commonMain.aws.smithy.kotlin.runtime.serde.xml.deserialization.XmlLexer.kt Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of serde-xml-jvm Show documentation
Show all versions of serde-xml-jvm Show documentation
XML serialization and deserialization for Smithy services generated by smithy-kotlin
/*
* Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
* SPDX-License-Identifier: Apache-2.0.
*/
package aws.smithy.kotlin.runtime.serde.xml.deserialization
import aws.smithy.kotlin.runtime.serde.DeserializationException
import aws.smithy.kotlin.runtime.serde.xml.XmlToken
import aws.smithy.kotlin.runtime.util.text.codePointToChars
private val decimalCharRef = "#([0-9]+)".toRegex()
private val hexCharRef = "#x([0-9a-fA-F]+)".toRegex()
// https://www.w3.org/TR/xml/#sec-predefined-ent
private val namedRefs = mapOf(
"lt" to '<',
"gt" to '>',
"amp" to '&',
"apos" to '\'',
"quot" to '"',
).mapValues { charArrayOf(it.value) }
private fun XmlToken.QualifiedName.isXmlns(): Boolean = (local == "xmlns" && prefix == null) || (prefix == "xmlns")
private fun XmlToken.QualifiedName.xmlnsPrefix(): String? = if (local == "xmlns") null else local
private typealias AttributeMap = Map
private fun AttributeMap.extractNsDeclarations(): Pair> {
val parts = toList().partition { it.first.isXmlns() }
val attributes = parts.second.toMap()
val nsDeclarations = parts.first.map { XmlToken.Namespace(it.second, it.first.xmlnsPrefix()) }
return attributes to nsDeclarations
}
/**
* A lexer that scans a [StringTextStream] and reads [XmlToken] elements.
*/
public class XmlLexer(internal val source: StringTextStream) {
private var state: LexerState = LexerState.Initial
public val endOfDocument: Boolean
get() = state == LexerState.EndOfDocument
/**
* Throws a [DeserializationException] with the given message and location string.
* @param msg The error message to include with the exception.
*/
@Suppress("NOTHING_TO_INLINE")
internal inline fun error(msg: String): Nothing = source.error(msg)
/**
* Parses the next token from the source.
* @return The next [XmlToken] in the source, or null if the end of the source has been reached.
*/
public fun parseNext(): XmlToken? = when (val currentState = this.state) {
LexerState.EndOfDocument -> null
is LexerState.Tag.EmptyTag -> {
state = currentState.parent ?: LexerState.EndOfDocument
// In this case we don't actually need to read more from the source because an empty tag is represented as
// a BeginElement (which was previously returned) and an EndElement (which will be returned now).
XmlToken.EndElement(currentState.depth, currentState.name)
}
is LexerState.Tag.OpenTag ->
if (source.peekMatches("<") && !source.peekMatches(" {
skipPreprocessingInstructions()
state = LexerState.BeforeRootTag
parseNext()
}
LexerState.BeforeRootTag -> {
skipSpace()
readTagToken()
}
}
/**
* Reads an attribute key and value from the source.
*/
private fun readAttribute(): Pair {
val name = readName()
skipSpace()
val equals = source.readOrThrow("trying to read attribute equals")
if (equals != '=') error("Unexpected '$equals' while trying to read attribute equals")
skipSpace()
val value = readQuoted()
return name to value
}
/**
* Reads a CDATA section from the source. This assumes that the source position is immediately after the `", "trying to read CDATA content")
source.advance(3, "trying to read end of CATA") // Skip trailing `]]>`
return body
}
/**
* Reads a tag or attribute name from the source.
*/
private fun readName(): XmlToken.QualifiedName = source.readWhileXmlName().qualify()
/**
* Reads a quoted string from the source. The quotes may be single (') or double (").
*/
private fun readQuoted(): String {
val quoteChar = source.readOrThrow("trying to read attribute value")
if (quoteChar != '\'' && quoteChar != '"') {
error("Unexpected '$quoteChar' while trying to read attribute value")
}
return buildString {
while (true) {
when (val c = source.readOrThrow("trying to read a string")) {
'&' -> append(readReference())
'<' -> error("Unexpected '<' while trying to read a string")
quoteChar -> break
else -> append(c)
}
}
}
}
/**
* Reads a character reference (e.g., `#x1a2b;`) or entity reference (e.g., `apos;`) from the source. This assumes
* that the leading `&` has already been consumed.
