R/rt_fund_pmc.R

In serghiou/rtransparent: Identifies indicators of transparency

# TODO: First find the paragraphs that contain intersting words and then
#    apply the obliteration and rest of functions and test if this saves
#    time!



#' Identify mentions of support
#'
#' Identifies mentions of "This work was funded by ..." and of "This work was
#'     completed using funds from ..."
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_1 <- function(article) {

  # synonyms <- .create_synonyms()
  # words <- c("This", "research", "is_have", "funded", "by")
  #
  # this_research <-
  #   synonyms %>%
  #   magrittr::extract(words[1:2]) %>%
  #   lapply(.bound) %>%
  #   lapply(.encase) %>%
  #   paste(collapse = " ")
  #
  # synonyms %>%
  #   magrittr::extract(words[3:5]) %>%
  #   lapply(.bound) %>%
  #   lapply(.encase) %>%
  #   # lapply(.max_words) %>%
  #   paste(collapse = synonyms$txt) %>%
  #   paste(this_research, .) %>%
  #   grep(article, perl = T)


  synonyms <- .create_synonyms()
  words <- c("This_singular", "research_singular", "is_singular", "funded_funding", "by")

  this_research <-
    synonyms %>%
    magrittr::extract(words[1:2]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = ".{0,15}")

  was_funded_by <-
    synonyms %>%
    magrittr::extract(words[3:5]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt)

  singular <-
    c(this_research, was_funded_by) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

  if (!length(singular)) {

    words <- c("These", "researches", "are", "funded_funding", "by")

    synonyms %>%
      magrittr::extract(words) %>%
      lapply(.bound, location = "end") %>%
      lapply(.encase) %>%
      # lapply(.max_words) %>%
      paste(collapse = synonyms$txt) %>%
      grep(article, perl = T)

  } else {

    return(singular)

  }
}


#' Identify mentions of support
#'
#' Returns the index with the elements of interest.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("funded_funding", "this_singular", "research_singular")

  singular <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound, location = "end") %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

  if (!length(singular)) {

    words <- c("funded", "these", "researches")

    synonyms %>%
      magrittr::extract(words) %>%
      lapply(.bound, location = "end") %>%
      lapply(.encase) %>%
      # lapply(.max_words) %>%
      paste(collapse = synonyms$txt) %>%
      grep(article, perl = T)

  } else {

    return(singular)

  }
}


#' Identify mentions of support
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_3 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("research", "is_have", "funded", "by")

  research_is <-
    synonyms %>%
    magrittr::extract(words[1:2]) %>%
    lapply(.bound, location = "end") %>%
    lapply(.encase) %>%
    paste(collapse = " ")

  synonyms %>%
    magrittr::extract(words[3:4]) %>%
    lapply(.bound, location = "end") %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt) %>%
    paste(research_is, ., sep = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of support
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_4 <- function(article) {

  # TODO: I have now made this function much more general than before, by
  #    removing the requirement for "is" from the start. It seems that this now
  #    became more sensitive with no loss in specificity. If it remains so in
  #    further testing, remove all previews functions, which are basically more
  #    specific versions of this!

  synonyms <- .create_synonyms()
  words <- c("funded", "by", "award")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)
}


get_support_5 <- function(article) {

  # TODO: This is trying to capture some phrases missed by get_support_4 b/c of
  #    the requirement for is, e.g. "Project supported by X". I am introducing
  #    a 3 word limit at the start to make it stricter - upon tests, there were
  #    very very few mistakes (1/200 FP) without introducing this restriction.

  synonyms <- .create_synonyms()
  words <- c("funded", "by", "foundation")

  funded_by_award <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt)

  start_of_sentence <- "(^\\s*|(:|\\.)\\s*|[A-Z][a-zA-Z]+\\s*)"
  .max_words(start_of_sentence, n_max = 4, space_first = F) %>%
    paste0(funded_by_award) %>%
    grep(article, perl = T)
}


#' Identify mentions of support
#'
#' Return the index of support statements such as: We gratefully acknowledge
#'     support from the UK Engineering and Physical Sciences Research Council.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_6 <- function(article) {

  synonyms <- .create_synonyms()
  words_1 <- c("acknowledge", "support_only", "foundation_award")
  words_2 <- c("support_only", "foundation_award", "acknowledged")

  .max_words(c("acknowledge", "support_only"), 2)

  acknowledge <-
    synonyms %>%
    magrittr::extract(words_1[1]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    lapply(.max_words)

  support_foundation <-
    synonyms %>%
    magrittr::extract(words_1[2:3]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt)

  a <-
    c(acknowledge, support_foundation) %>%
    paste(collapse = " ") %>%
    grep(article, perl = T)

  b <-
    synonyms %>%
    magrittr::extract(words_2) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

  return(unique(c(a, b)))
}


#' Identify mentions of support
#'
#' Return the index of support statements such as: Support was provided by the
#'     U.S. Department of Agriculture (USDA) Forest Service.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_7 <- function(article) {

  synonyms <- .create_synonyms()

  a <- "Support"
  b <- .encase(synonyms$received)
  d <- .encase(synonyms$by)
  grep(paste(a, b, d, sep = synonyms$txt), article, perl = T)

}


#' Identify mentions of support
#'
#' Return the index of support statements such as: We are grateful to the
#'     National Institute of Mental Health (MH091070: PI’s Dante Cicchetti and
#'     Sheree L. Toth) for support of this work.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_8 <- function(article) {

  # TODO: This function was onboarded to capture a tiny proportion of
  #    statements. Consider (1) activating this function only if no hit by the
  #    more prevalent functions, (2) placing it in Acknowledgements (all missed
  #    article that prompted this are in the acknowledgements) and (3) to
  #    change this function into processing sentence by sentence to cover for
  #    all permutations of these terms.

  synonyms <- .create_synonyms()
  words <- c("foundation", "provide", "funding_financial", "for_of", "research")
  max_words <- .max_words(" ", n_max = 3, space_first = F)

  # foundation <-
  #   synonyms %>%
  #   magrittr::extract(words[1]) %>%
  #   lapply(.bound) %>%
  #   lapply(.encase)
  #
  # funding_for_research <-
  #   synonyms %>%
  #   magrittr::extract(words[2:4]) %>%
  #   lapply(.bound) %>%
  #   lapply(.encase) %>%
  #   paste(collapse = max_words)
  #
  #   for_research <-
  #     synonyms %>%
  #     magrittr::extract(words[4:5]) %>%
  #     lapply(.bound) %>%
  #     lapply(.encase) %>%
  #     paste(collapse = max_words)

  # c(foundation, funding_for_research) %>%
  #   paste(collapse = synonyms$txt) %>%
  #   grep(article, perl = T)

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}


#' Identify mentions of support
#'
#' Return the index of support statements such as: The US Environmental
#'     Protection Agency’s (USEPA) Office of Research and Development funded and
#'     managed the research described in the present study.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_9 <- function(article) {

