Nothing
inst/biblioshiny/contentAnalysisServer.R
In bibliometrix: Comprehensive Science Mapping Analysis
## SERVER components for contentanalysis in Biblioshiny ----
#' Enhanced Server Logic for Content Analysis Tab with 3 Tabs
#'
#' @param input Shiny input object
#' @param output Shiny output object
#' @param session Shiny session object
#' @param values Reactive values object
#' Enhanced Server Logic for Content Analysis Tab with 3 Tabs
#'
#' @param input Shiny input object
#' @param output Shiny output object
#' @param session Shiny session object
#' @param values Reactive values object
content_analysis_server <- function(input, output, session, values) {
# Helper function for null coalescing
`%||%` <- function(lhs, rhs) {
if (!is.null(lhs) && length(lhs) > 0) lhs else rhs
}
# Cartella temporanea per i PDF
tmpdir <- tempdir()
# Rendi la cartella temporanea accessibile via HTTP
addResourcePath("tmpPDF", tmpdir)
# Preview visibility reactive value
preview_visible <- reactiveVal(TRUE)
# Check if text is extracted
output$text_extracted <- reactive({
return(!is.null(values$pdf_text) && nchar(values$pdf_text) > 0)
})
outputOptions(output, "text_extracted", suspendWhenHidden = FALSE)
# Preview visibility reactive output
output$preview_visible <- reactive({
return(preview_visible())
})
outputOptions(output, "preview_visible", suspendWhenHidden = FALSE)
# ===========================================
# PDF UPLOAD AND TEXT EXTRACTION
# ===========================================
# Check if PDF is uploaded
output$pdf_uploaded <- reactive({
return(!is.null(input$pdf_file))
})
outputOptions(output, "pdf_uploaded", suspendWhenHidden = FALSE)
# Display PDF info
output$pdf_info <- renderText({
if (!is.null(input$pdf_file)) {
file_size <- round(input$pdf_file$size / 1024 / 1024, 2)
paste("File:", input$pdf_file$name, "| Size:", file_size, "MB")
}
})
## PDF PREVIEW
output$pdf_viewer <- renderUI({
req(input$pdf_file)
# Verifica che sia un PDF
if (!grepl("\\.pdf$", input$pdf_file$name, ignore.case = TRUE)) {
return(
div(
style = "color: red; padding: 20px;",
"Errore: Il file selezionato non è un PDF"
)
)
}
# Copia il file nella cartella temporanea con nome univoco
pdf_name <- paste0(
"viewer_",
format(Sys.time(), "%Y%m%d_%H%M%S"),
"_",
input$pdf_file$name
)
pdf_path <- file.path(tmpdir, pdf_name)
file.copy(input$pdf_file$datapath, pdf_path, overwrite = TRUE)
# Crea l'iframe che punta al file tramite il path accessibile
tags$iframe(
src = paste0("tmpPDF/", pdf_name),
width = "100%",
height = "800px",
style = "border: 1px solid #ddd;"
)
})
# Pulizia alla chiusura della sessione
onSessionEnded(function() {
files <- list.files(
tmpdir,
pattern = "^viewer_.*\\.pdf$",
full.names = TRUE
)
file.remove(files)
})
## END PDF PREVIEW
# Extract text from PDF
observeEvent(input$extract_text, {
req(input$pdf_file)
# VALIDATION: Check if citation type is selected
if (
is.null(input$citation_type_import) ||
input$citation_type_import == "" ||
length(input$citation_type_import) == 0
) {
showNotification(
HTML(
"<strong>Citation Format Required</strong><br/>Please select the citation format used in your PDF before extracting text."
),
type = "warning",
duration = 5
)
return() # Stop execution
}
tryCatch(
{
if (is.na(input$Columns)) {
n_columns <- NULL
} else {
n_columns <- input$Columns
}
# Get citation type from input
citation_type <- input$citation_type_import
# Extract text with citation type parameter
pdf_text <- pdf2txt_auto(
input$pdf_file$datapath,
n_columns = n_columns,
citation_type = citation_type,
enable_ai_support = input$enable_ai_support,
ai_model = values$gemini_api_model,
api_key = NULL
)
values$pdf_text <- pdf_text[[1]]
# Extract PDF metadata (authors, title, journal, year, doi)
pdf_metadata_info <- tryCatch(
{
contentanalysis::extract_pdf_metadata(
input$pdf_file$datapath,
fields = "all"
)
},
error = function(e) {
NULL
}
)
# Store metadata in values
values$pdf_metadata <- pdf_metadata_info
# Check if AI support is enabled
if (input$enable_ai_support) {
if (!is.null(pdf_text)) {
showNotification(
"Text extracted successfully with AI enhancement!",
type = "message",
duration = 3
)
}
}
values$pdf_sections <- pdf_text
values$citation_type_used <- citation_type # Store for later use
pdf_metadata <- unlist(pdftools::pdf_info(input$pdf_file$datapath))
# Extract DOI
extracted_doi <- tryCatch(
{
extract_doi_from_pdf(input$pdf_file$datapath)
},
error = function(e) {
NULL
}
)
# Store DOI and update input field
if (
!is.null(extracted_doi) &&
!is.na(extracted_doi) &&
nzchar(trimws(extracted_doi))
) {
values$pdf_doi <- trimws(extracted_doi)
updateTextInput(
session,
"pdf_doi_input",
value = trimws(extracted_doi)
)
} else {
values$pdf_doi <- ""
updateTextInput(session, "pdf_doi_input", value = "")
}
# Create notification message based on citation type
notification_msg <- if (
!is.null(extracted_doi) && nzchar(trimws(extracted_doi))
) {
"PDF text extracted successfully! DOI detected and populated."
} else {
"PDF text extracted successfully! Please enter DOI manually if needed."
}
# Add citation type info to notification
citation_type_label <- switch(
citation_type,
"author_year" = "Author-year",
"numeric_bracketed" = "Numeric brackets",
"numeric_superscript" = "Numeric superscript",
"all" = "All formats"
)
if (citation_type == "numeric_superscript") {
notification_msg <- paste0(
notification_msg,
" Superscript citations converted to [n] format."
)
}
notification_msg <- paste0(
notification_msg,
" (Citation format: ",
citation_type_label,
")"
)
showNotification(
notification_msg,
type = "message",
duration = 5
)
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
},
error = function(e) {
showNotification(
paste("Error extracting PDF text:", e$message),
type = "error",
duration = 5
)
}
)
})
# Disable/Enable extract button based on citation type selection
observe({
if (
!is.null(input$citation_type_import) &&
input$citation_type_import != "" &&
length(input$citation_type_import) > 0
) {
# Citation type selected - ensure button is enabled
shinyjs::enable("extract_text")
shinyjs::removeCssClass("extract_text", "btn-secondary")
shinyjs::addCssClass("extract_text", "btn-info")
} else {
# No citation type selected - visual feedback but don't disable
# (We handle this in the observeEvent instead)
shinyjs::removeCssClass("extract_text", "btn-info")
shinyjs::addCssClass("extract_text", "btn-secondary")
}
})
# Display PDF metadata (authors, title, journal) - compact version for top-right
output$pdf_metadata_display <- renderUI({
if (!is.null(values$pdf_metadata)) {
metadata <- values$pdf_metadata
# Extract metadata fields with safe defaults
authors <- if (
!is.null(metadata$authors) &&
nzchar(metadata$authors) &&
!is.na(metadata$authors)
) {
if (nchar(metadata$authors) > 120) {
paste0(substr(metadata$authors, 1, 117), "...")
} else {
metadata$authors
}
} else {
"Authors not available"
}
title <- if (
!is.null(metadata$title) &&
nzchar(metadata$title) &&
!is.na(metadata$title)
) {
# Truncate title if too long
if (nchar(metadata$title) > 120) {
paste0(substr(metadata$title, 1, 117), "...")
} else {
metadata$title
}
} else {
"Title not available"
}
journal <- if (
!is.null(metadata$journal) &&
nzchar(metadata$journal) &&
!is.na(metadata$journal)
) {
metadata$journal
} else {
"Journal not available"
}
year <- if (
!is.null(metadata$year) &&
!is.na(metadata$year) &&
!is.na(metadata$year)
) {
as.character(metadata$year)
} else {
""
}
# Build compact display for top-right corner
div(
# style = "background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); padding: 12px 15px; border-radius: 6px; margin-bottom: 15px; box-shadow: 0 2px 4px rgba(0,0,0,0.1); color: white; text-align: left; font-size: 12px;",
style = "background: linear-gradient(135deg, #1D8FE1 0%, #5765B1 100%); padding: 12px 15px; border-radius: 6px; margin-bottom: 15px; box-shadow: 0 2px 4px rgba(0,0,0,0.1); color: white; text-align: left; font-size: 12px;",
# Title (most prominent)
div(
style = "margin-bottom: 6px; font-weight: 600; font-size: 13px; line-height: 1.3;",
title
),
# Authors
div(
style = "margin-bottom: 4px; opacity: 0.95;",
icon("user", style = "margin-right: 5px; font-size: 11px;"),
tags$span(authors, style = "font-size: 11px;")
),
# Journal and Year
div(
style = "opacity: 0.9;",
icon("book", style = "margin-right: 5px; font-size: 11px;"),
tags$span(
paste0(journal, if (nzchar(year)) paste0(" (", year, ")") else ""),
style = "font-size: 11px;"
)
)
)
}
})
# Check if metadata is available
output$metadata_available <- reactive({
return(!is.null(values$pdf_metadata))
})
outputOptions(output, "metadata_available", suspendWhenHidden = FALSE)
# Check if DOI was detected automatically
output$doi_detected <- reactive({
if (!is.null(values$pdf_doi) && nzchar(values$pdf_doi)) {
# Check if it's the same as what was automatically extracted
extracted_doi <- tryCatch(
{
if (!is.null(input$pdf_file)) {
extract_doi_from_pdf(input$pdf_file$datapath)
} else {
NULL
}
},
error = function(e) NULL
)
return(
!is.null(extracted_doi) &&
nzchar(extracted_doi) &&
extracted_doi == input$pdf_doi_input
)
}
return(FALSE)
})
outputOptions(output, "doi_detected", suspendWhenHidden = FALSE)
# Display DOI after analysis
output$doi_display_after_analysis <- renderText({
doi_value <- values$pdf_doi
if (!is.null(doi_value) && !is.na(doi_value) && nzchar(trimws(doi_value))) {
return(trimws(doi_value))
} else {
return("Not available")
}
})
# DOI status icon
output$doi_status_icon <- renderUI({
# Verifica se il campo DOI ha un valore
doi_value <- input$pdf_doi_input
if (is.null(doi_value) || !nzchar(trimws(doi_value))) {
# Nessun DOI inserito
return(
tagList(
icon(
"exclamation-triangle",
style = "color: #e74c3c; font-size: 18px; margin-right: 5px;"
),
tags$small(
"DOI not found",
style = "color: #e74c3c; font-weight: bold;"
)
)
)
}
# C'è un DOI, verifica se è stato auto-rilevato
auto_detected <- FALSE
if (!is.null(input$pdf_file)) {
extracted_doi <- tryCatch(
{
extract_doi_from_pdf(input$pdf_file$datapath)
},
error = function(e) {
NULL
}
)
# Verifica se il DOI estratto corrisponde a quello inserito
if (
!is.null(extracted_doi) &&
!is.na(extracted_doi) &&
nzchar(trimws(extracted_doi))
) {
auto_detected <- (trimws(extracted_doi) == trimws(doi_value))
}
}
# Restituisci l'icona appropriata
if (auto_detected) {
return(
tagList(
icon(
"check-circle",
style = "color: #27ae60; font-size: 18px; margin-right: 5px;"
),
tags$small(
"Auto-detected",
style = "color: #27ae60; font-weight: bold;"
)
)
)
} else {
return(
tagList(
icon(
"edit",
style = "color: #f39c12; font-size: 18px; margin-right: 5px;"
),
tags$small(
"Manually entered",
style = "color: #f39c12; font-weight: bold;"
)
)
)
}
})
# Update DOI value when user edits the field
observeEvent(
input$pdf_doi_input,
{
if (!is.null(input$pdf_doi_input)) {
values$pdf_doi <- trimws(input$pdf_doi_input)
}
},
ignoreInit = TRUE
)
# ===========================================
# LOAD PREVIOUSLY SAVED TEXT FILE
# ===========================================
# Check if text file is loaded
output$text_file_loaded <- reactive({
return(!is.null(input$load_text_file))
})
outputOptions(output, "text_file_loaded", suspendWhenHidden = FALSE)
# Display text file info
output$text_file_info <- renderText({
if (!is.null(input$load_text_file)) {
file_size <- round(input$load_text_file$size / 1024, 2)
paste("File:", input$load_text_file$name, "| Size:", file_size, "KB")
}
})
observeEvent(input$load_text_file, {
req(input$load_text_file)
tryCatch(
{
# Read the text file
file_content <- readLines(input$load_text_file$datapath, warn = FALSE)
if (length(file_content) == 0) {
showNotification(
"The text file is empty.",
type = "error",
duration = 5
)
return()
}
# Extract DOI from first line if present (format: "DOI: xxxxx")
first_line <- file_content[1]
extracted_doi <- ""
text_start_line <- 1
if (grepl("^DOI:\\s*", first_line, ignore.case = TRUE)) {
extracted_doi <- sub("^DOI:\\s*", "", first_line, ignore.case = TRUE)
extracted_doi <- trimws(extracted_doi)
text_start_line <- 2
}
# Extract CITATION_TYPE from second line if present (format: "CITATION_TYPE: xxxxx")
extracted_citation_type <- ""
if (length(file_content) >= 2) {
second_line <- file_content[2]
if (grepl("^CITATION_TYPE:\\s*", second_line, ignore.case = TRUE)) {
extracted_citation_type <- sub(
"^CITATION_TYPE:\\s*",
"",
second_line,
ignore.case = TRUE
)
extracted_citation_type <- trimws(extracted_citation_type)
text_start_line <- 3
}
}
# Get the text content (everything after metadata lines)
if (length(file_content) >= text_start_line) {
text_content <- reconstruct_from_txt(
file_content[text_start_line:length(file_content)]
)
} else {
text_content <- ""
}
# Store the text
values$pdf_text <- unlist(text_content[1])
values$pdf_sections <- text_content
# Update DOI field
if (nzchar(extracted_doi)) {
values$pdf_doi <- extracted_doi
updateTextInput(session, "pdf_doi_input", value = extracted_doi)
} else {
values$pdf_doi <- ""
updateTextInput(session, "pdf_doi_input", value = "")
}
# Update citation_type field if available
if (nzchar(extracted_citation_type)) {
values$citation_type_used <- extracted_citation_type
# Note: citation_type_import is only used during PDF import, not needed here
}
# Show notification
notification_msg <- "Text file loaded successfully!"
if (nzchar(extracted_doi)) {
notification_msg <- paste0(notification_msg, " DOI: ", extracted_doi)
}
if (nzchar(extracted_citation_type)) {
citation_type_label <- switch(
extracted_citation_type,
"author_year" = "Author-year",
"numeric_bracketed" = "Numeric brackets",
"numeric_superscript" = "Numeric superscript",
"all" = "All formats",
extracted_citation_type
)
notification_msg <- paste0(
notification_msg,
" | Citation format: ",
citation_type_label
)
}
showNotification(
notification_msg,
type = "message",
duration = 7
)
# Enable analysis button
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
},
error = function(e) {
showNotification(
paste("Error loading text file:", e$message),
type = "error",
duration = 5
)
}
)
})
# ===========================================
# SAVE EXTRACTED TEXT TO FILE
# ===========================================
output$save_text_file <- downloadHandler(
filename = function() {
# Generate filename with DOI or timestamp
if (!is.null(values$pdf_doi) && nzchar(values$pdf_doi)) {
# Clean DOI for filename (replace / with _)
clean_doi <- gsub("/", "_", values$pdf_doi)
clean_doi <- gsub("[^a-zA-Z0-9._-]", "", clean_doi)
paste0("extracted_text_", clean_doi, ".txt")
} else {
paste0("extracted_text_", format(Sys.time(), "%Y%m%d_%H%M%S"), ".txt")
}
},
content = function(file) {
# Prepare content with DOI and CITATION_TYPE in first lines
file_content <- ""
# Add DOI as first line if available
if (!is.null(values$pdf_doi) && nzchar(values$pdf_doi)) {
file_content <- paste0("DOI: ", values$pdf_doi, "\n")
} else {
file_content <- "DOI: \n"
}
# Add CITATION_TYPE as second line if available
if (
!is.null(values$citation_type_used) && nzchar(values$citation_type_used)
) {
file_content <- paste0(
file_content,
"CITATION_TYPE: ",
values$citation_type_used,
"\n"
)
} else {
file_content <- paste0(file_content, "CITATION_TYPE: \n")
}
if (length(values$pdf_sections) > 1) {
txt <- values$pdf_sections
section_names <- setdiff(names(txt), "Full_text")
# save txt in a text file adding a row with \n\nSECTION: section names\n\n before each section except Full_text
#file_content <- ""
for (section in section_names) {
file_content <- paste0(
file_content,
"\n\nSECTION LINE SEPARATION: ",
section,
"\n\n",
txt[[section]]
)
}
} else {
file_content <- paste0(file_content, unlist(values$pdf_sections[1]))
}
# # Add the extracted text
# if (!is.null(values$pdf_text)) {
# file_content <- paste0(file_content, values$pdf_text)
# }
# Write to file
writeLines(file_content, file, sep = "")
}
)
# Text length information
output$text_length_info <- renderText({
if (!is.null(values$pdf_text)) {
char_count <- nchar(values$pdf_text)
word_count <- length(strsplit(values$pdf_text, "\\s+")[[1]])
paste(
"Characters:",
format(char_count, big.mark = ","),
"| Words:",
format(word_count, big.mark = ",")
)
}
})
# Text preview output
output$text_preview <- renderText({
if (!is.null(values$pdf_text)) {
preview_length <- 80000
full_text <- values$pdf_text
if (nchar(full_text) > preview_length) {
truncated <- substr(full_text, 1, preview_length)
last_period <- max(c(
regexpr("\\. [A-Z]", truncated, perl = TRUE),
regexpr("\\.\n", truncated, perl = TRUE),
regexpr("\\?\n", truncated, perl = TRUE),
regexpr("!\n", truncated, perl = TRUE)
))
if (last_period > 100) {
truncated <- substr(truncated, 1, last_period)
}
paste0(
truncated,
"\n\n[...text continues for ",
format(nchar(full_text) - nchar(truncated), big.mark = ","),
" more characters...]"
