~alcinnz/hurl

ref: f605efd1872c427894e67aa40893cad9ce8ce577 hurl/hurl-xml/src/Network/MIME/XML.hs -rw-r--r-- 15.9 KiB
f605efd1 — Adrian Cochrane Release v0.1 decoupled from Rhapsode, with CSV/TSV & refined tablesorting! 1 year, 7 months ago
                                                                                
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
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
{-# LANGUAGE OverloadedStrings #-}
{-# LANGUAGE TemplateHaskell #-}
{-# LANGUAGE PatternSynonyms, ViewPatterns #-}
module Network.MIME.XML(Page(..), loadVisited,
    fetchDocument, pageForText, applyCSScharset, readStrict) where

import           Data.Text.Lazy (fromStrict)
import qualified Data.Text as Txt
import           Data.Text (Text)
import qualified Data.Text.IO as Txt
import           Data.Text.Encoding
import qualified Data.Text.Lazy as LTxt
import qualified Data.ByteString.Lazy as B
import qualified Text.HTML.DOM as HTML
import qualified Text.XML as XML
import           Text.XML (Document(..))
import           Network.URI
import           Network.URI.Fetch
import           Network.URI.Charset
import qualified Data.Map as M
import qualified Data.Set as Set
import           Data.Set (Set(..))
import           Data.List (intercalate)
import           Data.Time.Clock

-- For alternative styles
import qualified Data.CSS.Syntax.Tokens as CSSTok
import Stylist.Parse

import System.IO
import System.IO.Temp
import Data.Default.Class
import System.Directory
import System.FilePath ((</>))
import Data.FileEmbed
import Data.Maybe (fromMaybe)
import Text.Read (readMaybe)

import Network.MIME.XML.Table -- Apply table sorting here...
import Data.HTML2CSS (html2css)

data Page styles = Page {
    pageURL :: URI,
    css :: styles,
    initCSS :: styles,
    html :: Document,
    pageTitle :: String,
    pageMIME :: String,
    apps :: [Application],
    backStack :: [(String, URI)],
    forwardStack :: [(String, URI)],
    -- Probably don't need an MVar here, but let's be safe!
    visitedURLs :: Set Text,
    appName :: String
}

loadVisited :: String -> IO (Set Text)
loadVisited appname = do
    dir <- getXdgDirectory XdgData appname
    let path = dir </> "history.gmni"
    exists <- doesFileExist path

    if exists then do
        file <- readStrict path
        let hist = Set.fromList [Txt.pack uri | _:uri:_ <- map words $ lines file]
        return hist
    else return Set.empty

readStrict path = do s <- Prelude.readFile path; length s `seq` return s

utf8' bytes = convertCharset "utf-8" $ B.toStrict bytes
aCCEPT = ["text/xml", "application/xml", "text/html", "text/gemini",
    "text/csv", "text/tab-separated-values", "text/css", "text/*", "*/*"]

fetchDocument http referer URI { uriScheme = "action:", uriPath = "nocache" } =
    fetchDocument http { cachingEnabled = False } referer $ pageURL referer
fetchDocument http referer URI { uriScheme = "action:", uriPath = "novalidate" } =
    fetchDocument http { validateCertificates = False } referer $ pageURL referer
fetchDocument http referer URI { uriScheme = "action:", uriPath = "history/back" } =
        fetchURL' http aCCEPT (pageURL referer') >>= parseDocument' referer' http False
    where referer' = shiftHistory referer (-1)
fetchDocument http referer URI { uriScheme = "action:", uriPath = "history/forward" } =
        fetchURL' http aCCEPT (pageURL referer') >>= parseDocument' referer' http False
    where referer' = shiftHistory referer 1
fetchDocument http referer URI {
        uriScheme = "action:", uriPath = 'h':'i':'s':'t':'o':'r':'y':'/':x
    } | Just x' <- readMaybe x, referer' <- shiftHistory referer x' =
        fetchURL' http aCCEPT (pageURL referer') >>= parseDocument' referer http False
fetchDocument http referer URI { uriScheme = "app:", uriPath = appID } = do
    dispatchByApp http Application {
        name = "", icon = nullURI, description = "",
        appId = appID
      } (pageMIME referer) $ pageURL referer
    return referer -- TODO play an error or success sound
fetchDocument http referer@Page { pageURL = uri0 } uri@URI { uriFragment = anchor }
    | uri { uriFragment = "" } == uri0 { uriFragment = "" } = return referer {
        html = applySortDoc anchor $ html referer,
        pageURL = uri
    }
fetchDocument http referer uri = fetchURL' http aCCEPT uri >>= parseDocument' referer http True

shiftHistory :: Page style -> Integer -> Page style
shiftHistory self 0 = self
shiftHistory self@Page { backStack = (title, url):bs } delta | delta < 0 =
    shiftHistory self {
        backStack = bs,
        forwardStack = (pageTitle self, pageURL self):forwardStack self,
        pageTitle = title,
        pageURL = url
    } $ succ delta
shiftHistory self@Page { forwardStack = (title, url):fs } delta | delta > 0 =
    shiftHistory self {
        forwardStack = fs,
        backStack = (pageTitle self, pageURL self):backStack self,
        pageTitle = title,
        pageURL = url
    } $ pred delta
shiftHistory self _ = self -- Error case.