  # Only done for words other than "supported" b/c this is very generic

  synonyms <- .create_synonyms()
  words <- c("foundation", "funded", "research")

  foundation <-
    synonyms %>%
    magrittr::extract(words[1]) %>%
    lapply(.bound) %>%
    lapply(.encase)

  funding <-
    synonyms %>%
    magrittr::extract(words[2]) %>%
    lapply(grep, pattern = "upport", value = T, invert = T) %>%
    lapply(.bound) %>%
    lapply(.encase)

  research <-
    synonyms %>%
    magrittr::extract(words[3]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste0(".{0,20}\\.")  # only match to words at the end of the sentence.

  c(foundation, funding, research) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}


#' Identify mentions of support
#'
#' Return the index of support statements such as: We are grateful to the
#'     National Institute of Mental Health (MH091070: PI’s Dante Cicchetti and
#'     Sheree L. Toth) for support of this work.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_support_10 <- function(article) {

  # TODO: This function was onboarded to capture a tiny proportion of
  #    statements. Consider (1) activating this function only if no hit by the
  #    more prevalent functions, (2) placing it in Acknowledgements (all missed
  #    article that prompted this are in the acknowledgements) and (3) to
  #    change this function into processing sentence by sentence to cover for
  #    all permutations of these terms.

  synonyms <- .create_synonyms()
  words <- c("thank", "foundation", "funding_financial", "research")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}


#' Identify mentions of developed
#'
#' Return the index of statements such as: This publication was developed
#'     under Assistance Agreement No. 83563701-0 awarded by the U.S.
#'     Environmental Protection Agency to the University of Michigan.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_developed_1 <- function(article) {

  # TODO: This function was onboarded to capture a tiny proportion of
  #    statements. Consider (1) activating this function only if no hit by the
  #    more prevalent functions, (2) placing it in Acknowledgements (all missed
  #    article that prompted this are in the acknowledgements) and (3) to
  #    change this function into processing sentence by sentence to cover for
  #    all permutations of these terms.

  synonyms <- .create_synonyms()
  words <- c("This", "research", "is", "developed", "by", "foundation")

  this_research_is <-
    synonyms %>%
    magrittr::extract(words[1:3]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt)

  developed <- words[4]

  by_foundation <-
    synonyms %>%
    magrittr::extract(words[5:6]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt)

  c(this_research_is, developed, by_foundation) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}

#' Identify mentions of received
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_received_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("received", "funds_award_financial", "by")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = " ") %>%
    grep(article, perl = T)
}


#' Identify mentions of received
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_received_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("received", "funding_financial", "by", "foundation_award")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of recepient
#'
#' Returns the index of mentions such as: "Recipient of National Institutes of
#'     Health Grants AG044552, AI121621, and DK112365.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_recipient_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("recipient", "award")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of "the authors ... financial support
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_authors_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("This", "author", "funds_award_financial")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)


  # "[Ss]upport",
  # "[Ff]inancial assistance",
  # "\\b[Aa]id\\b",

  # The authors thank X for funding
  # The authors received no/did not receive financial support
  # The authors acknowledge the/disclosed receipt of support\\b/financial support of
  # The authors are supported by

  # The authors received no funds
  # The authors have no support or funding to report
}


#' Identify mentions of "the authors have no funding ..."
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_authors_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("This", "author", "have", "no", "funding_financial_award")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of "thank ... financial support
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_thank_1 <- function(article) {

  synonyms <- .create_synonyms()

  synonyms$financial <- c(synonyms$financial, "for supporting")
  words <- c("We", "thank", "financial")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of thank you statements
#'
#' Returns the index with the elements related to "The authors acknowledge
#'     Worldwide Cancer Research (AICR) 15-1002; Blueprint 282510; ..."
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_thank_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("thank")

  thank <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt)

  thank %>%
    paste("[0-9]{5}", sep = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of "funding for this study was..."
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_fund_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("funding_financial_award", "for", "research", "received")

  funding_for <-
    synonyms %>%
    magrittr::extract(words[1:2]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = " ")

  synonyms %>%
    magrittr::extract(words[3:4]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    paste(funding_for, ., sep = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Identify mentions of Funding titles
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_fund_2 <- function(article) {

  b <- integer()
  synonyms <- .create_synonyms()
  words <- c("funding_title")

  a <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.title) %>%
    lapply(.encase) %>%
    paste() %>%
    grep(article, perl = T)

  if (!!length(a)) {

    for (i in seq_along(a)) {

      if (is.na(article[a[i] + 1])) {

        b <- c(b, a[i])

      } else {

        if (nchar(article[a[i] + 1]) == 0) {

          b <- c(b, a[i], a[i] + 2)

        } else {

          b <- c(b, a[i], a[i] + 1)

        }
      }
    }

  } else {

    b <-
      synonyms %>%
      magrittr::extract(words) %>%
      lapply(.title, within_text = T) %>%
      lapply(.encase) %>%
      paste() %>%
      grep(article, perl = T)

  }
  return(b)
}


#' Identify mentions of Funding titles
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_fund_3 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("any_title")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.encase) %>%
    paste("[A-Z]") %>%
    grep(article, perl = T)

}


#' Identify mentions of funds in acknowledgements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_fund_acknow <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("funded", "funds_award_financial")

  funded_synonyms <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    unlist()

  c(funded_synonyms, "NIH (|\\()(R|P)[0-9]{2}", "awarded by") %>%
    .encase %>%
    grep(article, perl = T, ignore.case = T)

}


#' Identify mentions of funds in acknowledgements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_fund_acknow_new <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("acknowledge", "support_only", "grant|foundation|insititute|organization")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    unlist() %>%
    .encase %>%
    grep(article, perl = T, ignore.case = T)

}


# get_fund_acknow_new <- function(article) {
#
#   synonyms <- .create_synonyms()
#   words <- c("acknowledge")
#
#   a <- synonyms %>%
#     magrittr::extract(words) %>%
#     lapply(.bound) %>%
#     lapply(.encase)
#
#   grep(paste0(a, synonyms$txt, "[0-9]{3,10}"), article, perl = T)
#
# }


#' Identify mentions of "Supported by ..."
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_supported_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("Supported", "by")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = " ") %>%
    paste("[a-zA-Z]+") %>%
    grep(article, perl = T)

}

#' Identify mentions of Financial support titles
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_financial_1 <- function(article) {

  b <- integer()

  synonyms <- .create_synonyms()
  words <- c("financial_title")

  a <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.title) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste() %>%
    grep(article, perl = T)


  if (!!length(a)) {

    for (i in seq_along(a)) {

      if (is.na(article[a[i] + 1])) {

        b <- c(b, a[i])

      } else {

        if (nchar(article[a[i] + 1]) == 0) {

          b <- c(b, a[i], a[i] + 2)

        } else {

          b <- c(b, a[i], a[i] + 1)