)
} else {
full_text
}
}
})
# Toggle preview visibility
observeEvent(input$toggle_preview, {
preview_visible(!preview_visible())
if (preview_visible()) {
updateActionButton(session, "toggle_preview", "Hide Preview")
} else {
updateActionButton(session, "toggle_preview", "Show Preview")
}
})
# ===========================================
# CONTENT ANALYSIS
# ===========================================
observeEvent(input$run_analysis, {
req(values$pdf_text)
updateActionButton(
session,
"run_analysis",
label = "Processing...",
icon = icon("spinner", class = "fa-spin")
)
tryCatch(
{
custom_stops <- NULL
if (
!is.null(input$custom_stopwords) && nzchar(input$custom_stopwords)
) {
custom_stops <- trimws(strsplit(input$custom_stopwords, ",")[[1]])
}
# Determine citation segmentation
use_sections_cit <- switch(
input$citation_segmentation %||% "auto",
"auto" = "auto",
"sections" = TRUE,
"segments" = FALSE,
"auto" # default fallback
)
n_segs_cit <- input$n_segments_citations %||% 10
# Get citation type (use stored value from extraction or default)
citation_type <- values$citation_type_used
if (is.null(citation_type) || citation_type == "") {
citation_type <- input$citation_type_import
if (is.null(citation_type) || citation_type == "") {
citation_type <- "author_year" # ultimate fallback
}
}
# Run analysis with citation type parameter
values$analysis_results <- analyze_scientific_content(
text = values$pdf_sections,
doi = values$pdf_doi,
citation_type = citation_type,
window_size = input$window_size,
parse_multiple_citations = input$parse_multiple,
remove_stopwords = input$remove_stopwords,
custom_stopwords = custom_stops,
use_sections_for_citations = use_sections_cit,
n_segments_citations = n_segs_cit
)
# remove doi_pattern citations
values$analysis_results$citation_contexts <- values$analysis_results$citation_contexts %>%
dplyr::filter(!citation_type %in% c("doi_pattern"))
values$references_oa <- values$analysis_results$references_oa
section_colors <- colorlist()[
1:length(unique(values$analysis_results$citation_contexts$section))
]
names(section_colors) <- unique(
values$analysis_results$citation_contexts$section
)
values$analysis_results$section_colors <- section_colors
# Calculate readability indices
values$readability_indices <- calculate_readability_indices(
text = values$pdf_text,
detailed = TRUE
)
# Create network if we have data
if (
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
values$network_plot <- create_citation_network(
values$analysis_results,
max_distance = input$max_distance,
show_labels = TRUE
)
}
preview_visible(FALSE)
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
showNotification(
paste0(
"Content analysis completed successfully! Citation type: ",
switch(
citation_type,
"author_year" = "Author-Year",
"numeric_bracketed" = "Numeric Bracketed",
"numeric_superscript" = "Numeric Superscript",
"all" = "All Formats"
)
),
type = "message",
duration = 4
)
# Close the PDF import box
session$sendCustomMessage('togglePdfImport', list(action = 'close'))
},
error = function(e) {
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
showNotification(
paste("Error in content analysis:", e$message),
type = "error",
duration = 5
)
}
)
})
# Check if analysis is completed
output$analysis_completed <- reactive({
return(!is.null(values$analysis_results))
})
outputOptions(output, "analysis_completed", suspendWhenHidden = FALSE)
# Check if Gemini API is available
output$gemini_api_available <- reactive({
return(!is.null(values$geminiAPI) && values$geminiAPI == TRUE)
})
outputOptions(output, "gemini_api_available", suspendWhenHidden = FALSE)
# ===========================================
# TAB 1: DESCRIPTIVE STATISTICS OUTPUTS
# ===========================================
output$total_words <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$total_words_analyzed,
big.mark = ","
)
} else {
"0"
}
})
output$total_citations <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$citations_extracted,
big.mark = ","
)
} else {
"0"
}
})
output$narrative_citations <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$narrative_citations,
big.mark = ","
)
} else {
"0"
}
})
output$citation_density <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$citation_density_per_1000_words,
digits = 1
)
} else {
"0.0"
}
})
# Readability indices outputs
output$flesch_kincaid_grade <- renderText({
if (!is.null(values$readability_indices)) {
format(
round(values$readability_indices$flesch_kincaid_grade, 1),
nsmall = 1
)
} else {
"0.0"
}
})
output$flesch_reading_ease <- renderText({
if (!is.null(values$readability_indices)) {
format(
round(values$readability_indices$flesch_reading_ease, 1),
nsmall = 1
)
} else {
"0.0"
}
})
output$ari_index <- renderText({
if (!is.null(values$readability_indices)) {
format(
round(values$readability_indices$automated_readability_index, 1),
nsmall = 1
)
} else {
"0.0"
}
})
output$gunning_fog_index <- renderText({
if (!is.null(values$readability_indices)) {
format(round(values$readability_indices$gunning_fog_index, 1), nsmall = 1)
} else {
"0.0"
}
})
# Text Statistics - Matching height with readability
# Text Statistics - Matching structure with readability
output$text_stats <- renderUI({
if (!is.null(values$analysis_results)) {
stats <- values$analysis_results$text_analytics$basic_stats
# Get additional stats from readability indices
syllables_info <- ""
complex_info <- ""
if (!is.null(values$readability_indices)) {
syllables_info <- format(
values$readability_indices$n_syllables,
big.mark = ","
)
complex_info <- sprintf(
"Complex words: %s (%.1f%%)",
format(values$readability_indices$n_complex_words, big.mark = ","),
values$readability_indices$pct_complex_words
)
} else {
syllables_info <- "N/A"
complex_info <- "Complex words: N/A"
}
text_content <- sprintf(
"TEXT STATISTICS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔤 Characters: %s
📄 Words: %s
✏️ Sentences: %s
🗣 Syllables: %s
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
%s
Avg words/sentence: %.1f
Lexical diversity: %.3f",
format(stats$total_characters, big.mark = ","),
format(stats$total_words, big.mark = ","),
format(stats$total_sentences, big.mark = ","),
syllables_info,
complex_info,
round(stats$avg_words_per_sentence, 1),
round(values$analysis_results$summary$lexical_diversity, 3)
)
tags$pre(
style = "background-color: #f8f9fa; padding: 15px; border-radius: 6px; border: 1px solid #dee2e6; font-family: 'Courier New', monospace; font-size: 13px; line-height: 1.5; color: #333; margin: 0; min-height: 280px;",
text_content
)
} else {
tags$pre(
style = "background-color: #f8f9fa; padding: 40px; border-radius: 6px; border: 1px solid #dee2e6; text-align: center; color: #999; margin: 0; min-height: 280px; display: flex; align-items: center; justify-content: center;",
"No analysis data available"
)
}
})
# Readability Indices - Add min-height
output$readability_details_html <- renderUI({
if (!is.null(values$readability_indices)) {
indices <- values$readability_indices
# Helper to get interpretation symbol
get_symbol <- function(index_name, value) {
switch(
index_name,
"flesch_kincaid" = if (value <= 12) {
"✓"
} else if (value <= 16) {
"◆"
} else {
"▲"
},
"flesch_ease" = if (value >= 60) {
"✓"
} else if (value >= 30) {
"◆"
} else {
"▲"
},
"ari" = if (value <= 12) {
"✓"
} else if (value <= 16) {
"◆"
} else {
"▲"
},
"gunning" = if (value <= 12) {
"✓"
} else if (value <= 16) {
"◆"
} else {
"▲"
}
)
}
# Helper to get interpretation text
get_interp_text <- function(index_name, value) {
switch(
index_name,
"flesch_kincaid" = paste("Grade", round(value)),
"flesch_ease" = if (value >= 60) {
"Easy"
} else if (value >= 30) {
"Moderate"
} else {
"Difficult"
},
"ari" = paste("Grade", round(value)),
"gunning" = if (value <= 12) {
"Readable"
} else if (value <= 16) {
"Difficult"
} else {
"Very difficult"
}
)
}
text_content <- sprintf(
"READABILITY INDICES
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📚 Flesch-Kincaid: %5.1f %s %s
📖 Reading Ease: %5.1f %s %s
🤖 ARI Index: %5.1f %s %s
☁️ Gunning Fog: %5.1f %s %s
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Sentences: %s • Words: %s
Syllables: %s • Complex: %s (%.1f%%)
Avg sentence length: %.1f words",
indices$flesch_kincaid_grade,
get_symbol("flesch_kincaid", indices$flesch_kincaid_grade),
get_interp_text("flesch_kincaid", indices$flesch_kincaid_grade),
indices$flesch_reading_ease,
get_symbol("flesch_ease", indices$flesch_reading_ease),
get_interp_text("flesch_ease", indices$flesch_reading_ease),
indices$automated_readability_index,
get_symbol("ari", indices$automated_readability_index),
get_interp_text("ari", indices$automated_readability_index),
indices$gunning_fog_index,
get_symbol("gunning", indices$gunning_fog_index),
get_interp_text("gunning", indices$gunning_fog_index),
format(indices$n_sentences, big.mark = ","),
format(indices$n_words, big.mark = ","),
format(indices$n_syllables, big.mark = ","),
format(indices$n_complex_words, big.mark = ","),
indices$pct_complex_words,
indices$avg_sentence_length
)
tags$pre(
style = "background-color: #f8f9fa; padding: 15px; border-radius: 6px; border: 1px solid #dee2e6; font-family: 'Courier New', monospace; font-size: 13px; line-height: 1.5; color: #333; margin: 0; min-height: 280px;",
text_content
)
} else {
tags$pre(
style = "background-color: #f8f9fa; padding: 40px; border-radius: 6px; border: 1px solid #dee2e6; text-align: center; color: #999; margin: 0; min-height: 280px; display: flex; align-items: center; justify-content: center;",
"No readability data available"
)
}
})
# Citation types table
output$citation_types_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citation_metrics$type_distribution)
) {
values$analysis_results$citation_metrics$type_distribution
} else {
data.frame(
citation_type = character(0),
n = numeric(0),
percentage = numeric(0)
)
}
},
options = list(
pageLength = 10,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Citation sections table
output$citation_sections_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
!is.null(
values$analysis_results$citation_metrics$section_distribution
)
) {
values$analysis_results$citation_metrics$section_distribution %>%
filter(n > 0)
} else {
data.frame(
section = character(0),
n = numeric(0),
percentage = numeric(0)
)
}
},
options = list(
pageLength = 10,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Frequent words table
output$frequent_words_table <- DT::renderDataTable(
{
if (!is.null(values$analysis_results)) {
values$analysis_results$word_frequencies %>%
slice_head(n = 15) %>%
select(word, n)
} else {
data.frame(word = character(0), n = numeric(0))
}
},
options = list(
pageLength = 15,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Bigrams table
output$bigrams_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
"2gram" %in% names(values$analysis_results$ngrams)
) {
values$analysis_results$ngrams$`2gram` %>%
select(ngram, n) %>%
slice_head(n = 15)
} else {
data.frame(ngram = character(0), n = numeric(0))
}
},
options = list(
pageLength = 15,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Trigrams table
output$trigrams_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
"3gram" %in% names(values$analysis_results$ngrams)
) {
values$analysis_results$ngrams$`3gram` %>%
select(ngram, n) %>%
slice_head(n = 15)
} else {
data.frame(ngram = character(0), n = numeric(0))
}
},
options = list(
pageLength = 15,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Text statistics
# output$text_stats <- renderText({
# if (!is.null(values$analysis_results)) {
# stats <- values$analysis_results$text_analytics$basic_stats
# paste(
# "Characters:", format(stats$total_characters, big.mark = ","), "\n",
# "Words:", format(stats$total_words, big.mark = ","), "\n",
# "Sentences:", format(stats$total_sentences, big.mark = ","), "\n",
# "Avg words/sentence:", round(stats$avg_words_per_sentence, 1), "\n",
# "Lexical diversity:", round(values$analysis_results$summary$lexical_diversity, 3)
# )
# } else {
# "No analysis data available"
# }
# })
# ===========================================
# TAB 2: IN-CONTEXT CITATION ANALYSIS
# ===========================================
# ===========================================
# UPDATE CITATION GROUPING (without re-running full analysis)
# ===========================================
observeEvent(input$update_citation_grouping, {
req(values$pdf_text)
req(values$analysis_results)
tryCatch(
{
# Determine citation segmentation
use_sections_cit <- switch(
input$citation_segmentation %||% "auto",
"auto" = "auto",
"sections" = TRUE,
"segments" = FALSE,
"auto"
)
n_segs_cit <- input$n_segments_citations %||% 10
# Get citations without section column
citations_for_mapping <- values$analysis_results$citations %>%
select(-section, -segment_type)
# Re-map citations to new segments/sections
citations_remapped <- map_citations_to_segments(
citations_df = citations_for_mapping,
text = values$pdf_sections,
use_sections = use_sections_cit,
n_segments = n_segs_cit
)
# Update the citations in analysis_results
values$analysis_results$citations <- citations_remapped %>%
rename(section = segment)
# Update citation contexts with new section info
if (!is.null(values$analysis_results$citation_contexts)) {
# Remove old section column if exists
if ("section" %in% names(values$analysis_results$citation_contexts)) {
values$analysis_results$citation_contexts <- values$analysis_results$citation_contexts %>%
select(-section)
}
# Add new section mapping
values$analysis_results$citation_contexts <- values$analysis_results$citation_contexts %>%
left_join(
values$analysis_results$citations %>%
select(citation_id, section),
by = "citation_id"
)
section_colors <- colorlist()[
1:length(unique(values$analysis_results$citation_contexts$section))
]
names(section_colors) <- unique(
values$analysis_results$citation_contexts$section
)
values$analysis_results$section_colors <- section_colors
}
# Update citation metrics - section distribution
if (!is.null(values$analysis_results$citation_metrics)) {
# Determine expected sections/segments
sections_to_use <- NULL
if (
use_sections_cit == TRUE ||
(use_sections_cit == "auto" &&
length(setdiff(
names(values$pdf_sections),
c("Full_text", "References")
)) >
0)
) {
# Using sections
sections_to_use <- setdiff(
names(values$pdf_sections),
c("Full_text", "References")
)
} else {
# Using segments
sections_to_use <- paste0("Segment ", 1:n_segs_cit)
}
# Recalculate section distribution
if (!is.null(sections_to_use)) {
values$analysis_results$citation_metrics$section_distribution <-
values$analysis_results$citations %>%
mutate(section = factor(section, levels = sections_to_use)) %>%
count(section, sort = FALSE, .drop = FALSE) %>%
mutate(percentage = round(n / sum(n) * 100, 2))
} else {
values$analysis_results$citation_metrics$section_distribution <-
values$analysis_results$citations %>%
count(section, sort = TRUE) %>%
mutate(percentage = round(n / sum(n) * 100, 2))
}
}
# Recreate network with new grouping
if (
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
values$network_plot <- create_citation_network(
values$analysis_results,
max_distance = input$max_distance,
show_labels = TRUE
)
}
showNotification(
"Citation grouping updated successfully!",
type = "message",
duration = 2
)
},
error = function(e) {
showNotification(
paste("Error updating citation grouping:", e$message),
type = "error",
duration = 5
)
# Print error to console for debugging
cat("Error in update_citation_grouping:\n")
print(e)
}
)
})
# Custom HTML output for citation contexts
output$citation_contexts_html <- renderUI({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citation_contexts) &&
nrow(values$analysis_results$citation_contexts) > 0
) {
contexts <- values$analysis_results$citation_contexts
# Apply filters (REMOVED context_type_filter)
if (!is.null(input$context_search) && nzchar(input$context_search)) {
contexts <- contexts %>%
filter(
str_detect(
citation_text,
regex(input$context_search, ignore_case = TRUE)
) |
str_detect(
full_context,
regex(input$context_search, ignore_case = TRUE)
)
)
}
# REMOVED: type filter code
if (!is.null(input$context_min_words)) {
contexts <- contexts %>%
filter(context_word_count >= input$context_min_words)
}
# Create HTML for each citation context
if (nrow(contexts) > 0) {
citation_boxes <- lapply(1:nrow(contexts), function(i) {
context <- contexts[i, ]
section_colors <- values$analysis_results$section_colors
section_name <- if (
!is.null(context$section) && !is.na(context$section)
) {
context$section
} else {
"Unknown"
}
box_color <- section_colors[section_name]
if (is.na(box_color)) {
box_color <- "#CCCCCC"
}
has_reference <- FALSE
tooltip_text <- "No reference matched for this citation"
if ("ref_full_text" %in% names(context)) {
ref_value <- context[["ref_full_text"]]
if (
length(ref_value) > 0 &&
!is.na(ref_value) &&
is.character(ref_value) &&
nzchar(trimws(ref_value))
) {
has_reference <- TRUE
ref_text <- trimws(ref_value)
ref_text <- gsub("&", "&", ref_text)
ref_text <- gsub("<", "<", ref_text)
ref_text <- gsub(">", ">", ref_text)
ref_text <- gsub('"', """, ref_text)
ref_text <- gsub("'", "'", ref_text)
ref_text <- gsub("\n", "<br/>", ref_text)
tooltip_text <- paste0(
"<strong style='color: #4ECDC4;'>Reference:</strong><br/>",
ref_text
)
}
}
citation_id <- paste0("citation_", i, "_", sample(1:10000, 1))
div(
style = paste0(
"margin-bottom: 20px; padding: 15px; border-radius: 8px; ",
"box-shadow: 0 2px 4px rgba(0,0,0,0.1); ",
"border-left: 4px solid ",
box_color,
"; ",
"background-color: #fafafa;"
),
div(
style = "margin-bottom: 10px; font-size: 12px; color: #666;",
span(paste("Citation", i, "•"), style = "font-weight: bold;"),
span(
section_name,
style = paste0(
"color: ",
box_color,
"; font-weight: bold; margin-left: 8px;"
)
),
span(paste("• Position:", context$citation_position_in_text))
),
div(
style = "display: flex; align-items: center; font-family: 'Courier New', monospace;",
div(
style = "flex: 1; text-align: right; padding-right: 15px; color: #555; font-size: 14px;",
if (nzchar(context$words_before)) {
paste("...", context$words_before)
} else {
"[start of text]"
}
),
div(
id = citation_id,
class = "citation-with-tooltip",
style = paste0(
"background-color: ",
box_color,
"; color: white; ",
"padding: 8px 12px; border-radius: 6px; font-weight: bold; ",
"font-size: 14px; white-space: nowrap; max-width: 300px; ",
"overflow: hidden; text-overflow: ellipsis; ",
"cursor: ",
if (has_reference) "pointer" else "default",
"; position: relative;"
),
`data-toggle` = if (has_reference) "tooltip" else NULL,
`data-placement` = "top",
`data-html` = "true",
`data-original-title` = tooltip_text,
context$citation_text
),
div(
style = "flex: 1; text-align: left; padding-left: 15px; color: #555; font-size: 14px;",
if (nzchar(context$words_after)) {
paste(context$words_after, "...")
} else {
"[end of text]"
}
)
),
div(
style = "margin-top: 10px; font-size: 11px; color: #888;",
paste("Context words:", context$context_word_count, "•"),
if (!is.na(context$is_narrative) && context$is_narrative) {
span(
"Narrative citation",
style = "color: #e67e22; font-weight: bold;"
)
} else {
span(
"Parenthetical citation",
style = "color: #3498db; font-weight: bold;"
)
},
if (has_reference) {
span(
" • ",
icon("book", style = "color: #27ae60;"),
" Reference matched (hover to view)",
style = "color: #27ae60; font-weight: bold;"
)
} else {
span(
" • ",
icon("times-circle", style = "color: #e74c3c;"),
" No reference matched",
style = "color: #999; font-style: italic;"
)
}
)
)
})
tagList(
tags$script(HTML(
"
$(document).ready(function(){
$('.citation-with-tooltip').tooltip('dispose');
$('.citation-with-tooltip[data-toggle=\"tooltip\"]').tooltip({
container: 'body',
trigger: 'hover focus',
delay: { show: 200, hide: 100 },
html: true,
boundary: 'window'
});
setTimeout(function() {
$('.citation-with-tooltip[data-toggle=\"tooltip\"]').tooltip('dispose');
$('.citation-with-tooltip[data-toggle=\"tooltip\"]').tooltip({
container: 'body',
trigger: 'hover focus',
delay: { show: 200, hide: 100 },
html: true,
boundary: 'window'
});
}, 800);
});
"
)),
tags$style(HTML(
"
.tooltip {
font-family: Arial, sans-serif;
pointer-events: none;
}
.tooltip-inner {
max-width: 500px;
min-width: 250px;
text-align: left;
background-color: rgba(0, 0, 0, 0.95);
padding: 12px 16px;
font-size: 13px;
line-height: 1.6;
border-radius: 6px;
box-shadow: 0 4px 12px rgba(0,0,0,0.4);
}
.tooltip-inner strong {
color: #4ECDC4;
display: block;
margin-bottom: 6px;
}
.tooltip.show {
opacity: 1;
}
.tooltip.top .tooltip-arrow {
border-top-color: rgba(0, 0, 0, 0.95);
}
.citation-with-tooltip[data-toggle]:hover {
opacity: 0.9;
transform: scale(1.02);
transition: all 0.2s ease;
}
"
)),
citation_boxes
)
} else {
div(
style = "text-align: center; padding: 40px; color: #999;",
icon("search", style = "font-size: 24px; margin-bottom: 10px;"),
h4("No citations match your current filters"),
p("Try adjusting your search terms or filters.")
)
}
} else {
div(
style = "text-align: center; padding: 40px; color: #999;",
icon("quote-right", style = "font-size: 24px; margin-bottom: 10px;"),
h4("No citation contexts available"),
p("Run the analysis first to see in-context citations.")
)
}
})
# ===========================================
# TAB 3: NETWORK ANALYSIS
# ===========================================
# Network visualization
output$citation_network <- renderVisNetwork({
if (!is.null(values$network_plot)) {
values$network_plot
} else {
visNetwork(
nodes = data.frame(
id = 1,
label = "No citations found",
size = 20,
color = "#CCCCCC"
),
edges = data.frame(from = numeric(0), to = numeric(0))
) %>%
visOptions(highlightNearest = FALSE) %>%
visInteraction(dragNodes = FALSE, dragView = FALSE, zoomView = FALSE)
}
})
# Network information
output$network_info <- renderText({
if (!is.null(values$network_plot)) {
stats <- attr(values$network_plot, "stats")
if (!is.null(stats)) {
section_info <- ""
if (!is.null(stats$section_distribution)) {
section_info <- paste0(
"\n\nSection Distribution:\n",
paste(
stats$section_distribution$section,
": ",
stats$section_distribution$n,
collapse = "\n"
)
)
}
paste(
"Nodes:",
stats$n_nodes,
"\n",
"Edges:",
stats$n_edges,
"\n",
"Avg. Distance:",
stats$avg_distance,
"chars\n",
"Max Distance Filter:",
stats$max_distance,
"chars\n\n",
"Network Density:",
round(stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2), 3),
section_info
)
} else {
"Network statistics not available"
}
} else {
"No network data available"
}
})
# Strongest connections table
output$strongest_connections <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
values$analysis_results$network_data %>%
arrange(abs(distance)) %>%
slice_head(n = 8) %>%
select(citation1, citation2, distance) %>%
mutate(
citation1 = str_trunc(citation1, 20),
citation2 = str_trunc(citation2, 20),
distance = paste(abs(distance), "chars")
)
} else {
data.frame(
citation1 = character(0),
citation2 = character(0),
distance = character(0)
)
}
},
options = list(
pageLength = 8,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# ===========================================
# DOWNLOAD HANDLERS
# ===========================================
output$download_network <- downloadHandler(
filename = function() {
paste0("citation_network_", Sys.Date(), ".csv")
},
content = function(file) {
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$network_data)
) {
write.csv(values$analysis_results$network_data, file, row.names = FALSE)
} else {
write.csv(
data.frame(message = "No network data available"),
file,
row.names = FALSE
)
}
}
)
output$download_contexts <- downloadHandler(
filename = function() {
paste0("citation_contexts_", Sys.Date(), ".csv")
},
content = function(file) {
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citation_contexts)
) {
write.csv(
values$analysis_results$citation_contexts,
file,
row.names = FALSE
)
} else {
write.csv(
data.frame(message = "No citation contexts available"),
file,
row.names = FALSE
)
}
}
)
# ===========================================
# TAB 4: WORD TRENDS ANALYSIS
# ===========================================
# Populate word choices when analysis is complete
observe({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$word_frequencies) &&
nrow(values$analysis_results$word_frequencies) > 0
) {
# Get top 30 most frequent words
top_words <- values$analysis_results$word_frequencies %>%
slice_head(n = 30) %>%
pull(word)
# Get top bigrams if available
top_bigrams <- character(0)
if (
!is.null(values$analysis_results$ngrams) &&
"2gram" %in% names(values$analysis_results$ngrams)
) {
top_bigrams <- values$analysis_results$ngrams$`2gram` %>%
slice_head(n = 15) %>%
pull(ngram)
}
# Get top trigrams if available
top_trigrams <- character(0)
if (
!is.null(values$analysis_results$ngrams) &&
"3gram" %in% names(values$analysis_results$ngrams)
) {
top_trigrams <- values$analysis_results$ngrams$`3gram` %>%
slice_head(n = 10) %>%
pull(ngram)
}
# Combine all choices with labels
all_choices <- c(top_words, top_bigrams, top_trigrams)
# Create named list for optgroups
choices_list <- list(
"Top Words (1-gram)" = top_words
)
if (length(top_bigrams) > 0) {
choices_list[["Top Bigrams (2-gram)"]] <- top_bigrams
}
if (length(top_trigrams) > 0) {
choices_list[["Top Trigrams (3-gram)"]] <- top_trigrams
}
updateSelectizeInput(
session,
"trend_words",
choices = choices_list,
selected = NULL,
server = TRUE
)
}
})
# Check if sections are available
output$sections_available <- reactive({
if (!is.null(values$pdf_sections) && length(values$pdf_sections) > 0) {
section_names <- setdiff(
names(values$pdf_sections),
c("Full_text", "References")
)
return(length(section_names) > 0)
}
return(FALSE)
})
outputOptions(output, "sections_available", suspendWhenHidden = FALSE)
# Check if trends are available
output$trends_available <- reactive({
return(
!is.null(values$word_trends_data) && nrow(values$word_trends_data) > 0
)
})
outputOptions(output, "trends_available", suspendWhenHidden = FALSE)
# Update word trends when button is clicked
observeEvent(input$update_trends, {
# Validate inputs
if (is.null(values$pdf_text) || nchar(values$pdf_text) == 0) {
showNotification(
"Please extract text from PDF first.",
type = "warning",
duration = 3
)
return()
}
if (is.null(input$trend_words) || length(input$trend_words) == 0) {
showNotification(
"Please select at least one word to track.",
type = "warning",
duration = 3
)
return()
}
if (length(input$trend_words) > 10) {
showNotification(
"Maximum 10 words allowed. Please deselect some words.",
type = "warning",
duration = 3
)
return()
}
tryCatch(
{
# Determine use_sections parameter
use_sections_param <- switch(
input$segmentation_type,
"auto" = "auto",
"sections" = TRUE,
"segments" = FALSE
)
# Get number of segments
n_segs <- if (!is.null(input$n_segments)) input$n_segments else 10
# Determine which text to use
# If user explicitly chooses segments, use Full_text only
# Otherwise, use the sections structure
text_input <- if (input$segmentation_type == "segments") {
values$pdf_text # Just the full text as string
} else {
values$pdf_sections # List with sections
}
# Calculate word distribution
values$word_trends_data <- calculate_word_distribution(
text = text_input,
selected_words = input$trend_words,
use_sections = use_sections_param,
n_segments = n_segs,
remove_stopwords = FALSE # Don't remove stopwords for selected words
)
# Check if sections were requested but not available
sections_used <- attr(values$word_trends_data, "use_sections")
if (input$segmentation_type == "sections" && !sections_used) {
showNotification(
"Sections not available. Using equal-length segments instead.",
type = "warning",
duration = 4
)
}
showNotification(
"Word distribution calculated successfully!",
type = "message",
duration = 2
)
},
error = function(e) {
showNotification(
paste("Error calculating word distribution:", e$message),
type = "error",
duration = 5
)
}
)
})
# Render word trends plot
output$word_trends_plot <- renderPlotly({
req(values$word_trends_data)
tryCatch(
{
plot_word_distribution(
word_distribution_data = values$word_trends_data,
plot_type = input$trend_plot_type,
smooth = TRUE,
show_points = input$trend_show_points,
colors = NULL # Use automatic colors
)
},
error = function(e) {
showNotification(
paste("Error creating plot:", e$message),
type = "error",
duration = 5
)
return(NULL)
}
)
})
# Render word trends statistics table
output$word_trends_table <- DT::renderDataTable({
req(values$word_trends_data)
# Create summary statistics
stats_table <- values$word_trends_data %>%
group_by(word) %>%
summarise(
`Total Occurrences` = sum(count),
`Avg Frequency` = paste0(round(mean(relative_frequency) * 100, 3), "%"),
`Min Frequency` = paste0(round(min(relative_frequency) * 100, 3), "%"),
`Max Frequency` = paste0(round(max(relative_frequency) * 100, 3), "%"),
`Std Dev` = round(sd(relative_frequency) * 100, 3),
`Peak Segment` = segment_name[which.max(relative_frequency)],
.groups = "drop"
) %>%
arrange(desc(`Total Occurrences`))
DT::datatable(
stats_table,
options = list(
pageLength = 10,
dom = 'ftip',
ordering = TRUE,
searching = FALSE,
scrollX = TRUE
),
rownames = FALSE,
class = 'cell-border stripe'
) %>%
DT::formatStyle(
'word',
fontWeight = 'bold',
color = '#2E86AB'
)
})
# Download handler for word trends
output$download_word_trends <- downloadHandler(
filename = function() {
paste0("word_trends_", Sys.Date(), ".csv")
},
content = function(file) {
if (!is.null(values$word_trends_data)) {
# Prepare data for export
export_data <- values$word_trends_data %>%
select(
word,
segment_id,
segment_name,
count,
relative_frequency,
percentage,
total_words
) %>%
arrange(word, segment_id)
write.csv(export_data, file, row.names = FALSE)
} else {
write.csv(
data.frame(message = "No word trends data available"),
file,
row.names = FALSE
)
}
}
)
# ===========================================
# TAB 5: REFERENCES
# ===========================================
# Check if references are available
output$references_available <- reactive({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
return(nrow(values$analysis_results$parsed_references) > 0)
}
return(FALSE)
})
outputOptions(output, "references_available", suspendWhenHidden = FALSE)
# # Total references count
output$total_refs <- renderText({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
format(nrow(values$analysis_results$parsed_references), big.mark = ",")
} else {
"0"
}
})
# # PDF references count
output$pdf_refs <- renderText({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
refs <- values$analysis_results$parsed_references
pdf_refs <- sum(tolower(refs$ref_source) == "pdf", na.rm = TRUE)
format(pdf_refs, big.mark = ",")
} else {
"0"
}
})
# # Crossref references count
output$crossref_refs <- renderText({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
refs <- values$analysis_results$parsed_references
crossref_refs <- sum(tolower(refs$ref_source) == "crossref", na.rm = TRUE)
format(crossref_refs, big.mark = ",")
} else {
"0"
}
})
# Openalex references count
output$openalex_refs <- renderText({
req(values$references_oa)
if (is.null(values$references_oa) || nrow(values$references_oa) == 0) {
return("0")
}
nrow(values$references_oa)
})
# References HTML display
output$references_html <- renderUI({
req(values$analysis_results)
req(values$analysis_results$parsed_references)
refs <- values$analysis_results$parsed_references
# Apply search filter
if (!is.null(input$reference_search) && nzchar(input$reference_search)) {
search_term <- tolower(input$reference_search)
refs <- refs %>%
filter(
if_else(
!is.na(ref_full_text),
str_detect(tolower(ref_full_text), fixed(search_term)),
FALSE
) |
if_else(
!is.na(ref_authors),
str_detect(tolower(ref_authors), fixed(search_term)),
FALSE
) |
if_else(
!is.na(ref_year),
str_detect(tolower(ref_year), fixed(search_term)),
FALSE
) |
if_else(
!is.na(doi),
str_detect(tolower(doi), fixed(search_term)),
FALSE
)
)
}
if (nrow(refs) == 0) {
return(
div(
style = "text-align: center; padding: 40px; color: #999;",
icon("search", style = "font-size: 24px; margin-bottom: 10px;"),
h4("No references match your search"),
p("Try different search terms.")