parseDocument' ref@Page {visitedURLs = hist} sess saveHist resp@(URI {uriFragment = anch}, mime, _) = do
    page <- parseDocument ref sess resp >>= logHistory hist
    apps' <- appsForMIME sess mime
    return $ attachHistory page {
        pageMIME = mime,
        apps = apps',
        html = applySortDoc anch $ html page
    }
  where
    attachHistory x@Page { pageTitle = title, pageURL = url }
        | saveHist = x { backStack = (title, url):backStack ref, forwardStack = forwardStack ref }
        | otherwise = x
parseDocument :: StyleSheet s => Page s -> Session -> (URI, String, Either Text B.ByteString)
        -> IO (Page s)
parseDocument ref sess (uri, "html/x-error\t", resp) = parseDocument ref sess (uri, "text/html", resp)
parseDocument Page {initCSS = css', appName = name} _ (uri, "text/html", Left text) =
    pageForDoc css' name uri $ HTML.parseLT $ fromStrict text
parseDocument Page {initCSS = css', appName = name} _(uri, "text/html", Right bytes) =
    pageForDoc css' name uri $ HTML.parseLBS bytes
parseDocument Page {initCSS = css', appName = name} _
        (uri, 't':'e':'x':'t':'/':'g':'e':'m':'i':'n':'i':';':'l':'a':'n':'g':'=':lang, Left text) =
    pageForDoc css' name uri $ parseGemini (Just lang) text
parseDocument Page {initCSS = css', appName = name} _
        (uri, 't':'e':'x':'t':'/':'g':'e':'m':'i':'n':'i':';':'l':'a':'n':'g':'=':lang, Right bytes) =
    pageForDoc css' name uri $ parseGemini (Just lang) $ utf8' bytes
parseDocument Page {initCSS = css', appName = name} _ (uri, "text/gemini", Left text) =
    pageForDoc css' name uri $ parseGemini Nothing text
parseDocument Page {initCSS = css', appName = name} _ (uri, "text/gemini", Right bytes) =
    pageForDoc css' name uri $ parseGemini Nothing $ utf8' bytes
parseDocument a b (a', b'@"text/css", Right bytes) =
    parseDocument a b (a', b', Left $ applyCSScharset (map Txt.unpack charsets) $ B.toStrict bytes)
parseDocument referer@Page {pageURL = uri', initCSS = css', appName = name} _
    (uri, "text/css", Left text)
  | URI {uriAuthority = Just host} <- pageURL referer = do
    -- Save this per-domain setting
    dir <- (</> "domain") <$> getXdgDirectory XdgConfig name
    createDirectoryIfMissing True dir
    Txt.writeFile (dir </> uriRegName host) $
        CSSTok.serialize $ map absolutizeCSS $ CSSTok.tokenize text

    return ret
  | otherwise = return ret
 where
  ret = referer {
        css = parseForURL css' uri text
    }
  absolutizeCSS (CSSTok.Url text) | Just rel <- parseRelativeReference $ Txt.unpack text =
    CSSTok.Url $ Txt.pack $ uriToStr' $ relativeTo rel uri'
  absolutizeCSS tok = tok
parseDocument ref@Page {initCSS = css', appName = name} _ (uri, "text/csv", Left body) =
    pageForDoc css' name uri $ parseDelimitedToTable ',' body
parseDocument ref@Page {initCSS = css', appName = name} _ (uri, "text/tab-separated-values", Left body) =
    pageForDoc css' name uri $ parseDelimitedToTable '\t' body
parseDocument ref@Page {initCSS = css', appName = name} _ (uri, "text/csv", Right body) =
    pageForDoc css' name uri $ parseDelimitedToTable ',' $ utf8' body
parseDocument ref@Page {initCSS = css', appName = name} _
        (uri, "text/tab-separated-values", Right body) =
    pageForDoc css' name uri $ parseDelimitedToTable '\t' $ utf8' body

