat(sprintf("Preview: %.60s...\n", result$content))
+```

docs/snippets/r/advanced/vector_database_integration.md

@@ -0,0 +1,24 @@
+```r title="R"
+library(kreuzberg)
+
+config <- list(
+  chunking = list(max_characters = 1000L, overlap = 200L)
+)
+
+json <- extract_file_sync("document.pdf", "application/pdf", config)
+result <- jsonlite::fromJSON(json, simplifyVector = FALSE)
+
+for (i in seq_len(min(3L, length(result$chunks)))) {
+  chunk <- result$chunks[[i]]
+  vector_doc <- list(
+    id = sprintf("doc_%d", i),
+    text = chunk,
+    metadata = list(
+      source = "document.pdf",
+      chunk_index = i,
+      length = nchar(chunk)
+    )
+  )
+  cat(sprintf("Vector DB entry %d: %d chars\n", i, nchar(chunk)))
+}
+```