)
)
}
# Create HTML for each reference
reference_items <- lapply(1:nrow(refs), function(i) {
ref <- refs[i, ]
# Get reference text
ref_text <- if (!is.na(ref$ref_full_text) && nzchar(ref$ref_full_text)) {
ref$ref_full_text
} else {
"[Reference text not available]"
}
# Check if OpenAlex data is available for this reference
has_oa_data <- FALSE
oa_button <- NULL
if (
!is.null(values$references_oa) && !is.na(ref$doi) && nzchar(ref$doi)
) {
# Cerca corrispondenza nel dataset OpenAlex
ref_doi_clean <- tolower(trimws(ref$doi))
# Prova diverse varianti del DOI
oa_match_idx <- which(
tolower(gsub("https://doi.org/", "", values$references_oa$doi)) ==
ref_doi_clean |
tolower(values$references_oa$doi) ==
paste0("https://doi.org/", ref_doi_clean)
)
if (length(oa_match_idx) > 0) {
has_oa_data <- TRUE
oa_button <- tags$button(
class = "btn btn-info btn-xs",
style = "float: right; margin-left: 10px;",
onclick = sprintf("showOADetails(%d)", i),
icon("info-circle"),
" View Details"
)
}
}
# Determine source badge
source_badge <- if (
!is.na(ref$ref_source) && tolower(ref$ref_source) == "crossref"
) {
tags$span(
class = "label label-success",
style = "font-size: 11px; margin-left: 10px;",
icon("cloud-download-alt"),
" Crossref"
)
} else {
tags$span(
class = "label label-info",
style = "font-size: 11px; margin-left: 10px;",
icon("file-pdf"),
" PDF"
)
}
# OpenAlex badge if available
oa_badge <- if (has_oa_data) {
tags$span(
class = "label label-primary",
style = "font-size: 11px; margin-left: 5px;",
icon("database"),
" OpenAlex"
)
}
# DOI link if available
doi_link <- NULL
if ("doi" %in% names(ref) && !is.na(ref$doi) && nzchar(ref$doi)) {
doi_link <- tags$div(
style = "margin-top: 8px; font-size: 12px;",
icon("link", style = "color: #3498db;"),
tags$a(
href = paste0("https://doi.org/", ref$doi),
target = "_blank",
ref$doi,
style = "color: #3498db; margin-left: 5px;"
)
)
}
# Author and year display
author_year_info <- NULL
if (!is.na(ref$ref_first_author) && !is.na(ref$ref_year)) {
n_auth <- if (!is.na(ref$n_authors) && ref$n_authors > 1) {
paste0(" et al. (", ref$n_authors, " authors)")
} else {
""
}
author_year_info <- tags$div(
style = "margin-top: 8px; font-size: 12px; color: #666;",
icon("user", style = "color: #3498db;"),
tags$span(
style = "margin-left: 5px;",
paste0(ref$ref_first_author, n_auth, " • ", ref$ref_year)
)
)
}
# Citation count if available
citation_info <- NULL
if (
"citation_count" %in%
names(ref) &&
!is.na(ref$citation_count) &&
ref$citation_count > 0
) {
citation_info <- tags$span(
style = "margin-left: 15px; color: #e67e22; font-size: 12px;",
icon("quote-right"),
sprintf(
" Cited %d time%s in text",
ref$citation_count,
ifelse(ref$citation_count > 1, "s", "")
)
)
}
div(
class = "reference-item",
`data-ref-index` = i,
style = "padding: 15px; margin-bottom: 15px; border-left: 4px solid #2E86AB; background-color: #fafafa; border-radius: 5px; box-shadow: 0 1px 3px rgba(0,0,0,0.1); position: relative;",
# Reference number, badges and OA button
div(
style = "margin-bottom: 10px;",
tags$span(
style = "font-weight: bold; color: #2E86AB; font-size: 16px;",
paste0("[", i, "]")
),
source_badge,
oa_badge,
citation_info,
oa_button
),
# Full reference text
div(
style = "font-family: 'Georgia', serif; font-size: 14px; line-height: 1.6; color: #333; clear: both;",
ref_text
),
# Author and year info
author_year_info,
# DOI link
doi_link
)
})
tagList(
tags$style(HTML(
"
.reference-item:hover {
background-color: #f0f8ff !important;
transition: background-color 0.2s ease;
}
.oa-document-summary .info-box {
display: flex;
align-items: center;
border-radius: 4px;
padding: 10px;
color: white;
}
.oa-document-summary .info-box-icon {
font-size: 32px;
flex: 0 0 70px;
text-align: center;
}
.oa-document-summary .info-box-content {
flex: 1;
}
.oa-document-summary .info-box-text {
display: block;
font-size: 12px;
text-transform: uppercase;
opacity: 0.9;
}
.oa-document-summary .info-box-number {
display: block;
font-size: 20px;
font-weight: bold;
margin-top: 5px;
}
"
)),
reference_items
)
})
# Download handler for references
output$download_references <- downloadHandler(
filename = function() {
paste0("references_", Sys.Date(), ".txt")
},
content = function(file) {
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
refs <- values$analysis_results$parsed_references
# Create formatted text
ref_text <- paste0(
"BIBLIOGRAPHY\n",
"============\n",
"Generated: ",
Sys.time(),
"\n",
"Total references: ",
nrow(refs),
"\n",
"From PDF: ",
sum(tolower(refs$ref_source) == "pdf", na.rm = TRUE),
"\n",
"From Crossref: ",
sum(tolower(refs$ref_source) == "crossref", na.rm = TRUE),
"\n\n"
)
for (i in 1:nrow(refs)) {
ref <- refs[i, ]
source_label <- if (
!is.na(ref$ref_source) && tolower(ref$ref_source) == "crossref"
) {
"[Crossref]"
} else {
"[PDF]"
}
ref_text <- paste0(
ref_text,
"[",
i,
"] ",
source_label,
"\n",
ref$ref_full_text,
"\n"
)
if (!is.na(ref$ref_first_author) && !is.na(ref$ref_year)) {
n_auth <- if (!is.na(ref$n_authors) && ref$n_authors > 1) {
paste0(" (", ref$n_authors, " authors)")
} else {
""
}
ref_text <- paste0(
ref_text,
"Author: ",
ref$ref_first_author,
n_auth,
" • Year: ",
ref$ref_year,
"\n"
)
}
if ("doi" %in% names(ref) && !is.na(ref$doi) && nzchar(ref$doi)) {
ref_text <- paste0(ref_text, "DOI: ", ref$doi, "\n")
}
if ("citation_count" %in% names(ref) && !is.na(ref$citation_count)) {
ref_text <- paste0(
ref_text,
"Cited ",
ref$citation_count,
" times in text\n"
)
}
ref_text <- paste0(ref_text, "\n")
}
writeLines(ref_text, file)
} else {
writeLines("No references available", file)
}
}
)
# ===========================================
# TAB: BIBLIOAI SUMMARY - SERVER LOGIC
# ===========================================
# ===========================================
# BIBLIOAI SUMMARY INITIALIZATION
# ===========================================
# Initialize API key from environment on startup
observe(
{
api_key_env <- Sys.getenv("GEMINI_API_KEY")
if (nzchar(api_key_env) && is.null(values$gemini_api_key)) {
values$gemini_api_key <- api_key_env
}
},
priority = 100
)
# ===========================================
# API STATUS DISPLAY
# ===========================================
output$gemini_api_status <- renderUI({
api_key_env <- Sys.getenv("GEMINI_API_KEY")
if (!nzchar(api_key_env)) {
tags$span(
class = "label label-danger",
style = "font-size: 12px; padding: 6px 12px;",
icon("exclamation-circle"),
" API Key Not Configured"
)
} else {
tags$span(
class = "label label-success",
style = "font-size: 12px; padding: 6px 12px;",
icon("check-circle"),
" API Key Configured"
)
}
})
# ===========================================
# SUMMARY TYPE DESCRIPTION
# ===========================================
output$summary_type_description <- renderUI({
req(input$summary_type)
descriptions <- list(
short_abstract = "A concise, one-paragraph abstract (around 250 words) covering objectives, methods, findings, and conclusions.",
narrative_abstract = "An engaging narrative (500-600 words) that tells the story of the research from problem to implications.",
imrad_summary = "A structured summary following the IMRaD format: Introduction, Methods, Results, and Discussion.",
thematic_bibliography = "Groups the references by research themes and provides descriptions of each thematic area.",
research_questions = "Extracts research questions, context, and motivation behind the study.",
background_literature = "Summarizes the theoretical framework and previous research that grounds this study.",
methods_summary = "Detailed breakdown of research design, data collection, analysis techniques, and tools used.",
implications = "Focuses on theoretical contributions and practical applications derived from the findings.",
list_tables_figures = "Complete list of all tables and figures with their captions as they appear in the document."
)
description <- descriptions[[input$summary_type]]
div(
style = "background-color: #f0f8ff; padding: 12px; border-radius: 4px; border-left: 3px solid #3498db;",
tags$small(
icon("info-circle", style = "color: #3498db; margin-right: 5px;"),
description,
style = "color: #555; line-height: 1.6;"
)
)
})
# ===========================================
# GENERATE SUMMARY
# ===========================================
observeEvent(input$generate_summary, {
# Validate PDF file
if (is.null(input$pdf_file)) {
showNotification(
"Please upload a PDF file first.",
type = "warning",
duration = 4
)
return()
}
# Validate API key
api_key <- Sys.getenv("GEMINI_API_KEY")
if (!nzchar(api_key)) {
showNotification(
HTML(
"<strong>API Key Required</strong><br/>
Please configure your Gemini API key in Biblioshiny settings.<br/>
You can get a free API key at: <a href='https://aistudio.google.com/apikey' target='_blank'>Google AI Studio</a>"
),
type = "error",
duration = 8
)
return()
}
# Get model from Biblioshiny settings or use default
model <- values$gemini_api_model
if (is.null(model) || !nzchar(model)) {
model <- "2.0-flash-lite"
}
# Use "huge" as default output size
output_size <- "huge"
tryCatch(
{
# Generate prompt
prompt <- get_article_summary_prompt(input$summary_type)
# Show progress notification
progress_id <- showNotification(
"Sending request to Gemini AI...",
type = "message",
duration = NULL,
closeButton = FALSE
)
# Call Gemini AI
result <- gemini_ai(
docs = input$pdf_file$datapath,
prompt = prompt,
model = model,
api_key = api_key,
outputSize = output_size,
retry_503 = 5
)
# Remove progress notification
removeNotification(progress_id)
# Check if result is an error message
if (is.character(result) && length(result) == 1) {
if (
grepl("^❌", result) || grepl("Error", result, ignore.case = TRUE)
) {
showNotification(
HTML(paste0(
"<strong>AI Request Failed</strong><br/>",
gsub("❌ ", "", result)
)),
type = "error",
duration = 10
)
return()
}
}
# Store results
values$biblioai_summary <- result
values$biblioai_summary_type <- input$summary_type
values$biblioai_summary_timestamp <- Sys.time()
# Get summary type label
summary_labels <- list(
short_abstract = "Short Abstract",
narrative_abstract = "Narrative Abstract",
imrad_summary = "IMRaD Summary",
thematic_bibliography = "Thematic Bibliography",
research_questions = "Research Questions & Context",
background_literature = "Background & Literature",
methods_summary = "Methods Summary",
implications = "Implications & Conclusions",
list_tables_figures = "Tables & Figures List"
)
summary_label <- summary_labels[[input$summary_type]]
showNotification(
HTML(paste0(
"<strong>Summary Generated!</strong><br/>",
"Type: ",
summary_label
)),
type = "message",
duration = 5
)
},
error = function(e) {
removeNotification(progress_id)
showNotification(
HTML(paste0(
"<strong>Error Generating Summary</strong><br/>",
e$message
)),
type = "error",
duration = 8
)
cat("BiblioAI Summary Error:\n")
print(e)
}
)
})
# ===========================================
# SUMMARY AVAILABILITY
# ===========================================
output$summary_available <- reactive({
return(!is.null(values$biblioai_summary))
})
outputOptions(output, "summary_available", suspendWhenHidden = FALSE)
# ===========================================
# SUMMARY DISPLAY
# ===========================================
output$summary_type_display <- renderText({
req(values$biblioai_summary_type)
summary_labels <- list(
short_abstract = "Short Abstract (250 words)",
narrative_abstract = "Narrative Abstract (500-600 words)",
imrad_summary = "IMRaD Structure Summary",
thematic_bibliography = "Thematic Bibliography",
research_questions = "Research Questions & Context",
background_literature = "Background & Literature Review",
methods_summary = "Methodology Summary",
implications = "Implications & Conclusions",
list_tables_figures = "Tables & Figures List"
)
summary_labels[[values$biblioai_summary_type]]
})
output$summary_timestamp <- renderText({
req(values$biblioai_summary_timestamp)
format(values$biblioai_summary_timestamp, "%Y-%m-%d %H:%M:%S")
})
output$summary_content_display <- renderUI({
req(values$biblioai_summary)
# Combine multiple text elements if result is a vector
summary_text <- if (
is.character(values$biblioai_summary) &&
length(values$biblioai_summary) > 1
) {
paste(values$biblioai_summary, collapse = "\n\n")
} else {
values$biblioai_summary
}
# Convert markdown-style headers to HTML for better display
summary_html <- summary_text
# Convert **text** to bold
summary_html <- gsub(
"\\*\\*([^*]+)\\*\\*",
"<strong>\\1</strong>",
summary_html
)
# Convert *text* to italic
summary_html <- gsub("\\*([^*]+)\\*", "<em>\\1</em>", summary_html)
# Convert ### headers to h4
summary_html <- gsub(
"^### (.+)$",
"<h4 style='color: #2E86AB; margin-top: 20px; margin-bottom: 10px;'>\\1</h4>",
summary_html,
perl = TRUE
)
# Convert ## headers to h3
summary_html <- gsub(
"^## (.+)$",
"<h3 style='color: #2E86AB; margin-top: 25px; margin-bottom: 10px;'>\\1</h3>",
summary_html,
perl = TRUE
)
# Convert # headers to h2
summary_html <- gsub(
"^# (.+)$",
"<h2 style='color: #2E86AB; margin-top: 30px; margin-bottom: 15px;'>\\1</h2>",
summary_html,
perl = TRUE
)
# Convert bullet points
summary_html <- gsub("^- (.+)$", "<li>\\1</li>", summary_html, perl = TRUE)
summary_html <- gsub(
"^\\* (.+)$",
"<li>\\1</li>",
summary_html,
perl = TRUE
)
# Wrap consecutive <li> in <ul>
summary_html <- gsub(
"(<li>.*?</li>)\n(<li>)",
"\\1\\2",
summary_html,
perl = TRUE
)
summary_html <- gsub(
"(<li>.*?</li>)",
"<ul style='margin-left: 20px;'>\\1</ul>",
summary_html,
perl = TRUE
)
# Convert line breaks to <br> but not inside lists
summary_html <- gsub("\n(?!<)", "<br/>", summary_html, perl = TRUE)
HTML(summary_html)
})
# ===========================================
# DOWNLOAD SUMMARY
# ===========================================
output$download_summary <- downloadHandler(
filename = function() {
summary_type <- values$biblioai_summary_type
if (is.null(summary_type)) {
summary_type <- "summary"
}
paste0("biblioai_", summary_type, "_", Sys.Date(), ".txt")
},
content = function(file) {
if (!is.null(values$biblioai_summary)) {
# Create header
header <- paste0(
"====================================\n",
"BiblioAI Summary\n",
"====================================\n\n",
"Summary Type: ",
values$biblioai_summary_type,
"\n",
"Generated: ",
format(values$biblioai_summary_timestamp, "%Y-%m-%d %H:%M:%S"),
"\n",
"Document: ",
input$pdf_file$name,
"\n\n",
"====================================\n\n"
)
# Combine multiple text elements if result is a vector
summary_text <- if (
is.character(values$biblioai_summary) &&
length(values$biblioai_summary) > 1
) {
paste(values$biblioai_summary, collapse = "\n\n")
} else {
values$biblioai_summary
}
# Write to file
writeLines(paste0(header, summary_text), file, useBytes = TRUE)
} else {
writeLines("No summary available", file)
}
}
)
# ===========================================
# RESET SUMMARY ON NEW PDF OR RESET
# ===========================================
# Clear summary when new PDF is uploaded
observeEvent(input$pdf_file, {
values$biblioai_summary <- NULL
values$biblioai_summary_type <- NULL
values$biblioai_summary_timestamp <- NULL
})
# Clear summary on reset
observeEvent(input$reset_analysis, {
values$pdf_metadata <- NULL
values$biblioai_summary <- NULL
values$biblioai_summary_type <- NULL
values$biblioai_summary_timestamp <- NULL
# Reset UI inputs
updateSelectInput(session, "summary_type", selected = "short_abstract")
})
# Clear metadata when new PDF is uploaded
observeEvent(input$pdf_file, {
values$pdf_metadata <- NULL
})
# ===========================================
# OPENALEX MODAL HANDLER
# ===========================================
observeEvent(input$selected_oa_ref_index, {
req(values$analysis_results$parsed_references)
req(values$references_oa)
req(input$selected_oa_ref_index)
tryCatch(
{
ref_index <- input$selected_oa_ref_index
ref <- values$analysis_results$parsed_references[ref_index, ]
if (is.na(ref$doi) || !nzchar(ref$doi)) {
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
"Reference Details",
style = "font-weight: 600; font-size: 20px;"
)),
content = as.character(div(
style = "text-align: center; padding: 40px; color: #999;",
icon(
"exclamation-triangle",
style = "font-size: 32px; margin-bottom: 15px; color: #f39c12;"
),
h4("No DOI available"),
p("Cannot fetch OpenAlex data without a DOI.")
))
)
)
return()
}
# Trova i dati OpenAlex corrispondenti
ref_doi_clean <- tolower(trimws(ref$doi))
oa_match_idx <- which(
tolower(gsub("https://doi.org/", "", values$references_oa$doi)) ==
ref_doi_clean |
tolower(values$references_oa$doi) ==
paste0("https://doi.org/", ref_doi_clean)
)
if (length(oa_match_idx) == 0) {
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
"Reference Details",
style = "font-weight: 600; font-size: 20px;"
)),
content = as.character(div(
style = "text-align: center; padding: 40px; color: #999;",
icon(
"database",
style = "font-size: 32px; margin-bottom: 15px; color: #e74c3c;"
),
h4("OpenAlex data not found"),
p(paste("Could not find OpenAlex data for DOI:", ref$doi))
))
)
)
return()
}
# Prendi il primo match
oa_data <- values$references_oa[oa_match_idx[1], ]
# Estrai il titolo per l'header
doc_title <- if (!is.na(oa_data$title) && nzchar(oa_data$title)) {
oa_data$title
} else if (
!is.na(oa_data$display_name) && nzchar(oa_data$display_name)
) {
oa_data$display_name
} else {
"Reference Details"
}
# Tronca il titolo se troppo lungo
if (nchar(doc_title) > 120) {
doc_title <- paste0(substr(doc_title, 1, 117), "...")
