///|
fn pdf_ascii_hex_digit(value : Int) -> Byte {
if value < 10 {
(48 + value).to_byte()
} else {
(55 + value).to_byte()
}
}
///|
fn pdf_ascii_hex_value(value : Int) -> Int? {
match value {
value if value >= 48 && value <= 57 => Some(value - 48)
value if value >= 65 && value <= 70 => Some(value - 55)
value if value >= 97 && value <= 102 => Some(value - 87)
_ => None
}
}
///|
/// Encode bytes using PDF ASCIIHexDecode syntax.
///
/// The returned bytes are uppercase hexadecimal pairs followed by the `>`
/// end-of-data marker.
pub fn pdf_ascii_hex_encode_view(data : BytesView) -> PdfBytes {
let output = Array::make(data.length() * 2 + 1, b'>')
let mut position = 0
for byte in data {
let value = byte.to_int()
output[position] = pdf_ascii_hex_digit(value / 16)
output[position + 1] = pdf_ascii_hex_digit(value % 16)
position += 2
}
Bytes::from_array(output)
}
///|
/// Owned-byte wrapper for `pdf_ascii_hex_encode_view`.
pub fn pdf_ascii_hex_encode(data : PdfBytes) -> PdfBytes {
pdf_ascii_hex_encode_view(data)
}
///|
/// Decode PDF ASCIIHexDecode data.
///
/// Whitespace is ignored and a trailing odd nibble is padded with zero when the
/// `>` marker is reached. Invalid hex digits raise `PdfError::InvalidHexEscape`.
pub fn pdf_ascii_hex_decode_view(data : BytesView) -> PdfBytes raise PdfError {
let output = Array::make((data.length() + 1) / 2, b'\x00')
let mut position = 0
let mut high : Int? = None
let mut done = false
for byte in data {
if !done {
let value = byte.to_int()
if pdf_is_whitespace_byte(value) {
()
} else if value == 62 {
match high {
Some(high_nibble) => {
output[position] = (high_nibble * 16).to_byte()
position += 1
}
None => ()
}
done = true
} else {
match pdf_ascii_hex_value(value) {
Some(nibble) =>
match high {
Some(high_nibble) => {
output[position] = (high_nibble * 16 + nibble).to_byte()
position += 1
high = None
}
None => high = Some(nibble)
}
None => raise InvalidHexEscape
}
}
}
}
Bytes::from_array(output[:position])
}
///|
/// Owned-byte wrapper for `pdf_ascii_hex_decode_view`.
pub fn pdf_ascii_hex_decode(data : PdfBytes) -> PdfBytes raise PdfError {
pdf_ascii_hex_decode_view(data)
}
///|
let pdf_ascii85_zero : Byte = 122
///|
let pdf_ascii85_terminator : Byte = 126
///|
let pdf_ascii85_max_word : UInt64 = 4294967295
///|
fn pdf_ascii85_encode_word(
output : Array[Byte],
position : Int,
word : UInt64,
output_length : Int,
compress_zero : Bool,
) -> Int {
let mut current = position
if compress_zero && word == 0 {
output[current] = pdf_ascii85_zero
current += 1
} else {
let digits = Array::make(5, b'!')
let mut value = word
for index in 0..<5 {
digits[index] = ((value % 85).to_int() + 33).to_byte()
value = value / 85
}
for i in 0.. UInt64 {
let mut word : UInt64 = 0
for i in 0..<4 {
let byte = if i < count { data[start + i] } else { 0 }
word = word * 256 + byte.to_uint64()
}
word
}
///|
/// Encode bytes using PDF ASCII85Decode syntax.
///
/// Four zero bytes are compressed as `z`, partial final groups are shortened,
/// and the result ends with `~>`.
pub fn pdf_ascii85_encode_view(data : BytesView) -> PdfBytes {
let output = Array::make((data.length() + 3) / 4 * 5 + 2, b' ')
let mut position = 0
let mut index = 0
while index < data.length() {
let remaining = data.length() - index
let count = if remaining < 4 { remaining } else { 4 }
let word = pdf_ascii85_word_from_view(data, index, count)
position = pdf_ascii85_encode_word(
output,
position,
word,
count + 1,
count == 4,
)
index += count
}
output[position] = pdf_ascii85_terminator
output[position + 1] = b'>'
Bytes::from_array(output[:position + 2])
}
///|
/// Owned-byte wrapper for `pdf_ascii85_encode_view`.
pub fn pdf_ascii85_encode(data : PdfBytes) -> PdfBytes {
pdf_ascii85_encode_view(data)
}
///|
fn pdf_ascii85_decode_group(
digits : ArrayView[Int],
output : Array[Byte],
position : Int,
output_length : Int,
) -> Int raise PdfError {
let mut word : UInt64 = 0
for digit in digits {
word = word * 85 + digit.to_uint64()
}
if word > pdf_ascii85_max_word {
raise InvalidASCII85Data
}
let mut current = position
if output_length >= 1 {
output[current] = (word / 16777216).to_byte()
current += 1
}
if output_length >= 2 {
output[current] = (word / 65536 % 256).to_byte()
current += 1
}
if output_length >= 3 {
output[current] = (word / 256 % 256).to_byte()
current += 1
}
if output_length >= 4 {
output[current] = (word % 256).to_byte()
current += 1
}
current
}
///|
fn pdf_ascii85_finish_partial(
digits : Array[Int],
output : Array[Byte],
position : Int,
) -> Int raise PdfError {
if digits.length() == 1 {
raise InvalidASCII85Data
} else if digits.length() > 1 {
let output_length = digits.length() - 1
while digits.length() < 5 {
digits.push(84)
}
pdf_ascii85_decode_group(digits, output, position, output_length)
} else {
position
}
}
///|
/// Decode PDF ASCII85Decode data.
///
/// Whitespace is ignored. Missing terminators, malformed partial groups,
/// invalid `z` placement, and overflowing groups raise `PdfError::InvalidASCII85Data`.
pub fn pdf_ascii85_decode_view(data : BytesView) -> PdfBytes raise PdfError {
let output = Array::make(data.length() * 4, b'\x00')
let mut position = 0
let mut digits : Array[Int] = Array::new(capacity=5)
let mut done = false
for byte in data {
if !done {
let value = byte.to_int()
if pdf_is_whitespace_byte(value) {
()
} else if byte == pdf_ascii85_terminator {
position = pdf_ascii85_finish_partial(digits, output, position)
done = true
} else if byte == pdf_ascii85_zero {
if digits.length() == 0 {
for _ in 0..<4 {
output[position] = b'\x00'
position += 1
}
} else {
raise InvalidASCII85Data
}
} else if value >= 33 && value <= 117 {
digits.push(value - 33)
if digits.length() == 5 {
position = pdf_ascii85_decode_group(digits, output, position, 4)
digits = Array::new(capacity=5)
}
} else {
raise InvalidASCII85Data
}
}
}
if !done {
raise InvalidASCII85Data
}
Bytes::from_array(output[:position])
}
///|
/// Owned-byte wrapper for `pdf_ascii85_decode_view`.
pub fn pdf_ascii85_decode(data : PdfBytes) -> PdfBytes raise PdfError {
pdf_ascii85_decode_view(data)
}
///|
fn pdf_lzw_next_code(
bits : BitStream,
code_length : Int,
) -> Int? raise PdfError {
try bits.get_value_int(code_length) catch {
EndOfInput => None
_ => raise InvalidLZWData
} noraise {
value => Some(value)
}
}
///|
fn pdf_lzw_decode_code(
code : Int,
prefix_code : Array[Int],
append_character : Array[Int],
next_code : Int,
) -> Array[Int] raise PdfError {
if code < 0 || code >= next_code || code >= prefix_code.length() {
raise InvalidLZWData
}
let stack : Array[Int] = Array::new(capacity=16)
let mut current = code
let mut depth = 0
while current > 255 {
if current >= next_code || current >= prefix_code.length() || depth > 8192 {
raise InvalidLZWData
}
stack.push(append_character[current])
current = prefix_code[current]
depth += 1
}
if current < 0 || current > 255 {
raise InvalidLZWData
}
let output : Array[Int] = Array::new(capacity=stack.length() + 1)
output.push(current)
if stack.length() > 0 {
for offset in 0.. Int? raise PdfError {
let mut code = 256
while code == 256 {
match pdf_lzw_next_code(bits, code_length) {
None => break None
Some(value) => code = value
}
} nobreak {
Some(code)
}
}
///|
fn pdf_lzw_output_literal(
output : Array[Byte],
code : Int,
) -> Unit raise PdfError {
if code < 0 || code > 255 {
raise InvalidLZWData
}
output.push(code.to_byte())
}
///|
fn pdf_lzw_consumed_bytes(bits : BitStream) -> Int {
let count = bits.bits_read()
if count % 8 == 0 {
count / 8
} else {
count / 8 + 1
}
}
///|
fn pdf_lzw_decode_view_prefix(
data : BytesView,
early_change? : Int = 1,
) -> (PdfBytes, Int) raise PdfError {
let prefix_code = Array::make(8192, 0)
let append_character = Array::make(8192, 0)
let bits = bitstream_of_view(data)
let output : Array[Byte] = Array::new(capacity=data.length())
let mut code_length = 9
let mut next_code = 258
let mut old_code = 256
let mut character = 0
match pdf_lzw_read_non_clear_code(bits, code_length) {
None => raise InvalidLZWData
Some(257) => (Bytes::from_array(output), pdf_lzw_consumed_bytes(bits))
Some(first_code) => {
pdf_lzw_output_literal(output, first_code)
old_code = first_code
character = first_code
let mut consumed = 0
let mut finished = false
while !finished {
match pdf_lzw_next_code(bits, code_length) {
None => raise InvalidLZWData
Some(257) => {
consumed = pdf_lzw_consumed_bytes(bits)
finished = true
}
Some(256) => {
code_length = 9
next_code = 258
old_code = 256
match pdf_lzw_read_non_clear_code(bits, code_length) {
None => raise InvalidLZWData
Some(257) => {
consumed = pdf_lzw_consumed_bytes(bits)
finished = true
}
Some(code) => {
pdf_lzw_output_literal(output, code)
old_code = code
character = code
}
}
}
Some(new_code) => {
let chars = if new_code == next_code {
let expanded = pdf_lzw_decode_code(
old_code, prefix_code, append_character, next_code,
)
expanded.push(character)
expanded
} else if new_code < next_code {
pdf_lzw_decode_code(
new_code, prefix_code, append_character, next_code,
)
} else {
raise InvalidLZWData
}
character = chars[0]
for value in chars {
pdf_lzw_output_literal(output, value)
}
if next_code < prefix_code.length() {
prefix_code[next_code] = old_code
append_character[next_code] = character
next_code += 1
match next_code + early_change {
512 | 1024 | 2048 => code_length += 1
_ => ()
}
}
old_code = new_code
}
}
}
(Bytes::from_array(output), consumed)
}
}
}
///|
/// Decode PDF LZW data from a byte view.
///
/// `early_change` controls when code width grows; PDF streams default to `1`.
/// Malformed codes or bitstream errors raise `PdfError::InvalidLZWData`.
pub fn pdf_lzw_decode_view(
data : BytesView,
early_change? : Int = 1,
) -> PdfBytes raise PdfError {
let prefix_code = Array::make(8192, 0)
let append_character = Array::make(8192, 0)
let bits = bitstream_of_view(data)
let output : Array[Byte] = Array::new(capacity=data.length())
let mut code_length = 9
let mut next_code = 258
let mut old_code = 256
let mut character = 0
match pdf_lzw_read_non_clear_code(bits, code_length) {
None | Some(257) => Bytes::from_array(output)
Some(first_code) => {
pdf_lzw_output_literal(output, first_code)
old_code = first_code
character = first_code
let mut finished = false
while !finished {
match pdf_lzw_next_code(bits, code_length) {
None => finished = true
Some(257) => finished = true
Some(256) => {
code_length = 9
next_code = 258
old_code = 256
match pdf_lzw_read_non_clear_code(bits, code_length) {
None | Some(257) => finished = true
Some(code) => {
pdf_lzw_output_literal(output, code)
old_code = code
character = code
}
}
}
Some(new_code) => {
let chars = if new_code == next_code {
let expanded = pdf_lzw_decode_code(
old_code, prefix_code, append_character, next_code,
)
expanded.push(character)
expanded
} else if new_code < next_code {
pdf_lzw_decode_code(
new_code, prefix_code, append_character, next_code,
)
} else {
raise InvalidLZWData
}
character = chars[0]
for value in chars {
pdf_lzw_output_literal(output, value)
}
if next_code < prefix_code.length() {
prefix_code[next_code] = old_code
append_character[next_code] = character
next_code += 1
match next_code + early_change {
512 | 1024 | 2048 => code_length += 1
_ => ()
}
}
old_code = new_code
}
}
}
Bytes::from_array(output)
}
}
}
///|
/// Owned-byte wrapper for `pdf_lzw_decode_view`.
pub fn pdf_lzw_decode(
data : PdfBytes,
early_change? : Int = 1,
) -> PdfBytes raise PdfError {
pdf_lzw_decode_view(data, early_change~)
}
///|
fn pdf_filter_name(text : String) -> PdfName {
pdf_name_of_bytes(@ascii.encode(text))
}
///|
let pdf_filter_crypt_name : PdfName = pdf_filter_name("/Crypt")
///|
let pdf_filter_ascii_hex_decode_name : PdfName = pdf_filter_name(
"/ASCIIHexDecode",
)
///|
let pdf_filter_ascii_hex_short_name : PdfName = pdf_filter_name("/AHx")
///|
let pdf_filter_ascii85_decode_name : PdfName = pdf_filter_name("/ASCII85Decode")
///|
let pdf_filter_ascii85_short_name : PdfName = pdf_filter_name("/A85")
///|
let pdf_filter_run_length_decode_name : PdfName = pdf_filter_name(
"/RunLengthDecode",
)
///|
let pdf_filter_run_length_short_name : PdfName = pdf_filter_name("/RL")
///|
let pdf_filter_flate_decode_name : PdfName = pdf_filter_name("/FlateDecode")
///|
let pdf_filter_flate_short_name : PdfName = pdf_filter_name("/Fl")
///|
let pdf_filter_lzw_decode_name : PdfName = pdf_filter_name("/LZWDecode")
///|
let pdf_filter_lzw_short_name : PdfName = pdf_filter_name("/LZW")
///|
let pdf_filter_ccitt_fax_decode_name : PdfName = pdf_filter_name(
"/CCITTFaxDecode",
)
///|
let pdf_filter_ccitt_fax_short_name : PdfName = pdf_filter_name("/CCF")
///|
/// Stream encoding requested by high-level stream encode helpers.
///
/// CCITT variants carry image columns and rows because those values are needed
/// in `/DecodeParms` when writing fax-compressed image streams.
pub(all) enum PdfStreamEncoding {
PdfStreamASCIIHex
PdfStreamASCII85
PdfStreamRunLength
PdfStreamFlate
PdfStreamCCITT(Int, Int)
PdfStreamCCITTG4(Int, Int)
} derive(Debug, Eq, ToJson)
///|
/// PNG/TIFF predictor requested when encoding stream data.
///
/// Predictor values correspond to PDF `/Predictor` numbers.
pub(all) enum PdfStreamPredictor {
PdfStreamPredictorTIFF2
PdfStreamPredictorPNGNone
PdfStreamPredictorPNGSub
PdfStreamPredictorPNGUp
PdfStreamPredictorPNGAverage
PdfStreamPredictorPNGPaeth
PdfStreamPredictorPNGOptimum
} derive(Debug, Eq, ToJson)
///|
/// Return the PDF filter name for this stream encoding.
pub fn PdfStreamEncoding::filter_name(self : PdfStreamEncoding) -> PdfName {
match self {
PdfStreamASCIIHex => pdf_filter_ascii_hex_decode_name
PdfStreamASCII85 => pdf_filter_ascii85_decode_name
PdfStreamRunLength => pdf_filter_run_length_decode_name
PdfStreamFlate => pdf_filter_flate_decode_name
PdfStreamCCITT(_, _) | PdfStreamCCITTG4(_, _) =>
pdf_filter_ccitt_fax_decode_name
}
}
///|
/// Return the PDF `/Predictor` number for this predictor.
pub fn PdfStreamPredictor::number(self : PdfStreamPredictor) -> Int {
match self {
PdfStreamPredictorTIFF2 => 2
PdfStreamPredictorPNGNone => 10
PdfStreamPredictorPNGSub => 11
PdfStreamPredictorPNGUp => 12
PdfStreamPredictorPNGAverage => 13
PdfStreamPredictorPNGPaeth => 14
PdfStreamPredictorPNGOptimum => 15
}
}
///|
/// Encode 1-bit image data as CCITT Group 3 one-dimensional fax data.
pub fn pdf_ccitt_encode_view(
columns : Int,
rows : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
pdf_ccitt_encode_g3_1d_view(data, columns, rows)
}
///|
/// Owned-byte wrapper for `pdf_ccitt_encode_view`.
pub fn pdf_ccitt_encode(
columns : Int,
rows : Int,
data : PdfBytes,
) -> PdfBytes raise PdfError {
pdf_ccitt_encode_view(columns, rows, data)
}
///|
/// Encode 1-bit image data as CCITT Group 4 fax data.
pub fn pdf_ccitt_g4_encode_view(
columns : Int,
rows : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
pdf_ccitt_encode_g4_view(data, columns, rows)
}
///|
/// Owned-byte wrapper for `pdf_ccitt_g4_encode_view`.
pub fn pdf_ccitt_g4_encode(
columns : Int,
rows : Int,
data : PdfBytes,
) -> PdfBytes raise PdfError {
pdf_ccitt_g4_encode_view(columns, rows, data)
}
///|
fn pdf_filter_is_name(filter : PdfName, name : PdfName) -> Bool {
filter == name
}
///|
fn pdf_filter_is_one_of(
filter : PdfName,
first : PdfName,
second : PdfName,
) -> Bool {
pdf_filter_is_name(filter, first) || pdf_filter_is_name(filter, second)
}
///|
/// Decode bytes with a single PDF filter name.
