Delete .gitignore

Fix unicode encode error

.gitignore

@@ -1 +0,0 @@
-.vscode

Makefile

@@ -0,0 +1,22 @@
+all: linux freebsd openbsd macos windows rpi
+
+linux: 
+	GOOS=linux GOARCH=amd64 go build -a -o wrp-linux wrp.go
+
+freebsd: 
+	GOOS=freebsd GOARCH=amd64 go build -a -o wrp-freebsd wrp.go
+
+openbsd: 
+	GOOS=openbsd GOARCH=amd64 go build -a -o wrp-openbsd wrp.go
+
+macos: 
+	GOOS=darwin GOARCH=amd64 go build -a -o wrp-macos wrp.go
+
+windows: 
+	GOOS=windows GOARCH=amd64 go build -a -o wrp-windows.exe wrp.go
+
+rpi:
+	GOOS=linux GOARCH=arm go build -a -o wrp-linux-rpi wrp.go
+
+clean:
+	rm -rf wrp-linux wrp-freebsd wrp-openbsd wrp-macos wrp-windows.exe wrp-linux-rpi

README.md

@@ -2,31 +2,35 @@
 
 A HTTP proxy server that allows to use historical and obsolete web browsers on the modern web. It works by rendering the web page in to a GIF image associated with clickable imagemap of original web links.
 
-**You are looking at a GoLang / CDP branch of WRP.**
+## Current Status
 
-**This code is under active development and not fully usable yet.**
-
-## Done so far
-
-* basic browser-in-browser mode
-* screenshot and serve image+map via CDP
-* gif with Floyd–Steinberg dithering
-* random image addressing
-* resolve relative links
-* paginated scrolling
-* google search on input not starting with ^http
+* This is the new GoLang/ChdomeDP version.
+* It's still lacking some features of the [older version](/old) (such as real http proxy mode and image manipulation) but it surpases it in terms of stability and usability. 
+* It's beta quality but I can fix/maintain the code unlike the older version.
 
 ## Todo
 
-* ISMAP - underway
-* net/url: invalid control character in URL on Windows
-* configurable color palete and quantization
-* real http proxy support
-* option to encode as png/jpeg
-* padded box model coordinates
-* better http server shutdown
-* chromedp logging, timeout, non-headless flags
+* Configurable color palete and quantization.
+* Real http proxy support via [goproxy](https://github.com/elazarl/goproxy) - if you really need a real proxy, for now use the [old/](/old) version.
+* Padded box model coordinates.
+* Input boxes support. However today you can cheat by using headed mode and input your data on the WRP server.
 
-## Python version
+## Usage
 
-Check [master branch](https://github.com/tenox7/wrp/tree/master) for "stable" Python-Webkit version.
+1. [Download a WRP binary](https://github.com/tenox7/wrp/releases) and run on a  server/gateway. 
+2. Point your legacy browser to the IP address:port of WRP server.
+3. Type a search string or a http/https URL and click Go.
+4. Adjust your screen width/heigh/scale to fit in your old browser.
+5. For very very very old browsers such as Mosaic 2.x and IBM WebExplorer 1.x check the I checkbox to enable ISMAP mode. However this normally should not be needed.
+6. Scroll web page by clicking Up/Down. To go to top enter 0 and click Go.
+
+## Flags
+```
+-l  listen address:port, default :8080
+-h  headed mode, display browser window
+-d  chromedp debug logging
+```
+
+## More info and screenshots
+* http://virtuallyfun.superglobalmegacorp.com/2014/03/11/web-rendering-proxy-update/
+* http://virtuallyfun.superglobalmegacorp.com/2014/03/03/surfing-modern-web-with-ancient-browsers/

old/wrp.py

@@ -284,10 +284,10 @@ if sys.platform.startswith('linux') or sys.platform.startswith('freebsd'):
                 if ISMAP == True:
                     mapfile.write("rect %s %i,%i %i,%i\n".decode('utf-8', errors='ignore') % (turl, xmin, ymin, xmax, ymax))
                 else:
-                    httpout.write("<AREA SHAPE=\"RECT\""
+                    httpout.write(("<AREA SHAPE=\"RECT\""
                                   " COORDS=\"%i,%i,%i,%i\""
                                   " ALT=\"%s\" HREF=\"%s\">\n".decode('utf-8', errors='ignore')
-                                  % (xmin, ymin, xmax, ymax, turl, turl))
+                                  % (xmin, ymin, xmax, ymax, turl, turl)).encode("utf-8"))
 
