Can `pdftools` detect which radio button is selected on a fillable form?
Edit April 22, 2024 - I'm going to close this question after about three weeks. The short answer to the original question about whether the R package pdftools can parse pdf forms appears to be "no". Both @Felix_Jassler and @K-J provided useful "work-arounds". Personally, I found the combination of reticulate and the python module pypdf the best fit for my needs but that's a very different question.
The package has worked extremely well on processing "traditional" non fillable forms.
In my first attempts at using it with "fillable forms" I can't seem to find a way to distinguish between radio buttons at all (or checkboxes in an elegant way) that are selected and those that are not. I'm not sure if I'm missing some nuance, making a complete mistake, or whether the functions don't support it?
An example "blank" original form is here. For the reprex below I am focusing on a small segment of the form on page 1 that I have included as a screenshot after filling out and saving a few entries.
I would like to know if there is a way to distinguish the fact that "Long Term Care" is selected in the filled out form versus not selected in the original?
Thank you in advance. Below is what I hope is a reprex that will help, since I could not find an easy safe place to "post" the example filled out form I used dput to put the resulting data in the reprex obviously users can grab the original and save changes to their local filesystem if desired.
suppressPackageStartupMessages(library(dplyr))
library(stringr)
## Not sure if poppler version matters?
library(pdftools)
#> Using poppler version 23.04.0
## Download and save the original form as original.pdf
download.file("https://www.cdc.gov/infectioncontrol/pdf/icar/IPC-demo-LTC-508.pdf",
"original.pdf")
## Download and save the original form as example.pdf
# download.file("https://www.cdc.gov/infectioncontrol/pdf/icar/IPC-demo-LTC-508.pdf",
# "filled.pdf")
## Let's use just the first page for the reprex
## Improve human readability with the str_split
original_pageone_text <- str_split(pdf_text("original.pdf")[[1]], "\n")[[1]]
original_pageone_text
#> [1] " Infection Control Assessment and Response (ICAR) Tool for General Infection"
#> [2] " Prevention and Control (IPC) Across Settings"
#> [3] ""
#> [4] " Section 1: Facility Demographics and Infection Prevention and Control (IPC) Infrastructure"
#> [5] " Long–Term Care"
#> [6] "General Facility Demographics and IPC Infrastructure"
#> [7] ""
#> [8] "Date of Assessment:"
#> [9] ""
#> [10] "Facility Name:"
#> [11] ""
#> [12] "State/Territory: County:"
#> [13] ""
#> [14] "Zip Code: State/Territory-assigned Unique ID (if applicable):"
#> [15] ""
#> [16] "Facility type (Complete the demographic form that NHSN Facility Organization ID (if applicable):"
#> [17] "corresponds to the type of facility):"
#> [18] " CMS Facility ID (if applicable):"
#> [19] " ● Acute Care Hospital / Critical Access Hospital"
#> [20] " ● Long-term Care"
#> [21] " ● Outpatient/Ambulatory Care"
#> [22] " ● Other (specify):"
#> [23] ""
#> [24] ""
#> [25] ""
#> [26] "Facility Respondent Name(s) and Job Title(s):"
#> [27] ""
#> [28] ""
#> [29] ""
#> [30] "Rationale for assessment:"
#> [31] " ■ Requested by facility"
#> [32] " ■ Requested by accrediting agency/ licensing organization"
#> [33] " ■ Requested by state or local health department"
#> [34] " ■ HAI prevention focused:"
#> [35] " ■ CAUTI"
#> [36] " ■ CLABSI"
#> [37] " ■ SSI"
#> [38] " ■ CDI"
#> [39] " ■ Other (specify):"
#> [40] " ■ Prevention collaborative (specify partners):"
#> [41] " ■ Outbreak (specify):"
#> [42] " ■ Other (specify):"
#> [43] ""
#> [44] " Obtain a list of products used for cleaning and disinfection of environmental surfaces and"
#> [45] " non-critical patient/resident care equipment in the facility"
#> [46] ""
