MM-9072/MM-10185 Force-convert the encoding of OpenGraph metadata to UTF-8 (#8631)

* Force-convert non-UTF8 HTML to UTF8 before opengraph processing

* Split the force-encoding function

* Add benchmark Test for the forceHTMLEncodingToUTF8()

```
Running tool: /home/comet/go-v1.9.2/bin/go test -benchmem -run=^$ github.com/mattermost/mattermost-server/app -bench ^BenchmarkForceHTMLEncodingToUTF8$

[03:32:58 KST 2018/04/21] [INFO] (github.com/mattermost/mattermost-server/app.TestMain:28) -test.run used, not creating temporary containers
goos: linux
goarch: amd64
pkg: github.com/mattermost/mattermost-server/app
BenchmarkForceHTMLEncodingToUTF8/with_converting-4         	  100000	     11201 ns/op	   18704 B/op	      32 allocs/op
BenchmarkForceHTMLEncodingToUTF8/without_converting-4      	  300000	      3931 ns/op	    4632 B/op	      13 allocs/op
PASS
ok  	github.com/mattermost/mattermost-server/app	2.703s
Success: Benchmarks passed.
```

* Remove an unnecessary constraint

* Add pre-check if content-type header is already utf-8

* Move the checking for utf-8 into forceHTMLEncodingToUTF8() for testing

* Revert df3f347213faa0d023c26d201fa6531f46391086..HEAD, without Gopkg.lock

1 files changed, 140 insertions, 0 deletions

diff --git a/vendor/golang.org/x/text/encoding/traditionalchinese/maketables.go b/vendor/golang.org/x/text/encoding/traditionalchinese/maketables.go
new file mode 100644
index 000000000..cf7fdb31a
--- /dev/null
+++ b/vendor/golang.org/x/text/encoding/traditionalchinese/maketables.go
@@ -0,0 +1,140 @@
+// Copyright 2013 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build ignore
+
+package main
+
+// This program generates tables.go:
+//	go run maketables.go | gofmt > tables.go
+
+import (
+	"bufio"
+	"fmt"
+	"log"
+	"net/http"
+	"sort"
+	"strings"
+)
+
+func main() {
+	fmt.Printf("// generated by go run maketables.go; DO NOT EDIT\n\n")
+	fmt.Printf("// Package traditionalchinese provides Traditional Chinese encodings such as Big5.\n")
+	fmt.Printf(`package traditionalchinese // import "golang.org/x/text/encoding/traditionalchinese"` + "\n\n")
+
+	res, err := http.Get("http://encoding.spec.whatwg.org/index-big5.txt")
+	if err != nil {
+		log.Fatalf("Get: %v", err)
+	}
+	defer res.Body.Close()
+
+	mapping := [65536]uint32{}
+	reverse := [65536 * 4]uint16{}
+
+	scanner := bufio.NewScanner(res.Body)
+	for scanner.Scan() {
+		s := strings.TrimSpace(scanner.Text())
+		if s == "" || s[0] == '#' {
+			continue
+		}
+		x, y := uint16(0), uint32(0)
+		if _, err := fmt.Sscanf(s, "%d 0x%x", &x, &y); err != nil {
+			log.Fatalf("could not parse %q", s)
+		}
+		if x < 0 || 126*157 <= x {
+			log.Fatalf("Big5 code %d is out of range", x)
+		}
+		mapping[x] = y
+
+		// The WHATWG spec http://encoding.spec.whatwg.org/#indexes says that
+		// "The index pointer for code point in index is the first pointer
+		// corresponding to code point in index", which would normally mean
+		// that the code below should be guarded by "if reverse[y] == 0", but
+		// last instead of first seems to match the behavior of
+		// "iconv -f UTF-8 -t BIG5". For example, U+8005 者 occurs twice in
+		// http://encoding.spec.whatwg.org/index-big5.txt, as index 2148
+		// (encoded as "\x8e\xcd") and index 6543 (encoded as "\xaa\xcc")
+		// and "echo 者 | iconv -f UTF-8 -t BIG5 | xxd" gives "\xaa\xcc".
+		c0, c1 := x/157, x%157
+		if c1 < 0x3f {
+			c1 += 0x40
+		} else {
+			c1 += 0x62
+		}
+		reverse[y] = (0x81+c0)<<8 | c1
+	}
+	if err := scanner.Err(); err != nil {
+		log.Fatalf("scanner error: %v", err)
+	}
+
+	fmt.Printf("// decode is the decoding table from Big5 code to Unicode.\n")
+	fmt.Printf("// It is defined at http://encoding.spec.whatwg.org/index-big5.txt\n")
+	fmt.Printf("var decode = [...]uint32{\n")
+	for i, v := range mapping {
+		if v != 0 {
+			fmt.Printf("\t%d: 0x%08X,\n", i, v)
+		}
+	}
+	fmt.Printf("}\n\n")
+
+	// Any run of at least separation continuous zero entries in the reverse map will
+	// be a separate encode table.
+	const separation = 1024
+
+	intervals := []interval(nil)
+	low, high := -1, -1
+	for i, v := range reverse {
+		if v == 0 {
+			continue
+		}
+		if low < 0 {
+			low = i
+		} else if i-high >= separation {
+			if high >= 0 {
+				intervals = append(intervals, interval{low, high})
+			}
+			low = i
+		}
+		high = i + 1
+	}
+	if high >= 0 {
+		intervals = append(intervals, interval{low, high})
+	}
+	sort.Sort(byDecreasingLength(intervals))
+
+	fmt.Printf("const numEncodeTables = %d\n\n", len(intervals))
+	fmt.Printf("// encodeX are the encoding tables from Unicode to Big5 code,\n")
+	fmt.Printf("// sorted by decreasing length.\n")
+	for i, v := range intervals {
+		fmt.Printf("// encode%d: %5d entries for runes in [%6d, %6d).\n", i, v.len(), v.low, v.high)
+	}
+	fmt.Printf("\n")
+
+	for i, v := range intervals {
+		fmt.Printf("const encode%dLow, encode%dHigh = %d, %d\n\n", i, i, v.low, v.high)
+		fmt.Printf("var encode%d = [...]uint16{\n", i)
+		for j := v.low; j < v.high; j++ {
+			x := reverse[j]
+			if x == 0 {
+				continue
+			}
+			fmt.Printf("\t%d-%d: 0x%04X,\n", j, v.low, x)
+		}
+		fmt.Printf("}\n\n")
+	}
+}
+
+// interval is a half-open interval [low, high).
+type interval struct {
+	low, high int
+}
+
+func (i interval) len() int { return i.high - i.low }
+
+// byDecreasingLength sorts intervals by decreasing length.
+type byDecreasingLength []interval
+
+func (b byDecreasingLength) Len() int           { return len(b) }
+func (b byDecreasingLength) Less(i, j int) bool { return b[i].len() > b[j].len() }
+func (b byDecreasingLength) Swap(i, j int)      { b[i], b[j] = b[j], b[i] }