///
/// Recognizes long and abbreviated filter names. `/Crypt` is treated as an
/// identity filter at this byte-dispatch layer; unsupported filters raise
/// `PdfError::FilterNotSupported`.
pub fn pdf_decode_filter_view(
filter : PdfName,
data : BytesView,
) -> PdfBytes raise PdfError {
if pdf_filter_is_name(filter, pdf_filter_crypt_name) {
data.to_owned()
} else if pdf_filter_is_one_of(
filter, pdf_filter_ascii_hex_decode_name, pdf_filter_ascii_hex_short_name,
) {
pdf_ascii_hex_decode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_ascii85_decode_name, pdf_filter_ascii85_short_name,
) {
pdf_ascii85_decode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_run_length_decode_name, pdf_filter_run_length_short_name,
) {
pdf_run_length_decode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_flate_decode_name, pdf_filter_flate_short_name,
) {
pdf_flate_decode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_lzw_decode_name, pdf_filter_lzw_short_name,
) {
pdf_lzw_decode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_ccitt_fax_decode_name, pdf_filter_ccitt_fax_short_name,
) {
pdf_ccitt_decode_view(data, pdf_ccitt_default_params())
} else {
raise FilterNotSupported(filter)
}
}
///|
/// Decode owned bytes with a single PDF filter name.
///
/// Identity and Flate paths avoid an extra copy where possible.
pub fn pdf_decode_filter_bytes(
filter : PdfName,
data : PdfBytes,
) -> PdfBytes raise PdfError {
if pdf_filter_is_name(filter, pdf_filter_crypt_name) {
data
} else if pdf_filter_is_one_of(
filter, pdf_filter_flate_decode_name, pdf_filter_flate_short_name,
) {
pdf_flate_decode(data)
} else {
pdf_decode_filter_view(filter, data)
}
}
///|
/// Encode bytes with a single PDF filter name.
pub fn pdf_encode_filter_view(
filter : PdfName,
data : BytesView,
) -> PdfBytes raise PdfError {
pdf_encode_filter_view_with_optional_flate_level(filter, data, None)
}
///|
/// Encode bytes with a single PDF filter name and explicit Flate level.
///
/// `flate_level` follows the zlib range accepted by the Flate encoder.
pub fn pdf_encode_filter_view_with_flate_level(
filter : PdfName,
data : BytesView,
flate_level : Int,
) -> PdfBytes raise PdfError {
pdf_encode_filter_view_with_optional_flate_level(
filter,
data,
Some(flate_level),
)
}
///|
fn pdf_encode_filter_view_with_optional_flate_level(
filter : PdfName,
data : BytesView,
flate_level : Int?,
) -> PdfBytes raise PdfError {
if pdf_filter_is_name(filter, pdf_filter_crypt_name) {
data.to_owned()
} else if pdf_filter_is_one_of(
filter, pdf_filter_ascii_hex_decode_name, pdf_filter_ascii_hex_short_name,
) {
pdf_ascii_hex_encode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_ascii85_decode_name, pdf_filter_ascii85_short_name,
) {
pdf_ascii85_encode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_run_length_decode_name, pdf_filter_run_length_short_name,
) {
pdf_run_length_encode_view(data)
} else if pdf_filter_is_one_of(
filter, pdf_filter_flate_decode_name, pdf_filter_flate_short_name,
) {
match flate_level {
Some(level) => pdf_flate_encode_view_with_level(data, level)
None => pdf_flate_encode_view(data)
}
} else {
raise FilterNotSupported(filter)
}
}
///|
/// Encode owned bytes with a single PDF filter name.
pub fn pdf_encode_filter_bytes(
filter : PdfName,
data : PdfBytes,
) -> PdfBytes raise PdfError {
pdf_encode_filter_bytes_with_optional_flate_level(filter, data, None)
}
///|
/// Encode owned bytes with a single PDF filter name and explicit Flate level.
pub fn pdf_encode_filter_bytes_with_flate_level(
filter : PdfName,
data : PdfBytes,
flate_level : Int,
) -> PdfBytes raise PdfError {
pdf_encode_filter_bytes_with_optional_flate_level(
filter,
data,
Some(flate_level),
)
}
///|
fn pdf_encode_filter_bytes_with_optional_flate_level(
filter : PdfName,
data : PdfBytes,
flate_level : Int?,
) -> PdfBytes raise PdfError {
if pdf_filter_is_name(filter, pdf_filter_crypt_name) {
data
} else if pdf_filter_is_one_of(
filter, pdf_filter_flate_decode_name, pdf_filter_flate_short_name,
) {
match flate_level {
Some(level) => pdf_flate_encode_with_level(data, level)
None => pdf_flate_encode(data)
}
} else {
pdf_encode_filter_view_with_optional_flate_level(filter, data, flate_level)
}
}
///|
let pdf_filter_key_name : PdfName = pdf_filter_name("/Filter")
///|
let pdf_short_filter_key_name : PdfName = pdf_filter_name("/F")
///|
let pdf_decodeparms_key_name : PdfName = pdf_filter_name("/DecodeParms")
///|
let pdf_short_decodeparms_key_name : PdfName = pdf_filter_name("/DP")
///|
let pdf_codec_length_key_name : PdfName = pdf_filter_name("/Length")
///|
let pdf_predictor_key_name : PdfName = pdf_filter_name("/Predictor")
///|
let pdf_early_change_key_name : PdfName = pdf_filter_name("/EarlyChange")
///|
let pdf_ccitt_k_key_name : PdfName = pdf_filter_name("/K")
///|
let pdf_ccitt_end_of_line_key_name : PdfName = pdf_filter_name("/EndOfLine")
///|
let pdf_ccitt_encoded_byte_align_key_name : PdfName = pdf_filter_name(
"/EncodedByteAlign",
)
///|
let pdf_ccitt_rows_key_name : PdfName = pdf_filter_name("/Rows")
///|
let pdf_ccitt_end_of_block_key_name : PdfName = pdf_filter_name("/EndOfBlock")
///|
let pdf_ccitt_black_is_1_key_name : PdfName = pdf_filter_name("/BlackIs1")
///|
let pdf_ccitt_damaged_rows_before_error_key_name : PdfName = pdf_filter_name(
"/DamagedRowsBeforeError",
)
///|
let pdf_colors_key_name : PdfName = pdf_filter_name("/Colors")
///|
let pdf_bits_per_component_key_name : PdfName = pdf_filter_name(
"/BitsPerComponent",
)
///|
let pdf_columns_key_name : PdfName = pdf_filter_name("/Columns")
///|
fn pdf_filter_key() -> PdfName {
pdf_filter_key_name
}
///|
fn pdf_short_filter_key() -> PdfName {
pdf_short_filter_key_name
}
///|
fn pdf_decodeparms_key() -> PdfName {
pdf_decodeparms_key_name
}
///|
fn pdf_short_decodeparms_key() -> PdfName {
pdf_short_decodeparms_key_name
}
///|
fn pdf_codec_length_key() -> PdfName {
pdf_codec_length_key_name
}
///|
fn pdf_predictor_key() -> PdfName {
pdf_predictor_key_name
}
///|
fn pdf_early_change_key() -> PdfName {
pdf_early_change_key_name
}
///|
fn pdf_ccitt_k_key() -> PdfName {
pdf_ccitt_k_key_name
}
///|
fn pdf_ccitt_end_of_line_key() -> PdfName {
pdf_ccitt_end_of_line_key_name
}
///|
fn pdf_ccitt_encoded_byte_align_key() -> PdfName {
pdf_ccitt_encoded_byte_align_key_name
}
///|
fn pdf_ccitt_rows_key() -> PdfName {
pdf_ccitt_rows_key_name
}
///|
fn pdf_ccitt_end_of_block_key() -> PdfName {
pdf_ccitt_end_of_block_key_name
}
///|
fn pdf_ccitt_black_is_1_key() -> PdfName {
pdf_ccitt_black_is_1_key_name
}
///|
fn pdf_ccitt_damaged_rows_before_error_key() -> PdfName {
pdf_ccitt_damaged_rows_before_error_key_name
}
///|
fn pdf_colors_key() -> PdfName {
pdf_colors_key_name
}
///|
fn pdf_bits_per_component_key() -> PdfName {
pdf_bits_per_component_key_name
}
///|
fn pdf_columns_key() -> PdfName {
pdf_columns_key_name
}
///|
fn pdf_stream_dictionary_length(dictionary : PdfObject) -> Unit raise PdfError {
match dictionary.lookup_immediate(pdf_codec_length_key()) {
Some(PdfInteger(_)) => ()
_ => raise StreamLengthExpected
}
}
///|
fn pdf_stream_filter_value(dictionary : PdfObject) -> PdfObject? {
match dictionary.lookup_immediate(pdf_filter_key()) {
Some(value) => Some(value)
None => dictionary.lookup_immediate(pdf_short_filter_key())
}
}
///|
fn pdf_stream_first_filter(dictionary : PdfObject) -> PdfName? raise PdfError {
match pdf_stream_filter_value(dictionary) {
None => None
Some(PdfNameObject(name)) => Some(name)
Some(PdfArray([])) => None
Some(PdfArray([PdfNameObject(name), ..])) => Some(name)
_ => raise FilterExpected
}
}
///|
fn PdfDocument::pdf_stream_filter_entry_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> (PdfName, PdfObject)? {
match self.lookup_direct(pdf_filter_key(), dictionary) {