#> [47] "EPA registration number(s) for products used in patient/resident rooms:"
#> [48] ""
#> [49] "EPA registration number(s) for products used in common areas:"
#> [50] ""
#> [51] "EPA registration number(s) for products used on non-critical patient/resident care equipment (e.g., blood glucose meters):"
#> [52] ""
#> [53] ""
#> [54] ""
#> [55] ""
#> [56] "CS334433-M 12/14/2022"
#> [57] ""
## Let's focus on the radio buttons at 19 through 22 I plan on
## selecting 20 just in case it is the way the character is
## "displayed" on-screen we'll get the value as an integer
utf8ToInt(original_pageone_text[20]) #9679 is the radio button
#> [1] 32 9679 32 76 111 110 103 45 116 101 114 109 32 67 97
#> [16] 114 101
original_radio_button <- utf8ToInt(original_pageone_text[20])
## Open filled.pdf and manually enter a few text boxes
## Select long term care radio button and save
## On screen it "appears" different
## To keep it reproducible I am putting the results
## of dput(str_split(pdf_text("filled.pdf")[[1]], "\n")[[1]]) here
filled_pageone_text <-
structure(c(" Infection Control Assessment and Response (ICAR) Tool for General Infection",
" Prevention and Control (IPC) Across Settings",
"", " Section 1: Facility Demographics and Infection Prevention and Control (IPC) Infrastructure",
" Long–Term Care", "General Facility Demographics and IPC Infrastructure",
"", "Date of Assessment:", "", "Facility Name: Imaginary", "",
"State/Territory: County:",
"", "Zip Code: State/Territory-assigned Unique ID (if applicable):",
"", "Facility type (Complete the demographic form that NHSN Facility Organization ID (if applicable):",
"corresponds to the type of facility):", " CMS Facility ID (if applicable):",
" ● Acute Care Hospital / Critical Access Hospital", " ● Long-term Care",
" ● Outpatient/Ambulatory Care", " ● Other (specify):", "",
"", "", "Facility Respondent Name(s) and Job Title(s):", "",
"", "", "Rationale for assessment:", " ✓", " ■ Requested by facility",
" ✓", " ■ Requested by accrediting agency/ licensing organization",
" ■ Requested by state or local health department", " ✓",
" ■ HAI prevention focused:", " ✓", " ✓ CAUTI", " ■",
" ✓ CLABSI", " ■", " ■ SSI", " ■ CDI",
" ■ Other (specify):", " ■ Prevention collaborative (specify partners):",
" ✓", " ■ Outbreak (specify):", " ■ Other (specify):",
"", " Obtain a list of products used for cleaning and disinfection of environmental surfaces and",
" non-critical patient/resident care equipment in the facility",
"", "EPA registration number(s) for products used in patient/resident rooms:",
"", "EPA registration number(s) for products used in common areas:",
"", "EPA registration number(s) for products used on non-critical patient/resident care equipment (e.g., blood glucose meters):",
"", "", "", "", "CS334433-M 12/14/2022", ""))
filled_pageone_text
#> [1] " Infection Control Assessment and Response (ICAR) Tool for General Infection"
#> [2] " Prevention and Control (IPC) Across Settings"
#> [3] ""
#> [4] " Section 1: Facility Demographics and Infection Prevention and Control (IPC) Infrastructure"
#> [5] " Long–Term Care"
#> [6] "General Facility Demographics and IPC Infrastructure"
#> [7] ""
#> [8] "Date of Assessment:"
#> [9] ""
#> [10] "Facility Name: Imaginary"
#> [11] ""
#> [12] "State/Territory: County:"
#> [13] ""
#> [14] "Zip Code: State/Territory-assigned Unique ID (if applicable):"
#> [15] ""
#> [16] "Facility type (Complete the demographic form that NHSN Facility Organization ID (if applicable):"
#> [17] "corresponds to the type of facility):"
#> [18] " CMS Facility ID (if applicable):"
#> [19] " ● Acute Care Hospital / Critical Access Hospital"
#> [20] " ● Long-term Care"
#> [21] " ● Outpatient/Ambulatory Care"
#> [22] " ● Other (specify):"
#> [23] ""
#> [24] ""
#> [25] ""
#> [26] "Facility Respondent Name(s) and Job Title(s):"
#> [27] ""
#> [28] ""