parseDocument ref sess (uri, mime, body) | mime' /= mime = parseDocument ref sess (uri, mime', body)
    where mime' = takeWhile (/= ';') mime
parseDocument Page {initCSS = css', appName = name} _ (uri, _, Left text)
    | Right doc <- XML.parseText def $ fromStrict text = pageForDoc css' name uri doc
    | otherwise = pageForText css' name uri text
parseDocument Page {initCSS = css', appName = name} _ (uri, _, Right bytes)
    | Right doc <- XML.parseLBS def bytes = pageForDoc css' name uri doc
parseDocument Page {initCSS = css', appName = name} _ (uri, 't':'e':'x':'t':'/':_, Right bytes) =
    -- charset wasn't specified, so assume utf-8.
    pageForText css' name uri $ utf8' bytes
parseDocument Page {initCSS = css', appName = name} sess resp@(uri, mime, _) = do
    dir <- getCurrentDirectory -- TODO find Downloads directory.
    ret <- saveDownload nullURI {
        uriScheme = "file:",
        uriAuthority = Just (URIAuth "" "" "")
    } dir resp >>= dispatchByMIME sess mime
    pageForDoc css' name uri $ HTML.parseLT $ LTxt.pack $ fromMaybe "Unsupported filetype" ret

pageForText css' appname uri txt = pageForDoc css' appname uri XML.Document {
        XML.documentPrologue = XML.Prologue [] Nothing [],
        XML.documentRoot = XML.Element {
            XML.elementName = "pre",
            XML.elementAttributes = M.empty,
            XML.elementNodes = [XML.NodeContent txt]
        },
        XML.documentEpilogue = []
    }

pageForDoc :: StyleSheet s => s -> String -> URI -> Document -> IO (Page s)
pageForDoc css' appname uri doc = do
    -- See if the user has configured an alternate stylesheet for this domain.
    let authorStyle = return $ html2css doc uri css'
    styles <- case uriAuthority uri of
        Nothing -> authorStyle
        Just host -> do
            dir <- getXdgDirectory XdgConfig appname
            let path = dir </> "domain" </> uriRegName host
            hasAltStyle <- doesFileExist path
            if not hasAltStyle then authorStyle else parse css' <$> Txt.readFile path

    return Page {pageURL = uri, html = doc, css = styles,
        initCSS = css', appName = appname,
        -- These fields are all blank, to be filled in later by logHistory & parseDocument'
        pageTitle = "", pageMIME = "", apps = [],
        backStack = [], forwardStack = [], visitedURLs = Set.empty}

logHistory hist ret@Page {pageURL = url', html = doc, appName = name} = do
    dir <- getXdgDirectory XdgData name
    createDirectoryIfMissing True dir
    now <- getCurrentTime
    let title = Txt.unpack $ getTitle $ XML.documentRoot doc
    appendFile (dir </> "history.gmni") $ '\n' : intercalate " " [
        "=>", uriToStr' url', show now, title
      ]

    return ret { pageTitle = title, visitedURLs = Set.insert (Txt.pack $ uriToStr' url') hist}
  where
    getTitle (XML.Element "title" _ childs) = Txt.concat [txt | XML.NodeContent txt <- childs]
    getTitle (XML.Element "h1" _ childs) = Txt.concat [txt | XML.NodeContent txt <- childs]
    getTitle (XML.Element _ _ childs)
        | title:_ <- [getTitle el | XML.NodeElement el <- childs] = title
        | otherwise = ""

uriToStr' :: URI -> String
uriToStr' uri = uriToString id uri ""

--------
---- CSS charset sniffing
--------
applyCSScharset (charset:charsets) bytes
        | cssCharset (CSSTok.tokenize text) == Txt.pack charset = text
        | otherwise = applyCSScharset charsets bytes
    where
        text = convertCharset charset bytes
applyCSScharset _ bytes = convertCharset "utf-8" bytes
cssCharset toks | (CSSTok.AtKeyword "charset":toks') <- skipCSSspace toks,
        (CSSTok.String charset:_) <- skipCSSspace toks' = charset
    | otherwise = ""
skipCSSspace (CSSTok.Whitespace:toks) = skipCSSspace toks
skipCSSspace toks = toks

--------
---- Gemini implementation
--------
-- Copied from css-syntax.
pattern (:.) :: Char -> Txt.Text -> Txt.Text
pattern x :. xs <- (Txt.uncons -> Just (x, xs))

infixr 5 :.