Some(value) => Some((pdf_filter_key(), value))
None =>
match self.lookup_direct(pdf_short_filter_key(), dictionary) {
Some(value) => Some((pdf_short_filter_key(), value))
None => None
}
}
}
///|
fn PdfDocument::pdf_stream_first_filter_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> PdfName? raise PdfError {
match self.pdf_stream_first_filter_entry_direct(dictionary) {
Some((_, _, name)) => Some(name)
None => None
}
}
///|
fn PdfDocument::pdf_stream_first_filter_entry_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> (PdfName, PdfObject, PdfName)? raise PdfError {
match self.pdf_stream_filter_entry_direct(dictionary) {
None => None
Some((key, PdfNameObject(name) as value)) => Some((key, value, name))
Some((_, PdfArray([]))) => None
Some((key, PdfArray([first, ..]) as value)) =>
match self.direct(first) {
PdfNameObject(name) => Some((key, value, name))
_ => raise FilterExpected
}
_ => raise FilterExpected
}
}
///|
fn pdf_remove_first_filter_entry(
dictionary : PdfObject,
key : PdfName,
) -> PdfObject raise PdfError {
match dictionary.lookup_immediate(key) {
None => dictionary
Some(PdfNameObject(_)) | Some(PdfArray([_])) =>
dictionary.remove_dict_entry(key)
Some(PdfArray([_, .. rest])) =>
dictionary.replace_dict_entry(
key,
PdfArray([ for value in rest => value ]),
)
_ => raise FilterExpected
}
}
///|
fn pdf_remove_first_decodeparms_entry(
dictionary : PdfObject,
key : PdfName,
) -> PdfObject raise PdfError {
match dictionary.lookup_immediate(key) {
None => dictionary
Some(PdfDictionary(_)) | Some(PdfArray([])) =>
dictionary.remove_dict_entry(key)
Some(PdfArray([_, .. rest])) =>
dictionary.replace_dict_entry(
key,
PdfArray([ for value in rest => value ]),
)
_ => raise FilterExpected
}
}
///|
fn pdf_remove_first_decoder(dictionary : PdfObject) -> PdfObject raise PdfError {
let without_filter = if dictionary.lookup_immediate(pdf_filter_key()) != None {
pdf_remove_first_filter_entry(dictionary, pdf_filter_key())
} else {
pdf_remove_first_filter_entry(dictionary, pdf_short_filter_key())
}
let without_decodeparms = if without_filter.lookup_immediate(
pdf_decodeparms_key(),
) !=
None {
pdf_remove_first_decodeparms_entry(without_filter, pdf_decodeparms_key())
} else {
pdf_remove_first_decodeparms_entry(
without_filter,
pdf_short_decodeparms_key(),
)
}
without_decodeparms
}
///|
fn pdf_remove_first_filter_entry_value(
dictionary : PdfObject,
key : PdfName,
value : PdfObject,
) -> PdfObject raise PdfError {
match value {
PdfArray([]) | PdfArray([_]) => dictionary.remove_dict_entry(key)
PdfArray([_, .. rest]) =>
dictionary.replace_dict_entry(
key,
PdfArray([ for value in rest => value ]),
)
_ => dictionary.remove_dict_entry(key)
}
}
///|
fn pdf_remove_first_decodeparms_entry_value(
dictionary : PdfObject,
key : PdfName,
value : PdfObject,
) -> PdfObject raise PdfError {
match value {
PdfArray([_, .. rest]) =>
dictionary.replace_dict_entry(
key,
PdfArray([ for value in rest => value ]),
)
_ => dictionary.remove_dict_entry(key)
}
}
///|
fn PdfDocument::pdf_remove_first_decoder_direct(
self : PdfDocument,
dictionary : PdfObject,
filter_key : PdfName,
filter_value : PdfObject,
) -> PdfObject raise PdfError {
let without_filter = pdf_remove_first_filter_entry_value(
dictionary, filter_key, filter_value,
)
match self.pdf_decodeparms_entry_direct(without_filter) {
Some((key, value)) =>
pdf_remove_first_decodeparms_entry_value(without_filter, key, value)
None => without_filter
}
}
///|
fn pdf_stream_replace_length(
dictionary : PdfObject,
length : Int,
) -> PdfObject raise PdfError {
dictionary.replace_dict_entry(pdf_codec_length_key(), PdfInteger(length))
}
///|
fn pdf_predictor_scanline_width(
colors : Int,
bits_per_component : Int,
columns : Int,
) -> Int raise PdfError {
if colors <= 0 || bits_per_component <= 0 || columns <= 0 {
raise PredictorExpected
}
(colors * bits_per_component * columns + 7) / 8
}
///|
fn pdf_predictor_bytes_per_pixel(
colors : Int,
bits_per_component : Int,
) -> Int raise PdfError {
if colors <= 0 || bits_per_component <= 0 {
raise PredictorExpected
}
(colors * bits_per_component + 7) / 8
}
///|
fn pdf_predictor_get0(values : ArrayView[Int], index : Int) -> Int {
if index < 0 {
0
} else {
values[index]
}
}
///|
fn pdf_predictor_wrapping_byte(value : Int) -> Byte {
let remainder = value % 256
if remainder < 0 {
(remainder + 256).to_byte()
} else {
remainder.to_byte()
}
}
///|
fn pdf_paeth_predictor(a : Int, b : Int, c : Int) -> Int {
let p = a + b - c
let pa = (p - a).abs()
let pb = (p - b).abs()
let pc = (p - c).abs()
if pa <= pb && pa <= pc {
a
} else if pb <= pc {
b
} else {
c
}
}
///|
fn pdf_decode_tiff_predictor(
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
if bits_per_component != 8 {
raise PredictorNotSupported(2)
}
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let output : Array[Byte] = [ for byte in data => byte ]
for line in 0..<(data.length() / scanline_width) {
let line_start = line * scanline_width
let mut position = colors
while position < scanline_width {
let index = line_start + position
output[index] = ((
output[index - colors].to_int() + output[index].to_int()
) %
256).to_byte()
position += 1
}
}
Bytes::from_array(output)
}
///|
fn pdf_encode_tiff_predictor(
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
if bits_per_component != 8 {
raise PredictorNotSupported(2)
}
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let output : Array[Byte] = [ for byte in data => byte ]
for line in 0..<(data.length() / scanline_width) {
let line_start = line * scanline_width
let mut position = colors
while position < scanline_width {
let index = line_start + position
output[index] = pdf_predictor_wrapping_byte(
data[index].to_int() - data[index - colors].to_int(),
)
position += 1
}
}
Bytes::from_array(output)
}
///|
fn pdf_decode_png_current(
data : BytesView,
row_start : Int,
available : Int,
index : Int,
) -> Int {
if index < available {
data[row_start + index].to_int()
} else {
0
}
}
///|
fn pdf_decode_png_decoded_length(data : BytesView, scanline_width : Int) -> Int {
let mut rows = 0
let mut index = 0
while index < data.length() {
rows += 1
index += 1
let available = if index + scanline_width > data.length() {
data.length() - index
} else {
scanline_width
}
index += available
}
rows * scanline_width
}
///|
fn pdf_decode_png_push_scanline(
output : Array[Byte],
position : Int,
predictor : Int,
prior_decoded : ArrayView[Int],
decoded : Array[Int],
data : BytesView,
row_start : Int,
available : Int,
bytes_per_pixel : Int,
) -> Int raise PdfError {
let scanline_width = decoded.length()
let mut output_position = position
match predictor {
0 =>
for i in 0..
for i in 0..
for i in 0..
for i in 0..
for i in 0.. raise PredictorNotSupported(predictor)
}
output_position
}
///|
fn pdf_decode_png_predictor(
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let bytes_per_pixel = pdf_predictor_bytes_per_pixel(
colors, bits_per_component,
)
let output = Array::make(
pdf_decode_png_decoded_length(data, scanline_width),
b'\x00',
)
let mut position = 0
let mut prior_decoded : Array[Int] = [ for _ in 0.. 0 ]
let mut decoded : Array[Int] = [ for _ in 0.. 0 ]
let mut index = 0
while index < data.length() {
let predictor = data[index].to_int()
index += 1
let available = if index + scanline_width > data.length() {
data.length() - index
} else {
scanline_width
}
let row_start = index
index += available
position = pdf_decode_png_push_scanline(
output, position, predictor, prior_decoded, decoded, data, row_start, available,
bytes_per_pixel,
)
let previous = prior_decoded
prior_decoded = decoded
decoded = previous
}
Bytes::from_array(output[:position])
}
///|
/// Decode data after applying a PDF predictor.