#> [29] ""
#> [30] "Rationale for assessment:"
#> [31] " ✓"
#> [32] " ■ Requested by facility"
#> [33] " ✓"
#> [34] " ■ Requested by accrediting agency/ licensing organization"
#> [35] " ■ Requested by state or local health department"
#> [36] " ✓"
#> [37] " ■ HAI prevention focused:"
#> [38] " ✓"
#> [39] " ✓ CAUTI"
#> [40] " ■"
#> [41] " ✓ CLABSI"
#> [42] " ■"
#> [43] " ■ SSI"
#> [44] " ■ CDI"
#> [45] " ■ Other (specify):"
#> [46] " ■ Prevention collaborative (specify partners):"
#> [47] " ✓"
#> [48] " ■ Outbreak (specify):"
#> [49] " ■ Other (specify):"
#> [50] ""
#> [51] " Obtain a list of products used for cleaning and disinfection of environmental surfaces and"
#> [52] " non-critical patient/resident care equipment in the facility"
#> [53] ""
#> [54] "EPA registration number(s) for products used in patient/resident rooms:"
#> [55] ""
#> [56] "EPA registration number(s) for products used in common areas:"
#> [57] ""
#> [58] "EPA registration number(s) for products used on non-critical patient/resident care equipment (e.g., blood glucose meters):"
#> [59] ""
#> [60] ""
#> [61] ""
#> [62] ""
#> [63] "CS334433-M 12/14/2022"
#> [64] ""
## Same methodology as above
utf8ToInt(filled_pageone_text[20]) #9679 is the radio button
#> [1] 32 9679 32 76 111 110 103 45 116 101 114 109 32 67 97
#> [16] 114 101
filled_radio_button <- utf8ToInt(filled_pageone_text[20])
## Clearly different
identical(original_pageone_text, filled_pageone_text)
#> [1] FALSE
setdiff(filled_pageone_text, original_pageone_text)
#> [1] "Facility Name: Imaginary"
#> [2] "State/Territory: County:"
#> [3] "Zip Code: State/Territory-assigned Unique ID (if applicable):"
#> [4] "Facility type (Complete the demographic form that NHSN Facility Organization ID (if applicable):"
#> [5] " CMS Facility ID (if applicable):"
#> [6] " ✓"
#> [7] " ✓ CAUTI"
#> [8] " ■"
#> [9] " ✓ CLABSI"
identical(original_radio_button, filled_radio_button)
#> [1] TRUE
sessionInfo()
#> R version 4.3.2 (2023-10-31)
#> Platform: aarch64-apple-darwin20 (64-bit)
#> Running under: macOS Sonoma 14.2.1
#>
#> Matrix products: default
#> BLAS: /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRblas.0.dylib
#> LAPACK: /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRlapack.dylib; LAPACK version 3.11.0
#>
#> locale:
#> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
#>
#> time zone: America/New_York
#> tzcode source: internal
#>
#> attached base packages:
#> [1] stats graphics grDevices utils datasets methods base
#>
#> other attached packages:
#> [1] pdftools_3.4.0 stringr_1.5.1 dplyr_1.1.4
#>
#> loaded via a namespace (and not attached):
#> [1] vctrs_0.6.5 cli_3.6.2 knitr_1.45 rlang_1.1.3
#> [5] xfun_0.42 stringi_1.8.3 purrr_1.0.2 styler_1.10.2
#> [9] generics_0.1.3 glue_1.7.0 askpass_1.2.0 qpdf_1.3.2
#> [13] htmltools_0.5.7 fansi_1.0.6 rmarkdown_2.25 R.cache_0.16.0
#> [17] tibble_3.2.1 evaluate_0.23 fastmap_1.1.1 yaml_2.3.8
#> [21] lifecycle_1.0.4 compiler_4.3.2 fs_1.6.3 Rcpp_1.0.12
#> [25] pkgconfig_2.0.3 rstudioapi_0.15.0 R.oo_1.26.0 R.utils_2.12.3
#> [29] digest_0.6.34 R6_2.5.1 tidyselect_1.2.0 utf8_1.2.4
#> [33] reprex_2.1.0 pillar_1.9.0 magrittr_2.0.3 R.methodsS3_1.8.2
#> [37] tools_4.3.2 withr_3.0.0
Created on 2024-04-01 with reprex v2.1.0
Solution
Inspired by @KJ's comment, here's a somewhat laborious approach. We start by reading the binary of the PDF file.
filename <- "original.pdf"
# from the binary, remove all non-printable ascii characters
# then, convert to string
pdf_content <- (
readBin(filename, "raw", file.info(filename)$size)
|> Filter(f = function(x) x == 10 || (x >= 32 && x < 127))
|> rawToChar()
)
pdf_lines <- strsplit(pdf_content, "\n")[[1]]
From here, we could try to find some of the selected values by searching for tags starting with the /V attribute.
> lines <- which(grepl('<< /V', pdf_lines, fixed = TRUE))
> pdf_lines[lines]
[ ] ...