             if ISMAP != True:
                 httpout.write("</MAP>\n")

pywebkit/Changelog.md

@@ -1,36 +0,0 @@
-## [2.0] - 2017-05-10
-### Added
-- Support PyQt5 if available.
-- Sets title from original one.
-- Returns server errors as is.
-- Download non-HTML files as is.
-- For JavaScript capable browsers detect and automatically set view width.
-- Add support for configuring which image format to use.
-- Added support for PythonMagick. If found, allows to dither, color-reduce, or convert to grayscale or monochrome.
-- If PythonMagick is found, render as PNG and convert to user-requested format using it, for better quality.
-
-### Changed
-- Support www prepented to http://wrp.stop command.
-
-### Fixed
-- Prevent python crashes with non-ASCII character in URLs.
-
-## [1.4] - 2017-01-22
-### Added
-- Suport for ISMAP on Linux.
-- Use queues instead of globals in Linux.
-
-## [1.3] - 2017-01-21
-### Changed
-- Merged mac OS and Linux in a single executable.
-- Use queues instead of globals in Linux.
-
-### Fixed
-- Call PyQt to close application on http://wrp.stop
-
-## [1.2] - 2016-12-27
-### Added
-- Support for IMAP on mac OS.
-
-### Changed
-- Use queues instead of globals in mac OS.

wrp.go

@@ -19,7 +19,6 @@ import (
 	"math/rand"
 	"net/http"
 	"net/url"
-	"os"
 	"strconv"
 	"strings"
 	"time"
@@ -32,72 +31,96 @@ import (
 	"github.com/chromedp/chromedp"
 )
 
+// Ismap for server side processing
+type Ismap struct {
+	xmin int64
+	ymin int64
+	xmax int64
+	ymax int64
+	url  string
+}
+
 var (
-	ctx    context.Context
-	cancel context.CancelFunc
-	gifmap = make(map[string]bytes.Buffer)
+	version = "3.0"
+	ctx     context.Context
+	cancel  context.CancelFunc
+	gifmap  = make(map[string]bytes.Buffer)
+	ismap   = make(map[string][]Ismap)
 )
 
