Skip to content

fix(rss): properly detect charset and decode French and non-UTF8 full content #1323

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
heberjeur wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix(rss): properly detect charset and decode French and non-UTF8 full content #1323

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
82 changes: 49 additions & 33 deletions app/src/main/java/me/ash/reader/infrastructure/rss/RssHelper.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -90,7 +90,9 @@ constructor(

private fun toHttpContentType(contentType: String?): String =
contentType?.let {
if (it.contains("charset=", ignoreCase = true)) it else "$it; charset=UTF-8"
if (it.contains("charset=", ignoreCase = true)) {
it.replace(',', ';')
} else "$it; charset=UTF-8"
} ?: "text/xml; charset=UTF-8"

private fun parseFeed(body: ByteArray, httpContentType: String): SyndFeed =
Expand All @@ -112,44 +114,58 @@ constructor(
return (preferred ?: fallback)?.absUrl("href")?.takeIf { it.isNotBlank() }
}

fun detectHtmlCharset(contentTypeHeader: String?, bodyBytes: ByteArray): Charset {
// 1. Check HTTP Content-Type header (handles semicolons, commas, and quotes)
if (!contentTypeHeader.isNullOrBlank()) {
val match = Regex("""(?i)charset\s*=\s*["']?([a-zA-Z0-9_-]+)""").find(contentTypeHeader)
if (match != null) {
val charsetName = match.groupValues[1].trim('\'', '"', ';', ' ')
runCatching { return Charset.forName(charsetName) }
}
}

// 2. Check BOM (Byte Order Mark)
if (bodyBytes.size >= 3 && bodyBytes[0] == 0xEF.toByte() && bodyBytes[1] == 0xBB.toByte() && bodyBytes[2] == 0xBF.toByte()) {
return Charsets.UTF_8
}
if (bodyBytes.size >= 2 && bodyBytes[0] == 0xFE.toByte() && bodyBytes[1] == 0xFF.toByte()) {
return Charsets.UTF_16BE
}
if (bodyBytes.size >= 2 && bodyBytes[0] == 0xFF.toByte() && bodyBytes[1] == 0xFE.toByte()) {
return Charsets.UTF_16LE
}

// 3. Inspect HTML <head> for <meta charset="..."> or <meta http-equiv="content-type" content="...">
val previewLength = minOf(bodyBytes.size, 4096)
val asciiPreview = String(bodyBytes, 0, previewLength, Charsets.ISO_8859_1)

val metaCharsetMatch = Regex("""(?i)<meta[^>]+charset\s*=\s*["']?([a-zA-Z0-9_-]+)""").find(asciiPreview)
if (metaCharsetMatch != null) {
val charsetName = metaCharsetMatch.groupValues[1].trim('\'', '"', ';', ' ')
runCatching { return Charset.forName(charsetName) }
}

val metaHttpEquivMatch = Regex("""(?i)<meta[^>]+content\s*=\s*["'][^"']*charset=([a-zA-Z0-9_-]+)""").find(asciiPreview)
?: Regex("""(?i)http-equiv\s*=\s*["']?content-type["']?[^>]+content\s*=\s*["'][^"']*charset=([a-zA-Z0-9_-]+)""").find(asciiPreview)
if (metaHttpEquivMatch != null) {
val charsetName = metaHttpEquivMatch.groupValues[1].trim('\'', '"', ';', ' ')
runCatching { return Charset.forName(charsetName) }
}

// 4. Default to UTF-8
return Charsets.UTF_8
}

@Throws(Exception::class)
suspend fun parseFullContent(link: String, title: String): String {
return withContext(ioDispatcher) {
val response = response(okHttpClient, link)
if (response.commonIsSuccessful) {
val responseBody = response.body
val charset = responseBody.contentType()?.charset()
val content =
responseBody.source().use {
if (charset != null) {
return@use it.readString(charset)
}

val peekContent = it.peek().readString(Charsets.UTF_8)

val charsetFromMeta =
runCatching {
val element =
Jsoup.parse(peekContent, link)
.selectFirst("meta[http-equiv=content-type]")
return@runCatching if (element == null) Charsets.UTF_8
else {
element
.attr("content")
.substringAfter("charset=")
.removeSurrounding("\"")
.lowercase()
.let { Charset.forName(it) }
}
}
.getOrDefault(Charsets.UTF_8)

if (charsetFromMeta == Charsets.UTF_8) {
peekContent
} else {
it.readString(charsetFromMeta)
}
}
val contentTypeHeader = response.header("Content-Type")
val bytes = responseBody.bytes()
val charset = detectHtmlCharset(contentTypeHeader, bytes)
val content = String(bytes, charset)

val articleContent = Readability.parseToElement(content, link)
articleContent?.let {
Expand Down
56 changes: 56 additions & 0 deletions app/src/test/java/me/ash/reader/infrastructure/rss/RssHelperTest.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,4 +87,60 @@ class RssHelperTest {
"""
Assert.assertEquals(imageUrlString, rssHelper.findThumbnail(case))
}

@Test
fun testDetectHtmlCharsetFromHeaderWithComma() {
val bytes = "<html><body>Test</body></html>".toByteArray(Charsets.ISO_8859_1)
val charset = rssHelper.detectHtmlCharset("text/html, charset=iso-8859-1", bytes)
Assert.assertEquals(Charsets.ISO_8859_1, charset)
}

@Test
fun testDetectHtmlCharsetFromMetaCharset() {
val html = """
<!DOCTYPE html>
<html>
<head>
<meta charset="iso-8859-1">
<title>Test</title>
</head>
<body>Perplexity a lancé son abonnement</body>
</html>
""".trimIndent()
val bytes = html.toByteArray(Charsets.ISO_8859_1)
val charset = rssHelper.detectHtmlCharset("text/html", bytes)
Assert.assertEquals(Charsets.ISO_8859_1, charset)

val decoded = String(bytes, charset)
Assert.assertTrue(decoded.contains("Perplexity a lancé son abonnement"))
Assert.assertFalse(decoded.contains("\uFFFD"))
}

@Test
fun testDetectHtmlCharsetFromMetaHttpEquiv() {
val html = """
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
</head>
<body>Déjà vu été français</body>
</html>
""".trimIndent()
val win1252 = java.nio.charset.Charset.forName("windows-1252")
val bytes = html.toByteArray(win1252)
val charset = rssHelper.detectHtmlCharset(null, bytes)
Assert.assertEquals(win1252, charset)

val decoded = String(bytes, charset)
Assert.assertTrue(decoded.contains("Déjà vu été français"))
}

@Test
fun testFrenchCharactersDecodingAccuracy() {
val originalFrenchText = "Club des développeurs : Actualités, cours, tutoriels & événements d'ingénierie"
val bytes = originalFrenchText.toByteArray(Charsets.ISO_8859_1)
val charset = rssHelper.detectHtmlCharset("text/html, charset=iso-8859-1", bytes)
val decoded = String(bytes, charset)
Assert.assertEquals(originalFrenchText, decoded)
}
}
Loading