        }
      }
    }

  } else {

    b <-
      synonyms %>%
      magrittr::extract(words) %>%
      lapply(.title, within_text = T) %>%
      lapply(.encase) %>%
      # lapply(.max_words) %>%
      paste() %>%
      grep(article, perl = T)

  }
  return(b)
}


#' Identify mentions of Financial support titles followed by specific text
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_financial_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("financial_title", "No")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = " ") %>%
    grep(article, perl = T)

}


#' Identify mentions of Financial support titles followed by specific text
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_financial_3 <- function(article) {


  synonyms <- .create_synonyms()
  words <- c("financial_title", "this", "research")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    # lapply(.max_words) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}


#' Identify mentions of Disclosure statements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_disclosure_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("disclosure_title")

  a <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.title) %>%
    lapply(.encase) %>%
    paste(collapse = "|") %>%
    grep(article, perl = T)

  out <- integer()

  if (!!length(a)) {

    for (i in seq_along(a)) {

      if (is.na(article[a[i] + 1])) {

        b <- integer()
        d <- integer()

      } else {

        if (nchar(article[a[i] + 1]) == 0) {
          b <- a[i] + 2
        } else {
          b <- a[i] + 1
        }

        d <-
          synonyms$funding_financial_award %>%
          lapply(.bound) %>%
          paste0(collapse = "|") %>%
          grep(article[b], perl = T)

      }

      if (!!length(d)) out <- c(out, a[i], b)
    }
  }
  return(out)
}


#' Identify mentions of Disclosure statements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_disclosure_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("disclosure_title", "funding_financial_award")

  disclosure <-
    synonyms %>%
    magrittr::extract(words[1]) %>%
    lapply(.title, within_text = T)

  funding <-
    synonyms %>%
    magrittr::extract(words[2]) %>%
    lapply(.bound)

  c(disclosure, funding) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}


#' Identify mentions of Financial support titles
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_grant_1 <- function(article) {

  # TODO: This whole function takes a LOT of time to run, but it is ONLY
  #    activated for "(Grant [Ss]ponsor(|s):)|)Contract [Gg]rant
  #    [Ss]ponsor(|s):). Consider replacing it with just this!!!

  b <- integer()

  synonyms <- .create_synonyms()
  words <- c("grant_title")

  a <-
    synonyms %>%
    magrittr::extract(words) %>%
    lapply(.title) %>%
    lapply(.encase) %>%
    paste() %>%
    grep(article, perl = T)


  if (!!length(a)) {

    for (i in seq_along(a)) {

      if (is.na(article[a[i] + 1])) {

        b <- c(b, a[i])

      } else {

        if (nchar(article[a[i] + 1]) == 0) {

          b <- c(b, a[i], a[i] + 2)

        } else {

          b <- c(b, a[i], a[i] + 1)

        }
      }
    }

  } else {

    # This is done to avoid mentions such as "Grant AK, Brown AZ, ..."
    grant <- c("G(?i)rant ", "^[A-Z](?i)\\w+ grant ", "Contract grant ")

    support <-
      synonyms %>%
      magrittr::extract(c("support", "funder")) %>%
      lapply(paste0, "(?-i)") %>%
      lapply(.title, within_text = T) %>%
      unlist()

    b <-
      grant %>%
      lapply(paste0, support) %>%
      unlist() %>%
      .encase() %>%
      grep(article, perl = T)

  }
  return(b)
}


#' Identify mentions of Grant numbers in the Funding/Acknowledgements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_french_1 <- function(article) {

  # This study was financed by... - avoiding specifics b/c of UTF-8 characters
  grep("Cette.*tude.*financ.*par", article, perl = T, ignore.case = T)

}


#' Identify mentions of funding in frence
#'
#' Returns the index of mentions such as: "Remerciements Cette étude a été
#'     financée par l’Institut de Recherches Médicales et d’Etudes des Plantes
#'     Médicinales (IMPM)."
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_project_acknow <- function(article) {

  grep("project (no|num)", article, perl = T, ignore.case = T)

}


#' Get common phrases
#'
#' Returns the index with the elements of interest.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_common_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("no", "funding_financial_award", "is", "received")

  no_funding <-
    synonyms %>%
    magrittr::extract(words[1:2]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = " ")

  was_received <-
    synonyms %>%
    magrittr::extract(words[3:4]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = " ")

  no_funding %>%
    paste(was_received, sep = synonyms$txt) %>%
    grep(article, perl = T)
}



#' Get common phrases
#'
#' Returns the index with the elements of interest.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_common_2 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("No", "funding_financial_award", "received")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)
}


#' Get common phrases
#'
#' Identify statements of the following type: "All authors are required to
#'     disclose all affiliations, funding sources and financial or management
#'     relationships that could be perceived as potential sources of bias. The
#'     authors disclosed none.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_common_3 <- function(article) {

  grep("required to disclose.*disclosed none", article)

}


#' Get common phrases
#'
#' Identify statements of the following type: "There were no external funding
#'     sources for this study"
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_common_4 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("no", "funding_financial_award", "for", "this", "research")

  no_funding <-
    synonyms %>%
    magrittr::extract(words[1:2]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt)

  for_this <-
    synonyms %>%
    magrittr::extract(words[3:4]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = " ")

  research <-
    synonyms %>%
    magrittr::extract(words[5]) %>%
    lapply(.bound) %>%
    lapply(.encase)

  c(no_funding, for_this, research) %>%
    paste(collapse = synonyms$txt) %>%
    grep(article, perl = T)

}


#' Get common phrases
#'
#' Identify statements of the following type: "No specific sources of funding."
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_common_5 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("no", "sources", "funding_financial")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = .max_words(" ", space_first = F)) %>%
    grep(article, perl = T)

}


#' Identify mentions of "Acknowledgement and"
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_acknow_1 <- function(article) {

  b <- integer()

  txt_0 <- "^Acknowledg(|e)ment(|s)"
  txt_1 <- "(of|and)"
  txt_2 <- "([Ss]upport |\\b[Ff]unding|\\b[Ff]inancial)"

  total_txt <- c(txt_0, txt_1, txt_2)
  indicator_regex <- paste0(total_txt, collapse = " ")

  a <- grep(indicator_regex, article, perl = T, ignore.case = T)

  if (!!length(a)) {

    for (i in seq_along(a)) {

      if (is.na(article[a[i] + 1])) {

        b <- c(b, a)

      } else {

        if (nchar(article[a[i] + 1]) == 0) {

          b <- c(b, a[i], a[i] + 2)