-func pageServer(out http.ResponseWriter, r *http.Request) {
-	r.ParseForm()
-	u := r.FormValue("url")
+func pageServer(out http.ResponseWriter, req *http.Request) {
+	req.ParseForm()
+	u := req.FormValue("url")
 	var istr string
+	var ion string
 	var i bool
-	if r.FormValue("i") == "on" {
+	if req.FormValue("i") == "on" {
 		istr = "CHECKED"
 		i = true
+		ion = "&i=on"
 	} else {
 		istr = ""
 		i = false
 	}
-	p, _ := strconv.ParseInt(r.FormValue("p"), 10, 64)
-	if r.FormValue("pg") == ">>" {
+	p, _ := strconv.ParseInt(req.FormValue("p"), 10, 64)
+	if req.FormValue("pg") == "Dn" {
 		p++
-	} else if r.FormValue("pg") == "<<" {
+	} else if req.FormValue("pg") == "Up" {
 		p--
 	} else {
 		p = 0
 	}
-	w, _ := strconv.ParseInt(r.FormValue("w"), 10, 64)
+	w, _ := strconv.ParseInt(req.FormValue("w"), 10, 64)
 	if w < 10 {
 		w = 1024
 	}
-	h, _ := strconv.ParseInt(r.FormValue("h"), 10, 64)
+	h, _ := strconv.ParseInt(req.FormValue("h"), 10, 64)
 	if h < 10 {
 		h = 768
 	}
-	s, _ := strconv.ParseFloat(r.FormValue("s"), 64)
+	s, _ := strconv.ParseFloat(req.FormValue("s"), 64)
 	if s < 0.1 {
 		s = 1.0
 	}
-	log.Printf("%s Page Reqest for url=\"%s\" [%s]\n", r.RemoteAddr, u, r.URL.Path)
+	c, _ := strconv.ParseInt(req.FormValue("c"), 10, 64)
+	if c < 2 || c > 256 {
+		c = 256
+	}
+	log.Printf("%s Page Reqest for url=\"%s\" [%s]\n", req.RemoteAddr, u, req.URL.Path)
 	out.Header().Set("Content-Type", "text/html")
-	fmt.Fprintf(out, "<HTML>\n<HEAD><TITLE>WRP %s</TITLE></HEAD>\n<BODY BGCOLOR=\"#F0F0F0\">", u)
-	fmt.Fprintf(out, "<FORM ACTION=\"/\">URL/Search: <INPUT TYPE=\"TEXT\" NAME=\"url\" VALUE=\"%s\" SIZE=\"40\">", u)
-	fmt.Fprintf(out, "<INPUT TYPE=\"SUBMIT\" VALUE=\"Go\"><P>\n")
-	fmt.Fprintf(out, "ISMAP:<INPUT TYPE=\"CHECKBOX\" NAME=\"i\" %s> \n", istr)
-	fmt.Fprintf(out, "Width:<INPUT TYPE=\"TEXT\" NAME=\"w\" VALUE=\"%d\" SIZE=\"4\"> \n", w)
-	fmt.Fprintf(out, "Height:<INPUT TYPE=\"TEXT\" NAME=\"h\" VALUE=\"%d\" SIZE=\"4\"> \n", h)
-	fmt.Fprintf(out, "Scale:<INPUT TYPE=\"TEXT\" NAME=\"s\" VALUE=\"%1.2f\" SIZE=\"3\"> \n", s)
-	fmt.Fprintf(out, "Page:<INPUT TYPE=\"HIDDEN\" NAME=\"p\" VALUE=\"%d\"> \n", p)
-	fmt.Fprintf(out, "<INPUT TYPE=\"SUBMIT\" NAME=\"pg\" VALUE=\"<<\"> %d \n", p)
-	fmt.Fprintf(out, "<INPUT TYPE=\"SUBMIT\" NAME=\"pg\" VALUE=\">>\"> \n")
-	fmt.Fprintf(out, "</FORM><P>\n")
-	if len(u) > 4 {
+	fmt.Fprintf(out, "<!-- Web Rendering Proxy Version %s -->\n", version)
+	fmt.Fprintf(out, "<HTML>\n<HEAD><TITLE>WRP %s</TITLE></HEAD>\n<BODY BGCOLOR=\"#F0F0F0\">\n", u)
+	fmt.Fprintf(out, "<FORM ACTION=\"/\"><INPUT TYPE=\"TEXT\" NAME=\"url\" VALUE=\"%s\" SIZE=\"20\">", u)
+	fmt.Fprintf(out, "<INPUT TYPE=\"SUBMIT\" VALUE=\"Go\"> \n")
+	fmt.Fprintf(out, "<INPUT TYPE=\"SUBMIT\" NAME=\"pg\" VALUE=\"Up\"> \n")
+	fmt.Fprintf(out, "<INPUT TYPE=\"TEXT\" NAME=\"p\" VALUE=\"%d\" SIZE=\"2\"> \n", p)
+	fmt.Fprintf(out, "<INPUT TYPE=\"SUBMIT\" NAME=\"pg\" VALUE=\"Dn\"> \n")
+	fmt.Fprintf(out, "I <INPUT TYPE=\"CHECKBOX\" NAME=\"i\" %s> \n", istr)
+	fmt.Fprintf(out, "W <INPUT TYPE=\"TEXT\" NAME=\"w\" VALUE=\"%d\" SIZE=\"4\"> \n", w)
+	fmt.Fprintf(out, "H <INPUT TYPE=\"TEXT\" NAME=\"h\" VALUE=\"%d\" SIZE=\"4\"> \n", h)
+	fmt.Fprintf(out, "S <INPUT TYPE=\"TEXT\" NAME=\"s\" VALUE=\"%1.2f\" SIZE=\"3\"> \n", s)
+	fmt.Fprintf(out, "C <INPUT TYPE=\"TEXT\" NAME=\"c\" VALUE=\"%d\" SIZE=\"3\"> \n", c)
+	fmt.Fprintf(out, "</FORM><BR>\n")
+	if len(u) > 1 {
 		if strings.HasPrefix(u, "http") {
-			capture(u, w, h, s, p, i, out)
+			capture(u, w, h, s, int(c), p, i, req.RemoteAddr, out)
 		} else {
-			capture(fmt.Sprintf("http://www.google.com/search?q=%s", url.QueryEscape(u)), w, h, s, p, i, out)
+			capture(fmt.Sprintf("http://www.google.com/search?q=%s", url.QueryEscape(u)), w, h, s, int(c), p, i, req.RemoteAddr, out)
 		}
 	} else {
 		fmt.Fprintf(out, "No URL or search query specified")
 	}
-	fmt.Fprintf(out, "</BODY>\n</HTML>\n")
+	fmt.Fprintf(out, "\n<P><A HREF=\"/?url=https://github.com/tenox7/wrp/&w=%d&h=%d&s=%1.2f&c=%d%s\">Web Rendering Proxy Version %s</A></BODY>\n</HTML>\n", w, h, s, c, ion, version)
 }
 
