use straight google search instead of the API. added argparse for better help messages. added ability to result different positions, get the full answer, and get the link

README.md

@@ -6,26 +6,59 @@ Are you a hack programmer? Do you find yourself constantly Googling for how to d
 Suppose you want to know how to format a date in bash. Why open your browser and read through blogs when you can just...
 
     $ howdoi format string bash
-    > date +"%m-%d-%y"
+    > [foo@bar ~]$date --date "2012-02-13" +%s
+    > 1329055200
+    > [foo@bar ~]$date --date @1329055200
+    > Mon Feb 13 00:00:00 EST 2012
+    > [foo@bar ~]$date --date @1329055200 +"%Y-%m-%d"
+    > 2012-02-13
+
+howdoi will answer all sorts of queries
 
     $ howdoi print stack trace python
+    > import traceback
+    >
+    > try:
+    >     1/0
+    > except:
+    >     print '>>> traceback <<<'
+    >     traceback.print_exc()
+    >     print '>>> end of traceback <<<'
     > traceback.print_exc()
 
-    $ howdoi create tar
+    $ howdoi convert mp4 to animated gif
+    > video=/path/to/video.avi
+    > outdir=/path/to/output.gif
+    > mplayer "$video" \
+    >         -ao null \
+    >         -ss "00:01:00" \  # starting point
+    >         -endpos 10 \ # duration in second
+    >         -vo gif89a:fps=13:output=$outdir \
+    >         -vf scale=240:180
+
+    $ howdoi create tar archive
     > tar -cf backup.tar --exclude "www/subf3" www
 
 Usage:
 
-`howdoi query`
+    usage: howdoi [-h] [-p POS] [-f] [-l] QUERY [QUERY ...]
+
+    code search tool
+
+    positional arguments:
+      QUERY              the question to answer
+
+    optional arguments:
+      -h, --help         show this help message and exit
+      -p POS, --pos POS  return answer in specified position (default: 1)
+      -f, --full         return the full text of the answer
+      -l, --link         display only the answer link
 
 Extra notes:
 
-*   Requires [BeautifulSoup](http://www.crummy.com/software/BeautifulSoup/)
+*   Requires [PyQuery](http://pypi.python.org/pypi/pyquery)
 *   Special thanks to Rich Jones ([@miserlou](https://github.com/miserlou)) for the idea
 
 TODOs:
 
-*   Pick the longest code block instead of the first
-*   Flag for printing multiple answers instead of the first
-*   Flag for displaying the entire answer
-*   Flag for retrieving the page URL
+*   Be able to pick StackOverflow result based on different criteria (active, oldest, etc.)

howdoi

@@ -1,50 +0,0 @@
-#!/usr/bin/python
-
-##################################################
-#
-# howdoi - a code search tool.
-# written by Benjamin Gleitzman (gleitz@mit.edu)
-# inspired by Rich Jones (rich@anomos.info)
-#
-##################################################
-
-import urllib
-import urllib2
-import sys
-import json
-
-from BeautifulSoup import BeautifulSoup as bs
-
-SEARCH_URL = "https://www.googleapis.com/customsearch/v1?key=AIzaSyCo6SQ6XNvvS3fdJLcDNR4mpdIGGmVcXAk&cx=015163316206774170098:pj94ujarmcg&q={0}&alt=json"
-
-def get_result(url):
-    result = urllib2.urlopen(url)
-    return result.read()
-
-def get_instructions(query):
-    url = SEARCH_URL.format(urllib.quote(query))
-    result = get_result(url)
-    if not result:
-        return ''
-    else:
-        response = json.loads(result)
-        try:
-            link = response['items'][0]['link']
-            page = get_result(link)
-            soup = bs(page)
-            first_answer = soup.find("div", {"id": "answers"})
-            instructions = first_answer.find("code") or first_answer.find("pre")
-            return instructions.text
-        except:
-            return ''
-
-
-def howdoi(query):
-    instructions = get_instructions(query) or "Sorry, couldn't find any help with that topic"
-    print instructions
-
-if __name__ == "__main__":
-    if len(sys.argv) < 2:
-        print "USAGE: howdoi query (e.g. howdoi format date bash)"
-    else:
-        howdoi(" ".join(sys.argv[1:]))

howdoi.py

@@ -0,0 +1,94 @@
+#!/usr/bin/python
+
+##################################################
+#
+# howdoi - a code search tool.
+# written by Benjamin Gleitzman (gleitz@mit.edu)
+# inspired by Rich Jones (rich@anomos.info)
+#
+##################################################
+
+import urllib
+import urllib2
+import sys
+import json
+import argparse
+import re
+
+from pyquery import PyQuery as pq
+
+GOOGLE_SEARCH_URL = "https://www.google.com/search?q=site:stackoverflow.com%20{0}"
+DUCK_SEARCH_URL = "http://duckduckgo.com/html?q=site%3Astackoverflow.com%20{0}"
+USER_AGENT = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_2) AppleWebKit/537.17 (KHTML, like Gecko) Chrome/24.0.1309.0 Safari/537.17"
+
+def get_result(url):
+    print url
+    opener = urllib2.build_opener()
+    opener.addheaders = [('User-agent', USER_AGENT)]
+    result = opener.open(url)
+    return result.read()
+
+def is_question(link):
+    return re.search('questions/\d+/', link)
+
+def get_google_links(query):
+    url = GOOGLE_SEARCH_URL.format(urllib.quote(query))
+    result = get_result(url)
+    html = pq(result)
+    return [a.attrib['href'] for a in html('.l')]
+
+def get_duck_links(query):
+    url = DUCK_SEARCH_URL.format(urllib.quote(query))
+    result = get_result(url)
+    html = pq(result)
+    links = [l.find('a').attrib['href'] for l in html('.links_main')]
+
+def get_link_at_pos(links, pos):
+    pos = int(args['pos']) - 1
+    for link in links:
+        if is_question(link):
+            if pos == 0:
+                break
+            else:
+                pos = pos - 1
+                continue
+    return link
+
+def get_instructions(args):
+    links = get_google_links(args['query'])
+    if not links:
+        return ''
+
+    link = get_link_at_pos(links, args['pos'])
+    if args.get('link'):
+        return '> ' + link
+
+    link = link + '?answertab=votes'
+    page = get_result(link)
+    html = pq(page)
+    first_answer = html('.answer').eq(0)
+    instructions = first_answer.find('pre') or first_answer.find('code')
+    if args['full'] or not instructions:
+        text = first_answer.find('.post-text').eq(0).text()
+    else:
+        text = instructions.eq(0).text()
+    text = '> ' + text
+    text = text.replace('\n', '\n> ')
+    return text
+
+def howdoi(args):
+    args['query'] = ' '.join(args['query']).replace('?', '')
+    instructions = get_instructions(args) or 'Sorry, couldn\'t find any help with that topic'
+    print instructions
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description='code search tool')
+    parser.add_argument('query', metavar='QUERY', type=str, nargs='+',
+                        help='the question to answer')
+    parser.add_argument('-p','--pos', help='return answer in specified position (default: 1)', default=1)
+    parser.add_argument('-f','--full', help='return the full text of the answer',
+                        action='store_true')
+    parser.add_argument('-l','--link', help='display only the answer link',
+                        action='store_true')
+    args = vars(parser.parse_args())
+    howdoi(args)