el name text = XML.Element name M.empty [XML.NodeContent text]

parseGemini :: Maybe String -> Txt.Text -> XML.Document
parseGemini lang txt = XML.Document {
        XML.documentPrologue = XML.Prologue [] Nothing [],
        XML.documentRoot = XML.Element {
            XML.elementName = "body",
            XML.elementAttributes = M.fromList [
                ("lang", Txt.pack lang') | Just langs <- [lang], lang' <- [csv langs]],
            XML.elementNodes = map XML.NodeElement $ parseGemini' $ Txt.lines txt
        },
        XML.documentEpilogue = []
    }

csv (',':_) = ""
csv (c:rest) = c:csv rest
csv "" = ""

parseGemini' :: [Txt.Text] -> [XML.Element]
parseGemini' (('#':.'#':.'#' :. '#':.'#':.'#':.line):lines) =
    el "h6" line : parseGemini' lines
parseGemini' (('#':.'#':.'#' :. '#':.'#':.line):lines) =
    el "h5" line : parseGemini' lines
parseGemini' (('#':.'#':.'#' :. '#':.line):lines) =
    el "h4" line : parseGemini' lines
parseGemini' (('#':.'#':.'#':.line):lines) = el "h3" line : parseGemini' lines
parseGemini' (('#':.'#':.line):lines) = el "h2" line : parseGemini' lines
parseGemini' (('#':.line):lines) = el "h1" line : parseGemini' lines
-- Not properly structured, but still sounds fine...
parseGemini' (('*':.line):lines) = el "li" line : parseGemini' lines
parseGemini' (('>':.line):lines) = el "blockquote" line : parseGemini' lines

parseGemini' (('=':.'>':.line):lines)
    | (url:text@(_:_)) <- Txt.words line = (el "a" $ Txt.unwords text) {
            XML.elementAttributes = M.insert "href" url M.empty
        } : parseGemini' lines
    | otherwise = (el "a" $ Txt.strip line) {
            XML.elementAttributes = M.insert "href" (Txt.strip line) M.empty
        } : parseGemini' lines
parseGemini' (('`':.'`':.'`':.line):lines) = el "p" line : go lines
    where
        go (('`':.'`':.'`':._):lines) = parseGemini' lines
        go (_:lines) = go lines
        go [] = []
parseGemini' ("```":lines) = go [] lines
    where
        go texts (('`':.'`':.'`':._):lines) =
            el "pre" (Txt.unlines texts) : parseGemini' lines
        go texts (line:lines) = go (texts ++ [line]) lines
        go texts [] = []

parseGemini' (line:lines) = el "p" line : parseGemini' lines
parseGemini' [] = []

--------
---- TSV, CSV, etc
--------

parseDelimitedValues _ "" row rows = reverse (reverse row : rows)
parseDelimitedValues delim ('\r':.cs) row rows = parseDelimitedValues delim cs row rows
parseDelimitedValues delim ('\n':.cs) row rows = parseDelimitedValues delim cs [] (reverse row : rows)
parseDelimitedValues delim (c:.'"':.cs) row rows | c == delim =
        let (value, cs') = inner cs in parseDelimitedValues delim cs' (value:row) rows
    where
        inner (x:.y:.cs) | x == delim && y == delim = let (a, b) = inner cs in (delim `Txt.cons` a, b)
        inner (c:.cs) | c == delim = ("", cs)
            | otherwise = let (a, b) = inner cs in (c `Txt.cons` a, b)
        inner "" = ("", "")
parseDelimitedValues delim (c:.cs) row rows | c == delim =
    let (value, cs') = Txt.break (`elem` ['\r', '\n', delim]) cs
    in parseDelimitedValues delim cs' (value:row) rows
parseDelimitedValues delim cs row rows =
    let (value, cs') = Txt.break (`elem` ['\r', '\n', delim]) cs
    in parseDelimitedValues delim cs (value:row) rows

escapeDelimitedValues delim source = map (map inner) $ parseDelimitedValues delim source [] []
    where
        inner = Txt.strip . Txt.replace "\\\\" "\\" . Txt.replace "\\n" "\n" .
            Txt.replace "\\t" "\t" . Txt.replace "\\r" "\r"

parseDelimitedToTable delim source
    | (head:body) <- filter (not . null) $ escapeDelimitedValues delim source =
        XML.Document {
            XML.documentPrologue = XML.Prologue [] Nothing [],
            XML.documentRoot = XML.Element {
                XML.elementName = "table",
                XML.elementAttributes = M.empty,
                XML.elementNodes = rowToTr "th" head : map (rowToTr "td") body
            },
            XML.documentEpilogue = []
        }
    | otherwise = XML.Document { -- Empty TSV/CSV/etc
        XML.documentPrologue = XML.Prologue [] Nothing [],
        XML.documentRoot = XML.Element "table" M.empty [],
        XML.documentEpilogue = []
    }
rowToTr tagname values = XML.NodeElement $ XML.Element "tr" M.empty $ map inner values
    where
        inner = XML.NodeElement . XML.Element tagname M.empty . singleton . XML.NodeContent
        singleton a = [a]