')
+# Base Tests -- no rewrite
+>>> parse('', urlrewriter=no_base_canon_rewriter)
+
+
+
+
# HTML Entities
>>> parse('› > ?')
› > ?
@@ -102,8 +111,12 @@ ur"""
>>> parse('SomeTest
', head_insert = '')
SomeTest
-# don't rewrite rel=canonical
+# rel=canonical: rewrite (default)
>>> parse('')
+
+
+# rel=canonical: no_rewrite
+>>> parse('', urlrewriter=no_base_canon_rewriter)
# doctype
@@ -143,7 +156,12 @@ import pprint
urlrewriter = UrlRewriter('20131226101010/http://example.com/some/path/index.html', '/web/')
-def parse(data, head_insert = None):
+no_base_canon_rewriter = UrlRewriter('20131226101010/http://example.com/some/path/index.html',
+ '/web/',
+ rewrite_opts=dict(rewrite_rel_canon=False,
+ rewrite_base=False))
+
+def parse(data, head_insert=None, urlrewriter=urlrewriter):
parser = HTMLRewriter(urlrewriter, head_insert = head_insert)
#data = data.decode('utf-8')
result = parser.rewrite(data) + parser.close()
diff --git a/pywb/rewrite/test/test_regex_rewriters.py b/pywb/rewrite/test/test_regex_rewriters.py
index 92975a7f..19ea5eb6 100644
--- a/pywb/rewrite/test/test_regex_rewriters.py
+++ b/pywb/rewrite/test/test_regex_rewriters.py
@@ -3,7 +3,7 @@ r"""
# Custom Regex
#=================================================================
# Test https->http converter (other tests below in subclasses)
->>> RegexRewriter([(RegexRewriter.HTTPX_MATCH_STR, RegexRewriter.remove_https, 0)]).rewrite('a = https://example.com; b = http://example.com; c = https://some-url/path/https://embedded.example.com')
+>>> RegexRewriter(urlrewriter, [(RegexRewriter.HTTPX_MATCH_STR, RegexRewriter.remove_https, 0)]).rewrite('a = https://example.com; b = http://example.com; c = https://some-url/path/https://embedded.example.com')
'a = http://example.com; b = http://example.com; c = http://some-url/path/http://embedded.example.com'
diff --git a/pywb/rewrite/test/test_rewrite_live.py b/pywb/rewrite/test/test_rewrite_live.py
index b54138fa..3ea189a4 100644
--- a/pywb/rewrite/test/test_rewrite_live.py
+++ b/pywb/rewrite/test/test_rewrite_live.py
@@ -13,7 +13,7 @@ urlrewriter = UrlRewriter('20131226101010/http://example.com/some/path/index.htm
bn_urlrewriter = UrlRewriter('20131226101010bn_/http://example.com/some/path/index.html', '/pywb/')
def head_insert_func(rule, cdx):
- if rule.js_rewrite_location == True:
+ if rule.js_rewrite_location != 'urls':
return ''
else:
return ''
@@ -26,10 +26,10 @@ def test_local_1():
'com,example,test)/')
# wombat insert added
- assert '' in buff
+ assert '' in buff, buff
- # location rewritten
- assert 'window.WB_wombat_location = "/other.html"' in buff
+ # JS location and JS link rewritten
+ assert 'window.WB_wombat_location = "/pywb/20131226101010/http:\/\/example.com/dynamic_page.html"' in buff
# link rewritten
assert '"/pywb/20131226101010/http://example.com/some/path/another.html"' in buff
@@ -65,7 +65,7 @@ def test_local_no_head_banner_only():
# link NOT rewritten
assert '"another.html"' in buff
-def test_local_banner_only():
+def test_local_banner_only_no_rewrite():
status_headers, buff = get_rewritten(get_test_dir() + 'text_content/sample.html',
bn_urlrewriter,
head_insert_func,
@@ -74,13 +74,13 @@ def test_local_banner_only():
# wombat insert added
assert '' in buff
- # location NOT rewritten
- assert 'window.location = "/other.html"' in buff
+ # JS location NOT rewritten, JS link NOT rewritten
+ assert 'window.location = "http:\/\/example.com/dynamic_page.html"' in buff, buff
# link NOT rewritten
assert '"another.html"' in buff
-def test_local_2_no_js_location_rewrite():
+def test_local_2_link_only_rewrite():
status_headers, buff = get_rewritten(get_test_dir() + 'text_content/sample.html',
urlrewriter,
head_insert_func,
@@ -89,13 +89,28 @@ def test_local_2_no_js_location_rewrite():
# no wombat insert
assert '' not in buff
- # no location rewrite
- assert 'window.location = "/other.html"' in buff
+ # JS location NOT rewritten, JS link rewritten
+ assert 'window.location = "/pywb/20131226101010/http:\/\/example.com/dynamic_page.html"' in buff
# still link rewrite
assert '"/pywb/20131226101010/http://example.com/some/path/another.html"' in buff
+def test_local_2_js_loc_only_rewrite():
+ status_headers, buff = get_rewritten(get_test_dir() + 'text_content/sample.html',
+ urlrewriter,
+ head_insert_func,
+ 'example,example,test,loconly)/')
+
+ # wombat insert added
+ assert '' in buff
+
+ # JS location rewritten, JS link NOT rewritten
+ assert 'window.WB_wombat_location = "http:\/\/example.com/dynamic_page.html"' in buff
+
+ # still link rewrite in HTML
+ assert '"/pywb/20131226101010/http://example.com/some/path/another.html"' in buff
+
def test_example_1():
status_headers, buff = get_rewritten('http://example.com/', urlrewriter, req_headers={'Connection': 'close'})
diff --git a/pywb/rewrite/url_rewriter.py b/pywb/rewrite/url_rewriter.py
index aa87260c..a5cc7952 100644
--- a/pywb/rewrite/url_rewriter.py
+++ b/pywb/rewrite/url_rewriter.py
@@ -20,13 +20,14 @@ class UrlRewriter(object):
REL_SCHEME = ('//', r'\/\/', r'\\/\\/')
def __init__(self, wburl, prefix, full_prefix=None, rel_prefix=None,
- root_path=None, cookie_scope=None):
+ root_path=None, cookie_scope=None, rewrite_opts={}):
self.wburl = wburl if isinstance(wburl, WbUrl) else WbUrl(wburl)
self.prefix = prefix
self.full_prefix = full_prefix
self.rel_prefix = rel_prefix if rel_prefix else prefix
self.root_path = root_path if root_path else '/'
self.cookie_scope = cookie_scope
+ self.rewrite_opts = rewrite_opts
def rewrite(self, url, mod=None):
# if special protocol, no rewriting at all
diff --git a/pywb/rules.yaml b/pywb/rules.yaml
index aa80717e..b55d8278 100644
--- a/pywb/rules.yaml
+++ b/pywb/rules.yaml
@@ -120,6 +120,11 @@ rules:
# testing rules -- not for valid domain
#=================================================================
# this rule block is a non-existent prefix merely for testing
+ - url_prefix: 'example,example,test,loconly)/'
+
+ rewrite:
+ js_rewrite_location: location
+
- url_prefix: 'example,example,test)/'
canonicalize:
@@ -131,10 +136,10 @@ rules:
- id
rewrite:
- js_rewrite_location: False
+ js_rewrite_location: urls
- # all domain rules -- fallback to this dataset
+ # all domain rules -- fallback to this dataset
#=================================================================
# Applies to all urls -- should be last
- url_prefix: ''
diff --git a/pywb/static/wb.js b/pywb/static/wb.js
index 00592bbf..ea14c9fa 100644
--- a/pywb/static/wb.js
+++ b/pywb/static/wb.js
@@ -35,10 +35,14 @@ function init_banner() {
bid = PLAIN_BANNER_ID;
}
+ if (!document || !document.body) {
+ return;
+ }
+
if (document.getElementById(bid) != null) {
return;
}
-
+
_wb_js.create_banner_element(bid);
}
@@ -56,14 +60,14 @@ this.ts_to_date = function(ts, is_gmt)
if (ts.length < 14) {
return ts;
}
-
- var datestr = (ts.substring(0, 4) + "-" +
+
+ var datestr = (ts.substring(0, 4) + "-" +
ts.substring(4, 6) + "-" +
ts.substring(6, 8) + "T" +
ts.substring(8, 10) + ":" +
ts.substring(10, 12) + ":" +
ts.substring(12, 14) + "-00:00");
-
+
var date = new Date(datestr);
if (is_gmt) {
return date.toGMTString();
@@ -117,10 +121,21 @@ function notify_top() {
}
this.load = function() {
+ if (window._wb_js_inited) {
+ return;
+ }
+
+ window._wb_js_inited = true;
+
if ((window.self == window.top) && wbinfo) {
- if (wbinfo.top_url && (window.location.href != wbinfo.top_url) && wbinfo.mod != "bn_") {
+
+ var hash = window.location.hash;
+
+ var loc = window.location.href.replace(window.location.hash, "");
+
+ if (wbinfo.top_url && (loc != wbinfo.top_url) && wbinfo.mod != "bn_") {
// Auto-redirect to top frame
- window.location.replace(wbinfo.top_url);
+ window.location.replace(wbinfo.top_url + hash);
} else {
// Init Banner (no frame or top frame)
add_event("readystatechange", init_banner, document);
diff --git a/pywb/ui/head_insert.html b/pywb/ui/head_insert.html
index 4e53a5d0..812bcfe0 100644
--- a/pywb/ui/head_insert.html
+++ b/pywb/ui/head_insert.html
@@ -1,5 +1,5 @@
-{% if rule.js_rewrite_location and include_wombat %}
+{% if rule.js_rewrite_location != 'urls' and include_wombat %}
Test Content
diff --git a/setup.py b/setup.py
index b5ef2d26..4a36b078 100755
--- a/setup.py
+++ b/setup.py
@@ -34,7 +34,7 @@ class PyTest(TestCommand):
setup(
name='pywb',
- version='0.6.5',
+ version='0.6.6',
url='https://github.com/ikreymer/pywb',
author='Ilya Kreymer',
author_email='ikreymer@gmail.com',