}
# Crea l'HTML per il modal
modal_content <- create_oa_details_html(oa_data)
# Aggiorna il modal con il titolo del documento
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
doc_title,
style = "font-weight: 600; font-size: 20px; line-height: 1.4;"
)),
content = as.character(modal_content)
)
)
},
error = function(e) {
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
"Error",
style = "font-weight: 600; font-size: 20px;"
)),
content = as.character(div(
style = "text-align: center; padding: 40px; color: #e74c3c;",
icon(
"exclamation-circle",
style = "font-size: 32px; margin-bottom: 15px;"
),
h4("Error loading details"),
p(paste("Error:", e$message))
))
)
)
cat("Error in OA modal handler:", e$message, "\n")
print(e)
}
)
})
# ===========================================
# RESET FUNCTIONALITY
# ===========================================
observeEvent(input$reset_analysis, {
values$pdf_text <- NULL
values$analysis_results <- NULL
values$network_plot <- NULL
values$analysis_running <- FALSE
values$readability_indices <- NULL
values$word_trends_data <- NULL
values$pdf_doi <- NULL
values$citation_type_used <- NULL
updateSelectizeInput(
session,
"trend_words",
choices = NULL,
selected = NULL
)
preview_visible(TRUE)
tryCatch(
{
shinyjs::reset("pdf_file")
updateTextInput(session, "pdf_doi_input", value = "")
# MODIFIED: Reset to no selection
updateRadioButtons(
session,
"citation_type_import",
selected = character(0)
)
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
updateActionButton(session, "toggle_preview", "Hide Preview")
updateTextInput(session, "context_search", value = "")
updateSelectInput(session, "context_type_filter", selected = "")
},
error = function(e) {
cat("Reset UI elements failed:", e$message, "\n")
}
)
showNotification(
"Analysis reset successfully!",
type = "message",
duration = 3
)
# Re-open the PDF import box
session$sendCustomMessage('togglePdfImport', list(action = 'open'))
})
# ===========================================
# EXCEL RESULTS DOWNLOAD
# ===========================================
# Download all results as Excel file
output$download_all_results <- downloadHandler(
filename = function() {
paste0("content_analysis_results_", Sys.Date(), ".xlsx")
},
content = function(file) {
# Verifica che ci siano risultati disponibili
if (is.null(values$analysis_results)) {
showNotification(
"No analysis results available. Please run the analysis first.",
type = "error",
duration = 5
)
return()
}
tryCatch(
{
# Create workbook
wb_content <- openxlsx::createWorkbook()
# === SHEET 1: SUMMARY STATISTICS ===
openxlsx::addWorksheet(wb_content, "Summary_Statistics")
summary_data <- data.frame(
Metric = c(
"Total Words",
"Total Citations",
"Narrative Citations",
"Parenthetical Citations",
"Citation Density (per 1000 words)",
"Lexical Diversity",
"Total Characters",
"Total Sentences",
"Avg Words per Sentence"
),
Value = c(
values$analysis_results$summary$total_words_analyzed,
values$analysis_results$summary$citations_extracted,
values$analysis_results$summary$narrative_citations,
values$analysis_results$summary$citations_extracted -
values$analysis_results$summary$narrative_citations,
round(
values$analysis_results$summary$citation_density_per_1000_words,
2
),
round(values$analysis_results$summary$lexical_diversity, 4),
values$analysis_results$text_analytics$basic_stats$total_characters,
values$analysis_results$text_analytics$basic_stats$total_sentences,
round(
values$analysis_results$text_analytics$basic_stats$avg_words_per_sentence,
2
)
),
stringsAsFactors = FALSE
)
openxlsx::writeData(
wb_content,
"Summary_Statistics",
summary_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Summary_Statistics",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#2E86AB",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:2,
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Summary_Statistics",
cols = 1:2,
widths = c(35, 20)
)
# === SHEET 2: READABILITY INDICES ===
if (!is.null(values$readability_indices)) {
openxlsx::addWorksheet(wb_content, "Readability_Indices")
readability_data <- data.frame(
Index = c(
"Flesch-Kincaid Grade Level",
"Flesch Reading Ease",
"Automated Readability Index (ARI)",
"Gunning Fog Index",
"Average Sentence Length (words)",
"Number of Sentences",
"Number of Words",
"Number of Syllables",
"Number of Complex Words",
"Percentage Complex Words"
),
Value = c(
round(values$readability_indices$flesch_kincaid_grade, 2),
round(values$readability_indices$flesch_reading_ease, 2),
round(
values$readability_indices$automated_readability_index,
2
),
round(values$readability_indices$gunning_fog_index, 2),
round(values$readability_indices$avg_sentence_length, 2),
values$readability_indices$n_sentences,
values$readability_indices$n_words,
values$readability_indices$n_syllables,
values$readability_indices$n_complex_words,
paste0(
round(values$readability_indices$pct_complex_words, 2),
"%"
)
),
stringsAsFactors = FALSE
)
openxlsx::writeData(
wb_content,
"Readability_Indices",
readability_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Readability_Indices",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#27ae60",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:2,
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Readability_Indices",
cols = 1:2,
widths = c(40, 20)
)
}
# === SHEET 3: WORD FREQUENCIES ===
if (
!is.null(values$analysis_results$word_frequencies) &&
nrow(values$analysis_results$word_frequencies) > 0
) {
openxlsx::addWorksheet(wb_content, "Word_Frequencies")
top_words <- values$analysis_results$word_frequencies %>%
slice_head(n = 50) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Word_Frequencies",
top_words,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Word_Frequencies",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#3498db",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(top_words),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Word_Frequencies",
cols = 1:ncol(top_words),
widths = "auto"
)
}
# === SHEET 4: BIGRAMS ===
if (
!is.null(values$analysis_results$ngrams) &&
"2gram" %in% names(values$analysis_results$ngrams)
) {
openxlsx::addWorksheet(wb_content, "Bigrams")
bigrams_data <- values$analysis_results$ngrams$`2gram` %>%
slice_head(n = 50) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Bigrams",
bigrams_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Bigrams",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#9b59b6",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(bigrams_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Bigrams",
cols = 1:ncol(bigrams_data),
widths = "auto"
)
}
# === SHEET 5: TRIGRAMS ===
if (
!is.null(values$analysis_results$ngrams) &&
"3gram" %in% names(values$analysis_results$ngrams)
) {
openxlsx::addWorksheet(wb_content, "Trigrams")
trigrams_data <- values$analysis_results$ngrams$`3gram` %>%
slice_head(n = 50) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Trigrams",
trigrams_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Trigrams",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#e67e22",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(trigrams_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Trigrams",
cols = 1:ncol(trigrams_data),
widths = "auto"
)
}
# === SHEET 6: CITATION TYPES ===
if (
!is.null(values$analysis_results$citation_metrics$type_distribution)
) {
openxlsx::addWorksheet(wb_content, "Citation_Types")
types_data <- values$analysis_results$citation_metrics$type_distribution %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Citation_Types",
types_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citation_Types",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#e74c3c",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(types_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citation_Types",
cols = 1:ncol(types_data),
widths = "auto"
)
}
# === SHEET 7: CITATIONS BY SECTION ===
if (
!is.null(
values$analysis_results$citation_metrics$section_distribution
)
) {
openxlsx::addWorksheet(wb_content, "Citations_by_Section")
sections_data <- values$analysis_results$citation_metrics$section_distribution %>%
filter(n > 0) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Citations_by_Section",
sections_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citations_by_Section",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#16a085",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(sections_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citations_by_Section",
cols = 1:ncol(sections_data),
widths = "auto"
)
}
# === SHEET 8: CITATION CONTEXTS ===
if (
!is.null(values$analysis_results$citation_contexts) &&
nrow(values$analysis_results$citation_contexts) > 0
) {
openxlsx::addWorksheet(wb_content, "Citation_Contexts")
# Seleziona le colonne principali se esistono, altrimenti esporta tutte
contexts_cols <- c(
"citation_id",
"citation_text",
"section",
"citation_type",
"is_narrative",
"words_before",
"words_after",
"full_context",
"context_word_count",
"citation_position_in_text"
)
available_cols <- intersect(
contexts_cols,
names(values$analysis_results$citation_contexts)
)
if (length(available_cols) > 0) {
contexts_export <- values$analysis_results$citation_contexts %>%
select(all_of(available_cols)) %>%
as.data.frame(stringsAsFactors = FALSE)
} else {
# Se nessuna colonna specificata esiste, esporta tutte
contexts_export <- values$analysis_results$citation_contexts %>%
as.data.frame(stringsAsFactors = FALSE)
}
openxlsx::writeData(
wb_content,
"Citation_Contexts",
contexts_export,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citation_Contexts",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#2E86AB",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(contexts_export),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citation_Contexts",
cols = 1:ncol(contexts_export),
widths = "auto"
)
}
# === SHEET 9: CITATION NETWORK ===
if (
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
openxlsx::addWorksheet(wb_content, "Citation_Network")
# Esporta tutte le colonne disponibili nel network
network_export <- values$analysis_results$network_data %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Citation_Network",
network_export,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citation_Network",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#8e44ad",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(network_export),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citation_Network",
cols = 1:ncol(network_export),
widths = "auto"
)
}
# === SHEET 10: WORD TRENDS (if available) ===
if (
!is.null(values$word_trends_data) &&
nrow(values$word_trends_data) > 0
) {
openxlsx::addWorksheet(wb_content, "Word_Trends")
trends_export <- values$word_trends_data %>%
arrange(word, segment_id) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Word_Trends",
trends_export,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Word_Trends",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#27ae60",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(trends_export),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Word_Trends",
cols = 1:ncol(trends_export),
widths = "auto"
)
# Add summary statistics for word trends
openxlsx::addWorksheet(wb_content, "Word_Trends_Summary")
trends_summary <- values$word_trends_data %>%
group_by(word) %>%
summarise(
total_occurrences = sum(count),
avg_frequency = mean(relative_frequency),
min_frequency = min(relative_frequency),
max_frequency = max(relative_frequency),
std_dev = sd(relative_frequency),
peak_segment = segment_name[which.max(relative_frequency)],
.groups = "drop"
) %>%
arrange(desc(total_occurrences)) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Word_Trends_Summary",
trends_summary,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Word_Trends_Summary",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#16a085",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(trends_summary),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Word_Trends_Summary",
cols = 1:ncol(trends_summary),
widths = "auto"
)
}
# Save workbook
openxlsx::saveWorkbook(wb_content, file, overwrite = TRUE)
},
error = function(e) {
# Log detailed error for debugging
cat("Error in download_all_results:\n")
cat("Error message:", e$message, "\n")
cat("Stack trace:\n")
print(e)
showNotification(
paste("Error exporting results:", e$message),
type = "error",
duration = 8
)
}
)
}
)
# ===========================================
# UTILITY FUNCTIONS
# ===========================================
is_analysis_valid <- reactive({
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citations) &&
nrow(values$analysis_results$citations) > 0
})
is_network_valid <- reactive({
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
})
output$debug_info <- renderText({
if (!is.null(values$analysis_results)) {
paste(
"Debug Info:\n",
"PDF text length:",
nchar(values$pdf_text %||% ""),
"\n",
"Citations found:",
nrow(values$analysis_results$citations %||% data.frame()),
"\n",
"Network connections:",
nrow(values$analysis_results$network_data %||% data.frame()),
"\n",
"Analysis running:",
values$analysis_running %||% FALSE
)
} else {
"No analysis data available"
}
})
outputOptions(output, "debug_info", suspendWhenHidden = FALSE)
}
# Helper function for HTML card OpenAlex ----
create_oa_details_html <- function(oa_data) {
# Helper per gestire valori NULL/NA
safe_value <- function(x, default = "Not available") {
if (
is.null(x) ||
length(x) == 0 ||
all(is.na(x)) ||
!nzchar(trimws(as.character(x)))
) {
return(default)
}
return(x)
}
# === ESTRAI AUTORI E AFFILIAZIONI INLINE ===
authors_html <- NULL
if (!is.null(oa_data$authorships) && length(oa_data$authorships) > 0) {
authorships_df <- oa_data$authorships[[1]]
if (!is.null(authorships_df) && nrow(authorships_df) > 0) {
authors_items <- lapply(1:min(nrow(authorships_df), 15), function(i) {
auth <- authorships_df[i, ]
author_name <- safe_value(auth$display_name, "Unknown Author")
author_id <- safe_value(auth$id, "")
# Estrai affiliazioni
affiliation_text <- ""
if (!is.null(auth$institutions) && length(auth$institutions) > 0) {
insts_df <- auth$institutions[[1]]
if (
!is.null(insts_df) &&
nrow(insts_df) > 0 &&
"display_name" %in% names(insts_df)
) {
affiliation_text <- paste(insts_df$display_name, collapse = ", ")
}
}
# Se non ci sono istituzioni, prova con affiliation_raw
if (
!nzchar(affiliation_text) &&
!is.na(auth$affiliation_raw) &&
nzchar(auth$affiliation_raw)
) {
affiliation_text <- auth$affiliation_raw
}
# Layout inline: nome e affiliazione sulla stessa riga
div(
style = "margin-bottom: 8px; padding: 6px 10px; background-color: #f8f9fa; border-radius: 4px; display: flex; align-items: center; flex-wrap: wrap;",
tags$div(
style = "font-weight: 500; font-size: 14px; margin-right: 12px;",
# Link OpenAlex se disponibile
if (nzchar(author_id) && author_id != "Not available") {
tags$a(
href = author_id,
target = "_blank",
style = "color: #2E86AB; text-decoration: none;",
author_name,
icon(
"external-link-alt",
style = "margin-left: 4px; font-size: 10px;"
)
)
} else {
tags$span(style = "color: #2E86AB;", author_name)
}
),
if (nzchar(affiliation_text)) {
tags$span(
style = "color: #666; font-size: 13px; font-style: italic;",
icon("university", style = "margin-right: 5px; font-size: 11px;"),
affiliation_text
)
}
)
})
authors_html <- tagList(authors_items)
if (nrow(authorships_df) > 15) {
authors_html <- tagList(
authors_html,
tags$small(
style = "color: #999; font-style: italic; display: block; margin-top: 8px; padding-left: 10px;",
paste("... and", nrow(authorships_df) - 15, "more authors")
)
)
}
}
}
# === KEYWORDS - TUTTE ===
keywords_html <- NULL
if (!is.null(oa_data$keywords) && length(oa_data$keywords) > 0) {
keywords_data <- oa_data$keywords[[1]]
if (
!is.null(keywords_data) &&
is.data.frame(keywords_data) &&
nrow(keywords_data) > 0
) {
if ("display_name" %in% names(keywords_data)) {
# NESSUN LIMITE - tutte le keywords
keywords_html <- lapply(keywords_data$display_name, function(kw) {
tags$span(
class = "label label-primary",
style = "margin-right: 5px; margin-bottom: 5px; font-size: 11px; display: inline-block; background-color: #3498db; color: white; padding: 4px 10px; border-radius: 3px;",
kw
)
})
}
}
}
# === ABSTRACT ===
abstract_text <- safe_value(oa_data$abstract, "Abstract not available")
# === TOPICS - SOLO I PRIMI 5 ===
topics_html <- NULL
if (!is.null(oa_data$topics) && length(oa_data$topics) > 0) {
topics_data <- oa_data$topics[[1]]
if (
!is.null(topics_data) &&
is.data.frame(topics_data) &&
nrow(topics_data) > 0
) {
if ("display_name" %in% names(topics_data)) {
main_topics <- topics_data
if ("type" %in% names(topics_data)) {
main_topics <- topics_data[topics_data$type == "topic", ]
}
if (nrow(main_topics) > 0) {
# LIMITE A 5 TOPICS
topics_html <- lapply(1:min(5, nrow(main_topics)), function(i) {
tags$span(
class = "label label-info",
style = "margin-right: 5px; margin-bottom: 5px; font-size: 11px; display: inline-block; background-color: #5bc0de; color: white; padding: 4px 10px; border-radius: 3px;",
main_topics$display_name[i]
)
})
}
}
}
}
# === JOURNAL INFO ===
journal_name <- safe_value(
oa_data$source_display_name,
"Journal not available"
)
issn <- safe_value(oa_data$issn_l, "")
# === DOCUMENT TYPE ===
doc_type <- safe_value(oa_data$type, "Unknown")
# === PUBLICATION INFO ===
pub_year <- safe_value(oa_data$publication_year, "N/A")
citations <- safe_value(oa_data$cited_by_count, 0)
fwci <- safe_value(oa_data$fwci, NA)
# === OPEN ACCESS INFO ===
is_oa <- !is.na(oa_data$is_oa) && oa_data$is_oa
oa_status <- tolower(safe_value(oa_data$oa_status, "closed")) # Normalizza in lowercase
# === DOI ===
doi_url <- safe_value(oa_data$doi, "")
doi_clean <- gsub("https://doi.org/", "", doi_url)
# === PDF URL ===
pdf_url <- safe_value(oa_data$pdf_url, "")
has_pdf <- nzchar(pdf_url) && !is.na(pdf_url) && pdf_url != "Not available"
# === CREA HTML COMPLETO ===
tagList(
div(
class = "oa-document-summary",
# Metadata Cards Row
fluidRow(
style = "margin-bottom: 20px;",
# Card 1: Publication Year (invariata)
column(
3,
div(
style = "background: linear-gradient(135deg, #3498db 0%, #2980b9 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);",
div(
style = "display: flex; align-items: center;",
icon(
"calendar",
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"Publication Year"
),
tags$div(
style = "font-size: 24px; font-weight: bold; margin-top: 2px;",
pub_year
)
)
)
)
),
# Card 2: Citations (invariata)
column(
3,
div(
style = "background: linear-gradient(135deg, #27ae60 0%, #229954 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);",
div(
style = "display: flex; align-items: center;",
icon(
"quote-right",
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"Citations"
),
tags$div(
style = "font-size: 24px; font-weight: bold; margin-top: 2px;",
format(as.numeric(citations), big.mark = ",")
)
)
)
)
),
# Card 3: Open Access Status (MODIFICATA)
column(3, {
# Determina colore e icona in base a oa_status
oa_colors <- list(
closed = list(
bg = "#95a5a6 0%, #7f8c8d",
icon = "lock",
text = "Closed"
),
gold = list(
bg = "#f39c12 0%, #e67e22",
icon = "unlock",
text = "Gold OA"
),
green = list(
bg = "#27ae60 0%, #229954",
icon = "unlock",
text = "Green OA"
),
bronze = list(
bg = "#cd7f32 0%, #b8732d",
icon = "unlock",
text = "Bronze OA"
),
hybrid = list(
bg = "#16a085 0%, #138d75",
icon = "unlock",
text = "Hybrid OA"
)
)
# Se oa_status non è nei valori conosciuti, usa closed come default
current_status <- if (oa_status %in% names(oa_colors)) {
oa_status
} else {
"closed"
}
status_info <- oa_colors[[current_status]]
div(
style = paste0(
"background: linear-gradient(135deg, ",
status_info$bg,
" 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);"
),
div(
style = "display: flex; align-items: center;",
icon(
status_info$icon,
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"Access"
),
tags$div(
style = "font-size: 18px; font-weight: bold; margin-top: 2px;",
status_info$text
)
)
)
)
}),
# Card 4: FWCI (invariata)
column(
3,
div(
style = "background: linear-gradient(135deg, #f39c12 0%, #e67e22 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);",
div(
style = "display: flex; align-items: center;",
icon(
"chart-line",
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"FWCI"
),
tags$div(
style = "font-size: 24px; font-weight: bold; margin-top: 2px;",
if (!is.na(fwci)) round(as.numeric(fwci), 2) else "N/A"
)
)
)
)
)
),
# Authors Section - PIÙ COMPATTA
if (!is.null(authors_html)) {
div(
class = "box box-primary",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("users", style = "margin-right: 8px;"),
"Authors",
style = "color: #2E86AB; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
style = "max-height: 250px; overflow-y: auto;",
authors_html
)
)
},
# Journal Info
div(
class = "box box-success",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("book", style = "margin-right: 8px;"),
"Journal",
style = "color: #27ae60; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
style = "padding: 15px;",
# Container principale
div(
style = "display: flex; align-items: center; justify-content: space-between; flex-wrap: wrap; gap: 20px;",
# Colonna 1: Nome Journal e ISSN
div(
style = "flex: 1 1 300px; min-width: 250px;",
tags$div(
style = "font-size: 16px; font-weight: 600; color: #2c3e50; margin-bottom: 5px;",
journal_name
),
if (nzchar(issn) && issn != "Not available") {
tags$div(
style = "color: #666; font-size: 13px;",
tags$strong("ISSN: "),
issn
)
}
),
# Colonna 2: Document Type
div(
style = "flex: 0 0 auto; display: flex; align-items: center; gap: 8px;",
tags$span(
style = "color: #666; font-size: 13px; font-weight: 500;",
"Type:"
),
tags$span(
class = "label label-info",
style = "font-size: 12px; padding: 5px 12px; background-color: #3498db; text-transform: capitalize;",
safe_value(oa_data$type, "Unknown")
)
),
# Colonna 3: DOI
if (nzchar(doi_url) && doi_url != "Not available") {
div(
style = "flex: 0 0 auto; display: flex; align-items: center; gap: 8px;",
tags$span(
style = "color: #666; font-size: 13px; font-weight: 500;",
"DOI:"
),
tags$a(
href = doi_url,
target = "_blank",
style = "color: #3498db; text-decoration: none; font-family: 'Courier New', monospace; font-size: 13px; font-weight: 500;",
doi_clean,
icon(
"external-link-alt",
style = "margin-left: 5px; font-size: 11px;"
)
)
)
}
)
)
),
# Abstract
div(
class = "box box-info",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("align-left", style = "margin-right: 8px;"),
"Abstract",
style = "color: #3498db; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
tags$p(
style = "text-align: justify; line-height: 1.7; color: #333;",
abstract_text
)
)
),
# Keywords - TUTTE
if (!is.null(keywords_html) && length(keywords_html) > 0) {
div(
class = "box box-warning",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("tags", style = "margin-right: 8px;"),
"Keywords",
tags$small(
style = "color: #999; font-weight: normal; margin-left: 8px;",
paste0("(", length(keywords_html), ")")
),
style = "color: #f39c12; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
keywords_html
)
)
},
# Topics - SOLO PRIMI 5
if (!is.null(topics_html) && length(topics_html) > 0) {
div(
class = "box",
style = "margin-bottom: 15px; border-top: 3px solid #9b59b6;",
div(
class = "box-header with-border",
h5(
icon("lightbulb", style = "margin-right: 8px;"),
"Research Topics",
tags$small(
style = "color: #999; font-weight: normal; margin-left: 8px;",
"(Top 5)"
),
style = "color: #9b59b6; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
topics_html
)
)
},
# Action Button - PDF
if (has_pdf) {
div(
style = "margin-top: 20px; padding-top: 15px; border-top: 1px solid #ddd; text-align: center;",
tags$a(
class = "btn btn-danger btn-lg",
href = pdf_url,
target = "_blank",
icon("file-pdf"),
" Download Full Text PDF",
style = "font-weight: bold; padding: 12px 30px;"
)
)
}
)
)
}
reconstruct_from_txt <- function(file_lines) {
reconstructed_txt <- list()
current_section <- "Full_text"
section_content <- ""
for (line in file_lines) {
if (grepl("^SECTION LINE SEPARATION: ", line)) {
# Save the previous section content
if (current_section != "Full_text") {
reconstructed_txt[[current_section]] <- section_content
}
# Start a new section
current_section <- sub("^SECTION LINE SEPARATION: ", "", line)
section_content <- ""
} else {
section_content <- paste0(
section_content,
ifelse(section_content == "", "", "\n"),
line
)
}
}
# Save the last section content
reconstructed_txt[[current_section]] <- section_content
if (!"Full_text" %in% names(reconstructed_txt)) {
Full_text <- paste(reconstructed_txt, collapse = "\n\n")
reconstructed_txt <- c(Full_text = Full_text, reconstructed_txt)
}
return(reconstructed_txt)
}
Try the bibliometrix package in your browser
Any scripts or data that you put into this service are public.
bibliometrix documentation built on Nov. 8, 2025, 5:06 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
## SERVER components for contentanalysis in Biblioshiny ----
#' Enhanced Server Logic for Content Analysis Tab with 3 Tabs
#'
#' @param input Shiny input object
#' @param output Shiny output object
#' @param session Shiny session object
#' @param values Reactive values object
#' Enhanced Server Logic for Content Analysis Tab with 3 Tabs
#'
#' @param input Shiny input object
#' @param output Shiny output object
#' @param session Shiny session object
#' @param values Reactive values object
content_analysis_server <- function(input, output, session, values) {
# Helper function for null coalescing
`%||%` <- function(lhs, rhs) {
if (!is.null(lhs) && length(lhs) > 0) lhs else rhs
}
# Cartella temporanea per i PDF
tmpdir <- tempdir()
# Rendi la cartella temporanea accessibile via HTTP
addResourcePath("tmpPDF", tmpdir)
# Preview visibility reactive value
preview_visible <- reactiveVal(TRUE)
# Check if text is extracted
output$text_extracted <- reactive({
return(!is.null(values$pdf_text) && nchar(values$pdf_text) > 0)
})
outputOptions(output, "text_extracted", suspendWhenHidden = FALSE)
# Preview visibility reactive output
output$preview_visible <- reactive({
return(preview_visible())
})
outputOptions(output, "preview_visible", suspendWhenHidden = FALSE)
# ===========================================
# PDF UPLOAD AND TEXT EXTRACTION
# ===========================================
# Check if PDF is uploaded
output$pdf_uploaded <- reactive({
return(!is.null(input$pdf_file))
})
outputOptions(output, "pdf_uploaded", suspendWhenHidden = FALSE)
# Display PDF info
output$pdf_info <- renderText({
if (!is.null(input$pdf_file)) {
file_size <- round(input$pdf_file$size / 1024 / 1024, 2)
paste("File:", input$pdf_file$name, "| Size:", file_size, "MB")
}
})
## PDF PREVIEW
output$pdf_viewer <- renderUI({
req(input$pdf_file)
# Verifica che sia un PDF
if (!grepl("\\.pdf$", input$pdf_file$name, ignore.case = TRUE)) {
return(
div(
style = "color: red; padding: 20px;",
"Errore: Il file selezionato non è un PDF"
)
)
}
# Copia il file nella cartella temporanea con nome univoco
pdf_name <- paste0(
"viewer_",
format(Sys.time(), "%Y%m%d_%H%M%S"),
"_",
input$pdf_file$name
)
pdf_path <- file.path(tmpdir, pdf_name)
file.copy(input$pdf_file$datapath, pdf_path, overwrite = TRUE)
# Crea l'iframe che punta al file tramite il path accessibile
tags$iframe(
src = paste0("tmpPDF/", pdf_name),
width = "100%",
height = "800px",
style = "border: 1px solid #ddd;"
)
})
# Pulizia alla chiusura della sessione
onSessionEnded(function() {
files <- list.files(
tmpdir,
pattern = "^viewer_.*\\.pdf$",
full.names = TRUE
)
file.remove(files)
})
## END PDF PREVIEW
# Extract text from PDF
observeEvent(input$extract_text, {
req(input$pdf_file)
# VALIDATION: Check if citation type is selected
if (
is.null(input$citation_type_import) ||
input$citation_type_import == "" ||
length(input$citation_type_import) == 0
) {
showNotification(
HTML(
"<strong>Citation Format Required</strong><br/>Please select the citation format used in your PDF before extracting text."