 func imgServer(out http.ResponseWriter, req *http.Request) {
 	log.Printf("%s IMG Request for %s\n", req.RemoteAddr, req.URL.Path)
-	gifbuf := gifmap[req.URL.Path]
+	gifbuf, ok := gifmap[req.URL.Path]
+	if !ok || gifbuf.Bytes() == nil {
+		fmt.Fprintf(out, "Unable to find image %s\n", req.URL.Path)
+		log.Printf("Unable to find image %s\n", req.URL.Path)
+		return
+	}
 	defer delete(gifmap, req.URL.Path)
 	out.Header().Set("Content-Type", "image/gif")
 	out.Header().Set("Content-Length", strconv.Itoa(len(gifbuf.Bytes())))
@@ -106,32 +129,49 @@ func imgServer(out http.ResponseWriter, req *http.Request) {
 }
 
 func mapServer(out http.ResponseWriter, req *http.Request) {
-	log.Printf("%s MAP Request for %s [%v]\n", req.RemoteAddr, req.URL.Path, req.URL.Query())
+	log.Printf("%s ISMAP Request for %s [%+v]\n", req.RemoteAddr, req.URL.Path, req.URL.RawQuery)
+	var loc string
+	var x, y int64
+	n, err := fmt.Sscanf(req.URL.RawQuery, "%d,%d", &x, &y)
+	if err != nil || n != 2 {
+		fmt.Fprintf(out, "n=%d, err=%s\n", n, err)
+		log.Printf("%s ISMAP n=%d, err=%s\n", req.RemoteAddr, n, err)
+		return
+	}
+	is, ok := ismap[req.URL.Path]
+	if !ok || is == nil {
+		fmt.Fprintf(out, "Unable to find map %s\n", req.URL.Path)
+		log.Printf("Unable to find map %s\n", req.URL.Path)
+		return
+	}
+	defer delete(ismap, req.URL.Path)
+	for _, i := range is {
+		if x >= i.xmin && x <= i.xmax && y >= i.ymin && y <= i.ymax {
+			loc = i.url
+		}
+	}
+	if len(loc) < 1 {
+		loc = is[0].url
+	}
+	log.Printf("%s ISMAP Redirect to: http://%s%s\n", req.RemoteAddr, req.Context().Value(http.LocalAddrContextKey), loc)
+	http.Redirect(out, req, fmt.Sprintf("http://%s%s", req.Context().Value(http.LocalAddrContextKey), loc), 301)
 }
 