///
/// Predictor `1` is identity, `2` is TIFF horizontal differencing, and
/// `10..15` are PNG predictors. Invalid image geometry or unsupported
/// predictors raise `PdfError`.
pub fn pdf_decode_predictor_view(
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
match predictor {
1 => data.to_owned()
2 => pdf_decode_tiff_predictor(colors, bits_per_component, columns, data)
value if value >= 10 && value <= 15 =>
pdf_decode_png_predictor(colors, bits_per_component, columns, data)
_ => raise PredictorNotSupported(predictor)
}
}
///|
/// Owned-byte wrapper for `pdf_decode_predictor_view`.
///
/// Predictor `1` returns the original owned bytes unchanged.
pub fn pdf_decode_predictor(
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
data : PdfBytes,
) -> PdfBytes raise PdfError {
if predictor == 1 {
data
} else {
pdf_decode_predictor_view(
predictor, colors, bits_per_component, columns, data,
)
}
}
///|
fn pdf_encode_png_encoded_length(data : BytesView, scanline_width : Int) -> Int {
data.length() / scanline_width * (scanline_width + 1)
}
///|
fn pdf_encode_png_sub_predictor(
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let bytes_per_pixel = pdf_predictor_bytes_per_pixel(
colors, bits_per_component,
)
let output = Array::make(
pdf_encode_png_encoded_length(data, scanline_width),
b'\x00',
)
let mut position = 0
for row in 0..<(data.length() / scanline_width) {
position = pdf_encode_png_push_scanline(
output,
position,
1,
data,
row * scanline_width,
scanline_width,
bytes_per_pixel,
)
}
Bytes::from_array(output[:position])
}
///|
fn pdf_encode_png_up_predictor(
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let bytes_per_pixel = pdf_predictor_bytes_per_pixel(
colors, bits_per_component,
)
let output = Array::make(
pdf_encode_png_encoded_length(data, scanline_width),
b'\x00',
)
let mut position = 0
for row in 0..<(data.length() / scanline_width) {
position = pdf_encode_png_push_scanline(
output,
position,
2,
data,
row * scanline_width,
scanline_width,
bytes_per_pixel,
)
}
Bytes::from_array(output[:position])
}
///|
fn pdf_encode_png_predictor_value(
predictor : Int,
data : BytesView,
row_start : Int,
index : Int,
scanline_width : Int,
bytes_per_pixel : Int,
) -> Int raise PdfError {
let left = if index < bytes_per_pixel {
0
} else {
data[row_start + index - bytes_per_pixel].to_int()
}
let up = if row_start < scanline_width {
0
} else {
data[row_start - scanline_width + index].to_int()
}
let up_left = if row_start < scanline_width || index < bytes_per_pixel {
0
} else {
data[row_start - scanline_width + index - bytes_per_pixel].to_int()
}
match predictor {
0 => 0
1 => left
2 => up
3 => (left + up) / 2
4 => pdf_paeth_predictor(left, up, up_left)
_ => raise PredictorNotSupported(predictor)
}
}
///|
fn pdf_encode_png_push_scanline(
output : Array[Byte],
position : Int,
predictor : Int,
data : BytesView,
row_start : Int,
scanline_width : Int,
bytes_per_pixel : Int,
) -> Int raise PdfError {
let mut current = position
output[current] = predictor.to_byte()
current += 1
for i in 0.. Int raise PdfError {
let mut score = 0
for i in 0.. 127 { 256 - value } else { value }
}
score
}
///|
fn pdf_encode_png_tagged_predictor(
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let bytes_per_pixel = pdf_predictor_bytes_per_pixel(
colors, bits_per_component,
)
let output = Array::make(
pdf_encode_png_encoded_length(data, scanline_width),
b'\x00',
)
let mut position = 0
for row in 0..<(data.length() / scanline_width) {
position = pdf_encode_png_push_scanline(
output,
position,
predictor,
data,
row * scanline_width,
scanline_width,
bytes_per_pixel,
)
}
Bytes::from_array(output[:position])
}
///|
fn pdf_encode_png_optimum_predictor(
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
let scanline_width = pdf_predictor_scanline_width(
colors, bits_per_component, columns,
)
let bytes_per_pixel = pdf_predictor_bytes_per_pixel(
colors, bits_per_component,
)
let output = Array::make(
pdf_encode_png_encoded_length(data, scanline_width),
b'\x00',
)
let mut position = 0
for row in 0..<(data.length() / scanline_width) {
let row_start = row * scanline_width
let mut best_predictor = 0
let mut best_score = pdf_encode_png_predictor_score(
0, data, row_start, scanline_width, bytes_per_pixel,
)
for predictor in 1..<=4 {
let score = pdf_encode_png_predictor_score(
predictor, data, row_start, scanline_width, bytes_per_pixel,
)
if score < best_score {
best_predictor = predictor
best_score = score
}
}
position = pdf_encode_png_push_scanline(
output, position, best_predictor, data, row_start, scanline_width, bytes_per_pixel,
)
}
Bytes::from_array(output[:position])
}
///|
/// Encode data with a PDF predictor.
///
/// Predictor `1` is identity, `2` is TIFF horizontal differencing, `10..14`
/// use fixed PNG row filters, and `15` chooses the lowest-scoring PNG filter
/// per row.
pub fn pdf_encode_predictor_view(
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
data : BytesView,
) -> PdfBytes raise PdfError {
match predictor {
1 => data.to_owned()
2 => pdf_encode_tiff_predictor(colors, bits_per_component, columns, data)
10 =>
pdf_encode_png_tagged_predictor(
0, colors, bits_per_component, columns, data,
)
11 =>
pdf_encode_png_sub_predictor(colors, bits_per_component, columns, data)
12 => pdf_encode_png_up_predictor(colors, bits_per_component, columns, data)
13 =>
pdf_encode_png_tagged_predictor(
3, colors, bits_per_component, columns, data,
)
14 =>
pdf_encode_png_tagged_predictor(
4, colors, bits_per_component, columns, data,
)
15 =>
pdf_encode_png_optimum_predictor(
colors, bits_per_component, columns, data,
)
_ => raise PredictorNotSupported(predictor)
}
}
///|
/// Owned-byte wrapper for `pdf_encode_predictor_view`.
///
/// Predictor `1` returns the original owned bytes unchanged.