        } else {

          b <- c(b, a[i], a[i] + 1)
        }

      }
    }
  }
  return(b)
}


#' Identify acknowledgements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
get_acknow_2 <- function(article) {

  txt_0 <- "(^A(?i)cknowledg(|e)ment(|s)(?-i))"

  txt_1 <- "(^Acknowledg(|e)ment(|s)"
  txt_2 <- "(of|and)"
  txt_3 <- "([Ss]upport |\\b[Ff]unding|\\b[Ff]inancial))"

  total_txt <- c(txt_1, txt_2, txt_3)
  indicator_regex <- paste0(total_txt, collapse = " ")
  indicator_regex <- paste(txt_0, indicator_regex, sep = "|")

  grep(indicator_regex, article, perl = T)
}


#' Identify funding titles using XML labels
#'
#' Extract XML titles related to funding and all text children.
#'
#' @param article_xml The text as an xml_document.
#' @return The title and its related text as a string.
.get_fund_pmc_title <- function(article_xml) {

  synonyms <- .create_synonyms()
  b <- ""

  fund_titles <- .encase(synonyms$any_title)

  # back_xpath <-
  #   c(
  #     "back/ack//*[self::title or self::bold or self::italic]",
  #     "back/fn-group//*[self::title or self::bold or self::italic]",
  #     "back/notes//*[self::title or self::bold or self::italic]"
  #   ) %>%
  #   paste(collapse = " | ")

  # If I had not stripped the d1 namespace:
  # "back//fn-group//*[self::d1:title or self::d1:bold or self::d1:italic]"
  # back_matter <-
  #   article_xml %>%
  #   xml2::xml_find_all(back_xpath)

  back_matter <-
    article_xml %>%
    xml2::xml_find_all("back/*[not(name()='ref-list')]") %>%
    xml2::xml_find_all(".//*[self::title or self::bold or self::italic or self::sup]")

  a <-
    back_matter %>%
    xml2::xml_text() %>%
    stringr::str_which(fund_titles)

  if (!!length(a)) {

    b <-
      back_matter %>%
      magrittr::extract(a) %>%
      xml2::xml_parent() %>%
      xml2::xml_contents() %>%
      xml2::xml_text() %>%
      paste(collapse = " ")

    return(b)
  }


  front_matter <-
    article_xml %>%
    xml2::xml_find_all("front//fn//*[self::title or self::bold or self::italic or self::sup]")

  a <-
    front_matter %>%
    xml2::xml_text() %>%
    stringr::str_which(fund_titles)

  if (!!length(a)) {

    b <-
      front_matter %>%
      magrittr::extract(a) %>%
      xml2::xml_parent() %>%
      xml2::xml_contents() %>%
      xml2::xml_text() %>%
      paste(collapse = " ")

    return(b)
  }


  fund_titles <- .title_strict(synonyms$any_title) %>% .encase()

  body_matter <-
    article_xml %>%
    xml2::xml_find_all("body/sec//*[self::title or self::bold or self::italic or self::sup]")

  a <-
    body_matter %>%
    xml2::xml_text() %>%
    stringr::str_which(fund_titles)

  if (!!length(a)) {

    b <-
      body_matter %>%
      magrittr::extract(a) %>%
      xml2::xml_parent() %>%
      xml2::xml_contents() %>%
      xml2::xml_text() %>%
      paste(collapse = " ")

    return(b)
  }

  return(b)
}


#' Identify funding group elements in NLM XML files
#'
#' Identify and return the text found in funding group elements of NLM XML
#'     files. Articles with funding-group elements also have a Funding
#'     indication on PubMed.
#'
#' @param article_xml An NLM XML as an xml_document.
#' @return The text of interest as a list indicating whether it was found, the
#'     quoated institutes and the actual statement as a string.
.get_fund_pmc_group <- function(article_xml) {

  fund_pmc <- list(
    is_fund_group_pmc = FALSE,
    fund_statement_pmc = "",
    fund_institute_pmc = "",
    fund_source_pmc = ""
  )

  fund_group <-
    article_xml %>%
    xml2::xml_find_all("front/article-meta/funding-group")

  if (!!length(fund_group)) {

    fund_pmc$is_fund_group_pmc <- TRUE

    fund_pmc$fund_statement_pmc <-
      fund_group %>%
      xml2::xml_find_all(".//funding-statement") %>%
      xml2::xml_text() %>%
      paste(collapse = "")

    fund_pmc$fund_institute_pmc <-
      fund_group %>%
      xml2::xml_find_all(".//institution") %>%
      xml2::xml_contents() %>%
      xml2::xml_text() %>%
      paste(collapse = "; ")

    fund_pmc$fund_source_pmc <-
      fund_group %>%
      xml2::xml_find_all(".//funding-source") %>%
      xml2::xml_text() %>%
      paste(collapse = "; ")
  }


  if (with(fund_pmc, fund_institute_pmc == "N/A" & fund_statement_pmc == "")) {

    fund_pmc$is_fund_group_pmc <- FALSE

  }

  return(fund_pmc)
}


#' Identify funding source elements in NLM XML files
#'
#' Identify and return the text found in funding source elements of NLM XML
#'     files. Some funding sources are not included in the funding-group
#'     elements - these are the sources that this function attempts to capture.
#'     It seems that all articles with such elements also have a Funding
#'     statement on PubMed.
#'
#' @param article_xml An NLM XML as an xml_document.
#' @return The text of interest as a string.
.get_fund_pmc_source <- function(article_xml) {

  article_xml %>%
    xml2::xml_find_all("body//funding-source | back//funding-source") %>%
    xml2::xml_contents() %>%
    xml2::xml_text() %>%
    paste(collapse = "; ")

}


#' Avoid disclosures that are in fact COI statements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
negate_disclosure_1 <- function(article) {

  synonyms <- .create_synonyms()

  txt <- "[a-zA-Z0-9\\s,()-]*"  # order matters

  disclose_synonyms <- c(
    "[Dd]isclose(|s)(|:|\\.)",
    "[Dd]isclosure(|s)(|:|\\.)"
  )

  conflict_synonyms <- c(
    "conflict(|s) of interest",
    "conflicting interest",
    "conflicting financial interest",
    "conflicting of interest",
    "conflits d'int",
    "conflictos de Inter"
  )

  compete_synonyms <- c(
    "competing interest",
    "competing of interest",
    "competing financial interest"
  )

  and_synonyms <- c(
    "and",
    "&",
    "or"
  )

  not_synonyms <- c(
    "not"
  )

  funded_synonyms <- c(
    "\\bfunded",
    "\\bfinanced",
    "\\bsupported",
    "\\bsponsored",
    "\\bresourced"
  )

  disclose <- .encase(disclose_synonyms)
  conflict <- .encase(c(conflict_synonyms, compete_synonyms))
  and <- .encase(and_synonyms)
  not <- .encase(not_synonyms)
  funded <- .encase(funded_synonyms)

  regex <- paste(disclose, conflict, and, not, funded, sep = txt)
  a <- grepl(regex, article, perl = T)

  if (any(a)) {

    return(a)

  } else {

    funded <- .encase(c(funded_synonyms, synonyms$funding))
    regex <- paste(disclose, funded, conflict, sep = txt)
    grepl(regex, article, perl = T)
  }
}


