Merge branch 'pdfjs' of https://github.com/Kingdread/qutebrowser into…

… Kingdread-pdfjs

.eslintignore

@@ -0,0 +1 @@
+qutebrowser/3rdparty/pdfjs/*

MANIFEST.in

@@ -3,6 +3,7 @@ recursive-include qutebrowser/html *.html
 recursive-include qutebrowser/img *.svg *.png
 recursive-include qutebrowser/test *.py
 recursive-include qutebrowser/javascript *.js
+graft qutebrowser/3rdparty
 graft icons
 graft doc/img
 graft misc
@@ -27,6 +28,7 @@ exclude qutebrowser.rcc
 exclude .coveragerc
 exclude .pylintrc
 exclude .eslintrc
+exclude .eslintignore
 exclude doc/help
 exclude .appveyor.yml
 exclude .travis.yml

README.asciidoc

@@ -272,3 +272,13 @@ GNU General Public License for more details.
 
 You should have received a copy of the GNU General Public License
 along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+pdf.js
+------
+
+qutebrowser uses https://github.com/mozilla/pdf.js/[pdf.js] to display
+PDF files in the browser.
+
+pdf.js is distributed under the terms of the Apache License. You can
+find a copy of the license in `qutebrowser/pdfjs/LICENSE` or online
+http://www.apache.org/licenses/LICENSE-2.0.html[here].

qutebrowser/browser/network/qutescheme.py

@@ -31,11 +31,13 @@
 
 import functools
 import configparser
+import mimetypes
 
 from PyQt5.QtCore import pyqtSlot, QObject
 from PyQt5.QtNetwork import QNetworkReply
 
 import qutebrowser
+from qutebrowser.browser import pdfjs
 from qutebrowser.browser.network import schemehandler, networkreply
 from qutebrowser.utils import (version, utils, jinja, log, message, docutils,
                                objreg)
@@ -93,8 +95,11 @@ def createRequest(self, _op, request, _outgoing_data):
             return networkreply.ErrorNetworkReply(
                 request, str(e), QNetworkReply.ContentNotFoundError,
                 self.parent())
+        mimetype, _encoding = mimetypes.guess_type(request.url().fileName())
+        if mimetype is None:
+            mimetype = 'text/html'
         return networkreply.FixedDataNetworkReply(
-            request, data, 'text/html', self.parent())
+            request, data, mimetype, self.parent())
 
 
 class JSBridge(QObject):
@@ -201,3 +206,10 @@ def qute_settings(win_id, _request):
         win_id=win_id, title='settings', config=configdata,
         confget=config_getter)
     return html.encode('UTF-8', errors='xmlcharrefreplace')
+
+
+@add_handler('pdfjs')
+def qute_pdfjs(_win_id, request):
+    """Handler for qute://pdfjs. Return the pdf.js viewer."""
+    urlpath = request.url().path()
+    return pdfjs.get_pdfjs_res(urlpath)