pub fn pdf_encode_predictor(
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
data : PdfBytes,
) -> PdfBytes raise PdfError {
if predictor == 1 {
data
} else {
pdf_encode_predictor_view(
predictor, colors, bits_per_component, columns, data,
)
}
}
///|
fn pdf_decodeparms_value(dictionary : PdfObject) -> PdfObject? {
match dictionary.lookup_immediate(pdf_decodeparms_key()) {
Some(value) => Some(value)
None => dictionary.lookup_immediate(pdf_short_decodeparms_key())
}
}
///|
fn pdf_first_decodeparms(dictionary : PdfObject) -> PdfObject? raise PdfError {
match pdf_decodeparms_value(dictionary) {
None | Some(PdfNull) | Some(PdfArray([])) => None
Some(PdfDictionary(_) as value) => Some(value)
Some(PdfArray([PdfNull, ..])) => None
Some(PdfArray([PdfDictionary(_) as value, ..])) => Some(value)
_ => raise PredictorExpected
}
}
///|
fn PdfDocument::pdf_decodeparms_value_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> PdfObject? {
match self.pdf_decodeparms_entry_direct(dictionary) {
Some((_, value)) => Some(value)
None => None
}
}
///|
fn PdfDocument::pdf_decodeparms_entry_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> (PdfName, PdfObject)? {
match self.lookup_direct(pdf_decodeparms_key(), dictionary) {
Some(value) => Some((pdf_decodeparms_key(), value))
None =>
match self.lookup_direct(pdf_short_decodeparms_key(), dictionary) {
Some(value) => Some((pdf_short_decodeparms_key(), value))
None => None
}
}
}
///|
fn PdfDocument::pdf_first_decodeparms_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> PdfObject? raise PdfError {
match self.pdf_decodeparms_value_direct(dictionary) {
None | Some(PdfNull) | Some(PdfArray([])) => None
Some(PdfDictionary(_) as value) => Some(value)
Some(PdfArray([first, ..])) =>
match self.direct(first) {
PdfNull => None
PdfDictionary(_) as value => Some(value)
_ => raise PredictorExpected
}
_ => raise PredictorExpected
}
}
///|
fn pdf_decodeparms_int(
dictionary : PdfObject,
key : PdfName,
default : Int,
) -> Int raise PdfError {
match dictionary.lookup_immediate(key) {
None => default
Some(PdfInteger(value)) => value
_ => raise PredictorExpected
}
}
///|
fn PdfDocument::pdf_decodeparms_int_direct(
self : PdfDocument,
dictionary : PdfObject,
key : PdfName,
default : Int,
) -> Int raise PdfError {
match self.lookup_direct(key, dictionary) {
None => default
Some(PdfInteger(value)) => value
_ => raise PredictorExpected
}
}
///|
fn pdf_apply_first_predictor(
dictionary : PdfObject,
data : PdfBytes,
) -> PdfBytes raise PdfError {
match pdf_first_decodeparms(dictionary) {
None => data
Some(decodeparms) => {
let predictor = pdf_decodeparms_int(decodeparms, pdf_predictor_key(), 1)
if predictor == 1 {
data
} else {
pdf_decode_predictor(
predictor,
pdf_decodeparms_int(decodeparms, pdf_colors_key(), 1),
pdf_decodeparms_int(decodeparms, pdf_bits_per_component_key(), 8),
pdf_decodeparms_int(decodeparms, pdf_columns_key(), 1),
data,
)
}
}
}
}
///|
fn PdfDocument::pdf_apply_first_predictor_direct(
self : PdfDocument,
dictionary : PdfObject,
data : PdfBytes,
) -> PdfBytes raise PdfError {
match self.pdf_first_decodeparms_direct(dictionary) {
None => data
Some(decodeparms) => {
let predictor = self.pdf_decodeparms_int_direct(
decodeparms,
pdf_predictor_key(),
1,
)
if predictor == 1 {
data
} else {
pdf_decode_predictor(
predictor,
self.pdf_decodeparms_int_direct(decodeparms, pdf_colors_key(), 1),
self.pdf_decodeparms_int_direct(
decodeparms,
pdf_bits_per_component_key(),
8,
),
self.pdf_decodeparms_int_direct(decodeparms, pdf_columns_key(), 1),
data,
)
}
}
}
}
///|
fn pdf_lzw_early_change(dictionary : PdfObject) -> Int raise PdfError {
match pdf_first_decodeparms(dictionary) {
None => 1
Some(decodeparms) =>
pdf_decodeparms_int(decodeparms, pdf_early_change_key(), 1)
}
}
///|
fn PdfDocument::pdf_lzw_early_change_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> Int raise PdfError {
match self.pdf_first_decodeparms_direct(dictionary) {
None => 1
Some(decodeparms) =>
self.pdf_decodeparms_int_direct(decodeparms, pdf_early_change_key(), 1)
}
}
///|
fn pdf_ccitt_first_decodeparms(
dictionary : PdfObject,
) -> PdfObject? raise PdfError {
match pdf_decodeparms_value(dictionary) {
None | Some(PdfNull) | Some(PdfArray([])) | Some(PdfArray([PdfNull, ..])) =>
None
Some(PdfDictionary(_) as value) => Some(value)
Some(PdfArray([PdfDictionary(_) as value, ..])) => Some(value)
_ => raise CCITTDecodeParmsExpected
}
}
///|
fn PdfDocument::pdf_ccitt_first_decodeparms_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> PdfObject? raise PdfError {
match self.pdf_decodeparms_value_direct(dictionary) {
None | Some(PdfNull) | Some(PdfArray([])) => None
Some(PdfDictionary(_) as value) => Some(value)
Some(PdfArray([first, ..])) =>
match self.direct(first) {
PdfNull => None
PdfDictionary(_) as value => Some(value)
_ => raise CCITTDecodeParmsExpected
}
_ => raise CCITTDecodeParmsExpected
}
}
///|
fn pdf_ccitt_decodeparms_int(
dictionary : PdfObject,
key : PdfName,
default : Int,
) -> Int raise PdfError {
match dictionary.lookup_immediate(key) {
None => default
Some(PdfInteger(value)) => value
_ => raise CCITTDecodeParmsExpected
}
}
///|
fn PdfDocument::pdf_ccitt_decodeparms_int_direct(
self : PdfDocument,
dictionary : PdfObject,
key : PdfName,
default : Int,
) -> Int raise PdfError {
match self.lookup_direct(key, dictionary) {
None => default
Some(PdfInteger(value)) => value
_ => raise CCITTDecodeParmsExpected
}
}
///|
fn pdf_ccitt_decodeparms_bool(
dictionary : PdfObject,
key : PdfName,
default : Bool,
) -> Bool raise PdfError {
match dictionary.lookup_immediate(key) {
None => default
Some(PdfBoolean(value)) => value
_ => raise CCITTDecodeParmsExpected
}
}
///|
fn PdfDocument::pdf_ccitt_decodeparms_bool_direct(
self : PdfDocument,
dictionary : PdfObject,
key : PdfName,
default : Bool,
) -> Bool raise PdfError {
match self.lookup_direct(key, dictionary) {
None => default
Some(PdfBoolean(value)) => value
_ => raise CCITTDecodeParmsExpected
}
}
///|
fn pdf_ccitt_params(dictionary : PdfObject) -> PdfCCITTParams raise PdfError {
let defaults = pdf_ccitt_default_params()
match pdf_ccitt_first_decodeparms(dictionary) {
None => defaults
Some(decodeparms) =>
{
k: pdf_ccitt_decodeparms_int(decodeparms, pdf_ccitt_k_key(), defaults.k),
end_of_line: pdf_ccitt_decodeparms_bool(
decodeparms,
pdf_ccitt_end_of_line_key(),
defaults.end_of_line,
),
encoded_byte_align: pdf_ccitt_decodeparms_bool(
decodeparms,
pdf_ccitt_encoded_byte_align_key(),
defaults.encoded_byte_align,
),
columns: pdf_ccitt_decodeparms_int(
decodeparms,
pdf_columns_key(),
defaults.columns,
),
rows: pdf_ccitt_decodeparms_int(
decodeparms,
pdf_ccitt_rows_key(),
defaults.rows,
),
end_of_block: pdf_ccitt_decodeparms_bool(
decodeparms,
pdf_ccitt_end_of_block_key(),
defaults.end_of_block,
),
black_is_1: pdf_ccitt_decodeparms_bool(
decodeparms,
pdf_ccitt_black_is_1_key(),
defaults.black_is_1,
),
damaged_rows_before_error: pdf_ccitt_decodeparms_int(
decodeparms,
pdf_ccitt_damaged_rows_before_error_key(),
defaults.damaged_rows_before_error,
),
}
}
}
///|
fn PdfDocument::pdf_ccitt_params_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> PdfCCITTParams raise PdfError {
let defaults = pdf_ccitt_default_params()
match self.pdf_ccitt_first_decodeparms_direct(dictionary) {
None => defaults
Some(decodeparms) =>
{
k: self.pdf_ccitt_decodeparms_int_direct(
decodeparms,
pdf_ccitt_k_key(),
defaults.k,
),
end_of_line: self.pdf_ccitt_decodeparms_bool_direct(
decodeparms,
pdf_ccitt_end_of_line_key(),
defaults.end_of_line,
),
encoded_byte_align: self.pdf_ccitt_decodeparms_bool_direct(
decodeparms,
pdf_ccitt_encoded_byte_align_key(),
defaults.encoded_byte_align,
),
columns: self.pdf_ccitt_decodeparms_int_direct(
decodeparms,
pdf_columns_key(),
defaults.columns,
),
rows: self.pdf_ccitt_decodeparms_int_direct(
decodeparms,
pdf_ccitt_rows_key(),
defaults.rows,
),
end_of_block: self.pdf_ccitt_decodeparms_bool_direct(
decodeparms,
pdf_ccitt_end_of_block_key(),
defaults.end_of_block,
),
black_is_1: self.pdf_ccitt_decodeparms_bool_direct(
decodeparms,
pdf_ccitt_black_is_1_key(),
defaults.black_is_1,
),
damaged_rows_before_error: self.pdf_ccitt_decodeparms_int_direct(
decodeparms,
pdf_ccitt_damaged_rows_before_error_key(),
defaults.damaged_rows_before_error,
),
}
}
}
///|
fn pdf_decode_stream_filter_bytes(
filter : PdfName,
dictionary : PdfObject,
data : PdfBytes,
) -> PdfBytes raise PdfError {
if pdf_filter_is_one_of(
filter, pdf_filter_lzw_decode_name, pdf_filter_lzw_short_name,
) {
pdf_lzw_decode(data, early_change=pdf_lzw_early_change(dictionary))
} else if pdf_filter_is_one_of(
filter, pdf_filter_ccitt_fax_decode_name, pdf_filter_ccitt_fax_short_name,
) {
pdf_ccitt_decode_view(data, pdf_ccitt_params(dictionary))
} else {
pdf_decode_filter_bytes(filter, data)
}
}
///|
fn PdfDocument::pdf_decode_stream_filter_bytes_direct(
self : PdfDocument,
filter : PdfName,
dictionary : PdfObject,
data : PdfBytes,
) -> PdfBytes raise PdfError {
if pdf_filter_is_one_of(
filter, pdf_filter_lzw_decode_name, pdf_filter_lzw_short_name,
) {
pdf_lzw_decode(
data,
early_change=self.pdf_lzw_early_change_direct(dictionary),
)
} else if pdf_filter_is_one_of(
filter, pdf_filter_ccitt_fax_decode_name, pdf_filter_ccitt_fax_short_name,
) {
pdf_ccitt_decode_view(data, self.pdf_ccitt_params_direct(dictionary))
} else {
pdf_decode_filter_bytes(filter, data)
}
}
///|
fn pdf_decode_stream_onestage_value(
stream : PdfStream,
) -> PdfStream raise PdfError {
pdf_stream_dictionary_length(stream.dictionary)
let data = pdf_stream_data_bytes(stream.data)
match pdf_stream_first_filter(stream.dictionary) {
None => {
let dictionary = pdf_stream_replace_length(
stream.dictionary,
data.length(),
)
{ dictionary, data: StreamGot(data) }
}
Some(filter) => {
let decoded = pdf_decode_stream_filter_bytes(
filter,
stream.dictionary,
data,
)
let unpredicted = pdf_apply_first_predictor(stream.dictionary, decoded)
let dictionary = pdf_stream_replace_length(
pdf_remove_first_decoder(stream.dictionary),
unpredicted.length(),
)
{ dictionary, data: StreamGot(unpredicted) }
}
}
}
///|
fn PdfDocument::pdf_stream_dictionary_length_direct(
self : PdfDocument,
dictionary : PdfObject,
) -> Unit raise PdfError {
match self.lookup_direct(pdf_codec_length_key(), dictionary) {
Some(PdfInteger(_)) => ()
_ => raise StreamLengthExpected
}
}
///|
fn PdfDocument::pdf_decode_stream_onestage_value_direct(
self : PdfDocument,
stream : PdfStream,
) -> PdfStream raise PdfError {
self.pdf_stream_dictionary_length_direct(stream.dictionary)
let data = pdf_stream_data_bytes(stream.data)
match self.pdf_stream_first_filter_entry_direct(stream.dictionary) {
None => {
let dictionary = pdf_stream_replace_length(
stream.dictionary,
data.length(),
)
{ dictionary, data: StreamGot(data) }
}
Some((filter_key, filter_value, filter)) => {
let decoded = self.pdf_decode_stream_filter_bytes_direct(
filter,
stream.dictionary,
data,
)
let unpredicted = self.pdf_apply_first_predictor_direct(
stream.dictionary,
decoded,
)
let dictionary = pdf_stream_replace_length(
self.pdf_remove_first_decoder_direct(
stream.dictionary,
filter_key,
filter_value,
),
unpredicted.length(),
)
{ dictionary, data: StreamGot(unpredicted) }
}
}
}
///|
/// Decode exactly one filter stage from a stream object.
