Package 'striprtf'

Test if a file looks like an RTF

Description

Validate if a file looks like an RTF. The test should be seen as a minimal requirement; If failed, the file is highly likely that the file is invalid, while passed, there is still possibility that the file does not follw the rule of RTF files.

Usage

looks_rtf(con, n = 1000)

Arguments

con	A connection object or string of file name
n	Integer that specifies the length of contents to be tested. If smaller than 10, forced to 10.

Value

Logical.

---

Extract Text from RTF (Rich Text Format) File

Description

Parses an RTF file and extracts plain text as character vector.

Usage

read_rtf(
  file,
  verbose = FALSE,
  row_start = "*| ",
  row_end = "",
  cell_end = " | ",
  ignore_tables = FALSE,
  check_file = TRUE,
  ...
)

strip_rtf(
  text,
  verbose = FALSE,
  row_start = "*| ",
  row_end = "",
  cell_end = " | ",
  ignore_tables = FALSE
)

Arguments

file	Path to an RTF file. Must be character of length 1.
verbose	Logical. If TRUE, progress report is printed on console. While it can be informative when parsing a large file, this option itself makes the process slow.
row_start, row_end	strings to be added at the beginning and end of table rows
cell_end	string to be put at the end of table cells
ignore_tables	if TRUE, no special treatment for tables
check_file	if TRUE, conducts a quick check on the file if it is an RTF file. If the file fails to pass the check, returns NULL without parsing the file.
...	Addional arguments passed to readLines
text	Character of length 1. Expected to be contents of an RTF file.

Details

Rich text format (RTF) files are written as a text file consisting of ASCII characters. The specification has been developed by Microsoft. This function interprets the character strings and extracts plain texts of the file. Major part of the algorithm of this function comes from a stack overflow thread (https://stackoverflow.com/a/188877) and the references therein. This function is a translation of the above to R language, associated with C++ codes for enhancement.

An advance from the preceding implementation is that the function accomodates with various ANSI code pages. For example, RTF files created by Japanese version of Microsoft Word marks \ansicpg932, which indicates the code page 932 is used for letter-code conversion. The function detects the code page indication and convert the characters to UTF-8 where possible. Conversion tables are retrieved from here: (https://www.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/).

Value

Character vector of extracted text

References

• Original discussion thread: https://stackoverflow.com/a/188877

• Code page table: https://www.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/

Examples

read_rtf(system.file("extdata/king.rtf", package = "striprtf"))

---

Renamed Functions

Description

From ver 0.3.1, the functions are renamed as follows:

• striprtf –> read_rtf

• rtf2text –> strip_rtf

Usage

striprtf(file, verbose = FALSE, ...)

rtf2text(text, verbose = FALSE)

Arguments

file	Path to an RTF file. Must be character of length 1.
verbose	Logical. If TRUE, progress report is printed on console. While it can be informative when parsing a large file, this option itself makes the process slow.
...	Addional arguments passed to readLines
text	Character of length 1. Expected to be contents of an RTF file.

Value

Character vector of extracted text

---

Find letters not used in strings

Description

Returns letters not used in strings

Usage

unused_letters(
  s,
  n = 1,
  avoid_strifrtf_internal = TRUE,
  as_number = FALSE,
  as_vector = FALSE
)

Arguments

s	character vector
n	number of letters to return
avoid_strifrtf_internal	If TRUE, letters used in the package's internal process are also regarded as "used".
as_number	if TRUE, return unicode numbers instead of letters itself
as_vector	if FALSE (and as_number is FALSE), return a single concatenated character, otherwise returns a character vector

Details

This function can be useful when some special characters must be temporarily converted to another letter without being confused with the same letters used elsewhere.

Letters are first searched from \u0001 upto \uffff. Do not specify too large n; An error is raised if a sufficient number of unsed letters are not found.

Value

unsed characters, format depends on as_number and as_vector arguments

---