使用 R 和 Amazon Web Services 进行文档分析

textract <- paws::textract()
analyze_document <- function(bucket, file) {
  # Start analyzing the PDF.  resp <- textract$start_document_analysis(    DocumentLocation = list(      S3Object = list(Bucket = bucket, Name = file)    ),    FeatureTypes = "TABLES"  )
  # Check that the analysis is done and get the result.  count <- 0  while (count < 30 && (!exists("result") || result$JobStatus == "IN_PROGRESS")) {    Sys.sleep(1)    result <- textract$get_document_analysis(      JobId = resp$JobId    )     count <- count + 1  }  return(result)}
analysis <- analyze_document("my-bucket", "GS-1966-01-11.pdf")

# List of 13# $ BlockType      : chr "TABLE"# $ Confidence     : num 100# $ Text           : chr(0) # $ RowIndex       : int(0) # $ ColumnIndex    : int(0) # $ RowSpan        : int(0) # $ ColumnSpan     : int(0) # $ Geometry       :List of 2# .. <not shown># $ Id             : chr "c6841638-d3e0-414b-af12-b94ed34aac8a"# $ Relationships  :List of 1# ..$ :List of 2# .. ..$ Type: chr "CHILD"# .. ..$ Ids : chr [1:256] "e1866e80-0ef0-4bdd-a6fd-9508bb833c03" ...# $ EntityTypes    : list()# $ SelectionStatus: chr(0) # $ Page           : int 3

# List of 13# $ BlockType      : chr "CELL"# $ Confidence     : num 100# $ Text           : chr(0) # $ RowIndex       : int 1# $ ColumnIndex    : int 2# $ RowSpan        : int 1# $ ColumnSpan     : int 1# $ Geometry       :List of 2# .. <not shown># $ Id             : chr "132b0343-0172-4ddd-bf30-12f133f0f31d"# $ Relationships  :List of 1# ..$ :List of 2# .. ..$ Type: chr "CHILD"# .. ..$ Ids : chr "5734871f-5a7f-460e-a5f5-dd42bbc57a27"# ...

# List of 13# $ BlockType      : chr "WORD"# $ Confidence     : num 99.8# $ Text           : chr "Year"# ...

get_tables <- function(analysis) {  tables <- list()  blocks <- analysis$Blocks  names(blocks) <- sapply(blocks, function(x) x$Id)
  for (block in blocks) {
    if (block$BlockType == "TABLE") {      cells <- get_children(block, blocks)      rows <- max(sapply(cells, function(x) x$RowIndex))      cols <- max(sapply(cells, function(x) x$ColumnIndex))      table <- matrix(nrow = rows, ncol = cols)
      # 1.Go through a table's cells one-by-one      for (cell in cells) {
        # 2.Get the cell's contents        words <- get_children(cell, blocks)        text <- paste(sapply(words, function(x) x$Text), collapse = " ")
        # 3.Insert the cell contents into the matrix        row <- cell$RowIndex        col <- cell$ColumnIndex        table[row, col] <- text      }      tables <- c(tables, list(table))    }  }  return(tables)} 
tables <- get_tables(analysis)

# Connect to the database using an IAM authentication token.rds <- paws::rds()token <- rds$build_auth_token("myhost:5432", "us-east-1", "david")con <- DBI::dbConnect(    RPostgres::Postgres(),    host = "myhost", port = 5432, dbname = "mydb",    user = "david", password = token)

# Create rows for each table to insert into the database.database_rows <- data.frame(  document = "GS-1966-01-11.pdf",  table_num = 1:length(tables),  data = sapply(tables, jsonlite::toJSON))
# Store the tables in the database.DBI::dbAppendTable(con, name = "tables", value = database_rows)