-
-func haltServer(out http.ResponseWriter, req *http.Request) {
-	log.Printf("%s Shutdown request received [%s]\n", req.RemoteAddr, req.URL.Path)
-	out.Header().Set("Content-Type", "text/plain")
-	fmt.Fprintf(out, "WRP Shutdown")
-	out.(http.Flusher).Flush()
-	cancel()
-	os.Exit(0)
-}
-
-func capture(gourl string, w int64, h int64, s float64, p int64, ismap bool, out http.ResponseWriter) {
+func capture(gourl string, w int64, h int64, s float64, co int, p int64, i bool, c string, out http.ResponseWriter) {
 	var nodes []*cdp.Node
 	ctxx := chromedp.FromContext(ctx)
 	var pngbuf []byte
 	var gifbuf bytes.Buffer
 	var loc string
 	var res *runtime.RemoteObject
+	is := make([]Ismap, 0)
+	var ion string
 
-	log.Printf("Processing Caputure Request for %s\n", gourl)
+	log.Printf("%s Processing Caputure Request for %s\n", c, gourl)
 
 	// Run ChromeDP Magic
 	err := chromedp.Run(ctx,
-		emulation.SetDeviceMetricsOverride(w, h, s, false),
+		emulation.SetDeviceMetricsOverride(int64(float64(w)/s), int64(float64(h)/s), s, false),
 		chromedp.Navigate(gourl),
 		chromedp.Evaluate(fmt.Sprintf("window.scrollTo(0, %d);", p*int64(float64(h)*float64(0.9))), &res),
 		chromedp.Sleep(time.Second*1),
@@ -140,38 +180,42 @@ func capture(gourl string, w int64, h int64, s float64, p int64, ismap bool, out
 		chromedp.Nodes("a", &nodes, chromedp.ByQueryAll))
 
 	if err != nil {
-		log.Printf("%s", err)
+		log.Printf("%s %s", c, err)
 		fmt.Fprintf(out, "<BR>%s<BR>", err)
 		return
 	}
 
-	log.Printf("Landed on: %s, Nodes: %d\n", loc, len(nodes))
+	log.Printf("%s Landed on: %s, Nodes: %d\n", c, loc, len(nodes))
 
 	// Process Screenshot Image
 	bytes.NewReader(pngbuf).Seek(0, 0)
 	img, err := png.Decode(bytes.NewReader(pngbuf))
 	if err != nil {
-		log.Printf("Failed to decode screenshot: %s\n", err)
+		log.Printf("%s Failed to decode screenshot: %s\n", c, err)
 		fmt.Fprintf(out, "<BR>Unable to decode page screenshot:<BR>%s<BR>\n", err)
 		return
 	}
 	gifbuf.Reset()
-	err = gif.Encode(&gifbuf, img, nil)
+	err = gif.Encode(&gifbuf, img, &gif.Options{NumColors: co})
 	if err != nil {
-		log.Printf("Failed to encode GIF: %s\n", err)
+		log.Printf("%s Failed to encode GIF: %s\n", c, err)
 		fmt.Fprintf(out, "<BR>Unable to encode GIF:<BR>%s<BR>\n", err)
 		return
 	}
-	imgpath := fmt.Sprintf("/img/%04d.gif", rand.Intn(9999))
-	log.Printf("Encoded GIF image: %s, Size: %dKB\n", imgpath, len(gifbuf.Bytes())/1024)
+	seq := rand.Intn(9999)
+	imgpath := fmt.Sprintf("/img/%04d.gif", seq)
+	mappath := fmt.Sprintf("/map/%04d.map", seq)
+	log.Printf("%s Encoded GIF image: %s, Size: %dKB\n", c, imgpath, len(gifbuf.Bytes())/1024)
 	gifmap[imgpath] = gifbuf
 
 	// Process Nodes
 	base, _ := url.Parse(loc)
-	if ismap {
-		fmt.Fprintf(out, "<A HREF=\"/map/123.map\"><IMG SRC=\"%s\" ALT=\"wrp\" ISMAP></A>", imgpath)
+	if i {
+		fmt.Fprintf(out, "<A HREF=\"%s\"><IMG SRC=\"%s\" ALT=\"wrp\" BORDER=\"0\" ISMAP></A>", mappath, imgpath)
+		is = append(is, Ismap{xmin: -1, xmax: -1, ymin: -1, ymax: -1, url: fmt.Sprintf("/?url=%s&w=%d&h=%d&s=%1.2f&c=%d&i=on", loc, w, h, s, co)})
+		ion = "&i=on"
 	} else {
-		fmt.Fprintf(out, "<IMG SRC=\"%s\" ALT=\"wrp\" USEMAP=\"#map\">\n<MAP NAME=\"map\">\n", imgpath)
+		fmt.Fprintf(out, "<IMG SRC=\"%s\" ALT=\"wrp\" BORDER=\"0\" USEMAP=\"#map\">\n<MAP NAME=\"map\">\n", imgpath)
 	}
 