[20] "<< /V /Long-term#20Care /Kids [ 25 0 R 26 0 R 27 0 R 28 0 R ] /T (S1 GF 7) /FT /Btn >>"
Notice how the document also contains 20 radio button groups. In this context, the attributes mean the following:
/Vstands for Value./V /Long-term#20Careindicates that the selected input isLong-term Care. If no value is selected, you'll find/V /. Some characters are escaped, such as:- whitespace ->
#20 #->#23(->#28)->#29/->#2f
- whitespace ->
/Kidsindirectly references other objects in the PDF, here presumably to indicate which objects belong to the same group of radio buttons./Tis the title or name for the form field./FTis the function, or role of the object. The value/Btnindicates that this is a checkbox. (see more here)
With this approach, a simple solution would be to find all /V tags. Then you have to find out, which titled tag corresponds to which radio button group in the PDF. Here, I'll use some regex magic to extract all key-value pairs for the radio buttons (/T being the key, /V being the value).
# note we're parsing the entire pdf string, not the lines individually
findRadioButtonValues <- function(pdf_content) {
# replacing '\\s+' with a single whitespace is not really necessary in your document
# I'll just add it for save-guarding in case it becomes a problem in other documents
# Note also that I'm assuming that all radio buttons start with '<< /V'
values <- (
pdf_content
|> stringr::str_replace_all('\n', ' ')
|> stringr::str_replace_all('\\s+', ' ')
|> stringr::str_match_all('<< /V /(.*?) [^>]*?/T (\\([^)]*\\))[^>]*>>')
)
values <- values[[1]]
# matching contains 3 groups that are split into three columns
# column 1: entire tags, "<< ... >>" (can be omitted)
# column 2: value of tags
# column 3: title of tags
values[,c(3,2)]
# if you prefer a named list with appropriate key-value pairs:
# structure(as.list(values[,2]), names = values[,3])
}
Then, we can call the function and get back all 20 radio button values.
> findRadioButtonValues(pdf_content)
[,1] [,2]
[1,] "(LTC 12f)" ""
[2,] "(LTC 12d )" ""
[3,] "(LTC 12c)" ""
[4,] "(LTC 12)" ""
[5,] "(LTC 9a 1)" ""
[6,] "(LTC 9)" ""
[7,] "(LTC 8)" ""
[8,] "(LTC 4)" ""
[9,] "(LTC 3)" "Government#20#28not#20VA#29"
[10,] "(LTC 2)" "Dual#20Medicare#2fMedicaid"
[11,] "(S1 11)" ""
[12,] "(S1 10)" "Not#20Assessed"
[13,] "(S1 9)" "Yes"
[14,] "(S1 8)" "No"
[15,] "(S1 7)" "Unknown"
[16,] "(S1 6)" "Not#20Assessed"
[17,] "(S1 3a)" "No"
[18,] "(S1 2a)" "Unknown"
[19,] "(S1 1a)" "No"
[20,] "(S1 GF 7)" "Long-term#20Care"
- Split character separated values into rows and add a column with index
- Conditionally color cells based on different column in gt
- How to let a bslib navbar collapse at another Bootstrap breakpoint?
- Issue with predict.plm() when using lag() inside the regression
- How to shift the labels from right to left?
- Why is my MCP giving me an area size that is too small?
- How to plot the Hazard Ratio + CI over time of survival data in ggplot in R?
- About the betweenness function in igraph
- Reading through multiple files in chunks in R
- How to show graticule coordinates along axes in leaflet
- How can I append my dataset based on timestamp?
- Converting a BMP image to a matrix in R
- How to use shinycssloaders withSpinner with a plot output in a bslib card?
- Error: object ‘isFALSE’ is not exported by 'namespace:xfun' when running anything in a .Rmd file
- colourpicker-panel is partially hidden in a rank_list
- add geom_label and geom_density to barplot distribution
- Configuration failed because libcurl was not found
- How do I remove the whitespace around a plotted image?
- Change ggplot font size in middle of title
- R/Exams change font
- Error in plot.new() : figure margins too large in R
- Positioning of ggplot insets with packages 'ggpp' or 'ggpmisc'
- precision, recall and f-measure in R
- Creating one legend for multiple plots
- Error when trying to filter my dataset based on a vector of strings
- Relative frequencies of multiple variables in R
- R DataTables do not display Buttons and Length Menu simultaneously
- Calculate mean of matrices having different dimensions
- Creating Summary table using GTSUMMARY
- Calculating percentages for different combinations in a dataset