),
type = "warning",
duration = 5
)
return() # Stop execution
}
tryCatch(
{
if (is.na(input$Columns)) {
n_columns <- NULL
} else {
n_columns <- input$Columns
}
# Get citation type from input
citation_type <- input$citation_type_import
# Extract text with citation type parameter
pdf_text <- pdf2txt_auto(
input$pdf_file$datapath,
n_columns = n_columns,
citation_type = citation_type,
enable_ai_support = input$enable_ai_support,
ai_model = values$gemini_api_model,
api_key = NULL
)
values$pdf_text <- pdf_text[[1]]
# Extract PDF metadata (authors, title, journal, year, doi)
pdf_metadata_info <- tryCatch(
{
contentanalysis::extract_pdf_metadata(
input$pdf_file$datapath,
fields = "all"
)
},
error = function(e) {
NULL
}
)
# Store metadata in values
values$pdf_metadata <- pdf_metadata_info
# Check if AI support is enabled
if (input$enable_ai_support) {
if (!is.null(pdf_text)) {
showNotification(
"Text extracted successfully with AI enhancement!",
type = "message",
duration = 3
)
}
}
values$pdf_sections <- pdf_text
values$citation_type_used <- citation_type # Store for later use
pdf_metadata <- unlist(pdftools::pdf_info(input$pdf_file$datapath))
# Extract DOI
extracted_doi <- tryCatch(
{
extract_doi_from_pdf(input$pdf_file$datapath)
},
error = function(e) {
NULL
}
)
# Store DOI and update input field
if (
!is.null(extracted_doi) &&
!is.na(extracted_doi) &&
nzchar(trimws(extracted_doi))
) {
values$pdf_doi <- trimws(extracted_doi)
updateTextInput(
session,
"pdf_doi_input",
value = trimws(extracted_doi)
)
} else {
values$pdf_doi <- ""
updateTextInput(session, "pdf_doi_input", value = "")
}
# Create notification message based on citation type
notification_msg <- if (
!is.null(extracted_doi) && nzchar(trimws(extracted_doi))
) {
"PDF text extracted successfully! DOI detected and populated."
} else {
"PDF text extracted successfully! Please enter DOI manually if needed."
}
# Add citation type info to notification
citation_type_label <- switch(
citation_type,
"author_year" = "Author-year",
"numeric_bracketed" = "Numeric brackets",
"numeric_superscript" = "Numeric superscript",
"all" = "All formats"
)
if (citation_type == "numeric_superscript") {
notification_msg <- paste0(
notification_msg,
" Superscript citations converted to [n] format."
)
}
notification_msg <- paste0(
notification_msg,
" (Citation format: ",
citation_type_label,
")"
)
showNotification(
notification_msg,
type = "message",
duration = 5
)
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
},
error = function(e) {
showNotification(
paste("Error extracting PDF text:", e$message),
type = "error",
duration = 5
)
}
)
})
# Disable/Enable extract button based on citation type selection
observe({
if (
!is.null(input$citation_type_import) &&
input$citation_type_import != "" &&
length(input$citation_type_import) > 0
) {
# Citation type selected - ensure button is enabled
shinyjs::enable("extract_text")
shinyjs::removeCssClass("extract_text", "btn-secondary")
shinyjs::addCssClass("extract_text", "btn-info")
} else {
# No citation type selected - visual feedback but don't disable
# (We handle this in the observeEvent instead)
shinyjs::removeCssClass("extract_text", "btn-info")
shinyjs::addCssClass("extract_text", "btn-secondary")
}
})
# Display PDF metadata (authors, title, journal) - compact version for top-right
output$pdf_metadata_display <- renderUI({
if (!is.null(values$pdf_metadata)) {
metadata <- values$pdf_metadata
# Extract metadata fields with safe defaults
authors <- if (
!is.null(metadata$authors) &&
nzchar(metadata$authors) &&
!is.na(metadata$authors)
) {
if (nchar(metadata$authors) > 120) {
paste0(substr(metadata$authors, 1, 117), "...")
} else {
metadata$authors
}
} else {
"Authors not available"
}
title <- if (
!is.null(metadata$title) &&
nzchar(metadata$title) &&
!is.na(metadata$title)
) {
# Truncate title if too long
if (nchar(metadata$title) > 120) {
paste0(substr(metadata$title, 1, 117), "...")
} else {
metadata$title
}
} else {
"Title not available"
}
journal <- if (
!is.null(metadata$journal) &&
nzchar(metadata$journal) &&
!is.na(metadata$journal)
) {
metadata$journal
} else {
"Journal not available"
}
year <- if (
!is.null(metadata$year) &&
!is.na(metadata$year) &&
!is.na(metadata$year)
) {
as.character(metadata$year)
} else {
""
}
# Build compact display for top-right corner
div(
# style = "background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); padding: 12px 15px; border-radius: 6px; margin-bottom: 15px; box-shadow: 0 2px 4px rgba(0,0,0,0.1); color: white; text-align: left; font-size: 12px;",
style = "background: linear-gradient(135deg, #1D8FE1 0%, #5765B1 100%); padding: 12px 15px; border-radius: 6px; margin-bottom: 15px; box-shadow: 0 2px 4px rgba(0,0,0,0.1); color: white; text-align: left; font-size: 12px;",
# Title (most prominent)
div(
style = "margin-bottom: 6px; font-weight: 600; font-size: 13px; line-height: 1.3;",
title
),
# Authors
div(
style = "margin-bottom: 4px; opacity: 0.95;",
icon("user", style = "margin-right: 5px; font-size: 11px;"),
tags$span(authors, style = "font-size: 11px;")
),
# Journal and Year
div(
style = "opacity: 0.9;",
icon("book", style = "margin-right: 5px; font-size: 11px;"),
tags$span(
paste0(journal, if (nzchar(year)) paste0(" (", year, ")") else ""),
style = "font-size: 11px;"
)
)
)
}
})
# Check if metadata is available
output$metadata_available <- reactive({
return(!is.null(values$pdf_metadata))
})
outputOptions(output, "metadata_available", suspendWhenHidden = FALSE)
# Check if DOI was detected automatically
output$doi_detected <- reactive({
if (!is.null(values$pdf_doi) && nzchar(values$pdf_doi)) {
# Check if it's the same as what was automatically extracted
extracted_doi <- tryCatch(
{
if (!is.null(input$pdf_file)) {
extract_doi_from_pdf(input$pdf_file$datapath)
} else {
NULL
}
},
error = function(e) NULL
)
return(
!is.null(extracted_doi) &&
nzchar(extracted_doi) &&
extracted_doi == input$pdf_doi_input
)
}
return(FALSE)
})
outputOptions(output, "doi_detected", suspendWhenHidden = FALSE)
# Display DOI after analysis
output$doi_display_after_analysis <- renderText({
doi_value <- values$pdf_doi
if (!is.null(doi_value) && !is.na(doi_value) && nzchar(trimws(doi_value))) {
return(trimws(doi_value))
} else {
return("Not available")
}
})
# DOI status icon
output$doi_status_icon <- renderUI({
# Verifica se il campo DOI ha un valore
doi_value <- input$pdf_doi_input
if (is.null(doi_value) || !nzchar(trimws(doi_value))) {
# Nessun DOI inserito
return(
tagList(
icon(
"exclamation-triangle",
style = "color: #e74c3c; font-size: 18px; margin-right: 5px;"
),
tags$small(
"DOI not found",
style = "color: #e74c3c; font-weight: bold;"
)
)
)
}
# C'è un DOI, verifica se è stato auto-rilevato
auto_detected <- FALSE
if (!is.null(input$pdf_file)) {
extracted_doi <- tryCatch(
{
extract_doi_from_pdf(input$pdf_file$datapath)
},
error = function(e) {
NULL
}
)
# Verifica se il DOI estratto corrisponde a quello inserito
if (
!is.null(extracted_doi) &&
!is.na(extracted_doi) &&
nzchar(trimws(extracted_doi))
) {
auto_detected <- (trimws(extracted_doi) == trimws(doi_value))
}
}
# Restituisci l'icona appropriata
if (auto_detected) {
return(
tagList(
icon(
"check-circle",
style = "color: #27ae60; font-size: 18px; margin-right: 5px;"
),
tags$small(
"Auto-detected",
style = "color: #27ae60; font-weight: bold;"
)
)
)
} else {
return(
tagList(
icon(
"edit",
style = "color: #f39c12; font-size: 18px; margin-right: 5px;"
),
tags$small(
"Manually entered",
style = "color: #f39c12; font-weight: bold;"
)
)
)
}
})
# Update DOI value when user edits the field
observeEvent(
input$pdf_doi_input,
{
if (!is.null(input$pdf_doi_input)) {
values$pdf_doi <- trimws(input$pdf_doi_input)
}
},
ignoreInit = TRUE
)
# ===========================================
# LOAD PREVIOUSLY SAVED TEXT FILE
# ===========================================
# Check if text file is loaded
output$text_file_loaded <- reactive({
return(!is.null(input$load_text_file))
})
outputOptions(output, "text_file_loaded", suspendWhenHidden = FALSE)
# Display text file info
output$text_file_info <- renderText({
if (!is.null(input$load_text_file)) {
file_size <- round(input$load_text_file$size / 1024, 2)
paste("File:", input$load_text_file$name, "| Size:", file_size, "KB")
}
})
observeEvent(input$load_text_file, {
req(input$load_text_file)
tryCatch(
{
# Read the text file
file_content <- readLines(input$load_text_file$datapath, warn = FALSE)
if (length(file_content) == 0) {
showNotification(
"The text file is empty.",
type = "error",
duration = 5
)
return()
}
# Extract DOI from first line if present (format: "DOI: xxxxx")
first_line <- file_content[1]
extracted_doi <- ""
text_start_line <- 1
if (grepl("^DOI:\\s*", first_line, ignore.case = TRUE)) {
extracted_doi <- sub("^DOI:\\s*", "", first_line, ignore.case = TRUE)
extracted_doi <- trimws(extracted_doi)
text_start_line <- 2
}
# Extract CITATION_TYPE from second line if present (format: "CITATION_TYPE: xxxxx")
extracted_citation_type <- ""
if (length(file_content) >= 2) {
second_line <- file_content[2]
if (grepl("^CITATION_TYPE:\\s*", second_line, ignore.case = TRUE)) {
extracted_citation_type <- sub(
"^CITATION_TYPE:\\s*",
"",
second_line,
ignore.case = TRUE
)
extracted_citation_type <- trimws(extracted_citation_type)
text_start_line <- 3
}
}
# Get the text content (everything after metadata lines)
if (length(file_content) >= text_start_line) {
text_content <- reconstruct_from_txt(
file_content[text_start_line:length(file_content)]
)
} else {
text_content <- ""
}
# Store the text
values$pdf_text <- unlist(text_content[1])
values$pdf_sections <- text_content
# Update DOI field
if (nzchar(extracted_doi)) {
values$pdf_doi <- extracted_doi
updateTextInput(session, "pdf_doi_input", value = extracted_doi)
} else {
values$pdf_doi <- ""
updateTextInput(session, "pdf_doi_input", value = "")
}
# Update citation_type field if available
if (nzchar(extracted_citation_type)) {
values$citation_type_used <- extracted_citation_type
# Note: citation_type_import is only used during PDF import, not needed here
}
# Show notification
notification_msg <- "Text file loaded successfully!"
if (nzchar(extracted_doi)) {
notification_msg <- paste0(notification_msg, " DOI: ", extracted_doi)
}
if (nzchar(extracted_citation_type)) {
citation_type_label <- switch(
extracted_citation_type,
"author_year" = "Author-year",
"numeric_bracketed" = "Numeric brackets",
"numeric_superscript" = "Numeric superscript",
"all" = "All formats",
extracted_citation_type
)
notification_msg <- paste0(
notification_msg,
" | Citation format: ",
citation_type_label
)
}
showNotification(
notification_msg,
type = "message",
duration = 7
)
# Enable analysis button
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
},
error = function(e) {
showNotification(
paste("Error loading text file:", e$message),
type = "error",
duration = 5
)
}
)
})
# ===========================================
# SAVE EXTRACTED TEXT TO FILE
# ===========================================
output$save_text_file <- downloadHandler(
filename = function() {
# Generate filename with DOI or timestamp
if (!is.null(values$pdf_doi) && nzchar(values$pdf_doi)) {
# Clean DOI for filename (replace / with _)
clean_doi <- gsub("/", "_", values$pdf_doi)
clean_doi <- gsub("[^a-zA-Z0-9._-]", "", clean_doi)
paste0("extracted_text_", clean_doi, ".txt")
} else {
paste0("extracted_text_", format(Sys.time(), "%Y%m%d_%H%M%S"), ".txt")
}
},
content = function(file) {
# Prepare content with DOI and CITATION_TYPE in first lines
file_content <- ""
# Add DOI as first line if available
if (!is.null(values$pdf_doi) && nzchar(values$pdf_doi)) {
file_content <- paste0("DOI: ", values$pdf_doi, "\n")
} else {
file_content <- "DOI: \n"
}
# Add CITATION_TYPE as second line if available
if (
!is.null(values$citation_type_used) && nzchar(values$citation_type_used)
) {
file_content <- paste0(
file_content,
"CITATION_TYPE: ",
values$citation_type_used,
"\n"
)
} else {
file_content <- paste0(file_content, "CITATION_TYPE: \n")
}
if (length(values$pdf_sections) > 1) {
txt <- values$pdf_sections
section_names <- setdiff(names(txt), "Full_text")
# save txt in a text file adding a row with \n\nSECTION: section names\n\n before each section except Full_text
#file_content <- ""
for (section in section_names) {
file_content <- paste0(
file_content,
"\n\nSECTION LINE SEPARATION: ",
section,
"\n\n",
txt[[section]]
)
}
} else {
file_content <- paste0(file_content, unlist(values$pdf_sections[1]))
}
# # Add the extracted text
# if (!is.null(values$pdf_text)) {
# file_content <- paste0(file_content, values$pdf_text)
# }
# Write to file
writeLines(file_content, file, sep = "")
}
)
# Text length information
output$text_length_info <- renderText({
if (!is.null(values$pdf_text)) {
char_count <- nchar(values$pdf_text)
word_count <- length(strsplit(values$pdf_text, "\\s+")[[1]])
paste(
"Characters:",
format(char_count, big.mark = ","),
"| Words:",
format(word_count, big.mark = ",")
)
}
})
# Text preview output
output$text_preview <- renderText({
if (!is.null(values$pdf_text)) {
preview_length <- 80000
full_text <- values$pdf_text
if (nchar(full_text) > preview_length) {
truncated <- substr(full_text, 1, preview_length)
last_period <- max(c(
regexpr("\\. [A-Z]", truncated, perl = TRUE),
regexpr("\\.\n", truncated, perl = TRUE),
regexpr("\\?\n", truncated, perl = TRUE),
regexpr("!\n", truncated, perl = TRUE)
))
if (last_period > 100) {
truncated <- substr(truncated, 1, last_period)
}
paste0(
truncated,
"\n\n[...text continues for ",
format(nchar(full_text) - nchar(truncated), big.mark = ","),
" more characters...]"