 	for _, n := range nodes {
@@ -183,37 +227,61 @@ func capture(gourl string, w int64, h int64, s float64, p int64, ismap bool, out
 		if err != nil {
 			continue
 		}
-		target := fmt.Sprintf("/?url=%s&w=%d&h=%d&s=%1.2f&", tgt, w, h, s) // no page# here
+		target := fmt.Sprintf("/?url=%s&w=%d&h=%d&s=%1.2f&c=%d%s", tgt, w, h, s, co, ion) // no page# here
 
 		if len(b.Content) > 6 && len(target) > 7 {
-			if ismap {
-
+			if i {
+				is = append(is, Ismap{
+					xmin: int64(b.Content[0] * s), ymin: int64(b.Content[1] * s),
+					xmax: int64(b.Content[4] * s), ymax: int64(b.Content[5] * s),
+					url: target})
 			} else {
-			fmt.Fprintf(out, "<AREA SHAPE=\"RECT\" COORDS=\"%.f,%.f,%.f,%.f\" ALT=\"%s\" TITLE=\"%s\" HREF=\"%s\">\n",
-				b.Content[0]*s, b.Content[1]*s, b.Content[4]*s, b.Content[5]*s, n.AttributeValue("href"), n.AttributeValue("href"), target)
+				fmt.Fprintf(out, "<AREA SHAPE=\"RECT\" COORDS=\"%.f,%.f,%.f,%.f\" ALT=\"%s\" TITLE=\"%s\" HREF=\"%s\">\n",
+					b.Content[0]*s, b.Content[1]*s, b.Content[4]*s, b.Content[5]*s, n.AttributeValue("href"), n.AttributeValue("href"), target)
 			}
 		}
 	}
 
-	if !ismap {
+	if i {
+		log.Printf("%s Encoded ISMAP %s\n", c, mappath)
+	} else {
 		fmt.Fprintf(out, "</MAP>\n")
 	}
 	out.(http.Flusher).Flush()
-	log.Printf("Done with caputure for %s\n", gourl)
+	log.Printf("%s Done with caputure for %s\n", c, gourl)
+	ismap[mappath] = is
 }
 
 func main() {
-	ctx, cancel = chromedp.NewContext(context.Background())
-	defer cancel()
 	var addr string
+	var head, headless bool
+	var debug bool
 	flag.StringVar(&addr, "l", ":8080", "Listen address:port, default :8080")
+	flag.BoolVar(&head, "h", false, "Headed mode - display browser window")
+	flag.BoolVar(&debug, "d", false, "Debug ChromeDP")
 	flag.Parse()
+	if head {
+		headless = false
+	} else {
+		headless = true
+	}
+	opts := append(chromedp.DefaultExecAllocatorOptions[:],
+		chromedp.Flag("headless", headless),
+	)
+	actx, cancel := chromedp.NewExecAllocator(context.Background(), opts...)
+	defer cancel()
+	if debug {
+		ctx, cancel = chromedp.NewContext(actx, chromedp.WithDebugf(log.Printf))
+	} else {
+		ctx, cancel = chromedp.NewContext(actx)
+	}
+	defer cancel()
 	rand.Seed(time.Now().UnixNano())
 	http.HandleFunc("/", pageServer)
 	http.HandleFunc("/img/", imgServer)
 	http.HandleFunc("/map/", mapServer)
 	http.HandleFunc("/favicon.ico", http.NotFound)
-	http.HandleFunc("/halt", haltServer)
+	log.Printf("Web Rendering Proxy Version %s\n", version)
 	log.Printf("Starting WRP http server on %s\n", addr)
 	http.ListenAndServe(addr, nil)
 }