///|
fn pdf_metadata_add_unique(values : Array[String], value : String) -> Unit {
if value != "" && !values.contains(value) {
values.push(value)
}
}
///|
fn pdf_metadata_namespace_prefixes(
data : BytesView,
xml_uri : String,
fallback : String,
) -> Array[String] {
let prefixes : Array[String] = []
pdf_metadata_add_unique(prefixes, fallback)
let namespace_bytes = @ascii.encode(xml_uri)
let xmlns : ReadOnlyArray[Int] = [120, 109, 108, 110, 115, 58]
let mut search_start = 0
while search_start < data.length() {
let namespace_start = pdf_metadata_find_bytes_from(
data, search_start, namespace_bytes,
)
if namespace_start < 0 {
search_start = data.length()
} else {
let mut attribute_start = namespace_start
while attribute_start > 0 &&
!pdf_metadata_is_xml_separator(data[attribute_start - 1]) {
attribute_start -= 1
}
if @core.pdf_view_matches_ascii_at(data, attribute_start, xmlns) {
let prefix_start = attribute_start + xmlns.length()
let mut prefix_end = prefix_start
while prefix_end < namespace_start &&
data[prefix_end].to_int() != 61 &&
data[prefix_end].to_int() != 34 &&
data[prefix_end].to_int() != 39 {
prefix_end += 1
}
match pdf_metadata_decode_xml_text(data[prefix_start:prefix_end]) {
Some(prefix) => pdf_metadata_add_unique(prefixes, prefix)
None => ()
}
}
search_start = namespace_start + namespace_bytes.length()
}
}
prefixes
}
///|
fn pdf_metadata_extract_quoted_xml_value(
data : BytesView,
start : Int,
quote : Int,
) -> String? {
let end = pdf_metadata_find_byte(data, start, quote)
if end < 0 {
None
} else {
pdf_metadata_decode_xml_text(data[start:end])
}
}
///|
fn pdf_metadata_extract_xmp_attribute(
data : BytesView,
prefix : String,
name : String,
) -> String? {
let double_pattern = @ascii.encode(prefix + ":" + name + "=\"")
let double_start = pdf_metadata_find_bytes_from(data, 0, double_pattern)
if double_start >= 0 {
pdf_metadata_extract_quoted_xml_value(
data,
double_start + double_pattern.length(),
34,
)
} else {
let single_pattern = @ascii.encode(prefix + ":" + name + "='")
let single_start = pdf_metadata_find_bytes_from(data, 0, single_pattern)
if single_start >= 0 {
pdf_metadata_extract_quoted_xml_value(
data,
single_start + single_pattern.length(),
39,
)
} else {
None
}
}
}
///|
fn pdf_metadata_combine_list_items(values : ArrayView[String]) -> String {
let output = StringBuilder()
for value in values {
if !output.is_empty() {
output.write_string(", ")
}
output.write_string(value)
}
output.to_string()
}
///|
fn pdf_metadata_extract_rdf_list_items_for_prefix(
value : BytesView,
prefix : String,
) -> Array[String] {
let items : Array[String] = []
let open_pattern = @ascii.encode("<" + prefix + ":li")
let close_pattern = @ascii.encode("")
let mut search_start = 0
while search_start < value.length() {
let item_start = pdf_metadata_find_bytes_from(
value, search_start, open_pattern,
)
if item_start < 0 {
search_start = value.length()
} else {
let name_end = item_start + open_pattern.length()
if name_end >= value.length() ||
!pdf_metadata_xml_tag_boundary(value[name_end]) {
search_start = name_end
} else {
let item_tag_end = pdf_metadata_find_byte(value, name_end, 62)
if item_tag_end < 0 {
search_start = value.length()
} else if value[item_tag_end - 1].to_int() == 47 {
search_start = item_tag_end + 1
} else {
let item_value_start = item_tag_end + 1
let item_value_end = pdf_metadata_find_bytes_from(
value, item_value_start, close_pattern,
)
if item_value_end < 0 {
search_start = value.length()
} else {
match
pdf_metadata_decode_xml_text(
value[item_value_start:item_value_end],
) {
Some(item) => items.push(item)
None => ()
}
search_start = item_value_end + close_pattern.length()
}
}
}
}
}
items
}
///|
fn pdf_metadata_extract_rdf_list_items(
data : BytesView,
value : BytesView,
) -> String? {
let items : Array[String] = []
let prefixes = pdf_metadata_namespace_prefixes(
data, @metadata.pdf_metadata_rdf_namespace, "rdf",
)
for prefix in prefixes {
for item in pdf_metadata_extract_rdf_list_items_for_prefix(value, prefix) {
items.push(item)
}
}
if items.length() == 0 {
None
} else {
Some(pdf_metadata_combine_list_items(items))
}
}
///|
fn pdf_metadata_extract_xmp_element(
data : BytesView,
prefix : String,
name : String,
) -> String? {
let open_pattern = @ascii.encode("<" + prefix + ":" + name)
let tag_start = pdf_metadata_find_bytes_from(data, 0, open_pattern)
if tag_start < 0 {
None
} else {
let tag_end = pdf_metadata_find_byte(
data,
tag_start + open_pattern.length(),
62,
)
if tag_end < 0 || data[tag_end - 1].to_int() == 47 {
None
} else {
let close_pattern = @ascii.encode("")
let value_start = tag_end + 1
let value_end = pdf_metadata_find_bytes_from(
data, value_start, close_pattern,
)
if value_end < 0 {
None
} else {
let value = data[value_start:value_end]
match pdf_metadata_extract_rdf_list_items(data, value) {
Some(items) => Some(items)
None => pdf_metadata_decode_xml_text(value)
}
}
}
}
}
///|
fn pdf_metadata_get_xmp_data(
data : BytesView,
xml_uri : String,
fallback_prefix : String,
name : String,
) -> String? {
let prefixes = pdf_metadata_namespace_prefixes(data, xml_uri, fallback_prefix)
for prefix in prefixes {
match pdf_metadata_extract_xmp_attribute(data, prefix, name) {
Some(value) => break Some(value)
None =>
match pdf_metadata_extract_xmp_element(data, prefix, name) {
Some(value) => break Some(value)
None => ()
}
}
} nobreak {
None
}
}
///|
/// Return the first XMP metadata value matching an XML namespace URI and local
/// name. This is the tree-free counterpart of cpdfmetadata's `get_data_for`,
/// covering namespace aliases, attributes, simple element text, and RDF lists.
pub fn pdf_metadata_get_data_for_bytes(
xml_uri : String,
name : String,
metadata : BytesView,
) -> String? {
pdf_metadata_get_xmp_data(
metadata,
xml_uri,
pdf_metadata_namespace_fallback_prefix(xml_uri),
name,
)
}