)
} else {
full_text
}
}
})
# Toggle preview visibility
observeEvent(input$toggle_preview, {
preview_visible(!preview_visible())
if (preview_visible()) {
updateActionButton(session, "toggle_preview", "Hide Preview")
} else {
updateActionButton(session, "toggle_preview", "Show Preview")
}
})
# ===========================================
# CONTENT ANALYSIS
# ===========================================
observeEvent(input$run_analysis, {
req(values$pdf_text)
updateActionButton(
session,
"run_analysis",
label = "Processing...",
icon = icon("spinner", class = "fa-spin")
)
tryCatch(
{
custom_stops <- NULL
if (
!is.null(input$custom_stopwords) && nzchar(input$custom_stopwords)
) {
custom_stops <- trimws(strsplit(input$custom_stopwords, ",")[[1]])
}
# Determine citation segmentation
use_sections_cit <- switch(
input$citation_segmentation %||% "auto",
"auto" = "auto",
"sections" = TRUE,
"segments" = FALSE,
"auto" # default fallback
)
n_segs_cit <- input$n_segments_citations %||% 10
# Get citation type (use stored value from extraction or default)
citation_type <- values$citation_type_used
if (is.null(citation_type) || citation_type == "") {
citation_type <- input$citation_type_import
if (is.null(citation_type) || citation_type == "") {
citation_type <- "author_year" # ultimate fallback
}
}
# Run analysis with citation type parameter
values$analysis_results <- analyze_scientific_content(
text = values$pdf_sections,
doi = values$pdf_doi,
citation_type = citation_type,
window_size = input$window_size,
parse_multiple_citations = input$parse_multiple,
remove_stopwords = input$remove_stopwords,
custom_stopwords = custom_stops,
use_sections_for_citations = use_sections_cit,
n_segments_citations = n_segs_cit
)
# remove doi_pattern citations
values$analysis_results$citation_contexts <- values$analysis_results$citation_contexts %>%
dplyr::filter(!citation_type %in% c("doi_pattern"))
values$references_oa <- values$analysis_results$references_oa
section_colors <- colorlist()[
1:length(unique(values$analysis_results$citation_contexts$section))
]
names(section_colors) <- unique(
values$analysis_results$citation_contexts$section
)
values$analysis_results$section_colors <- section_colors
# Calculate readability indices
values$readability_indices <- calculate_readability_indices(
text = values$pdf_text,
detailed = TRUE
)
# Create network if we have data
if (
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
values$network_plot <- create_citation_network(
values$analysis_results,
max_distance = input$max_distance,
show_labels = TRUE
)
}
preview_visible(FALSE)
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
showNotification(
paste0(
"Content analysis completed successfully! Citation type: ",
switch(
citation_type,
"author_year" = "Author-Year",
"numeric_bracketed" = "Numeric Bracketed",
"numeric_superscript" = "Numeric Superscript",
"all" = "All Formats"
)
),
type = "message",
duration = 4
)
# Close the PDF import box
session$sendCustomMessage('togglePdfImport', list(action = 'close'))
},
error = function(e) {
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
showNotification(
paste("Error in content analysis:", e$message),
type = "error",
duration = 5
)
}
)
})
# Check if analysis is completed
output$analysis_completed <- reactive({
return(!is.null(values$analysis_results))
})
outputOptions(output, "analysis_completed", suspendWhenHidden = FALSE)
# Check if Gemini API is available
output$gemini_api_available <- reactive({
return(!is.null(values$geminiAPI) && values$geminiAPI == TRUE)
})
outputOptions(output, "gemini_api_available", suspendWhenHidden = FALSE)
# ===========================================
# TAB 1: DESCRIPTIVE STATISTICS OUTPUTS
# ===========================================
output$total_words <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$total_words_analyzed,
big.mark = ","
)
} else {
"0"
}
})
output$total_citations <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$citations_extracted,
big.mark = ","
)
} else {
"0"
}
})
output$narrative_citations <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$narrative_citations,
big.mark = ","
)
} else {
"0"
}
})
output$citation_density <- renderText({
if (!is.null(values$analysis_results)) {
format(
values$analysis_results$summary$citation_density_per_1000_words,
digits = 1
)
} else {
"0.0"
}
})
# Readability indices outputs
output$flesch_kincaid_grade <- renderText({
if (!is.null(values$readability_indices)) {
format(
round(values$readability_indices$flesch_kincaid_grade, 1),
nsmall = 1
)
} else {
"0.0"
}
})
output$flesch_reading_ease <- renderText({
if (!is.null(values$readability_indices)) {
format(
round(values$readability_indices$flesch_reading_ease, 1),
nsmall = 1
)
} else {
"0.0"
}
})
output$ari_index <- renderText({
if (!is.null(values$readability_indices)) {
format(
round(values$readability_indices$automated_readability_index, 1),
nsmall = 1
)
} else {
"0.0"
}
})
output$gunning_fog_index <- renderText({
if (!is.null(values$readability_indices)) {
format(round(values$readability_indices$gunning_fog_index, 1), nsmall = 1)
} else {
"0.0"
}
})
# Text Statistics - Matching height with readability
# Text Statistics - Matching structure with readability
output$text_stats <- renderUI({
if (!is.null(values$analysis_results)) {
stats <- values$analysis_results$text_analytics$basic_stats
# Get additional stats from readability indices
syllables_info <- ""
complex_info <- ""
if (!is.null(values$readability_indices)) {
syllables_info <- format(
values$readability_indices$n_syllables,
big.mark = ","
)
complex_info <- sprintf(
"Complex words: %s (%.1f%%)",
format(values$readability_indices$n_complex_words, big.mark = ","),
values$readability_indices$pct_complex_words
)
} else {
syllables_info <- "N/A"
complex_info <- "Complex words: N/A"
}
text_content <- sprintf(
"TEXT STATISTICS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔤 Characters: %s
📄 Words: %s
✏️ Sentences: %s
🗣 Syllables: %s
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
%s
Avg words/sentence: %.1f
Lexical diversity: %.3f",
format(stats$total_characters, big.mark = ","),
format(stats$total_words, big.mark = ","),
format(stats$total_sentences, big.mark = ","),
syllables_info,
complex_info,
round(stats$avg_words_per_sentence, 1),
round(values$analysis_results$summary$lexical_diversity, 3)
)
tags$pre(
style = "background-color: #f8f9fa; padding: 15px; border-radius: 6px; border: 1px solid #dee2e6; font-family: 'Courier New', monospace; font-size: 13px; line-height: 1.5; color: #333; margin: 0; min-height: 280px;",
text_content
)
} else {
tags$pre(
style = "background-color: #f8f9fa; padding: 40px; border-radius: 6px; border: 1px solid #dee2e6; text-align: center; color: #999; margin: 0; min-height: 280px; display: flex; align-items: center; justify-content: center;",
"No analysis data available"
)
}
})
# Readability Indices - Add min-height
output$readability_details_html <- renderUI({
if (!is.null(values$readability_indices)) {
indices <- values$readability_indices
# Helper to get interpretation symbol
get_symbol <- function(index_name, value) {
switch(
index_name,
"flesch_kincaid" = if (value <= 12) {
"✓"
} else if (value <= 16) {
"◆"
} else {
"▲"
},
"flesch_ease" = if (value >= 60) {
"✓"
} else if (value >= 30) {
"◆"
} else {
"▲"
},
"ari" = if (value <= 12) {
"✓"
} else if (value <= 16) {
"◆"
} else {
"▲"
},
"gunning" = if (value <= 12) {
"✓"
} else if (value <= 16) {
"◆"
} else {
"▲"
}
)
}
# Helper to get interpretation text
get_interp_text <- function(index_name, value) {
switch(
index_name,
"flesch_kincaid" = paste("Grade", round(value)),
"flesch_ease" = if (value >= 60) {
"Easy"
} else if (value >= 30) {
"Moderate"
} else {
"Difficult"
},
"ari" = paste("Grade", round(value)),
"gunning" = if (value <= 12) {
"Readable"
} else if (value <= 16) {
"Difficult"
} else {
"Very difficult"
}
)
}
text_content <- sprintf(
"READABILITY INDICES
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📚 Flesch-Kincaid: %5.1f %s %s
📖 Reading Ease: %5.1f %s %s
🤖 ARI Index: %5.1f %s %s
☁️ Gunning Fog: %5.1f %s %s
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Sentences: %s • Words: %s
Syllables: %s • Complex: %s (%.1f%%)
Avg sentence length: %.1f words",
indices$flesch_kincaid_grade,
get_symbol("flesch_kincaid", indices$flesch_kincaid_grade),
get_interp_text("flesch_kincaid", indices$flesch_kincaid_grade),
indices$flesch_reading_ease,
get_symbol("flesch_ease", indices$flesch_reading_ease),
get_interp_text("flesch_ease", indices$flesch_reading_ease),
indices$automated_readability_index,
get_symbol("ari", indices$automated_readability_index),
get_interp_text("ari", indices$automated_readability_index),
indices$gunning_fog_index,
get_symbol("gunning", indices$gunning_fog_index),
get_interp_text("gunning", indices$gunning_fog_index),
format(indices$n_sentences, big.mark = ","),
format(indices$n_words, big.mark = ","),
format(indices$n_syllables, big.mark = ","),
format(indices$n_complex_words, big.mark = ","),
indices$pct_complex_words,
indices$avg_sentence_length
)
tags$pre(
style = "background-color: #f8f9fa; padding: 15px; border-radius: 6px; border: 1px solid #dee2e6; font-family: 'Courier New', monospace; font-size: 13px; line-height: 1.5; color: #333; margin: 0; min-height: 280px;",
text_content
)
} else {
tags$pre(
style = "background-color: #f8f9fa; padding: 40px; border-radius: 6px; border: 1px solid #dee2e6; text-align: center; color: #999; margin: 0; min-height: 280px; display: flex; align-items: center; justify-content: center;",
"No readability data available"
)
}
})
# Citation types table
output$citation_types_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citation_metrics$type_distribution)
) {
values$analysis_results$citation_metrics$type_distribution
} else {
data.frame(
citation_type = character(0),
n = numeric(0),
percentage = numeric(0)
)
}
},
options = list(
pageLength = 10,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Citation sections table
output$citation_sections_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
!is.null(
values$analysis_results$citation_metrics$section_distribution
)
) {
values$analysis_results$citation_metrics$section_distribution %>%
filter(n > 0)
} else {
data.frame(
section = character(0),
n = numeric(0),
percentage = numeric(0)
)
}
},
options = list(
pageLength = 10,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Frequent words table
output$frequent_words_table <- DT::renderDataTable(
{
if (!is.null(values$analysis_results)) {
values$analysis_results$word_frequencies %>%
slice_head(n = 15) %>%
select(word, n)
} else {
data.frame(word = character(0), n = numeric(0))
}
},
options = list(
pageLength = 15,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Bigrams table
output$bigrams_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
"2gram" %in% names(values$analysis_results$ngrams)
) {
values$analysis_results$ngrams$`2gram` %>%
select(ngram, n) %>%
slice_head(n = 15)
} else {
data.frame(ngram = character(0), n = numeric(0))
}
},
options = list(
pageLength = 15,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Trigrams table
output$trigrams_table <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
"3gram" %in% names(values$analysis_results$ngrams)
) {
values$analysis_results$ngrams$`3gram` %>%
select(ngram, n) %>%
slice_head(n = 15)
} else {
data.frame(ngram = character(0), n = numeric(0))
}
},
options = list(
pageLength = 15,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# Text statistics
# output$text_stats <- renderText({
# if (!is.null(values$analysis_results)) {
# stats <- values$analysis_results$text_analytics$basic_stats
# paste(
# "Characters:", format(stats$total_characters, big.mark = ","), "\n",
# "Words:", format(stats$total_words, big.mark = ","), "\n",
# "Sentences:", format(stats$total_sentences, big.mark = ","), "\n",
# "Avg words/sentence:", round(stats$avg_words_per_sentence, 1), "\n",
# "Lexical diversity:", round(values$analysis_results$summary$lexical_diversity, 3)
# )
# } else {
# "No analysis data available"
# }
# })
# ===========================================
# TAB 2: IN-CONTEXT CITATION ANALYSIS
# ===========================================
# ===========================================
# UPDATE CITATION GROUPING (without re-running full analysis)
# ===========================================
observeEvent(input$update_citation_grouping, {
req(values$pdf_text)
req(values$analysis_results)
tryCatch(
{
# Determine citation segmentation
use_sections_cit <- switch(
input$citation_segmentation %||% "auto",
"auto" = "auto",
"sections" = TRUE,
"segments" = FALSE,
"auto"
)
n_segs_cit <- input$n_segments_citations %||% 10
# Get citations without section column
citations_for_mapping <- values$analysis_results$citations %>%
select(-section, -segment_type)
# Re-map citations to new segments/sections
citations_remapped <- map_citations_to_segments(
citations_df = citations_for_mapping,
text = values$pdf_sections,
use_sections = use_sections_cit,
n_segments = n_segs_cit
)
# Update the citations in analysis_results
values$analysis_results$citations <- citations_remapped %>%
rename(section = segment)
# Update citation contexts with new section info
if (!is.null(values$analysis_results$citation_contexts)) {
# Remove old section column if exists
if ("section" %in% names(values$analysis_results$citation_contexts)) {
values$analysis_results$citation_contexts <- values$analysis_results$citation_contexts %>%
select(-section)
}
# Add new section mapping
values$analysis_results$citation_contexts <- values$analysis_results$citation_contexts %>%
left_join(
values$analysis_results$citations %>%
select(citation_id, section),
by = "citation_id"
)
section_colors <- colorlist()[
1:length(unique(values$analysis_results$citation_contexts$section))
]
names(section_colors) <- unique(
values$analysis_results$citation_contexts$section
)
values$analysis_results$section_colors <- section_colors
}
# Update citation metrics - section distribution
if (!is.null(values$analysis_results$citation_metrics)) {
# Determine expected sections/segments
sections_to_use <- NULL
if (
use_sections_cit == TRUE ||
(use_sections_cit == "auto" &&
length(setdiff(
names(values$pdf_sections),
c("Full_text", "References")
)) >
0)
) {
# Using sections
sections_to_use <- setdiff(
names(values$pdf_sections),
c("Full_text", "References")
)
} else {
# Using segments
sections_to_use <- paste0("Segment ", 1:n_segs_cit)
}
# Recalculate section distribution
if (!is.null(sections_to_use)) {
values$analysis_results$citation_metrics$section_distribution <-
values$analysis_results$citations %>%
mutate(section = factor(section, levels = sections_to_use)) %>%
count(section, sort = FALSE, .drop = FALSE) %>%
mutate(percentage = round(n / sum(n) * 100, 2))
} else {
values$analysis_results$citation_metrics$section_distribution <-
values$analysis_results$citations %>%
count(section, sort = TRUE) %>%
mutate(percentage = round(n / sum(n) * 100, 2))
}
}
# Recreate network with new grouping
if (
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
values$network_plot <- create_citation_network(
values$analysis_results,
max_distance = input$max_distance,
show_labels = TRUE
)
}
showNotification(
"Citation grouping updated successfully!",
type = "message",
duration = 2
)
},
error = function(e) {
showNotification(
paste("Error updating citation grouping:", e$message),
type = "error",
duration = 5
)
# Print error to console for debugging
cat("Error in update_citation_grouping:\n")
print(e)
}
)
})
# Custom HTML output for citation contexts
output$citation_contexts_html <- renderUI({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citation_contexts) &&
nrow(values$analysis_results$citation_contexts) > 0
) {
contexts <- values$analysis_results$citation_contexts
# Apply filters (REMOVED context_type_filter)
if (!is.null(input$context_search) && nzchar(input$context_search)) {
contexts <- contexts %>%
filter(
str_detect(
citation_text,
regex(input$context_search, ignore_case = TRUE)
) |
str_detect(
full_context,
regex(input$context_search, ignore_case = TRUE)
)
)
}
# REMOVED: type filter code
if (!is.null(input$context_min_words)) {
contexts <- contexts %>%
filter(context_word_count >= input$context_min_words)
}
# Create HTML for each citation context
if (nrow(contexts) > 0) {
citation_boxes <- lapply(1:nrow(contexts), function(i) {
context <- contexts[i, ]
section_colors <- values$analysis_results$section_colors
section_name <- if (
!is.null(context$section) && !is.na(context$section)
) {
context$section
} else {
"Unknown"
}
box_color <- section_colors[section_name]
if (is.na(box_color)) {
box_color <- "#CCCCCC"
}
has_reference <- FALSE
tooltip_text <- "No reference matched for this citation"
if ("ref_full_text" %in% names(context)) {
ref_value <- context[["ref_full_text"]]
if (
length(ref_value) > 0 &&
!is.na(ref_value) &&
is.character(ref_value) &&
nzchar(trimws(ref_value))
) {
has_reference <- TRUE
ref_text <- trimws(ref_value)
ref_text <- gsub("&", "&", ref_text)
ref_text <- gsub("<", "<", ref_text)
ref_text <- gsub(">", ">", ref_text)
ref_text <- gsub('"', """, ref_text)
ref_text <- gsub("'", "'", ref_text)
ref_text <- gsub("\n", "<br/>", ref_text)
tooltip_text <- paste0(
"<strong style='color: #4ECDC4;'>Reference:</strong><br/>",
ref_text
)
}
}
citation_id <- paste0("citation_", i, "_", sample(1:10000, 1))
div(
style = paste0(
"margin-bottom: 20px; padding: 15px; border-radius: 8px; ",
"box-shadow: 0 2px 4px rgba(0,0,0,0.1); ",
"border-left: 4px solid ",
box_color,
"; ",
"background-color: #fafafa;"
),
div(
style = "margin-bottom: 10px; font-size: 12px; color: #666;",
span(paste("Citation", i, "•"), style = "font-weight: bold;"),
span(
section_name,
style = paste0(
"color: ",
box_color,
"; font-weight: bold; margin-left: 8px;"
)
),
span(paste("• Position:", context$citation_position_in_text))
),
div(
style = "display: flex; align-items: center; font-family: 'Courier New', monospace;",
div(
style = "flex: 1; text-align: right; padding-right: 15px; color: #555; font-size: 14px;",
if (nzchar(context$words_before)) {
paste("...", context$words_before)
} else {
"[start of text]"
}
),
div(
id = citation_id,
class = "citation-with-tooltip",
style = paste0(
"background-color: ",
box_color,
"; color: white; ",
"padding: 8px 12px; border-radius: 6px; font-weight: bold; ",
"font-size: 14px; white-space: nowrap; max-width: 300px; ",
"overflow: hidden; text-overflow: ellipsis; ",
"cursor: ",
if (has_reference) "pointer" else "default",
"; position: relative;"
),
`data-toggle` = if (has_reference) "tooltip" else NULL,
`data-placement` = "top",
`data-html` = "true",
`data-original-title` = tooltip_text,
context$citation_text
),
div(
style = "flex: 1; text-align: left; padding-left: 15px; color: #555; font-size: 14px;",
if (nzchar(context$words_after)) {
paste(context$words_after, "...")
} else {
"[end of text]"
}
)
),
div(
style = "margin-top: 10px; font-size: 11px; color: #888;",
paste("Context words:", context$context_word_count, "•"),
if (!is.na(context$is_narrative) && context$is_narrative) {
span(
"Narrative citation",
style = "color: #e67e22; font-weight: bold;"
)
} else {
span(
"Parenthetical citation",
style = "color: #3498db; font-weight: bold;"
)
},
if (has_reference) {
span(
" • ",
icon("book", style = "color: #27ae60;"),
" Reference matched (hover to view)",
style = "color: #27ae60; font-weight: bold;"
)
} else {
span(
" • ",
icon("times-circle", style = "color: #e74c3c;"),
" No reference matched",
style = "color: #999; font-style: italic;"
)
}
)
)
})
tagList(
tags$script(HTML(
"
$(document).ready(function(){
$('.citation-with-tooltip').tooltip('dispose');
$('.citation-with-tooltip[data-toggle=\"tooltip\"]').tooltip({
container: 'body',
trigger: 'hover focus',
delay: { show: 200, hide: 100 },
html: true,
boundary: 'window'
});
setTimeout(function() {
$('.citation-with-tooltip[data-toggle=\"tooltip\"]').tooltip('dispose');
$('.citation-with-tooltip[data-toggle=\"tooltip\"]').tooltip({
container: 'body',
trigger: 'hover focus',
delay: { show: 200, hide: 100 },
html: true,
boundary: 'window'
});
}, 800);
});
"
)),
tags$style(HTML(
"
.tooltip {
font-family: Arial, sans-serif;
pointer-events: none;
}
.tooltip-inner {
max-width: 500px;
min-width: 250px;
text-align: left;
background-color: rgba(0, 0, 0, 0.95);
padding: 12px 16px;
font-size: 13px;
line-height: 1.6;
border-radius: 6px;
box-shadow: 0 4px 12px rgba(0,0,0,0.4);
}
.tooltip-inner strong {
color: #4ECDC4;
display: block;
margin-bottom: 6px;
}
.tooltip.show {
opacity: 1;
}
.tooltip.top .tooltip-arrow {
border-top-color: rgba(0, 0, 0, 0.95);
}
.citation-with-tooltip[data-toggle]:hover {
opacity: 0.9;
transform: scale(1.02);
transition: all 0.2s ease;
}
"
)),
citation_boxes
)
} else {
div(
style = "text-align: center; padding: 40px; color: #999;",
icon("search", style = "font-size: 24px; margin-bottom: 10px;"),
h4("No citations match your current filters"),
p("Try adjusting your search terms or filters.")
)
}
} else {
div(
style = "text-align: center; padding: 40px; color: #999;",
icon("quote-right", style = "font-size: 24px; margin-bottom: 10px;"),
h4("No citation contexts available"),
p("Run the analysis first to see in-context citations.")
)
}
})
# ===========================================
# TAB 3: NETWORK ANALYSIS
# ===========================================
# Network visualization
output$citation_network <- renderVisNetwork({
if (!is.null(values$network_plot)) {
values$network_plot
} else {
visNetwork(
nodes = data.frame(
id = 1,
label = "No citations found",
size = 20,
color = "#CCCCCC"
),
edges = data.frame(from = numeric(0), to = numeric(0))
) %>%
visOptions(highlightNearest = FALSE) %>%
visInteraction(dragNodes = FALSE, dragView = FALSE, zoomView = FALSE)
}
})
# Network information
output$network_info <- renderText({
if (!is.null(values$network_plot)) {
stats <- attr(values$network_plot, "stats")
if (!is.null(stats)) {
section_info <- ""
if (!is.null(stats$section_distribution)) {
section_info <- paste0(
"\n\nSection Distribution:\n",
paste(
stats$section_distribution$section,
": ",
stats$section_distribution$n,
collapse = "\n"
)
)
}
paste(
"Nodes:",
stats$n_nodes,
"\n",
"Edges:",
stats$n_edges,
"\n",
"Avg. Distance:",
stats$avg_distance,
"chars\n",
"Max Distance Filter:",
stats$max_distance,
"chars\n\n",
"Network Density:",
round(stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2), 3),
section_info
)
} else {
"Network statistics not available"
}
} else {
"No network data available"
}
})
# Strongest connections table
output$strongest_connections <- DT::renderDataTable(
{
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
values$analysis_results$network_data %>%
arrange(abs(distance)) %>%
slice_head(n = 8) %>%
select(citation1, citation2, distance) %>%
mutate(
citation1 = str_trunc(citation1, 20),
citation2 = str_trunc(citation2, 20),
distance = paste(abs(distance), "chars")
)
} else {
data.frame(
citation1 = character(0),
citation2 = character(0),
distance = character(0)
)
}
},
options = list(
pageLength = 8,
dom = 't',
ordering = FALSE,
searching = FALSE
)
)
# ===========================================
# DOWNLOAD HANDLERS
# ===========================================
output$download_network <- downloadHandler(
filename = function() {
paste0("citation_network_", Sys.Date(), ".csv")
},
content = function(file) {
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$network_data)
) {
write.csv(values$analysis_results$network_data, file, row.names = FALSE)
} else {
write.csv(
data.frame(message = "No network data available"),
file,
row.names = FALSE
)
}
}
)
output$download_contexts <- downloadHandler(
filename = function() {
paste0("citation_contexts_", Sys.Date(), ".csv")
},
content = function(file) {
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citation_contexts)
) {
write.csv(
values$analysis_results$citation_contexts,
file,
row.names = FALSE
)
} else {
write.csv(
data.frame(message = "No citation contexts available"),
file,
row.names = FALSE
)
}
}
)
# ===========================================
# TAB 4: WORD TRENDS ANALYSIS
# ===========================================
# Populate word choices when analysis is complete
observe({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$word_frequencies) &&
nrow(values$analysis_results$word_frequencies) > 0
) {
# Get top 30 most frequent words
top_words <- values$analysis_results$word_frequencies %>%
slice_head(n = 30) %>%
pull(word)
# Get top bigrams if available
top_bigrams <- character(0)
if (
!is.null(values$analysis_results$ngrams) &&
"2gram" %in% names(values$analysis_results$ngrams)
) {
top_bigrams <- values$analysis_results$ngrams$`2gram` %>%
slice_head(n = 15) %>%
pull(ngram)
}
# Get top trigrams if available
top_trigrams <- character(0)
if (
!is.null(values$analysis_results$ngrams) &&
"3gram" %in% names(values$analysis_results$ngrams)
) {
top_trigrams <- values$analysis_results$ngrams$`3gram` %>%
slice_head(n = 10) %>%
pull(ngram)
}
# Combine all choices with labels
all_choices <- c(top_words, top_bigrams, top_trigrams)
# Create named list for optgroups
choices_list <- list(
"Top Words (1-gram)" = top_words
)
if (length(top_bigrams) > 0) {
choices_list[["Top Bigrams (2-gram)"]] <- top_bigrams
}
if (length(top_trigrams) > 0) {
choices_list[["Top Trigrams (3-gram)"]] <- top_trigrams
}
updateSelectizeInput(
session,
"trend_words",
choices = choices_list,
selected = NULL,
server = TRUE
)
}
})
# Check if sections are available
output$sections_available <- reactive({
if (!is.null(values$pdf_sections) && length(values$pdf_sections) > 0) {
section_names <- setdiff(
names(values$pdf_sections),
c("Full_text", "References")
)
return(length(section_names) > 0)
}
return(FALSE)
})
outputOptions(output, "sections_available", suspendWhenHidden = FALSE)
# Check if trends are available
output$trends_available <- reactive({
return(
!is.null(values$word_trends_data) && nrow(values$word_trends_data) > 0
)
})
outputOptions(output, "trends_available", suspendWhenHidden = FALSE)
# Update word trends when button is clicked
observeEvent(input$update_trends, {
# Validate inputs
if (is.null(values$pdf_text) || nchar(values$pdf_text) == 0) {
showNotification(
"Please extract text from PDF first.",
type = "warning",
duration = 3
)
return()
}
if (is.null(input$trend_words) || length(input$trend_words) == 0) {
showNotification(
"Please select at least one word to track.",
type = "warning",
duration = 3
)
return()
}
if (length(input$trend_words) > 10) {
showNotification(
"Maximum 10 words allowed. Please deselect some words.",
type = "warning",
duration = 3
)
return()
}
tryCatch(
{
# Determine use_sections parameter
use_sections_param <- switch(
input$segmentation_type,
"auto" = "auto",
"sections" = TRUE,
"segments" = FALSE
)
# Get number of segments
n_segs <- if (!is.null(input$n_segments)) input$n_segments else 10
# Determine which text to use
# If user explicitly chooses segments, use Full_text only
# Otherwise, use the sections structure
text_input <- if (input$segmentation_type == "segments") {
values$pdf_text # Just the full text as string
} else {
values$pdf_sections # List with sections
}
# Calculate word distribution
values$word_trends_data <- calculate_word_distribution(
text = text_input,
selected_words = input$trend_words,
use_sections = use_sections_param,
n_segments = n_segs,
remove_stopwords = FALSE # Don't remove stopwords for selected words
)
# Check if sections were requested but not available
sections_used <- attr(values$word_trends_data, "use_sections")
if (input$segmentation_type == "sections" && !sections_used) {
showNotification(
"Sections not available. Using equal-length segments instead.",
type = "warning",
duration = 4
)
}
showNotification(
"Word distribution calculated successfully!",
type = "message",
duration = 2
)
},
error = function(e) {
showNotification(
paste("Error calculating word distribution:", e$message),
type = "error",
duration = 5
)
}
)
})
# Render word trends plot
output$word_trends_plot <- renderPlotly({
req(values$word_trends_data)
tryCatch(
{
plot_word_distribution(
word_distribution_data = values$word_trends_data,
plot_type = input$trend_plot_type,
smooth = TRUE,
show_points = input$trend_show_points,
colors = NULL # Use automatic colors
)
},
error = function(e) {
showNotification(
paste("Error creating plot:", e$message),
type = "error",
duration = 5
)
return(NULL)
}
)
})
# Render word trends statistics table
output$word_trends_table <- DT::renderDataTable({
req(values$word_trends_data)
# Create summary statistics
stats_table <- values$word_trends_data %>%
group_by(word) %>%
summarise(
`Total Occurrences` = sum(count),
`Avg Frequency` = paste0(round(mean(relative_frequency) * 100, 3), "%"),
`Min Frequency` = paste0(round(min(relative_frequency) * 100, 3), "%"),
`Max Frequency` = paste0(round(max(relative_frequency) * 100, 3), "%"),
`Std Dev` = round(sd(relative_frequency) * 100, 3),
`Peak Segment` = segment_name[which.max(relative_frequency)],
.groups = "drop"
) %>%
arrange(desc(`Total Occurrences`))
DT::datatable(
stats_table,
options = list(
pageLength = 10,
dom = 'ftip',
ordering = TRUE,
searching = FALSE,
scrollX = TRUE
),
rownames = FALSE,
class = 'cell-border stripe'
) %>%
DT::formatStyle(
'word',
fontWeight = 'bold',
color = '#2E86AB'
)
})
# Download handler for word trends
output$download_word_trends <- downloadHandler(
filename = function() {
paste0("word_trends_", Sys.Date(), ".csv")
},
content = function(file) {
if (!is.null(values$word_trends_data)) {
# Prepare data for export
export_data <- values$word_trends_data %>%
select(
word,
segment_id,
segment_name,
count,
relative_frequency,
percentage,
total_words
) %>%
arrange(word, segment_id)
write.csv(export_data, file, row.names = FALSE)
} else {
write.csv(
data.frame(message = "No word trends data available"),
file,
row.names = FALSE
)
}
}
)
# ===========================================
# TAB 5: REFERENCES
# ===========================================
# Check if references are available
output$references_available <- reactive({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
return(nrow(values$analysis_results$parsed_references) > 0)
}
return(FALSE)
})
outputOptions(output, "references_available", suspendWhenHidden = FALSE)
# # Total references count
output$total_refs <- renderText({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
format(nrow(values$analysis_results$parsed_references), big.mark = ",")
} else {
"0"
}
})
# # PDF references count
output$pdf_refs <- renderText({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
refs <- values$analysis_results$parsed_references
pdf_refs <- sum(tolower(refs$ref_source) == "pdf", na.rm = TRUE)
format(pdf_refs, big.mark = ",")
} else {
"0"
}
})
# # Crossref references count
output$crossref_refs <- renderText({
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
refs <- values$analysis_results$parsed_references
crossref_refs <- sum(tolower(refs$ref_source) == "crossref", na.rm = TRUE)
format(crossref_refs, big.mark = ",")
} else {
"0"
}
})
# Openalex references count
output$openalex_refs <- renderText({
req(values$references_oa)
if (is.null(values$references_oa) || nrow(values$references_oa) == 0) {
return("0")
}
nrow(values$references_oa)
})
# References HTML display
output$references_html <- renderUI({
req(values$analysis_results)
req(values$analysis_results$parsed_references)
refs <- values$analysis_results$parsed_references
# Apply search filter
if (!is.null(input$reference_search) && nzchar(input$reference_search)) {
search_term <- tolower(input$reference_search)
refs <- refs %>%
filter(
if_else(
!is.na(ref_full_text),
str_detect(tolower(ref_full_text), fixed(search_term)),
FALSE
) |
if_else(
!is.na(ref_authors),
str_detect(tolower(ref_authors), fixed(search_term)),
FALSE
) |
if_else(
!is.na(ref_year),
str_detect(tolower(ref_year), fixed(search_term)),
FALSE
) |
if_else(
!is.na(doi),
str_detect(tolower(doi), fixed(search_term)),
FALSE
)
)
}
if (nrow(refs) == 0) {
return(
div(
style = "text-align: center; padding: 40px; color: #999;",
icon("search", style = "font-size: 24px; margin-bottom: 10px;"),
h4("No references match your search"),
p("Try different search terms.")
