1_parse.R

# 20/06/2015
# Dom Bennett
# Parsing abstracts into 'abstracts/'

# DIRS
data.dir <- '0_data'
out.dir <- '1_parsed'
if (!file.exists (out.dir)) {
  dir.create(out.dir)
}

# PARAMETERS
slot <- 'title'

# EXTRACT PREDICTS ABSTRACTS
predicts.pubs <- readRDS(file.path (data.dir, "predicts",
                                    'predicts_bib-2015-06-12-02-45-49.rds'))
if (slot == 'title') {
  output <- predicts.pubs[,'Source_title']
} else {
  output <- predicts.pubs[,'Abstract']
}
write.table (output,
             file.path (out.dir, 'predicts.txt'),
             quote=FALSE, row.names=FALSE, col.names=FALSE)
rm(predicts.pubs)

# EXTRACT NON-PREDICTS ABSTRACTS
# read in as lines -- if 'incomplete final line error' add return at bottom of file
# if (slot == 'abstracts') {
#   raw.data <- read.delim (file.path (data.dir, 'non_predicts_abstracts.txt'))[,1]
# } else {
#   raw.data <- read.delim (file.path (data.dir, 'non_predicts_title.txt'))[,1]
# }
# lines <- ''
# add <- FALSE
# for (line in raw.data) {
#   # go through lines removing just abstracts
#   if (grepl ('^AB', line)) {
#     add <- TRUE
#     line <- sub ('AB', '', line)
#   }
#   if (grepl ('^TC', line)) {
#     add <- FALSE
#   }
#   if (add) {
#     lines <- paste (lines, line)
#   }
# }
# write.table (lines, file=file.path (out.dir, 'non_predicts.txt'))
# rm (raw.data)

# EXTRACT LPI ABSTRACTS
# only needs moving
if (slot == 'abstracts') {
  raw.data <- read.delim (file.path (data.dir, 'lpi_abstracts.txt'))[,1]
} else {
  raw.data <- read.delim (file.path (data.dir, 'lpi_titles.txt'))[,1]
}
write.table (raw.data, file.path (out.dir, 'lpi.txt'),
             quote=FALSE, row.names=FALSE, col.names=FALSE)

# EXTRACT PUBMED ABSTRACTS
# USE COPY AND PASTE
# raw.data <- read.delim (file.path (data.dir, 'pubmed_abstracts.txt'))[,1]
# write.table (raw.data, file.path (out.dir, 'pubmed_abstracts.txt'),
#              quote=FALSE, row.names=FALSE, col.names=FALSE)