parseDoc: parse a document and place content in a DocSet

In vjcitn/ssrch: a simple search engine

Description

parse a document and place content in a DocSet

Usage

parseDoc(csv, DocSetInstance = new("DocSet"), doctitle = NA_character_,
  docabst = NA_character_, rec_id_field = "experiment.accession",
  exclude_fields = c("study.accession"),
  substrings_to_omit = c("http://purl.obolibrary.org/obo/"),
  patterns_to_kill = "....-..-..|.*...,...",
  token_fixups = list(c("t''", "t'"), c(":$", "")), max_tok_nchar = 25,
  min_tok_nchar = 4, cleanFields = list("..*id$", ".name$", "_name$",
  "checksum", "isolate", "filename", "^ID$", "barcode", "Sample.Name"))

Arguments

csv	a character(1) CSV file path
DocSetInstance	if missing, DocSet is initialized in this function, otherwise the instance is updated with new content
doctitle	character(1) document title
docabst	character(1) a string: the document abstract
rec_id_field	character(1) field in CSV identifying records
exclude_fields	character vector of fields to ignore while parsing
substrings_to_omit	character vector of strings to remove from candidate keywords via gsub
patterns_to_kill	character(1) regexp that identifies tokens to be omitted from keyword set
token_fixups	a list if character(2) vectors that will be
max_tok_nchar	numeric(1) defaults to 25, tokens with more characters will be truncated to this length and suffixed with ellipsis
min_tok_nchar	numeric(1) defaults to 4, tokens shorter than this are not in index used with gsub() to repair irregularities. For example ‘c("t”", "t’")‘ will transform 'Burkitt”s' to 'Burkitt’s'
cleanFields	list of regular expressions identifying fields to ignore

Value

instance of DocSet

Note

The expected use case has 'DocSetInstance' being updated in a loop. Sharing of environments across multiple DocSetInstances can occur and unexpected behaviors may ensue. Note also that many of the parameter defaults to parseDoc are for the use case of processing SRA metadata.

Examples

myob = ssrch::docset_cancer68
td = tempdir()
alld = ls(docs2kw(myob))
r1 = retrieve_doc(alld[1], myob)
expo = write.csv(r1, paste0(td, "/expo.csv"))
pd = parseDoc(paste0(td, "/expo.csv"), doctitle=ssrch::titles68[alld[1]],
    docabst="qwerty")
pd
searchDocs("quer", pd) # query will fail
searchDocs("qwer", pd) # should succeed

vjcitn/ssrch documentation built on Oct. 25, 2019, 11:29 p.m.