1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
|
package org.terst.feedreader.data
import android.text.Html
import android.util.Xml
import org.xmlpull.v1.XmlPullParser
import java.io.InputStream
import java.time.OffsetDateTime
import java.time.ZonedDateTime
import java.time.format.DateTimeFormatter
import java.time.format.DateTimeFormatterBuilder
import java.time.temporal.ChronoField
import java.util.Locale
data class ParsedItem(
val guid: String?,
val title: String,
val link: String,
val author: String?,
val publishedAt: Long?,
val contentHtml: String?,
val imageUrl: String? = null
)
data class ParsedFeed(
val title: String,
val siteUrl: String?,
val items: List<ParsedItem>
)
/**
* Hand-rolled RSS 2.0 + Atom parser -- deliberately not a third-party library
* (rome, etc.), per the "simple design" scope: only the fields this app
* actually uses (title/link/guid/author/date/content), nothing else.
*/
object FeedParser {
// Some feeds (kottke.org among them) double-encode titles: the XML pull parser already
// resolves the outer XML escaping (& -> &), but what's left can itself still be
// HTML-entity-encoded text (e.g. a literal "“" or "’") that the feed author
// encoded as if for direct HTML embedding. Titles are shown as plain text (Compose Text,
// not a WebView), so that residual entity needs a second, explicit decode pass or it
// shows up on-screen literally as "“" instead of the curly quote it's meant to be.
private fun decodeTitleEntities(text: String): String =
Html.fromHtml(text, Html.FROM_HTML_MODE_LEGACY).toString()
fun parse(input: InputStream): ParsedFeed {
val parser = Xml.newPullParser()
parser.setFeature(XmlPullParser.FEATURE_PROCESS_NAMESPACES, false)
parser.setInput(input, null)
// Peek the root element to decide which dialect we're in.
var eventType = parser.eventType
while (eventType != XmlPullParser.START_TAG) {
eventType = parser.next()
}
return when (parser.name.lowercase()) {
"rss" -> parseRss(parser)
"feed" -> parseAtom(parser)
else -> throw IllegalArgumentException("Unrecognized feed root element: ${parser.name}")
}
}
private fun parseRss(parser: XmlPullParser): ParsedFeed {
var feedTitle = ""
var siteUrl: String? = null
val items = mutableListOf<ParsedItem>()
var eventType = parser.next()
var depth = 1
// Track whether we're directly inside <channel> (title/link there) vs inside an <item>.
var inItem = false
var itemGuid: String? = null
var itemTitle = ""
var itemLink = ""
var itemAuthor: String? = null
var itemDate: Long? = null
var itemContent: String? = null
var itemDescription: String? = null
var itemImageUrl: String? = null
while (eventType != XmlPullParser.END_DOCUMENT && depth > 0) {
when (eventType) {
XmlPullParser.START_TAG -> {
depth++
when (parser.name.lowercase()) {
"item" -> {
inItem = true
itemGuid = null; itemTitle = ""; itemLink = ""; itemAuthor = null
itemDate = null; itemContent = null; itemDescription = null; itemImageUrl = null
}
"title" -> {
val text = decodeTitleEntities(parser.nextText())
if (inItem) itemTitle = text else feedTitle = text
}
"link" -> {
val text = parser.nextText()
if (inItem) itemLink = text else if (siteUrl == null) siteUrl = text
}
"guid" -> if (inItem) itemGuid = parser.nextText()
"author" -> if (inItem) itemAuthor = parser.nextText()
"dc:creator" -> if (inItem) itemAuthor = parser.nextText()
"pubdate" -> if (inItem) itemDate = parseRfc822Date(parser.nextText())
"content:encoded" -> if (inItem) itemContent = parser.nextText()
"description" -> if (inItem) itemDescription = parser.nextText()
"media:thumbnail", "media:content" -> if (inItem && itemImageUrl == null) {
val url = parser.getAttributeValue(null, "url")
val type = parser.getAttributeValue(null, "type")
if (url != null && (type == null || type.startsWith("image/"))) itemImageUrl = url
}
"enclosure" -> if (inItem && itemImageUrl == null) {
val url = parser.getAttributeValue(null, "url")
val type = parser.getAttributeValue(null, "type")