///
/// The first filter and matching decode parameters are removed from the stream
/// dictionary, and `/Length` is refreshed. Non-stream objects raise
/// `PdfError::ParseStreamExpected`.
pub fn pdf_decode_stream_onestage(
object : PdfObject,
) -> PdfObject raise PdfError {
match object {
PdfStreamObject(stream) =>
PdfStreamObject(pdf_decode_stream_onestage_value(stream))
_ => raise ParseStreamExpected
}
}
///|
fn PdfDocument::pdf_stream_has_filter_value_direct(
self : PdfDocument,
stream : PdfStream,
) -> Bool raise PdfError {
self.pdf_stream_first_filter_direct(stream.dictionary) != None
}
///|
fn PdfDocument::pdf_decode_stream_value_direct(
self : PdfDocument,
stream : PdfStream,
) -> PdfStream raise PdfError {
let mut current = stream
let mut stages = 0
while self.pdf_stream_has_filter_value_direct(current) {
if stages >= 1024 {
raise FilterExpected
}
current = self.pdf_decode_stream_onestage_value_direct(current)
stages += 1
}
self.pdf_decode_stream_onestage_value_direct(current)
}
///|
/// Decode all supported filter stages from a stream object in this document.
///
/// Indirect filter and decode-parameter objects are resolved through the
/// document. Unsupported filters raise `PdfError::FilterNotSupported`.
pub fn PdfDocument::pdf_decode_stream(
self : PdfDocument,
object : PdfObject,
) -> PdfObject raise PdfError {
match self.direct(object) {
PdfStreamObject(stream) =>
PdfStreamObject(self.pdf_decode_stream_value_direct(stream))
_ => raise ParseStreamExpected
}
}
///|
fn PdfDocument::pdf_decode_stream_until_unknown_value_direct(
self : PdfDocument,
stream : PdfStream,
) -> PdfStream raise PdfError {
let mut current = stream
let mut stages = 0
let mut stopped = false
while !stopped && self.pdf_stream_has_filter_value_direct(current) {
if stages >= 1024 {
raise FilterExpected
}
try self.pdf_decode_stream_onestage_value_direct(current) catch {
FilterNotSupported(_) => stopped = true
error => raise error
} noraise {
decoded => {
current = decoded
stages += 1
}
}
}
if stopped {
current
} else {
self.pdf_decode_stream_onestage_value_direct(current)
}
}
///|
/// Decode supported filter stages until the first unsupported filter.
///
/// The unsupported filter and remaining filters stay in the stream dictionary;
/// malformed supported filters still raise.
pub fn PdfDocument::pdf_decode_stream_until_unknown(
self : PdfDocument,
object : PdfObject,
) -> PdfObject raise PdfError {
match self.direct(object) {
PdfStreamObject(stream) =>
PdfStreamObject(self.pdf_decode_stream_until_unknown_value_direct(stream))
_ => raise ParseStreamExpected
}
}
///|
fn pdf_decode_stream_until_unknown_value(
stream : PdfStream,
) -> PdfStream raise PdfError {
pdf_document_empty().pdf_decode_stream_until_unknown_value_direct(stream)
}
///|
/// Decode supported filter stages on every parsed stream in the document.
///
/// Streams already marked as decrypted stay in that state. Objects are updated
/// in place and logged as altered when their stream data changes.
pub fn PdfDocument::pdf_decode_streams_until_unknown(
self : PdfDocument,
) -> Unit raise PdfError {
for number in self.objects.object_numbers() {
let entry = self.objects.get_entry(number).unwrap()
match entry.data {
ObjectParsed(PdfStreamObject(stream)) => {
let decoded = self.pdf_decode_stream_until_unknown_value_direct(stream)
self.objects.set_entry(number, {
data: ObjectParsed(PdfStreamObject(decoded)),
generation: entry.generation,
})
self.objects.log.push((number, ObjectAltered))
}
ObjectParsedAlreadyDecrypted(PdfStreamObject(stream)) => {
let decoded = self.pdf_decode_stream_until_unknown_value_direct(stream)
self.objects.set_entry(number, {
data: ObjectParsedAlreadyDecrypted(PdfStreamObject(decoded)),
generation: entry.generation,
})
self.objects.log.push((number, ObjectAltered))
}
_ => ()
}
}
}
///|
fn pdf_stream_has_filter_value(stream : PdfStream) -> Bool raise PdfError {
pdf_stream_first_filter(stream.dictionary) != None
}
///|
fn pdf_decode_stream_value(stream : PdfStream) -> PdfStream raise PdfError {
let mut current = stream
let mut stages = 0
while pdf_stream_has_filter_value(current) {
if stages >= 1024 {
raise FilterExpected
}
current = pdf_decode_stream_onestage_value(current)
stages += 1
}
pdf_decode_stream_onestage_value(current)
}
///|
/// Decode all supported filter stages from a direct stream object.
pub fn pdf_decode_stream(object : PdfObject) -> PdfObject raise PdfError {
match object {
PdfStreamObject(stream) => PdfStreamObject(pdf_decode_stream_value(stream))
_ => raise ParseStreamExpected
}
}
///|
fn pdf_add_filter_to_dictionary(
dictionary : PdfObject,
filter : PdfName,
) -> PdfObject raise PdfError {
let filter_object = PdfNameObject(filter)
match dictionary.lookup_immediate(pdf_filter_key()) {
None => dictionary.add_dict_entry(pdf_filter_key(), filter_object)
Some(PdfNameObject(existing)) =>
dictionary.replace_dict_entry(
pdf_filter_key(),
PdfArray([filter_object, PdfNameObject(existing)]),
)
Some(PdfArray(filters)) =>
dictionary.replace_dict_entry(
pdf_filter_key(),
PdfArray([filter_object, ..filters]),
)
_ => raise FilterExpected
}
}
///|
/// Encode a stream object with one filter.
///
/// The new filter is prepended to existing filters and `/Length` is refreshed.