#' Identify "Funding disclosure. Nothing to declare" statements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
negate_disclosure_2 <- function(article) {

  synonyms <- .create_synonyms()

  Disclosure_synonyms <- c(
    "F(?i)inancial disclosure(|s)(?-i)(|:|\\.)",
    "F(?i)inancial declaration(|s)(?-i)(|:|\\.)",
    "Disclosure(|:|\\.)",
    "Declaration(|:|\\.)"
  )

  disclosure_synonyms <- c(
    "financial disclosure(|s)",
    "financial declaration(|s)",
    "disclosure",
    "declaration"
  )

  disclose_synonyms <- c(
    "to disclose",
    "to declare",
    "to report"
  )

  Disclosure <- .encase(Disclosure_synonyms)
  disclosure <- .encase(disclosure_synonyms)
  disclose   <- .encase(disclose_synonyms)
  no <- .encase(synonyms$No)
  no_1 <- "(no|not have any)"
  no_2 <- "(nil|nothing)"

  regex_1 <-  # Financial disclosure: Nothing to declare
    paste(Disclosure, no) %>%
    .encase()
  regex_2 <-  # Financial disclosure: The authors have no financial disclosures
    paste(Disclosure, paste(no_1, disclosure), sep = synonyms$txt) %>%
    .encase()
  regex_3 <-  # Financial disclosure: The author has nothing to declare
    paste(Disclosure, paste(no_2, disclose), sep = synonyms$txt) %>%
    .encase()

  regex <- paste(regex_1, regex_2, regex_3, sep = "|")
  grepl(regex, article, perl = T)

}


#' Avoid financial that is part of COI statements
#'
#' Returns the index with the elements of interest. More generic than _1.
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
negate_conflict_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("conflict_title")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.title) %>%
    lapply(.encase) %>%
    paste() %>%
    grepl(article, perl = T)
}


#' Avoid mentions that no funding information was provided
#'
#'
#' @param article A List with paragraphs of interest.
#' @return The index of the paragraph of interest.
negate_absence_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("No", "info", "of", "funding_financial_award", "is", "received")

  synonyms %>%
    magrittr::extract(words) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    grepl(article, perl = T)

}


#' Remove mentions of COIs that may cause FPs
#'
#' Returns the text without potentially misleading mentions of COIs.
#'
#' @param article A List with paragraphs of interest.
#' @return The list of paragraphs without mentions of financial COIs.
obliterate_conflict_1 <- function(article) {

  # Good for finding, but not for substituting b/c it's  a lookahead
  # words <- c(
  #   # positive lookahead makes these phrases interchangeable
  #   "(?=[a-zA-Z0-9\\s,()-]*(financial|support))",
  #   "(?=[a-zA-Z0-9\\s,()-]*(conflict|competing))"
  # )

  synonyms <- .create_synonyms()

  financial_1 <- "(funding|financial|support)"
  financial_2 <- "(financial|support)"
  relationship <- synonyms$relationship %>% .bound() %>% .encase()
  conflict <- synonyms$conflict %>% .bound() %>% .encase()
  financial_interest <- "(financial(?:\\s+\\w+){0,3} interest)"

  regex_1 <- paste(financial_1, relationship, conflict, sep = synonyms$txt)
  regex_2 <- paste(conflict, relationship, financial_2, sep = synonyms$txt)
  regex_3 <- paste(relationship, financial_interest,  sep = synonyms$txt)

  c(regex_1, regex_2, regex_3) %>%
    lapply(.encase) %>%
    paste(collapse = "|") %>%
    gsub("", article, perl = T)

}


#' Remove COI statements with language relevant to funding statements
#'
#' Returns the text without mentions of COIs related to industry ties.
#'
#' @param article A List with paragraphs of interest.
#' @return The list of paragraphs without mentions of financial COIs.
obliterate_conflict_2 <- function(article) {

  # Good for finding, but not for substituting b/c it's  a lookahead
  # words <- c(
  #   # positive lookahead makes these phrases interchangeable
  #   "(?=[a-zA-Z0-9\\s,()-]*(financial|support))",
  #   "(?=[a-zA-Z0-9\\s,()-]*(conflict|competing))"
  # )
  synonyms <- .create_synonyms()

  industry_words <- c(
    "\\bconsult(ant|ing)\\b",
    "\\bhonorari(um|a)\\b",
    "\\bspeaker\\b",
    "\\badvisory board(|s)\\b",
    "\\bfee(|s)\\b"
  )

  industry_words %>%
    .encase %>%
    paste("(\\.|;|$)", sep = ".*") %>%
    gsub("", article, perl = T, ignore.case = T)

}


#' Remove mentions of commonly misleading funding statements
#'
#' Returns the text without potentially misleading phrases related to funding.
#'
#' @param article A List with paragraphs of interest.
#' @return The list of paragraphs without mentions of financial COIs.
obliterate_misleading_fund_1 <- function(article) {

  misleading_fund <- c(
    "travel grant(|s)",
    "grant(|s) for travel(|ling)",
    "Fund(|s) Collection"
  )

  misleading_fund %>%
    .encase %>%
    gsub("", article, perl = T)

}



#' Remove disclosures with inappropriate sentences
#'
#' Returns the text without potentially misleading disclsoures mentions. This
#'     is intended to solve problems with disclosures, such as: Disclosure.
#'     Authors have no conflict of interests, and the work was not supported
#'     or funded by any drug company. This project was funded by the Deanship
#'     of Scientific Research, King Abdulaziz University, Jeddah, Saudi Arabia
#'     (Grant No. 4/165/1431);
#'
#' @param article A List with paragraphs of interest.
#' @return The list of paragraphs without mentions of financial COIs.
obliterate_disclosure_1 <- function(article) {

  synonyms <- .create_synonyms()
  words <- c("disclosure_title", "conflict", "and", "not", "funded")

  # disclosure_title <-
  #   synonyms %>%
  #   magrittr::extract(words[1]) %>%
  #   lapply(.title) %>%
  #   lapply(.encase)

  synonyms %>%
    magrittr::extract(words[2:5]) %>%
    lapply(.bound) %>%
    lapply(.encase) %>%
    paste(collapse = synonyms$txt) %>%
    paste0(synonyms$txt, ., synonyms$txt, "($|.)") %>%
    gsub("", article, perl = T)

  # disclosure_title %>%
  #   paste0("(", synonyms$txt, conflict_funded, synonyms$txt, ")") %>%
  #   gsub("\\1", article, perl = T)


  # if (any(a)) {
  #
  #   return(a)
  #
  # } else {
  #
  #   funded <- .encase(c(funded_synonyms, synonyms$funding))
  #   regex <- paste(disclose, funded, conflict, sep = txt)
  #   grepl(regex, article, perl = T)
  # }