if (url != null && type?.startsWith("image/") == true) itemImageUrl = url
}
}
}
XmlPullParser.END_TAG -> {
depth--
if (parser.name.lowercase() == "item") {
inItem = false
val content = itemContent ?: itemDescription
items += ParsedItem(
guid = itemGuid ?: itemLink.ifBlank { null },
title = itemTitle,
link = itemLink,
author = itemAuthor,
publishedAt = itemDate,
contentHtml = content,
imageUrl = itemImageUrl ?: extractFirstImageUrl(content)
)
}
}
}
if (depth > 0) eventType = parser.next()
}
return ParsedFeed(feedTitle, siteUrl, items)
}
private fun parseAtom(parser: XmlPullParser): ParsedFeed {
var feedTitle = ""
var siteUrl: String? = null
val items = mutableListOf<ParsedItem>()
var eventType = parser.next()
var depth = 1
var inEntry = false
var entryId: String? = null
var entryTitle = ""
var entryLink = ""
var entryAuthor: String? = null
var entryDate: Long? = null
var entryContent: String? = null
var entrySummary: String? = null
var entryImageUrl: String? = null
while (eventType != XmlPullParser.END_DOCUMENT && depth > 0) {
when (eventType) {
XmlPullParser.START_TAG -> {
depth++
when (parser.name.lowercase()) {
"entry" -> {
inEntry = true
entryId = null; entryTitle = ""; entryLink = ""; entryAuthor = null
entryDate = null; entryContent = null; entrySummary = null; entryImageUrl = null
}
"title" -> {
val text = decodeTitleEntities(parser.nextText())
if (inEntry) entryTitle = text else feedTitle = text
}
"link" -> {
val rel = parser.getAttributeValue(null, "rel")
val href = parser.getAttributeValue(null, "href")
val type = parser.getAttributeValue(null, "type")
if (href != null && (rel == null || rel == "alternate")) {
if (inEntry) entryLink = href else if (siteUrl == null) siteUrl = href
}
if (inEntry && rel == "enclosure" && type?.startsWith("image/") == true && href != null) {
entryImageUrl = href
}
}
"id" -> if (inEntry) entryId = parser.nextText()
"name" -> if (inEntry && entryAuthor == null) entryAuthor = parser.nextText()
"updated" -> if (inEntry) entryDate = parseIso8601Date(parser.nextText())
"published" -> if (inEntry) entryDate = parseIso8601Date(parser.nextText()) ?: entryDate
"content" -> if (inEntry) entryContent = parser.nextText()
"summary" -> if (inEntry) entrySummary = parser.nextText()
"media:thumbnail" -> if (inEntry && entryImageUrl == null) {
entryImageUrl = parser.getAttributeValue(null, "url")
}
}
}
XmlPullParser.END_TAG -> {
depth--
if (parser.name.lowercase() == "entry") {
inEntry = false
val content = entryContent ?: entrySummary
items += ParsedItem(
guid = entryId ?: entryLink.ifBlank { null },
title = entryTitle,
link = entryLink,
author = entryAuthor,
publishedAt = entryDate,
contentHtml = content,
imageUrl = entryImageUrl ?: extractFirstImageUrl(content)
)
}
}
}
if (depth > 0) eventType = parser.next()
}
return ParsedFeed(feedTitle, siteUrl, items)
}
private val imgTagRegex = Regex("""<img[^>]+src=["']([^"']+)["']""", RegexOption.IGNORE_CASE)
// Fallback for feeds with no explicit media:thumbnail/enclosure image --
// most blog feeds just embed the post's lead image directly in the content.
private fun extractFirstImageUrl(html: String?): String? =
html?.let { imgTagRegex.find(it)?.groupValues?.get(1) }
private val rfc822Formatters = listOf(
DateTimeFormatter.RFC_1123_DATE_TIME,
DateTimeFormatterBuilder()
.appendPattern("EEE, d MMM yyyy HH:mm:ss ")
.optionalStart().appendPattern("zzz").optionalEnd()
.optionalStart().appendOffset("+HHMM", "+0000").optionalEnd()
.parseDefaulting(ChronoField.OFFSET_SECONDS, 0)
.toFormatter(Locale.US)
)
private fun parseRfc822Date(raw: String): Long? {
val trimmed = raw.trim()
for (formatter in rfc822Formatters) {
try {
return ZonedDateTime.parse(trimmed, formatter).toInstant().toEpochMilli()
} catch (_: Exception) {
// try the next formatter
}
}
return null
}
private fun parseIso8601Date(raw: String): Long? = try {
OffsetDateTime.parse(raw.trim()).toInstant().toEpochMilli()
} catch (_: Exception) {
null
}
}
|