)
)
}
# Create HTML for each reference
reference_items <- lapply(1:nrow(refs), function(i) {
ref <- refs[i, ]
# Get reference text
ref_text <- if (!is.na(ref$ref_full_text) && nzchar(ref$ref_full_text)) {
ref$ref_full_text
} else {
"[Reference text not available]"
}
# Check if OpenAlex data is available for this reference
has_oa_data <- FALSE
oa_button <- NULL
if (
!is.null(values$references_oa) && !is.na(ref$doi) && nzchar(ref$doi)
) {
# Cerca corrispondenza nel dataset OpenAlex
ref_doi_clean <- tolower(trimws(ref$doi))
# Prova diverse varianti del DOI
oa_match_idx <- which(
tolower(gsub("https://doi.org/", "", values$references_oa$doi)) ==
ref_doi_clean |
tolower(values$references_oa$doi) ==
paste0("https://doi.org/", ref_doi_clean)
)
if (length(oa_match_idx) > 0) {
has_oa_data <- TRUE
oa_button <- tags$button(
class = "btn btn-info btn-xs",
style = "float: right; margin-left: 10px;",
onclick = sprintf("showOADetails(%d)", i),
icon("info-circle"),
" View Details"
)
}
}
# Determine source badge
source_badge <- if (
!is.na(ref$ref_source) && tolower(ref$ref_source) == "crossref"
) {
tags$span(
class = "label label-success",
style = "font-size: 11px; margin-left: 10px;",
icon("cloud-download-alt"),
" Crossref"
)
} else {
tags$span(
class = "label label-info",
style = "font-size: 11px; margin-left: 10px;",
icon("file-pdf"),
" PDF"
)
}
# OpenAlex badge if available
oa_badge <- if (has_oa_data) {
tags$span(
class = "label label-primary",
style = "font-size: 11px; margin-left: 5px;",
icon("database"),
" OpenAlex"
)
}
# DOI link if available
doi_link <- NULL
if ("doi" %in% names(ref) && !is.na(ref$doi) && nzchar(ref$doi)) {
doi_link <- tags$div(
style = "margin-top: 8px; font-size: 12px;",
icon("link", style = "color: #3498db;"),
tags$a(
href = paste0("https://doi.org/", ref$doi),
target = "_blank",
ref$doi,
style = "color: #3498db; margin-left: 5px;"
)
)
}
# Author and year display
author_year_info <- NULL
if (!is.na(ref$ref_first_author) && !is.na(ref$ref_year)) {
n_auth <- if (!is.na(ref$n_authors) && ref$n_authors > 1) {
paste0(" et al. (", ref$n_authors, " authors)")
} else {
""
}
author_year_info <- tags$div(
style = "margin-top: 8px; font-size: 12px; color: #666;",
icon("user", style = "color: #3498db;"),
tags$span(
style = "margin-left: 5px;",
paste0(ref$ref_first_author, n_auth, " • ", ref$ref_year)
)
)
}
# Citation count if available
citation_info <- NULL
if (
"citation_count" %in%
names(ref) &&
!is.na(ref$citation_count) &&
ref$citation_count > 0
) {
citation_info <- tags$span(
style = "margin-left: 15px; color: #e67e22; font-size: 12px;",
icon("quote-right"),
sprintf(
" Cited %d time%s in text",
ref$citation_count,
ifelse(ref$citation_count > 1, "s", "")
)
)
}
div(
class = "reference-item",
`data-ref-index` = i,
style = "padding: 15px; margin-bottom: 15px; border-left: 4px solid #2E86AB; background-color: #fafafa; border-radius: 5px; box-shadow: 0 1px 3px rgba(0,0,0,0.1); position: relative;",
# Reference number, badges and OA button
div(
style = "margin-bottom: 10px;",
tags$span(
style = "font-weight: bold; color: #2E86AB; font-size: 16px;",
paste0("[", i, "]")
),
source_badge,
oa_badge,
citation_info,
oa_button
),
# Full reference text
div(
style = "font-family: 'Georgia', serif; font-size: 14px; line-height: 1.6; color: #333; clear: both;",
ref_text
),
# Author and year info
author_year_info,
# DOI link
doi_link
)
})
tagList(
tags$style(HTML(
"
.reference-item:hover {
background-color: #f0f8ff !important;
transition: background-color 0.2s ease;
}
.oa-document-summary .info-box {
display: flex;
align-items: center;
border-radius: 4px;
padding: 10px;
color: white;
}
.oa-document-summary .info-box-icon {
font-size: 32px;
flex: 0 0 70px;
text-align: center;
}
.oa-document-summary .info-box-content {
flex: 1;
}
.oa-document-summary .info-box-text {
display: block;
font-size: 12px;
text-transform: uppercase;
opacity: 0.9;
}
.oa-document-summary .info-box-number {
display: block;
font-size: 20px;
font-weight: bold;
margin-top: 5px;
}
"
)),
reference_items
)
})
# Download handler for references
output$download_references <- downloadHandler(
filename = function() {
paste0("references_", Sys.Date(), ".txt")
},
content = function(file) {
if (
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$parsed_references)
) {
refs <- values$analysis_results$parsed_references
# Create formatted text
ref_text <- paste0(
"BIBLIOGRAPHY\n",
"============\n",
"Generated: ",
Sys.time(),
"\n",
"Total references: ",
nrow(refs),
"\n",
"From PDF: ",
sum(tolower(refs$ref_source) == "pdf", na.rm = TRUE),
"\n",
"From Crossref: ",
sum(tolower(refs$ref_source) == "crossref", na.rm = TRUE),
"\n\n"
)
for (i in 1:nrow(refs)) {
ref <- refs[i, ]
source_label <- if (
!is.na(ref$ref_source) && tolower(ref$ref_source) == "crossref"
) {
"[Crossref]"
} else {
"[PDF]"
}
ref_text <- paste0(
ref_text,
"[",
i,
"] ",
source_label,
"\n",
ref$ref_full_text,
"\n"
)
if (!is.na(ref$ref_first_author) && !is.na(ref$ref_year)) {
n_auth <- if (!is.na(ref$n_authors) && ref$n_authors > 1) {
paste0(" (", ref$n_authors, " authors)")
} else {
""
}
ref_text <- paste0(
ref_text,
"Author: ",
ref$ref_first_author,
n_auth,
" • Year: ",
ref$ref_year,
"\n"
)
}
if ("doi" %in% names(ref) && !is.na(ref$doi) && nzchar(ref$doi)) {
ref_text <- paste0(ref_text, "DOI: ", ref$doi, "\n")
}
if ("citation_count" %in% names(ref) && !is.na(ref$citation_count)) {
ref_text <- paste0(
ref_text,
"Cited ",
ref$citation_count,
" times in text\n"
)
}
ref_text <- paste0(ref_text, "\n")
}
writeLines(ref_text, file)
} else {
writeLines("No references available", file)
}
}
)
# ===========================================
# TAB: BIBLIOAI SUMMARY - SERVER LOGIC
# ===========================================
# ===========================================
# BIBLIOAI SUMMARY INITIALIZATION
# ===========================================
# Initialize API key from environment on startup
observe(
{
api_key_env <- Sys.getenv("GEMINI_API_KEY")
if (nzchar(api_key_env) && is.null(values$gemini_api_key)) {
values$gemini_api_key <- api_key_env
}
},
priority = 100
)
# ===========================================
# API STATUS DISPLAY
# ===========================================
output$gemini_api_status <- renderUI({
api_key_env <- Sys.getenv("GEMINI_API_KEY")
if (!nzchar(api_key_env)) {
tags$span(
class = "label label-danger",
style = "font-size: 12px; padding: 6px 12px;",
icon("exclamation-circle"),
" API Key Not Configured"
)
} else {
tags$span(
class = "label label-success",
style = "font-size: 12px; padding: 6px 12px;",
icon("check-circle"),
" API Key Configured"
)
}
})
# ===========================================
# SUMMARY TYPE DESCRIPTION
# ===========================================
output$summary_type_description <- renderUI({
req(input$summary_type)
descriptions <- list(
short_abstract = "A concise, one-paragraph abstract (around 250 words) covering objectives, methods, findings, and conclusions.",
narrative_abstract = "An engaging narrative (500-600 words) that tells the story of the research from problem to implications.",
imrad_summary = "A structured summary following the IMRaD format: Introduction, Methods, Results, and Discussion.",
thematic_bibliography = "Groups the references by research themes and provides descriptions of each thematic area.",
research_questions = "Extracts research questions, context, and motivation behind the study.",
background_literature = "Summarizes the theoretical framework and previous research that grounds this study.",
methods_summary = "Detailed breakdown of research design, data collection, analysis techniques, and tools used.",
implications = "Focuses on theoretical contributions and practical applications derived from the findings.",
list_tables_figures = "Complete list of all tables and figures with their captions as they appear in the document."
)
description <- descriptions[[input$summary_type]]
div(
style = "background-color: #f0f8ff; padding: 12px; border-radius: 4px; border-left: 3px solid #3498db;",
tags$small(
icon("info-circle", style = "color: #3498db; margin-right: 5px;"),
description,
style = "color: #555; line-height: 1.6;"
)
)
})
# ===========================================
# GENERATE SUMMARY
# ===========================================
observeEvent(input$generate_summary, {
# Validate PDF file
if (is.null(input$pdf_file)) {
showNotification(
"Please upload a PDF file first.",
type = "warning",
duration = 4
)
return()
}
# Validate API key
api_key <- Sys.getenv("GEMINI_API_KEY")
if (!nzchar(api_key)) {
showNotification(
HTML(
"<strong>API Key Required</strong><br/>
Please configure your Gemini API key in Biblioshiny settings.<br/>
You can get a free API key at: <a href='https://aistudio.google.com/apikey' target='_blank'>Google AI Studio</a>"
),
type = "error",
duration = 8
)
return()
}
# Get model from Biblioshiny settings or use default
model <- values$gemini_api_model
if (is.null(model) || !nzchar(model)) {
model <- "2.0-flash-lite"
}
# Use "huge" as default output size
output_size <- "huge"
tryCatch(
{
# Generate prompt
prompt <- get_article_summary_prompt(input$summary_type)
# Show progress notification
progress_id <- showNotification(
"Sending request to Gemini AI...",
type = "message",
duration = NULL,
closeButton = FALSE
)
# Call Gemini AI
result <- gemini_ai(
docs = input$pdf_file$datapath,
prompt = prompt,
model = model,
api_key = api_key,
outputSize = output_size,
retry_503 = 5
)
# Remove progress notification
removeNotification(progress_id)
# Check if result is an error message
if (is.character(result) && length(result) == 1) {
if (
grepl("^❌", result) || grepl("Error", result, ignore.case = TRUE)
) {
showNotification(
HTML(paste0(
"<strong>AI Request Failed</strong><br/>",
gsub("❌ ", "", result)
)),
type = "error",
duration = 10
)
return()
}
}
# Store results
values$biblioai_summary <- result
values$biblioai_summary_type <- input$summary_type
values$biblioai_summary_timestamp <- Sys.time()
# Get summary type label
summary_labels <- list(
short_abstract = "Short Abstract",
narrative_abstract = "Narrative Abstract",
imrad_summary = "IMRaD Summary",
thematic_bibliography = "Thematic Bibliography",
research_questions = "Research Questions & Context",
background_literature = "Background & Literature",
methods_summary = "Methods Summary",
implications = "Implications & Conclusions",
list_tables_figures = "Tables & Figures List"
)
summary_label <- summary_labels[[input$summary_type]]
showNotification(
HTML(paste0(
"<strong>Summary Generated!</strong><br/>",
"Type: ",
summary_label
)),
type = "message",
duration = 5
)
},
error = function(e) {
removeNotification(progress_id)
showNotification(
HTML(paste0(
"<strong>Error Generating Summary</strong><br/>",
e$message
)),
type = "error",
duration = 8
)
cat("BiblioAI Summary Error:\n")
print(e)
}
)
})
# ===========================================
# SUMMARY AVAILABILITY
# ===========================================
output$summary_available <- reactive({
return(!is.null(values$biblioai_summary))
})
outputOptions(output, "summary_available", suspendWhenHidden = FALSE)
# ===========================================
# SUMMARY DISPLAY
# ===========================================
output$summary_type_display <- renderText({
req(values$biblioai_summary_type)
summary_labels <- list(
short_abstract = "Short Abstract (250 words)",
narrative_abstract = "Narrative Abstract (500-600 words)",
imrad_summary = "IMRaD Structure Summary",
thematic_bibliography = "Thematic Bibliography",
research_questions = "Research Questions & Context",
background_literature = "Background & Literature Review",
methods_summary = "Methodology Summary",
implications = "Implications & Conclusions",
list_tables_figures = "Tables & Figures List"
)
summary_labels[[values$biblioai_summary_type]]
})
output$summary_timestamp <- renderText({
req(values$biblioai_summary_timestamp)
format(values$biblioai_summary_timestamp, "%Y-%m-%d %H:%M:%S")
})
output$summary_content_display <- renderUI({
req(values$biblioai_summary)
# Combine multiple text elements if result is a vector
summary_text <- if (
is.character(values$biblioai_summary) &&
length(values$biblioai_summary) > 1
) {
paste(values$biblioai_summary, collapse = "\n\n")
} else {
values$biblioai_summary
}
# Convert markdown-style headers to HTML for better display
summary_html <- summary_text
# Convert **text** to bold
summary_html <- gsub(
"\\*\\*([^*]+)\\*\\*",
"<strong>\\1</strong>",
summary_html
)
# Convert *text* to italic
summary_html <- gsub("\\*([^*]+)\\*", "<em>\\1</em>", summary_html)
# Convert ### headers to h4
summary_html <- gsub(
"^### (.+)$",
"<h4 style='color: #2E86AB; margin-top: 20px; margin-bottom: 10px;'>\\1</h4>",
summary_html,
perl = TRUE
)
# Convert ## headers to h3
summary_html <- gsub(
"^## (.+)$",
"<h3 style='color: #2E86AB; margin-top: 25px; margin-bottom: 10px;'>\\1</h3>",
summary_html,
perl = TRUE
)
# Convert # headers to h2
summary_html <- gsub(
"^# (.+)$",
"<h2 style='color: #2E86AB; margin-top: 30px; margin-bottom: 15px;'>\\1</h2>",
summary_html,
perl = TRUE
)
# Convert bullet points
summary_html <- gsub("^- (.+)$", "<li>\\1</li>", summary_html, perl = TRUE)
summary_html <- gsub(
"^\\* (.+)$",
"<li>\\1</li>",
summary_html,
perl = TRUE
)
# Wrap consecutive <li> in <ul>
summary_html <- gsub(
"(<li>.*?</li>)\n(<li>)",
"\\1\\2",
summary_html,
perl = TRUE
)
summary_html <- gsub(
"(<li>.*?</li>)",
"<ul style='margin-left: 20px;'>\\1</ul>",
summary_html,
perl = TRUE
)
# Convert line breaks to <br> but not inside lists
summary_html <- gsub("\n(?!<)", "<br/>", summary_html, perl = TRUE)
HTML(summary_html)
})
# ===========================================
# DOWNLOAD SUMMARY
# ===========================================
output$download_summary <- downloadHandler(
filename = function() {
summary_type <- values$biblioai_summary_type
if (is.null(summary_type)) {
summary_type <- "summary"
}
paste0("biblioai_", summary_type, "_", Sys.Date(), ".txt")
},
content = function(file) {
if (!is.null(values$biblioai_summary)) {
# Create header
header <- paste0(
"====================================\n",
"BiblioAI Summary\n",
"====================================\n\n",
"Summary Type: ",
values$biblioai_summary_type,
"\n",
"Generated: ",
format(values$biblioai_summary_timestamp, "%Y-%m-%d %H:%M:%S"),
"\n",
"Document: ",
input$pdf_file$name,
"\n\n",
"====================================\n\n"
)
# Combine multiple text elements if result is a vector
summary_text <- if (
is.character(values$biblioai_summary) &&
length(values$biblioai_summary) > 1
) {
paste(values$biblioai_summary, collapse = "\n\n")
} else {
values$biblioai_summary
}
# Write to file
writeLines(paste0(header, summary_text), file, useBytes = TRUE)
} else {
writeLines("No summary available", file)
}
}
)
# ===========================================
# RESET SUMMARY ON NEW PDF OR RESET
# ===========================================
# Clear summary when new PDF is uploaded
observeEvent(input$pdf_file, {
values$biblioai_summary <- NULL
values$biblioai_summary_type <- NULL
values$biblioai_summary_timestamp <- NULL
})
# Clear summary on reset
observeEvent(input$reset_analysis, {
values$pdf_metadata <- NULL
values$biblioai_summary <- NULL
values$biblioai_summary_type <- NULL
values$biblioai_summary_timestamp <- NULL
# Reset UI inputs
updateSelectInput(session, "summary_type", selected = "short_abstract")
})
# Clear metadata when new PDF is uploaded
observeEvent(input$pdf_file, {
values$pdf_metadata <- NULL
})
# ===========================================
# OPENALEX MODAL HANDLER
# ===========================================
observeEvent(input$selected_oa_ref_index, {
req(values$analysis_results$parsed_references)
req(values$references_oa)
req(input$selected_oa_ref_index)
tryCatch(
{
ref_index <- input$selected_oa_ref_index
ref <- values$analysis_results$parsed_references[ref_index, ]
if (is.na(ref$doi) || !nzchar(ref$doi)) {
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
"Reference Details",
style = "font-weight: 600; font-size: 20px;"
)),
content = as.character(div(
style = "text-align: center; padding: 40px; color: #999;",
icon(
"exclamation-triangle",
style = "font-size: 32px; margin-bottom: 15px; color: #f39c12;"
),
h4("No DOI available"),
p("Cannot fetch OpenAlex data without a DOI.")
))
)
)
return()
}
# Trova i dati OpenAlex corrispondenti
ref_doi_clean <- tolower(trimws(ref$doi))
oa_match_idx <- which(
tolower(gsub("https://doi.org/", "", values$references_oa$doi)) ==
ref_doi_clean |
tolower(values$references_oa$doi) ==
paste0("https://doi.org/", ref_doi_clean)
)
if (length(oa_match_idx) == 0) {
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
"Reference Details",
style = "font-weight: 600; font-size: 20px;"
)),
content = as.character(div(
style = "text-align: center; padding: 40px; color: #999;",
icon(
"database",
style = "font-size: 32px; margin-bottom: 15px; color: #e74c3c;"
),
h4("OpenAlex data not found"),
p(paste("Could not find OpenAlex data for DOI:", ref$doi))
))
)
)
return()
}
# Prendi il primo match
oa_data <- values$references_oa[oa_match_idx[1], ]
# Estrai il titolo per l'header
doc_title <- if (!is.na(oa_data$title) && nzchar(oa_data$title)) {
oa_data$title
} else if (
!is.na(oa_data$display_name) && nzchar(oa_data$display_name)
) {
oa_data$display_name
} else {
"Reference Details"
}
# Tronca il titolo se troppo lungo
if (nchar(doc_title) > 120) {
doc_title <- paste0(substr(doc_title, 1, 117), "...")