}


#' Identify and extract Funding statements in PMC XML files.
#'
#' Takes a PMC XML file as a list of strings and returns data related to the
#'     presence of a Funding statement, including whether a Funding statement
#'     exists. If a Funding statement exists, it extracts it. This is a modified
#'     version of the `rt_fund_pmc` designed for integration with `rt_all_pmc`.
#'
#' @param article_ls A PMC XML as a list of strings.
#' @param pmc_fund_ls A list of results from the `.get_fund_pmc` function.
#' @return A dataframe of results.
.rt_fund_pmc <- function(article_ls, pmc_fund_ls) {

  # TODO Update to match format of rt_coi_pmc.

  index <- integer()

  # Way faster than index_any[["reg_title_pmc"]] <- NA
  index_any <- list(
    support_1 = NA,
    support_3 = NA,
    support_4 = NA,
    support_5 = NA,
    support_6 = NA,
    support_7 = NA,
    support_8 = NA,
    support_9 = NA,
    support_10 = NA,
    developed_1 = NA,
    received_1 = NA,
    received_2 = NA,
    recipient_1 = NA,
    authors_1 = NA,
    authors_2 = NA,
    thank_1 = NA,
    thank_2 = NA,
    fund_1 = NA,
    fund_2 = NA,
    fund_3 = NA,
    supported_1 = NA,
    financial_1 = NA,
    financial_2 = NA,
    financial_3 = NA,
    grant_1 = NA,
    french_1 = NA,
    common_1 = NA,
    common_2 = NA,
    common_3 = NA,
    common_4 = NA,
    common_5 = NA,
    acknow_1 = NA,
    disclosure_1 = NA,
    disclosure_2 = NA
  )

  index_ack <- list(
    fund_ack = NA,
    project_ack = NA
  )

  relevance_ls <- list(
    is_relevant_fund = NA
  )

  out <- list(
    is_fund_pred = FALSE,
    fund_text = "",
    is_explicit_fund = NA
  )

  # True only if a fund-statement or fund-title was found
  if (pmc_fund_ls$is_fund_pred) {

    out$is_fund_pred <- TRUE
    out$fund_text <- pmc_fund_ls$fund_text

    if (!is.na(pmc_fund_ls$is_fund_pmc_title)) {

      out$is_explicit_fund <- TRUE

    }

    return(c(relevance_ls, out, index_any, index_ack))
  }

  # TODO Consider adding unique
  article <-
    article_ls[c("ack", "body", "footnotes")] %>%
    unlist() # %>%
  # unique()


  # Check relevance
  # TODO Consider adding Department
  fund_regex <- "fund|support|financ|receive|grant|none|sponsor|fellowship|Association|Institute|National|Foundation"
  article %<>% purrr::keep(~ str_detect(.x, regex(fund_regex, ignore_case = T)))

  relevance_ls$is_relevant_fund <- !!length(article)

  # Check for relevance
  if (!relevance_ls$is_relevant_fund) {

    return(c(relevance_ls, out, index_any, index_ack))

  }


  article_processed <-
    article %>%
    iconv(from = 'UTF-8', to = 'ASCII//TRANSLIT', sub = "") %>%   # keep first
    trimws() %>%
    .obliterate_fullstop_1() %>%
    .obliterate_semicolon_1() %>%  # adds minimal overhead
    .obliterate_comma_1() %>%   # adds minimal overhead
    .obliterate_apostrophe_1() %>%
    .obliterate_punct_1() %>%
    .obliterate_line_break_1() %>%
    obliterate_conflict_1() %>%
    obliterate_conflict_2() %>%
    obliterate_disclosure_1() %>%   # Adds 30s overhead!
    obliterate_misleading_fund_1()


  # Identify sequences of interest
  index_any$support_1 <- get_support_1(article_processed)
  index_any$support_3 <- get_support_3(article_processed)
  index_any$support_4 <- get_support_4(article_processed)
  index_any$support_5 <- get_support_5(article_processed)
  index_any$support_6 <- get_support_6(article_processed)
  index_any$support_7 <- get_support_7(article_processed)
  index_any$support_8 <- get_support_8(article_processed)
  index_any$support_9 <- get_support_9(article_processed)
  index_any$support_10 <- get_support_10(article_processed)
  index_any$developed_1 <- get_developed_1(article_processed)
  index_any$received_1 <- get_received_1(article_processed)
  index_any$received_2 <- get_received_2(article_processed)
  index_any$recipient_1 <- get_recipient_1(article_processed)
  index_any$authors_1 <- get_authors_1(article_processed)
  index_any$authors_2 <- get_authors_2(article_processed)
  index_any$thank_1 <- get_thank_1(article_processed)
  index_any$thank_2 <- get_thank_2(article_processed)
  index_any$fund_1 <- get_fund_1(article_processed)
  index_any$fund_2 <- get_fund_2(article_processed)
  index_any$fund_3 <- get_fund_3(article_processed)
  index_any$supported_1 <- get_supported_1(article_processed)
  index_any$financial_1 <- get_financial_1(article_processed)
  index_any$financial_2 <- get_financial_2(article_processed)
  index_any$financial_3 <- get_financial_3(article_processed)
  index_any$grant_1 <- get_grant_1(article_processed)
  index_any$french_1 <- get_french_1(article_processed)
  index_any$common_1 <- get_common_1(article_processed)
  index_any$common_2 <- get_common_2(article_processed)
  index_any$common_3 <- get_common_3(article_processed)
  index_any$common_4 <- get_common_4(article_processed)
  index_any$common_5 <- get_common_5(article_processed)
  index_any$acknow_1 <- get_acknow_1(article_processed)
  index_any$disclosure_1 <- get_disclosure_1(article_processed)
  index_any$disclosure_2 <- get_disclosure_2(article_processed)

  index <- unlist(index_any) %>% unique() %>% sort()

  # Remove potential mistakes
  if (!!length(index)) {

    # Funding info can be within COI statements, as per Ioannidis
    # Comment out until problems arise
    # if (length(unlist(index_any[c("authors_2")]))) {
    #   is_coi <- negate_conflict_1(article_processed[min(index) - 1])
    #   index <- index[!is_coi]
    # }


    # Difficult to make it work properly because it does not
    # disclosures <- c("disclosure_1", "disclosure_2")
    # if (!!length(unlist(index_any[disclosures]))) {
    #
    #   for (i in seq_along(disclosures)) {
    #
    #     ind <- index_any[[disclosures[i]]]
    #     is_coi_disclosure <- negate_disclosure_1(article_processed[ind])
    #     index_any[[disclosures[i]]] <- ind[!is_coi_disclosure]
    #
    #   }
    #
    # }

    is_absent <- negate_absence_1(article_processed[index])
    index <- index[!is_absent]