qutebrowser/browser/pdfjs.py

@@ -0,0 +1,175 @@
+# vim: ft=python fileencoding=utf-8 sts=4 sw=4 et:
+
+# Copyright 2015 Daniel Schadt
+#
+# This file is part of qutebrowser.
+#
+# qutebrowser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# qutebrowser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with qutebrowser.  If not, see <http://www.gnu.org/licenses/>.
+
+"""pdf.js integration for qutebrowser."""
+
+import os
+
+from PyQt5.QtCore import QUrl
+
+from qutebrowser.browser import webelem
+from qutebrowser.utils import utils
+
+
+class PDFJSNotFound(Exception):
+
+    """Raised when no pdf.js installation is found."""
+
+    pass
+
+
+def generate_pdfjs_page(url):
+    """Return the html content of a page that displays url with pdfjs.
+
+    Returns a string.
+
+    Args:
+        url: The url of the pdf as QUrl.
+    """
+    viewer = get_pdfjs_res('web/viewer.html').decode('utf-8')
+    script = _generate_pdfjs_script(url)
+    html_page = viewer.replace(
+        '</body>', '</body><script>{}</script>'.format(script)
+    )
+    return html_page
+
+
+def _generate_pdfjs_script(url):
+    """Generate the script that shows the pdf with pdf.js.
+
+    Args:
+        url: The url of the pdf page as QUrl.
+    """
+    return (
+        'PDFJS.verbosity = PDFJS.VERBOSITY_LEVELS.info;\n'
+        'PDFView.open("{url}");\n'
+    ).format(url=webelem.javascript_escape(url.toString(QUrl.FullyEncoded)))
+
+
+def fix_urls(asset):
+    """Take a html page and replace each relative URL wth an absolute.
+
+    This is specialized for pdf.js files and not a general purpose function.
+
+    Args:
+        asset: js file or html page as string.
+    """
+    new_urls = {
+        'viewer.css': 'qute://pdfjs/web/viewer.css',
+        'compatibility.js': 'qute://pdfjs/web/compatibility.js',
+        'locale/locale.properties':
+            'qute://pdfjs/web/locale/locale.properties',
+        'l10n.js': 'qute://pdfjs/web/l10n.js',
+        '../build/pdf.js': 'qute://pdfjs/build/pdf.js',
+        'debugger.js': 'qute://pdfjs/web/debugger.js',
+        'viewer.js': 'qute://pdfjs/web/viewer.js',
+        'compressed.tracemonkey-pldi-09.pdf': '',
+        './images/': 'qute://pdfjs/web/images/',
+        '../build/pdf.worker.js': 'qute://pdfjs/build/pdf.worker.js',
+        '../web/cmaps/': 'qute://pdfjs/web/cmaps/',
+    }
+    for original, new in new_urls.items():
+        asset = asset.replace(original, new)
+    return asset
+
+
+SYSTEM_PDFJS_PATHS = [
+    '/usr/share/pdf.js/',  # Debian pdf.js-common
+    '/usr/share/javascript/pdf/',  # Debian libjs-pdf
+    os.path.expanduser('~/.local/share/qutebrowser/pdfjs/'),  # fallback
+]
+
+
+def get_pdfjs_res(path):
+    """Get a pdf.js resource in binary format.
+
+    Args:
+        path: The path inside the pdfjs directory.
+    """
+    path = path.lstrip('/')
+    content = None
+
+    # First try a system wide installation
+    # System installations might strip off the 'build/' or 'web/' prefixes.
+    # qute expects them, so we need to adjust for it.
+    names_to_try = [path, _remove_prefix(path)]
+    for system_path in SYSTEM_PDFJS_PATHS:
+        content = _read_from_system(system_path, names_to_try)
+        if content is not None:
+            break
+
+    # Fallback to bundled pdf.js
+    if content is None:
+        res_path = '3rdparty/pdfjs/{}'.format(path)
+        try:
+            content = utils.read_file(res_path, binary=True)
+        except FileNotFoundError:
+            raise PDFJSNotFound
+
+    try:
+        # Might be script/html or might be binary
+        text_content = content.decode('utf-8')
+    except UnicodeDecodeError:
+        return content
+    text_content = fix_urls(text_content)
+    return text_content.encode('utf-8')
+
+
+def _remove_prefix(path):
+    """Remove the web/ or build/ prefix of a pdfjs-file-path.
+
+    Args:
+        path: Path as string where the prefix should be stripped off.
+    """
+    prefixes = {'web/', 'build/'}
+    if any(path.startswith(prefix) for prefix in prefixes):
+        return path.split('/', maxsplit=1)[1]
+    # Return the unchanged path if no prefix is found
+    return path
+
+
+def _read_from_system(system_path, names):
+    """Try to read a file with one of the given names in system_path.
+
+    Each file in names is considered equal, the first file that is found
+    is read and its binary content returned.
+
+    Returns None if no file could be found
+
+    Args:
+        system_path: The folder where the file should be searched.
+        names: List of possible file names.
+    """
+    for name in names:
+        try:
+            with open(os.path.join(system_path, name), 'rb') as f:
+                return f.read()
+        except OSError:
+            continue
+    return None
+
+
+def is_available():
+    """Return true if a pdfjs installation is available."""
+    try:
+        get_pdfjs_res('build/pdf.js')
+    except PDFJSNotFound:
+        return False
+    else:
+        return True

qutebrowser/browser/webpage.py

@@ -30,7 +30,7 @@
 from PyQt5.QtWebKitWidgets import QWebPage
 
 from qutebrowser.config import config
-from qutebrowser.browser import http, tabhistory
+from qutebrowser.browser import http, tabhistory, pdfjs
 from qutebrowser.browser.network import networkmanager
 from qutebrowser.utils import (message, usertypes, log, jinja, qtutils, utils,
                                objreg, debug)
@@ -218,6 +218,19 @@ def _ask(self, text, mode, default=None):
         q.deleteLater()
         return q.answer
 
+    def _show_pdfjs(self, reply):
+        """Show the reply with pdfjs."""
+        try:
+            page = pdfjs.generate_pdfjs_page(reply.url()).encode('utf-8')
+        except pdfjs.PDFJSNotFound:
+            # pylint: disable=no-member
+            # WORKAROUND for https://bitbucket.org/logilab/pylint/issue/490/
+            page = (jinja.env.get_template('no_pdfjs.html')
+                    .render(url=reply.url().toDisplayString())
+                    .encode('utf-8'))
+        self.mainFrame().setContent(page, 'text/html', reply.url())
+        reply.deleteLater()
+
     def shutdown(self):
         """Prepare the web page for being deleted."""
         self._is_shutting_down = True
@@ -305,6 +318,10 @@ def on_unsupported_content(self, reply):
             else:
                 reply.finished.connect(functools.partial(
                     self.display_content, reply, 'image/jpeg'))
+        elif (mimetype in {'application/pdf', 'application/x-pdf'} and
+              config.get('content', 'enable-pdfjs')):
+            # Use pdf.js to display the page
+            self._show_pdfjs(reply)
         else:
             # Unknown mimetype, so download anyways.
             download_manager.fetch(reply,

qutebrowser/config/configdata.py

@@ -841,6 +841,11 @@ def data(readonly=False):
              "required to exactly match the requested domain.\n\n"
              "Local domains are always exempt from hostblocking."),
 
+            ('enable-pdfjs', SettingValue(typ.Bool(), 'false'),
+             "Enable pdf.js to view PDF files in the browser.\n\n"
+             "Note that the files can still be downloaded by clicking"
+             " the download button in the pdf.js viewer."),
+
             readonly=readonly
         )),