}
# Crea l'HTML per il modal
modal_content <- create_oa_details_html(oa_data)
# Aggiorna il modal con il titolo del documento
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
doc_title,
style = "font-weight: 600; font-size: 20px; line-height: 1.4;"
)),
content = as.character(modal_content)
)
)
},
error = function(e) {
session$sendCustomMessage(
"updateOAModal",
list(
title = as.character(span(
"Error",
style = "font-weight: 600; font-size: 20px;"
)),
content = as.character(div(
style = "text-align: center; padding: 40px; color: #e74c3c;",
icon(
"exclamation-circle",
style = "font-size: 32px; margin-bottom: 15px;"
),
h4("Error loading details"),
p(paste("Error:", e$message))
))
)
)
cat("Error in OA modal handler:", e$message, "\n")
print(e)
}
)
})
# ===========================================
# RESET FUNCTIONALITY
# ===========================================
observeEvent(input$reset_analysis, {
values$pdf_text <- NULL
values$analysis_results <- NULL
values$network_plot <- NULL
values$analysis_running <- FALSE
values$readability_indices <- NULL
values$word_trends_data <- NULL
values$pdf_doi <- NULL
values$citation_type_used <- NULL
updateSelectizeInput(
session,
"trend_words",
choices = NULL,
selected = NULL
)
preview_visible(TRUE)
tryCatch(
{
shinyjs::reset("pdf_file")
updateTextInput(session, "pdf_doi_input", value = "")
# MODIFIED: Reset to no selection
updateRadioButtons(
session,
"citation_type_import",
selected = character(0)
)
updateActionButton(
session,
"run_analysis",
label = "Start",
icon = icon("play")
)
updateActionButton(session, "toggle_preview", "Hide Preview")
updateTextInput(session, "context_search", value = "")
updateSelectInput(session, "context_type_filter", selected = "")
},
error = function(e) {
cat("Reset UI elements failed:", e$message, "\n")
}
)
showNotification(
"Analysis reset successfully!",
type = "message",
duration = 3
)
# Re-open the PDF import box
session$sendCustomMessage('togglePdfImport', list(action = 'open'))
})
# ===========================================
# EXCEL RESULTS DOWNLOAD
# ===========================================
# Download all results as Excel file
output$download_all_results <- downloadHandler(
filename = function() {
paste0("content_analysis_results_", Sys.Date(), ".xlsx")
},
content = function(file) {
# Verifica che ci siano risultati disponibili
if (is.null(values$analysis_results)) {
showNotification(
"No analysis results available. Please run the analysis first.",
type = "error",
duration = 5
)
return()
}
tryCatch(
{
# Create workbook
wb_content <- openxlsx::createWorkbook()
# === SHEET 1: SUMMARY STATISTICS ===
openxlsx::addWorksheet(wb_content, "Summary_Statistics")
summary_data <- data.frame(
Metric = c(
"Total Words",
"Total Citations",
"Narrative Citations",
"Parenthetical Citations",
"Citation Density (per 1000 words)",
"Lexical Diversity",
"Total Characters",
"Total Sentences",
"Avg Words per Sentence"
),
Value = c(
values$analysis_results$summary$total_words_analyzed,
values$analysis_results$summary$citations_extracted,
values$analysis_results$summary$narrative_citations,
values$analysis_results$summary$citations_extracted -
values$analysis_results$summary$narrative_citations,
round(
values$analysis_results$summary$citation_density_per_1000_words,
2
),
round(values$analysis_results$summary$lexical_diversity, 4),
values$analysis_results$text_analytics$basic_stats$total_characters,
values$analysis_results$text_analytics$basic_stats$total_sentences,
round(
values$analysis_results$text_analytics$basic_stats$avg_words_per_sentence,
2
)
),
stringsAsFactors = FALSE
)
openxlsx::writeData(
wb_content,
"Summary_Statistics",
summary_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Summary_Statistics",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#2E86AB",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:2,
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Summary_Statistics",
cols = 1:2,
widths = c(35, 20)
)
# === SHEET 2: READABILITY INDICES ===
if (!is.null(values$readability_indices)) {
openxlsx::addWorksheet(wb_content, "Readability_Indices")
readability_data <- data.frame(
Index = c(
"Flesch-Kincaid Grade Level",
"Flesch Reading Ease",
"Automated Readability Index (ARI)",
"Gunning Fog Index",
"Average Sentence Length (words)",
"Number of Sentences",
"Number of Words",
"Number of Syllables",
"Number of Complex Words",
"Percentage Complex Words"
),
Value = c(
round(values$readability_indices$flesch_kincaid_grade, 2),
round(values$readability_indices$flesch_reading_ease, 2),
round(
values$readability_indices$automated_readability_index,
2
),
round(values$readability_indices$gunning_fog_index, 2),
round(values$readability_indices$avg_sentence_length, 2),
values$readability_indices$n_sentences,
values$readability_indices$n_words,
values$readability_indices$n_syllables,
values$readability_indices$n_complex_words,
paste0(
round(values$readability_indices$pct_complex_words, 2),
"%"
)
),
stringsAsFactors = FALSE
)
openxlsx::writeData(
wb_content,
"Readability_Indices",
readability_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Readability_Indices",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#27ae60",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:2,
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Readability_Indices",
cols = 1:2,
widths = c(40, 20)
)
}
# === SHEET 3: WORD FREQUENCIES ===
if (
!is.null(values$analysis_results$word_frequencies) &&
nrow(values$analysis_results$word_frequencies) > 0
) {
openxlsx::addWorksheet(wb_content, "Word_Frequencies")
top_words <- values$analysis_results$word_frequencies %>%
slice_head(n = 50) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Word_Frequencies",
top_words,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Word_Frequencies",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#3498db",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(top_words),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Word_Frequencies",
cols = 1:ncol(top_words),
widths = "auto"
)
}
# === SHEET 4: BIGRAMS ===
if (
!is.null(values$analysis_results$ngrams) &&
"2gram" %in% names(values$analysis_results$ngrams)
) {
openxlsx::addWorksheet(wb_content, "Bigrams")
bigrams_data <- values$analysis_results$ngrams$`2gram` %>%
slice_head(n = 50) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Bigrams",
bigrams_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Bigrams",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#9b59b6",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(bigrams_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Bigrams",
cols = 1:ncol(bigrams_data),
widths = "auto"
)
}
# === SHEET 5: TRIGRAMS ===
if (
!is.null(values$analysis_results$ngrams) &&
"3gram" %in% names(values$analysis_results$ngrams)
) {
openxlsx::addWorksheet(wb_content, "Trigrams")
trigrams_data <- values$analysis_results$ngrams$`3gram` %>%
slice_head(n = 50) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Trigrams",
trigrams_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Trigrams",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#e67e22",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(trigrams_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Trigrams",
cols = 1:ncol(trigrams_data),
widths = "auto"
)
}
# === SHEET 6: CITATION TYPES ===
if (
!is.null(values$analysis_results$citation_metrics$type_distribution)
) {
openxlsx::addWorksheet(wb_content, "Citation_Types")
types_data <- values$analysis_results$citation_metrics$type_distribution %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Citation_Types",
types_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citation_Types",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#e74c3c",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(types_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citation_Types",
cols = 1:ncol(types_data),
widths = "auto"
)
}
# === SHEET 7: CITATIONS BY SECTION ===
if (
!is.null(
values$analysis_results$citation_metrics$section_distribution
)
) {
openxlsx::addWorksheet(wb_content, "Citations_by_Section")
sections_data <- values$analysis_results$citation_metrics$section_distribution %>%
filter(n > 0) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Citations_by_Section",
sections_data,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citations_by_Section",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#16a085",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(sections_data),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citations_by_Section",
cols = 1:ncol(sections_data),
widths = "auto"
)
}
# === SHEET 8: CITATION CONTEXTS ===
if (
!is.null(values$analysis_results$citation_contexts) &&
nrow(values$analysis_results$citation_contexts) > 0
) {
openxlsx::addWorksheet(wb_content, "Citation_Contexts")
# Seleziona le colonne principali se esistono, altrimenti esporta tutte
contexts_cols <- c(
"citation_id",
"citation_text",
"section",
"citation_type",
"is_narrative",
"words_before",
"words_after",
"full_context",
"context_word_count",
"citation_position_in_text"
)
available_cols <- intersect(
contexts_cols,
names(values$analysis_results$citation_contexts)
)
if (length(available_cols) > 0) {
contexts_export <- values$analysis_results$citation_contexts %>%
select(all_of(available_cols)) %>%
as.data.frame(stringsAsFactors = FALSE)
} else {
# Se nessuna colonna specificata esiste, esporta tutte
contexts_export <- values$analysis_results$citation_contexts %>%
as.data.frame(stringsAsFactors = FALSE)
}
openxlsx::writeData(
wb_content,
"Citation_Contexts",
contexts_export,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citation_Contexts",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#2E86AB",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(contexts_export),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citation_Contexts",
cols = 1:ncol(contexts_export),
widths = "auto"
)
}
# === SHEET 9: CITATION NETWORK ===
if (
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
) {
openxlsx::addWorksheet(wb_content, "Citation_Network")
# Esporta tutte le colonne disponibili nel network
network_export <- values$analysis_results$network_data %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Citation_Network",
network_export,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Citation_Network",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#8e44ad",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(network_export),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Citation_Network",
cols = 1:ncol(network_export),
widths = "auto"
)
}
# === SHEET 10: WORD TRENDS (if available) ===
if (
!is.null(values$word_trends_data) &&
nrow(values$word_trends_data) > 0
) {
openxlsx::addWorksheet(wb_content, "Word_Trends")
trends_export <- values$word_trends_data %>%
arrange(word, segment_id) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Word_Trends",
trends_export,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Word_Trends",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#27ae60",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(trends_export),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Word_Trends",
cols = 1:ncol(trends_export),
widths = "auto"
)
# Add summary statistics for word trends
openxlsx::addWorksheet(wb_content, "Word_Trends_Summary")
trends_summary <- values$word_trends_data %>%
group_by(word) %>%
summarise(
total_occurrences = sum(count),
avg_frequency = mean(relative_frequency),
min_frequency = min(relative_frequency),
max_frequency = max(relative_frequency),
std_dev = sd(relative_frequency),
peak_segment = segment_name[which.max(relative_frequency)],
.groups = "drop"
) %>%
arrange(desc(total_occurrences)) %>%
as.data.frame(stringsAsFactors = FALSE)
openxlsx::writeData(
wb_content,
"Word_Trends_Summary",
trends_summary,
startRow = 1
)
openxlsx::addStyle(
wb_content,
"Word_Trends_Summary",
style = openxlsx::createStyle(
textDecoration = "bold",
fgFill = "#16a085",
fontColour = "#FFFFFF"
),
rows = 1,
cols = 1:ncol(trends_summary),
gridExpand = TRUE
)
openxlsx::setColWidths(
wb_content,
"Word_Trends_Summary",
cols = 1:ncol(trends_summary),
widths = "auto"
)
}
# Save workbook
openxlsx::saveWorkbook(wb_content, file, overwrite = TRUE)
},
error = function(e) {
# Log detailed error for debugging
cat("Error in download_all_results:\n")
cat("Error message:", e$message, "\n")
cat("Stack trace:\n")
print(e)
showNotification(
paste("Error exporting results:", e$message),
type = "error",
duration = 8
)
}
)
}
)
# ===========================================
# UTILITY FUNCTIONS
# ===========================================
is_analysis_valid <- reactive({
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$citations) &&
nrow(values$analysis_results$citations) > 0
})
is_network_valid <- reactive({
!is.null(values$analysis_results) &&
!is.null(values$analysis_results$network_data) &&
nrow(values$analysis_results$network_data) > 0
})
output$debug_info <- renderText({
if (!is.null(values$analysis_results)) {
paste(
"Debug Info:\n",
"PDF text length:",
nchar(values$pdf_text %||% ""),
"\n",
"Citations found:",
nrow(values$analysis_results$citations %||% data.frame()),
"\n",
"Network connections:",
nrow(values$analysis_results$network_data %||% data.frame()),
"\n",
"Analysis running:",
values$analysis_running %||% FALSE
)
} else {
"No analysis data available"
}
})
outputOptions(output, "debug_info", suspendWhenHidden = FALSE)
}
# Helper function for HTML card OpenAlex ----
create_oa_details_html <- function(oa_data) {
# Helper per gestire valori NULL/NA
safe_value <- function(x, default = "Not available") {
if (
is.null(x) ||
length(x) == 0 ||
all(is.na(x)) ||
!nzchar(trimws(as.character(x)))
) {
return(default)
}
return(x)
}
# === ESTRAI AUTORI E AFFILIAZIONI INLINE ===
authors_html <- NULL
if (!is.null(oa_data$authorships) && length(oa_data$authorships) > 0) {
authorships_df <- oa_data$authorships[[1]]
if (!is.null(authorships_df) && nrow(authorships_df) > 0) {
authors_items <- lapply(1:min(nrow(authorships_df), 15), function(i) {
auth <- authorships_df[i, ]
author_name <- safe_value(auth$display_name, "Unknown Author")
author_id <- safe_value(auth$id, "")
# Estrai affiliazioni
affiliation_text <- ""
if (!is.null(auth$institutions) && length(auth$institutions) > 0) {
insts_df <- auth$institutions[[1]]
if (
!is.null(insts_df) &&
nrow(insts_df) > 0 &&
"display_name" %in% names(insts_df)
) {
affiliation_text <- paste(insts_df$display_name, collapse = ", ")
}
}
# Se non ci sono istituzioni, prova con affiliation_raw
if (
!nzchar(affiliation_text) &&
!is.na(auth$affiliation_raw) &&
nzchar(auth$affiliation_raw)
) {
affiliation_text <- auth$affiliation_raw
}
# Layout inline: nome e affiliazione sulla stessa riga
div(
style = "margin-bottom: 8px; padding: 6px 10px; background-color: #f8f9fa; border-radius: 4px; display: flex; align-items: center; flex-wrap: wrap;",
tags$div(
style = "font-weight: 500; font-size: 14px; margin-right: 12px;",
# Link OpenAlex se disponibile
if (nzchar(author_id) && author_id != "Not available") {
tags$a(
href = author_id,
target = "_blank",
style = "color: #2E86AB; text-decoration: none;",
author_name,
icon(
"external-link-alt",
style = "margin-left: 4px; font-size: 10px;"
)
)
} else {
tags$span(style = "color: #2E86AB;", author_name)
}
),
if (nzchar(affiliation_text)) {
tags$span(
style = "color: #666; font-size: 13px; font-style: italic;",
icon("university", style = "margin-right: 5px; font-size: 11px;"),
affiliation_text
)
}
)
})
authors_html <- tagList(authors_items)
if (nrow(authorships_df) > 15) {
authors_html <- tagList(
authors_html,
tags$small(
style = "color: #999; font-style: italic; display: block; margin-top: 8px; padding-left: 10px;",
paste("... and", nrow(authorships_df) - 15, "more authors")
)
)
}
}
}
# === KEYWORDS - TUTTE ===
keywords_html <- NULL
if (!is.null(oa_data$keywords) && length(oa_data$keywords) > 0) {
keywords_data <- oa_data$keywords[[1]]
if (
!is.null(keywords_data) &&
is.data.frame(keywords_data) &&
nrow(keywords_data) > 0
) {
if ("display_name" %in% names(keywords_data)) {
# NESSUN LIMITE - tutte le keywords
keywords_html <- lapply(keywords_data$display_name, function(kw) {
tags$span(
class = "label label-primary",
style = "margin-right: 5px; margin-bottom: 5px; font-size: 11px; display: inline-block; background-color: #3498db; color: white; padding: 4px 10px; border-radius: 3px;",
kw
)
})
}
}
}
# === ABSTRACT ===
abstract_text <- safe_value(oa_data$abstract, "Abstract not available")
# === TOPICS - SOLO I PRIMI 5 ===
topics_html <- NULL
if (!is.null(oa_data$topics) && length(oa_data$topics) > 0) {
topics_data <- oa_data$topics[[1]]
if (
!is.null(topics_data) &&
is.data.frame(topics_data) &&
nrow(topics_data) > 0
) {
if ("display_name" %in% names(topics_data)) {
main_topics <- topics_data
if ("type" %in% names(topics_data)) {
main_topics <- topics_data[topics_data$type == "topic", ]
}
if (nrow(main_topics) > 0) {
# LIMITE A 5 TOPICS
topics_html <- lapply(1:min(5, nrow(main_topics)), function(i) {
tags$span(
class = "label label-info",
style = "margin-right: 5px; margin-bottom: 5px; font-size: 11px; display: inline-block; background-color: #5bc0de; color: white; padding: 4px 10px; border-radius: 3px;",
main_topics$display_name[i]
)
})
}
}
}
}
# === JOURNAL INFO ===
journal_name <- safe_value(
oa_data$source_display_name,
"Journal not available"
)
issn <- safe_value(oa_data$issn_l, "")
# === DOCUMENT TYPE ===
doc_type <- safe_value(oa_data$type, "Unknown")
# === PUBLICATION INFO ===
pub_year <- safe_value(oa_data$publication_year, "N/A")
citations <- safe_value(oa_data$cited_by_count, 0)
fwci <- safe_value(oa_data$fwci, NA)
# === OPEN ACCESS INFO ===
is_oa <- !is.na(oa_data$is_oa) && oa_data$is_oa
oa_status <- tolower(safe_value(oa_data$oa_status, "closed")) # Normalizza in lowercase
# === DOI ===
doi_url <- safe_value(oa_data$doi, "")
doi_clean <- gsub("https://doi.org/", "", doi_url)
# === PDF URL ===
pdf_url <- safe_value(oa_data$pdf_url, "")
has_pdf <- nzchar(pdf_url) && !is.na(pdf_url) && pdf_url != "Not available"
# === CREA HTML COMPLETO ===
tagList(
div(
class = "oa-document-summary",
# Metadata Cards Row
fluidRow(
style = "margin-bottom: 20px;",
# Card 1: Publication Year (invariata)
column(
3,
div(
style = "background: linear-gradient(135deg, #3498db 0%, #2980b9 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);",
div(
style = "display: flex; align-items: center;",
icon(
"calendar",
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"Publication Year"
),
tags$div(
style = "font-size: 24px; font-weight: bold; margin-top: 2px;",
pub_year
)
)
)
)
),
# Card 2: Citations (invariata)
column(
3,
div(
style = "background: linear-gradient(135deg, #27ae60 0%, #229954 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);",
div(
style = "display: flex; align-items: center;",
icon(
"quote-right",
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"Citations"
),
tags$div(
style = "font-size: 24px; font-weight: bold; margin-top: 2px;",
format(as.numeric(citations), big.mark = ",")
)
)
)
)
),
# Card 3: Open Access Status (MODIFICATA)
column(3, {
# Determina colore e icona in base a oa_status
oa_colors <- list(
closed = list(
bg = "#95a5a6 0%, #7f8c8d",
icon = "lock",
text = "Closed"
),
gold = list(
bg = "#f39c12 0%, #e67e22",
icon = "unlock",
text = "Gold OA"
),
green = list(
bg = "#27ae60 0%, #229954",
icon = "unlock",
text = "Green OA"
),
bronze = list(
bg = "#cd7f32 0%, #b8732d",
icon = "unlock",
text = "Bronze OA"
),
hybrid = list(
bg = "#16a085 0%, #138d75",
icon = "unlock",
text = "Hybrid OA"
)
)
# Se oa_status non è nei valori conosciuti, usa closed come default
current_status <- if (oa_status %in% names(oa_colors)) {
oa_status
} else {
"closed"
}
status_info <- oa_colors[[current_status]]
div(
style = paste0(
"background: linear-gradient(135deg, ",
status_info$bg,
" 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);"
),
div(
style = "display: flex; align-items: center;",
icon(
status_info$icon,
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"Access"
),
tags$div(
style = "font-size: 18px; font-weight: bold; margin-top: 2px;",
status_info$text
)
)
)
)
}),
# Card 4: FWCI (invariata)
column(
3,
div(
style = "background: linear-gradient(135deg, #f39c12 0%, #e67e22 100%); border-radius: 8px; padding: 15px; color: white; min-height: 85px; display: flex; flex-direction: column; justify-content: center; box-shadow: 0 2px 4px rgba(0,0,0,0.1);",
div(
style = "display: flex; align-items: center;",
icon(
"chart-line",
style = "font-size: 28px; margin-right: 12px; opacity: 0.9;"
),
div(
tags$div(
style = "font-size: 10px; text-transform: uppercase; opacity: 0.85; letter-spacing: 0.5px;",
"FWCI"
),
tags$div(
style = "font-size: 24px; font-weight: bold; margin-top: 2px;",
if (!is.na(fwci)) round(as.numeric(fwci), 2) else "N/A"
)
)
)
)
)
),
# Authors Section - PIÙ COMPATTA
if (!is.null(authors_html)) {
div(
class = "box box-primary",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("users", style = "margin-right: 8px;"),
"Authors",
style = "color: #2E86AB; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
style = "max-height: 250px; overflow-y: auto;",
authors_html
)
)
},
# Journal Info
div(
class = "box box-success",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("book", style = "margin-right: 8px;"),
"Journal",
style = "color: #27ae60; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
style = "padding: 15px;",
# Container principale
div(
style = "display: flex; align-items: center; justify-content: space-between; flex-wrap: wrap; gap: 20px;",
# Colonna 1: Nome Journal e ISSN
div(
style = "flex: 1 1 300px; min-width: 250px;",
tags$div(
style = "font-size: 16px; font-weight: 600; color: #2c3e50; margin-bottom: 5px;",
journal_name
),
if (nzchar(issn) && issn != "Not available") {
tags$div(
style = "color: #666; font-size: 13px;",
tags$strong("ISSN: "),
issn
)
}
),
# Colonna 2: Document Type
div(
style = "flex: 0 0 auto; display: flex; align-items: center; gap: 8px;",
tags$span(
style = "color: #666; font-size: 13px; font-weight: 500;",
"Type:"
),
tags$span(
class = "label label-info",
style = "font-size: 12px; padding: 5px 12px; background-color: #3498db; text-transform: capitalize;",
safe_value(oa_data$type, "Unknown")
)
),
# Colonna 3: DOI
if (nzchar(doi_url) && doi_url != "Not available") {
div(
style = "flex: 0 0 auto; display: flex; align-items: center; gap: 8px;",
tags$span(
style = "color: #666; font-size: 13px; font-weight: 500;",
"DOI:"
),
tags$a(
href = doi_url,
target = "_blank",
style = "color: #3498db; text-decoration: none; font-family: 'Courier New', monospace; font-size: 13px; font-weight: 500;",
doi_clean,
icon(
"external-link-alt",
style = "margin-left: 5px; font-size: 11px;"
)
)
)
}
)
)
),
# Abstract
div(
class = "box box-info",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("align-left", style = "margin-right: 8px;"),
"Abstract",
style = "color: #3498db; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
tags$p(
style = "text-align: justify; line-height: 1.7; color: #333;",
abstract_text
)
)
),
# Keywords - TUTTE
if (!is.null(keywords_html) && length(keywords_html) > 0) {
div(
class = "box box-warning",
style = "margin-bottom: 15px;",
div(
class = "box-header with-border",
h5(
icon("tags", style = "margin-right: 8px;"),
"Keywords",
tags$small(
style = "color: #999; font-weight: normal; margin-left: 8px;",
paste0("(", length(keywords_html), ")")
),
style = "color: #f39c12; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
keywords_html
)
)
},
# Topics - SOLO PRIMI 5
if (!is.null(topics_html) && length(topics_html) > 0) {
div(
class = "box",
style = "margin-bottom: 15px; border-top: 3px solid #9b59b6;",
div(
class = "box-header with-border",
h5(
icon("lightbulb", style = "margin-right: 8px;"),
"Research Topics",
tags$small(
style = "color: #999; font-weight: normal; margin-left: 8px;",
"(Top 5)"
),
style = "color: #9b59b6; margin: 0; font-weight: 600;"
)
),
div(
class = "box-body",
topics_html
)
)
},
# Action Button - PDF
if (has_pdf) {
div(
style = "margin-top: 20px; padding-top: 15px; border-top: 1px solid #ddd; text-align: center;",
tags$a(
class = "btn btn-danger btn-lg",
href = pdf_url,
target = "_blank",
icon("file-pdf"),
" Download Full Text PDF",
style = "font-weight: bold; padding: 12px 30px;"
)
)
}
)
)
}
reconstruct_from_txt <- function(file_lines) {
reconstructed_txt <- list()
current_section <- "Full_text"
section_content <- ""
for (line in file_lines) {
if (grepl("^SECTION LINE SEPARATION: ", line)) {
# Save the previous section content
if (current_section != "Full_text") {
reconstructed_txt[[current_section]] <- section_content
}
# Start a new section
current_section <- sub("^SECTION LINE SEPARATION: ", "", line)
section_content <- ""
} else {
section_content <- paste0(
section_content,
ifelse(section_content == "", "", "\n"),
line
)
}
}
# Save the last section content
reconstructed_txt[[current_section]] <- section_content
if (!"Full_text" %in% names(reconstructed_txt)) {
Full_text <- paste(reconstructed_txt, collapse = "\n\n")
reconstructed_txt <- c(Full_text = Full_text, reconstructed_txt)
}
return(reconstructed_txt)
}
Try the bibliometrix package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.