    # Currently removed b/c I made the disclosure functions more robust to
    #     statements like "Financial disclosure. Nothing to disclose.
    # disclosures <- unique(unlist(index_any[c("disclosure_1", "disclosure_2")]))
    # if (!!length(disclosures)) {
    #
    #   if (length(disclosures) == 1) {
    #
    #     is_disclosure <- negate_disclosure_2(paragraphs[index])
    #     index <- index[!is_disclosure]
    #
    #   } else {
    #
    #     disclosure_text <- paste(article_processed[disclosures], collapse = " ")
    #     is_disclosure <- negate_disclosure_2(disclosure_text)
    #     index <- setdiff(index, disclosures)
    #
    #   }
    # }
  }

  if (!!length(index)) {

    out$is_explicit_fund <- !!length(unlist(index_any))
    out$is_fund_pred <- !!length(index)
    out$fund_text <- article[index] %>% paste(collapse = " ")

    index_any %<>% purrr::map(function(x) !!length(x))

    return(c(relevance_ls, out, index_any, index_ack))
  }


  # Identify potentially missed signals
  i <- which(article %in% c(article_ls$ack, article_ls$footnotes))

  if (!!length(i)) {

    index_ack$fund_ack <- get_fund_acknow(article_processed[i])
    index_ack$project_ack <- get_project_acknow(article_processed[i])

    index <- i[unlist(index_ack) %>% unique() %>% sort()]
    index_ack %<>% purrr::map(function(x) !!length(x))

  }


  out$is_fund_pred <- !!length(index)
  out$fund_text <- article[index] %>% paste(collapse = " ")

  index_any %<>% purrr::map(function(x) !!length(x))

  if (out$is_fund_pred) {

    out$is_explicit_fund <- FALSE

  }

  # Placed here to give a chance to the title to populate the statement field
  if (!out$is_fund_pred & pmc_fund_ls$is_fund_pmc_group) {

    out$is_fund_pred <- TRUE
    out$fund_text <- pmc_fund_ls$fund_pmc_source
    out$is_explicit_fund <- FALSE

    return(c(relevance_ls, out, index_any, index_ack))

  }

  if (!out$is_fund_pred & pmc_fund_ls$is_fund_pmc_anysource) {

    out$is_fund_pred <- TRUE
    out$fund_text <- pmc_fund_ls$fund_pmc_anysource
    out$is_explicit_fund <- FALSE

  }

  return(c(relevance_ls, out, index_any, index_ack))
}



#' Identify and extract Funding statements in PMC XML files.
#'
#' Takes a PMC XML file and returns data related to the
#'     presence of a Funding statement, including whether a Funding statement
#'     exists. If a Funding statement exists, it extracts it.
#'
#' @param filename The name of the PMC XML as a string.
#' @param remove_ns TRUE if an XML namespace exists, else FALSE (default).
#' @return A dataframe of results. It returns all unique article identifiers,
#'     whether this article was deemed relevant to funding (e.g. was the word
#'     "fund" found within the text), whether a funding statement was found,
#'     whether a statement within the PMC tags dedicated to funding was found,
#'     the text identified, whether this text is explicit (i.e. whether it
#'     clearly indicated that funding was received) and whether each of the
#'     labeling functions identified the text or not. The functions are
#'     returned to add flexibility in how this package is used; for example,
#'     future definitions of Funding may differ from the one we used.
#' @examples
#' \dontrun{
#' # Path to PMC XML.
#' filepath <- "../inst/extdata/00003-PMID26637448-PMC4737611.xml"
#'
#' # Identify and extract meta-data and indicators of transparency.
#' results_table <- rt_fund_pmc(filepath, remove_ns = T)
#' }
#' @export
rt_fund_pmc <- function(filename, remove_ns = F) {

  # TODO Update to match format of rt_coi_pmc.

  xpath <- c(
    "front/article-meta/article-id[@pub-id-type = 'pmid']",
    "front/article-meta/article-id[@pub-id-type = 'pmc']",
    "front/article-meta/article-id[@pub-id-type = 'pmc-uid']",
    "front/article-meta/article-id[@pub-id-type = 'doi']"
  )

  index <- integer()

  # Way faster than index_any[["reg_title_pmc"]] <- NA
  index_any <- list(
    fund_group_pmc = NA,
    fund_title_pmc = NA,
    support_1 = NA,
    support_3 = NA,
    support_4 = NA,
    support_5 = NA,
    support_6 = NA,
    support_7 = NA,
    support_8 = NA,
    support_9 = NA,
    support_10 = NA,
    developed_1 = NA,
    received_1 = NA,
    received_2 = NA,
    recipient_1 = NA,
    authors_1 = NA,
    authors_2 = NA,
    thank_1 = NA,
    thank_2 = NA,
    fund_1 = NA,
    fund_2 = NA,
    fund_3 = NA,
    supported_1 = NA,
    financial_1 = NA,
    financial_2 = NA,
    financial_3 = NA,
    grant_1 = NA,
    french_1 = NA,
    common_1 = NA,
    common_2 = NA,
    common_3 = NA,
    common_4 = NA,
    common_5 = NA,
    acknow_1 = NA,
    disclosure_1 = NA,
    disclosure_2 = NA
  )

  index_ack <- list(
    fund_ack = NA,
    project_ack = NA
  )

  out <- list(
    pmid = NA,
    pmcid_pmc = NA,
    pmcid_uid = NA,
    doi = NA,
    is_relevant = NA,
    is_fund_pred = FALSE,
    is_fund_group_pmc = NA,
    fund_statement_pmc = "",
    fund_institute_pmc = "",
    fund_source_pmc = "",
    fund_anysource_pmc = "",
    fund_text = "",
    is_explicit = NA
  )


  # A lot of the PMC XML files are malformed
  article_xml <- tryCatch(.get_xml(filename, remove_ns), error = function(e) e)

  if (inherits(article_xml, "error")) {

    return(tibble::tibble(filename, is_success = F))

  }


  # Extract IDs
  out %<>% purrr::list_modify(!!!purrr::map(xpath, ~ .get_text(article_xml, .x, T)))


  # Capture fund-group elements
  fund_group_pmc <- .get_fund_pmc_group(article_xml)

  if (fund_group_pmc$is_fund_group_pmc) {

    index_any$fund_group_pmc <- TRUE
    out %<>% purrr::list_modify(!!!fund_group_pmc)

    out$is_relevant <- TRUE
    out$is_fund_pred <- TRUE

    if (nchar(fund_group_pmc$fund_statement_pmc) > 0) {
      return(tibble::as_tibble(c(out, index_any, index_ack)))
    }
  }
  out$is_fund_group_pmc <- FALSE
  index_any$fund_group_pmc <- FALSE


  # Capture missed fund-source elements
  out$fund_anysource_pmc <- .get_fund_pmc_source(article_xml)


  # Go through titles
  title_txt <- .get_fund_pmc_title(article_xml)
  is_title <- !!length(title_txt)

  if (is_title) {

    index_any$fund_title_pmc <- TRUE
    out$fund_text <- title_txt

    out$is_relevant <- TRUE
    out$is_explicit <- TRUE
    out$is_fund_pred <- TRUE

    return(tibble::as_tibble(c(out, index_any, index_ack)))