/// When `only_if_smaller` is true, the original stream is returned if encoded
/// data is not smaller than the original stream data.
pub fn pdf_encode_stream_with_filter(
filter : PdfName,
object : PdfObject,
only_if_smaller? : Bool = false,
flate_level? : Int? = None,
) -> PdfObject raise PdfError {
match object {
PdfStreamObject(stream) => {
let data = pdf_stream_data_bytes(stream.data)
let encoded = match flate_level {
Some(level) =>
pdf_encode_filter_bytes_with_flate_level(filter, data, level)
None => pdf_encode_filter_bytes(filter, data)
}
if only_if_smaller && !pdf_encoded_stream_is_smaller(encoded, data) {
object
} else {
let dictionary = pdf_stream_replace_length(
pdf_add_filter_to_dictionary(stream.dictionary, filter),
encoded.length(),
)
pdf_stream(dictionary, StreamGot(encoded))
}
}
_ => raise ParseStreamExpected
}
}
///|
fn pdf_encoded_stream_is_smaller(encoded : PdfBytes, data : PdfBytes) -> Bool {
encoded.length() + 20 < data.length()
}
///|
fn pdf_predictor_decodeparms(
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
) -> PdfObject {
pdf_dictionary([
(pdf_predictor_key(), PdfInteger(predictor)),
(pdf_colors_key(), PdfInteger(colors)),
(pdf_bits_per_component_key(), PdfInteger(bits_per_component)),
(pdf_columns_key(), PdfInteger(columns)),
])
}
///|
fn pdf_ccitt_encoding_decodeparms(
k : Int,
columns : Int,
rows : Int,
) -> PdfObject {
pdf_dictionary([
(pdf_ccitt_k_key(), PdfInteger(k)),
(pdf_columns_key(), PdfInteger(columns)),
(pdf_ccitt_rows_key(), PdfInteger(rows)),
])
}
///|
fn pdf_add_decodeparms_to_dictionary(
dictionary : PdfObject,
decodeparms : PdfObject,
) -> PdfObject raise PdfError {
match dictionary.lookup_immediate(pdf_decodeparms_key()) {
None => dictionary.add_dict_entry(pdf_decodeparms_key(), decodeparms)
Some(PdfDictionary(_) as existing) =>
dictionary.replace_dict_entry(
pdf_decodeparms_key(),
PdfArray([decodeparms, existing]),
)
Some(PdfArray(values)) =>
dictionary.replace_dict_entry(
pdf_decodeparms_key(),
PdfArray([decodeparms, ..values]),
)
_ => raise PredictorExpected
}
}
///|
fn pdf_encode_stream_with_ccitt(
columns : Int,
rows : Int,
k : Int,
object : PdfObject,
only_if_smaller : Bool,
) -> PdfObject raise PdfError {
match object {
PdfStreamObject(stream) => {
let data = pdf_stream_data_bytes(stream.data)
let encoded = if k < 0 {
pdf_ccitt_encode_g4_view(data, columns, rows)
} else {
pdf_ccitt_encode_g3_1d_view(data, columns, rows)
}
if only_if_smaller && !pdf_encoded_stream_is_smaller(encoded, data) {
object
} else {
let dictionary = pdf_stream_replace_length(
pdf_add_filter_to_dictionary(
pdf_add_decodeparms_to_dictionary(
stream.dictionary,
pdf_ccitt_encoding_decodeparms(k, columns, rows),
),
pdf_filter_ccitt_fax_decode_name,
),
encoded.length(),
)
pdf_stream(dictionary, StreamGot(encoded))
}
}
_ => raise ParseStreamExpected
}
}
///|
/// Encode a stream after first applying predictor encoding.
///
/// `/DecodeParms` are added for the predictor, the filter is prepended, and
/// `/Length` is refreshed. `only_if_smaller` compares encoded data against the
/// original stream data.
pub fn pdf_encode_stream_with_filter_and_predictor(
filter : PdfName,
predictor : Int,
colors : Int,
bits_per_component : Int,
columns : Int,
object : PdfObject,
only_if_smaller? : Bool = false,
flate_level? : Int? = None,
) -> PdfObject raise PdfError {
match object {
PdfStreamObject(stream) => {
let data = pdf_stream_data_bytes(stream.data)
let predicted = pdf_encode_predictor(
predictor, colors, bits_per_component, columns, data,
)
let encoded = match flate_level {
Some(level) =>
pdf_encode_filter_bytes_with_flate_level(filter, predicted, level)
None => pdf_encode_filter_bytes(filter, predicted)
}
if only_if_smaller && !pdf_encoded_stream_is_smaller(encoded, data) {
object
} else {
let dictionary = pdf_stream_replace_length(
pdf_add_filter_to_dictionary(
pdf_add_decodeparms_to_dictionary(
stream.dictionary,
pdf_predictor_decodeparms(
predictor, colors, bits_per_component, columns,
),
),
filter,
),
encoded.length(),
)
pdf_stream(dictionary, StreamGot(encoded))
}
}
_ => raise ParseStreamExpected
}
}
///|
/// Encode a stream using a typed encoding description.
///
/// CCITT encodings write fax `/DecodeParms`; other encodings may optionally
/// apply a predictor before the filter. CCITT does not accept predictors.
pub fn pdf_encode_stream_with_encoding(
encoding : PdfStreamEncoding,
object : PdfObject,
only_if_smaller? : Bool = false,
predictor? : PdfStreamPredictor? = None,
predictor_columns? : Int = 1,
flate_level? : Int? = None,
) -> PdfObject raise PdfError {
let filter = encoding.filter_name()
match encoding {
PdfStreamCCITT(columns, rows) =>
match predictor {
None =>
pdf_encode_stream_with_ccitt(
columns, rows, 0, object, only_if_smaller,
)
Some(_) => raise PredictorExpected
}
PdfStreamCCITTG4(columns, rows) =>
match predictor {
None =>
pdf_encode_stream_with_ccitt(
columns, rows, -1, object, only_if_smaller,
)
Some(_) => raise PredictorExpected
}
_ =>
match predictor {
None =>
pdf_encode_stream_with_filter(
filter,
object,
only_if_smaller~,
flate_level~,
)
Some(predictor) =>
pdf_encode_stream_with_filter_and_predictor(
filter,
predictor.number(),
1,
8,
predictor_columns,
object,
only_if_smaller~,
flate_level~,
)
}
}
}
///|
fn pdf_run_length_flush_literal(
output : Array[Byte],
position : Int,
data : BytesView,
start : Int,
length : Int,
) -> Int {
let mut current = position
if length > 0 {
output[current] = (length - 1).to_byte()
current += 1
for i in 0.. PdfBytes {
let output = Array::make(
data.length() + (data.length() + 127) / 128 + 1,
b'\x00',
)
let mut position = 0
let mut literal_start = 0
let mut literal_length = 0
let mut index = 0
while index < data.length() {
let value = data[index]
let mut run_length = 1
while index + run_length < data.length() &&
data[index + run_length] == value &&
run_length < 128 {
run_length += 1
}
if run_length >= 3 {
position = pdf_run_length_flush_literal(
output, position, data, literal_start, literal_length,
)
literal_length = 0
output[position] = (257 - run_length).to_byte()
output[position + 1] = value
position += 2
index += run_length
} else {
if literal_length == 0 {
literal_start = index
}
if literal_length + run_length > 128 {
position = pdf_run_length_flush_literal(
output, position, data, literal_start, literal_length,
)
literal_start = index
literal_length = 0
}
literal_length += run_length
index += run_length
}
}
position = pdf_run_length_flush_literal(
output, position, data, literal_start, literal_length,
)
output[position] = 128
Bytes::from_array(output[:position + 1])
}
///|
/// Owned-byte wrapper for `pdf_run_length_encode_view`.
pub fn pdf_run_length_encode(data : PdfBytes) -> PdfBytes {
pdf_run_length_encode_view(data)
}
///|
fn pdf_run_length_decoded_length(data : BytesView) -> Int {
let mut length = 0
let mut index = 0
let mut done = false
while !done && index < data.length() {
let control = data[index].to_int()
index += 1
if control < 128 {
let wanted = control + 1
let available = if index + wanted > data.length() {
data.length() - index
} else {
wanted
}
length += available
index += available
} else if control > 128 {
if index < data.length() {
length += 257 - control
index += 1
} else {
done = true
}
} else {
done = true
}
}
length
}
///|
/// Decode PDF RunLengthDecode data.
///
/// Decoding stops at the `128` end marker or end of input. Truncated trailing
/// runs are decoded as far as bytes are available, matching the tolerant reader
/// behavior used by CamlPDF.
pub fn pdf_run_length_decode_view(data : BytesView) -> PdfBytes {
let output = Array::make(pdf_run_length_decoded_length(data), b'\x00')
let mut position = 0
let mut index = 0
let mut done = false
while !done && index < data.length() {
let control = data[index].to_int()
index += 1
if control < 128 {
let wanted = control + 1
let available = if index + wanted > data.length() {
data.length() - index
} else {
wanted
}
for i in 0.. 128 {
if index < data.length() {
let value = data[index]
index += 1
for _ in 0..<(257 - control) {
output[position] = value
position += 1
}
} else {
done = true
}
} else {
done = true
}
}
Bytes::from_array(output[:position])
}
///|
/// Owned-byte wrapper for `pdf_run_length_decode_view`.
pub fn pdf_run_length_decode(data : PdfBytes) -> PdfBytes {
pdf_run_length_decode_view(data)
}