requirements: Add 'lxml' as a dependency required for parsing HTML.

We need to parse rendered HTML content of messages while preparing
content for mobile push notifications and for doing so we need to
use lxml's HTML parser.

requirements/common.txt

@@ -181,3 +181,6 @@ pyldap==2.4.37
 
 # Install Python Social Auth
 social-auth-app-django==1.2.0
+
+# Needed for messages' rendered content parsing in push notifications.
+lxml==4.0.0

requirements/dev_lock.txt

@@ -72,7 +72,7 @@ ipython==6.2.0
 jedi==0.10.2              # via ipython
 jinja2==2.9.6
 jmespath==0.9.3           # via boto3, botocore
-lxml==3.8.0               # via parsel, premailer, scrapy
+lxml==4.0.0
 markdown-include==0.5.1
 markdown==2.6.9           # via markdown-include
 markupsafe==1.0

requirements/prod_lock.txt

@@ -51,7 +51,7 @@ ipython-genutils==0.2.0   # via traitlets
 ipython==6.2.0
 jedi==0.10.2              # via ipython
 jinja2==2.9.6
-lxml==3.8.0               # via premailer
+lxml==4.0.0
 markdown-include==0.5.1
 markdown==2.6.9           # via markdown-include
 markupsafe==1.0

version.py

@@ -1,2 +1,2 @@
 ZULIP_VERSION = "1.6.0+git"
-PROVISION_VERSION = '10.3'
+PROVISION_VERSION = '10.4'