  }


  # Extract article text into a vector
  # .xml_preprocess(article_xml)  # 5x faster to obliterate within each section
  ack <- .xml_ack(article_xml)
  body <- .xml_body(article_xml, get_last_two = T)
  footnotes <- .xml_footnotes(article_xml) %>% .obliterate_contribs()
  article <- c(footnotes, body, ack)


  # Check relevance
  rel_regex <- "fund|support|financ|receive|grant|none|sponsor|fellowship"
  article %<>% purrr::keep(~ str_detect(.x, regex(rel_regex, ignore_case = T)))

  out$is_relevant <- !!length(article)

  # Check for relevance
  if (!out$is_relevant) {

    return(tibble::as_tibble(c(out, index_any, index_ack)))

  }


  # Text pre-processing
  article_processed <-
    article %>%
    iconv(from = 'UTF-8', to = 'ASCII//TRANSLIT', sub = "") %>%   # keep first
    .obliterate_fullstop_1() %>%
    .obliterate_semicolon_1() %>%  # adds minimal overhead
    .obliterate_comma_1() %>%   # adds minimal overhead
    .obliterate_apostrophe_1() %>%
    .obliterate_punct_1() %>%
    .obliterate_line_break_1() %>%
    obliterate_conflict_1() %>%
    obliterate_conflict_2() %>%
    obliterate_disclosure_1() %>%   # Adds 30s overhead!
    obliterate_misleading_fund_1()


  # Identify sequences of interest
  index_any$fund_group_pmc <- integer()
  index_any$fund_title_pmc <- integer()
  index_any$support_1 <- get_support_1(article_processed)
  index_any$support_3 <- get_support_3(article_processed)
  index_any$support_4 <- get_support_4(article_processed)
  index_any$support_5 <- get_support_5(article_processed)
  index_any$support_6 <- get_support_6(article_processed)
  index_any$support_7 <- get_support_7(article_processed)
  index_any$support_8 <- get_support_8(article_processed)
  index_any$support_9 <- get_support_9(article_processed)
  index_any$support_10 <- get_support_10(article_processed)
  index_any$developed_1 <- get_developed_1(article_processed)
  index_any$received_1 <- get_received_1(article_processed)
  index_any$received_2 <- get_received_2(article_processed)
  index_any$recipient_1 <- get_recipient_1(article_processed)
  index_any$authors_1 <- get_authors_1(article_processed)
  index_any$authors_2 <- get_authors_2(article_processed)
  index_any$thank_1 <- get_thank_1(article_processed)
  index_any$thank_2 <- get_thank_2(article_processed)
  index_any$fund_1 <- get_fund_1(article_processed)
  index_any$fund_2 <- get_fund_2(article_processed)
  index_any$fund_3 <- get_fund_3(article_processed)
  index_any$supported_1 <- get_supported_1(article_processed)
  index_any$financial_1 <- get_financial_1(article_processed)
  index_any$financial_2 <- get_financial_2(article_processed)
  index_any$financial_3 <- get_financial_3(article_processed)
  index_any$grant_1 <- get_grant_1(article_processed)
  index_any$french_1 <- get_french_1(article_processed)
  index_any$common_1 <- get_common_1(article_processed)
  index_any$common_2 <- get_common_2(article_processed)
  index_any$common_3 <- get_common_3(article_processed)
  index_any$common_4 <- get_common_4(article_processed)
  index_any$common_5 <- get_common_5(article_processed)
  index_any$acknow_1 <- get_acknow_1(article_processed)
  index_any$disclosure_1 <- get_disclosure_1(article_processed)
  index_any$disclosure_2 <- get_disclosure_2(article_processed)

  index <- unlist(index_any) %>% unique() %>% sort()

  # Remove potential mistakes
  if (!!length(index)) {

    # Funding info can be within COI statements, as per Ioannidis
    # Comment out until problems arise
    # if (length(unlist(index_any[c("authors_2")]))) {
    #   is_coi <- negate_conflict_1(article_processed[min(index) - 1])
    #   index <- index[!is_coi]
    # }


    # Difficult to make it work properly because it does not
    # disclosures <- c("disclosure_1", "disclosure_2")
    # if (!!length(unlist(index_any[disclosures]))) {
    #
    #   for (i in seq_along(disclosures)) {
    #
    #     ind <- index_any[[disclosures[i]]]
    #     is_coi_disclosure <- negate_disclosure_1(article_processed[ind])
    #     index_any[[disclosures[i]]] <- ind[!is_coi_disclosure]
    #
    #   }
    #
    # }

    is_absent <- negate_absence_1(article_processed[index])
    index <- index[!is_absent]

    # Currently removed b/c I made the disclosure functions more robust to
    #     statements like "Financial disclosure. Nothing to disclose.
    # disclosures <- unique(unlist(index_any[c("disclosure_1", "disclosure_2")]))
    # if (!!length(disclosures)) {
    #
    #   if (length(disclosures) == 1) {
    #
    #     is_disclosure <- negate_disclosure_2(paragraphs[index])
    #     index <- index[!is_disclosure]
    #
    #   } else {
    #
    #     disclosure_text <- paste(article_processed[disclosures], collapse = " ")
    #     is_disclosure <- negate_disclosure_2(disclosure_text)
    #     index <- setdiff(index, disclosures)
    #
    #   }
    # }
  }

  if (!!length(index)) {

    out$is_explicit <- !!length(unlist(index_any))
    out$is_fund_pred <- !!length(index)
    out$fund_text <- article[index] %>% paste(collapse = " ")

    index_any %<>% purrr::map(function(x) !!length(x))

    return(tibble::as_tibble(c(out, index_any, index_ack)))
  }


  # Identify potentially missed signals
  i <- which(article %in% c(ack, footnotes))

  if (!!length(i)) {

    index_ack$fund_ack <- get_fund_acknow(article_processed[i])
    index_ack$project_ack <- get_project_acknow(article_processed[i])

    index <- i[unlist(index_ack) %>% unique() %>% sort()]
    index_ack %<>% purrr::map(function(x) !!length(x))

  }


  out$is_fund_pred <- !!length(index)
  out$fund_text <- article[index] %>% paste(collapse = " ")

  index_any %<>% purrr::map(function(x) !!length(x))

  if (out$is_fund_pred) {

    out$is_explicit <- FALSE

  }

  # Placed here to give a chance to the title to populate the statement field
  if (!out$is_fund_pred & fund_group_pmc$is_fund_group_pmc) {

    index_any$fund_group_pmc <- TRUE
    out$is_fund_pred <- TRUE

    return(tibble::as_tibble(c(out, index_any, index_ack)))

  }

  if (!out$is_fund_pred & nchar(out$fund_anysource_pmc) > 0) {

    out$is_fund_pred <- TRUE

  }

  return(tibble::as_tibble(c(out, index_any, index_ack)))
}