*/
private fun readReference(): CharArray {
val ref = source.readUntil(";", "trying to read a char/entity reference")
source.advance(1, "trying to read the end of a char/entity reference")
val decimalMatch = decimalCharRef.matchEntire(ref)
if (decimalMatch != null) {
val code = decimalMatch.groupValues[1].toInt()
return Char.codePointToChars(code)
}
val hexMatch = hexCharRef.matchEntire(ref)
if (hexMatch != null) {
val code = hexMatch.groupValues[1].toInt(radix = 16)
return Char.codePointToChars(code)
}
return namedRefs.getOrElse(ref) { error("Unknown reference '$ref'") }
}
/**
* Reads a tag token from the source.
*/
private fun readTagToken(): XmlToken {
val lt = source.readOrThrow("looking for the start of a tag")
if (lt != '<') error("Unexpected character '$lt' while looking for the start of a tag")
if (source.advanceIf("!--")) {
skipComment()
return parseNext()!!
}
val token = if (source.advanceIf("/")) {
val openTagState = state as LexerState.Tag.OpenTag
val expectedName = openTagState.name
val actualName = readName()
if (actualName != expectedName) {
error("Unexpected '/$actualName' tag while looking for '/$expectedName' tag")
}
skipSpace()
val ch = source.readOrThrow("looking for the end of a tag")
if (ch != '>') error("Unexpected character '$ch' while looking for the end of a tag")
state = openTagState.parent ?: LexerState.EndOfDocument
XmlToken.EndElement(openTagState.depth, actualName)
} else {
val openTagState = (state as? LexerState.Tag.OpenTag)?.copy(seenChildren = true)
val name = readName()
skipSpace()
val allAttributes = mutableMapOf()
var selfClosingTag = false
while (true) {
when (source.readOrThrow("looking for the end of a tag")) {
'/' -> {
selfClosingTag = true
break
}
'>' -> break
else -> {
source.rewind(1, "looking for the beginning of an attribute")
allAttributes += readAttribute()
}
}
skipSpace()
}
val (attributes, nsDeclarations) = allAttributes.extractNsDeclarations()
val nextState = if (selfClosingTag) {
val gt = source.readOrThrow("looking for the end of a tag")
if (gt != '>') error("Unexpected characters while looking for the end of a tag")
LexerState.Tag.EmptyTag(name, openTagState)
} else {
LexerState.Tag.OpenTag(name, openTagState, false)
}
state = nextState
XmlToken.BeginElement(nextState.depth, name, attributes, nsDeclarations)
}
return token
}
/**
* Reads a text token from the source.
*/
private fun readTextToken(): XmlToken {
var isBlank = true
val text = buildString {
while (true) {
when (val nextCh = source.readOrThrow("reading text node")) {
' ', '\t', '\r', '\n' -> append(nextCh)
'<' -> when {
source.advanceIf("!--") -> skipComment()
source.advanceIf("![CDATA[") -> {
append(readCdata())
isBlank = false
}
else -> {
source.rewind(1, "looking for the beginning of a tag")
break
}
}
'&' -> {
isBlank = false
append(readReference())
}
else -> {
isBlank = false
append(nextCh)
}
}
}
}
val openTagState = state as LexerState.Tag.OpenTag
val openTagIsMostRecent = openTagState.seenChildren
val closeTagIsNext = source.peekMatches("`).
*/
private fun skipComment() {
source.readThrough("-->", "looking for the end of a comment")
}
/**
* Skips preprocessing instructions (e.g., ``) if any are found. Also skips spaces before/after
* preprocessing instructions.
*/
private fun skipPreprocessingInstructions() {
skipSpace()
while (source.advanceIf("")) {
readAttribute()
skipSpace()
}
skipSpace()
}
}
/**
* Skips whitespaces.
*/
private fun skipSpace() {
source.advanceWhileSpace()
}
/**
* Parses a string name into an [XmlToken.QualifiedName].
*/
private fun String.qualify(): XmlToken.QualifiedName {
val parts = split(':')
if (parts.any(String::isEmpty)) error("Cannot understand qualified name '$this'")
return when (parts.size) {
1 -> XmlToken.QualifiedName(parts[0])
2 -> XmlToken.QualifiedName(parts[1], parts[0])
else -> error("Cannot understand qualified name '$this'")
}
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy