searx

commit: bacc9a3df156605c49fb93ede6e6a730110d15db
parent e202c6fbec7c9dff67acdfe387699010d61da32b
Author: Kirill Isakov <ukwt@ya.ru>
Date:   Mon, 28 Mar 2016 19:15:03 +0600

Add paging support to XPath & Erowid engines

Diffstat:
M
searx/engines/xpath.py
16
+++++++++++++++-
M
searx/settings.yml
5
++++-

2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/searx/engines/xpath.py b/searx/engines/xpath.py
@@ -11,6 +11,14 @@ title_xpath = None
 suggestion_xpath = ''
 results_xpath = ''
 
+# parameters for engines with paging support
+#
+# number of results on each page
+# (only needed if the site requires not a page number, but an offset)
+page_size = 1
+# number of the first page (usually 0 or 1)
+first_page_num = 1
+
 
 '''
 if xpath_results is list, extract the text from each result and concat the list
@@ -76,8 +84,14 @@ def normalize_url(url):
 
 def request(query, params):
     query = urlencode({'q': query})[2:]
-    params['url'] = search_url.format(query=query)
+
+    fp = {'query': query}
+    if paging and search_url.find('{pageno}') >= 0:
+        fp['pageno'] = (params['pageno'] + first_page_num - 1) * page_size
+
+    params['url'] = search_url.format(**fp)
     params['query'] = query
+
     return params
 
 
diff --git a/searx/settings.yml b/searx/settings.yml
@@ -84,7 +84,10 @@ engines:
 
   - name : erowid
     engine : xpath
-    search_url : https://www.erowid.org/search.php?q={query}
+    paging : True
+    first_page_num : 0
+    page_size : 30
+    search_url : https://www.erowid.org/search.php?q={query}&s={pageno}
     url_xpath : //dl[@class="results-list"]/dt[@class="result-title"]/a/@href
     title_xpath : //dl[@class="results-list"]/dt[@class="result-title"]/a/text()
     content_xpath : //dl[@class="results-list"]/dd[@class="result-details"]