code/arxiv-metadata.R

#!/usr/bin/env Rscript

library(tidyverse)
library(xml2)
library(lubridate)
library(optparse)
library(glue)

# Define command line options
option_list <- list(
  make_option(c("-i", "--input"), type = "character", default = NULL,
              help = "Directory containing ArXiv XML files", metavar = "character"),
  make_option(c("-o", "--output"), type = "character", default = NULL,
              help = "Output CSV file path", metavar = "character")
)

# Parse command line arguments
opt_parser <- OptionParser(option_list = option_list)
opt <- parse_args(opt_parser)

# Check if required arguments are provided
if (is.null(opt$input)) {
  stop("Error: Input directory (-i, --input) is required")
}

if (is.null(opt$output)) {
  stop("Error: Output file (-o, --output) is required")
}

# List all XML files
xml_files <- list.files(path = opt$input, pattern = "\\.xml$", full.names = TRUE)

# Function to parse one XML file and extract all metadata
extract_metadata <- function(file_path) {
  # Read the XML
  doc <- read_xml(file_path)

  # Get file ID from filename (without .xml extension)
  file_id <- tools::file_path_sans_ext(basename(file_path))

  # Define namespaces properly
  ns <- c(atom = "http://www.w3.org/2005/Atom")

  # Find all entries
  entries <- xml_find_all(doc, "//atom:entry", ns)

  # For each entry, extract all metadata
  map_dfr(entries, function(entry) {
    # Extract authors
    authors <- xml_find_all(entry, ".//atom:author/atom:name", ns) %>%
      map_chr(~ xml_text(.)) %>%
      list()

    # Extract entry ID and version
    entry_id <- xml_text(xml_find_first(entry, ".//atom:id", ns))
    version_number <- as.numeric(str_extract(entry_id, "v(\\d+)$", group = 1))

    # Extract dates
    updated_date <- xml_text(xml_find_first(entry, ".//atom:updated", ns))
    published_date <- xml_text(xml_find_first(entry, ".//atom:published", ns))

    # Convert to Date format (yyyy-mm-dd)
    updated_date <- as.Date(ymd_hms(updated_date))
    published_date <- as.Date(ymd_hms(published_date))

    # Create tibble with all data
    tibble(
      id = file_id,
      href = glue("https://arxiv.org/abs/{file_id}"),
      updated_at = updated_date,
      published_at = published_date,
      count_versions = version_number
    )
  })
}

# Apply to all XML files and combine results
results <- map_dfr(xml_files, extract_metadata)

# Print the result
print(results)

# Save to CSV file
write_csv(results, opt$output)