def doc_categorize (repo, response, params):
from uplib.basicPlugins import show_abstract, _is_sensible_browser
from uplib.basicPlugins import show_title, STANDARD_BACKGROUND_COLOR, STANDARD_TOOLS_COLOR, STANDARD_LEGEND_COLOR
from uplib.basicPlugins import __issue_javascript_head_boilerplate as issue_javascript_head_boilerplate
from uplib.basicPlugins import __issue_menu_definition as issue_menu_definition
from uplib.basicPlugins import __issue_title_styles as issue_title_styles
global _CONFIGURATION
if _CONFIGURATION is None:
_CONFIGURATION = { "exclusions": [
re.compile(x.strip()) for x in configurator.default_configurator().get("categorize-excluded-categories", "").split(",") if x.strip()]}
def figure_size(count, avgsize):
if avgsize < 0.0001:
return 0.0001
return math.sqrt(math.log((count * (math.e - 1))/avgsize + 1))
doc_id = params.get("doc_id")
if not doc_id:
response.error(HTTPCodes.BAD_REQUEST, "No doc_id parameter specified.")
return
doc = repo.valid_doc_id(doc_id) and repo.get_document(doc_id)
if not doc:
response.error(HTTPCodes.BAD_REQUEST, "Invalid doc_id parameter '%s' specified." % doc_id)
return
fp = response.open()
title = (doc.get_metadata("title") or doc.id).encode("UTF-8", "strict")
fp.write("<head><title>Categorizing '%s'</title>\n" % htmlescape(title))
fp.write('<meta http-equiv="Content-Script-Type" content="text/javascript">\n')
fp.write('<link REL="SHORTCUT ICON" HREF="/favicon.ico">\n')
fp.write('<link REL="ICON" type="image/ico" HREF="/favicon.ico">\n')
issue_javascript_head_boilerplate(fp)
issue_title_styles(fp)
fp.write('</head><body bgcolor="%s">\n' % STANDARD_BACKGROUND_COLOR)
issue_menu_definition(fp)
show_abstract(repo, doc, fp, _is_sensible_browser(response.user_agent), showpagesearch=False)
fp.write("<hr />\n")
doccats = [x.lower() for x in doc.get_category_strings()]
for cat in doccats[:]:
if cat.find('/') >= 0:
parts = cat.split('/')
for i in range(1, len(parts)):
doccats.append('/'.join(parts[:i]))
tags = find_likely_tags(doc)
if tags:
# try to remove duplicates
stags = min(10, len(tags))
# tagnames = [tag[0].split('/')[0] for tag in tags[:stags] if tag[0].find('/') >= 0]
# count = 0
# i = 0
# while tagnames and (i < stags):
# if tags[i][0] in tagnames:
# del tags[i]
# stags = min(10, len(tags))
# tagnames = [tag[0].split('/')[0] for tag in tags[:stags] if tag[0].find('/') >= 0]
# else:
# i += 1
fp.write("<center><small><i>Likely categories</i></small><br />")
count = 0
topscore = _adjust_score(*tags[0][1][:2])
exclusions = _CONFIGURATION and _CONFIGURATION.get("exclusions")
for name, (score, ndocs, ascore) in tags:
if count > stags:
break
skip = False
for exclusion in exclusions:
if exclusion.match(name.lower()):
skip = True
break
if skip:
continue
if count > 0:
fp.write(" · ")
#size = max(0.5, (2/topscore) * ascore)
size = 1
color = (name.lower() in doccats) and "red" or "black"
action = '/'.join(response.request_path.split('/')[:3]) + '/doc_%s_category?doc_id=%s&tag=%s' % (
(name.lower() in doccats) and "remove" or "add", doc.id, urllib.quote_plus(name))
fp.write('<a style="font-size: %fem; color: %s;" href="%s" title="%s the \'%s\' category (score=%.3f)">%s</a>' % (
size, color, action,
(name.lower() in doccats) and "remove" or "add",
htmlescape(name), ascore, htmlescape(name)))
count += 1
fp.write("</center></p><hr />\n")
fp.write('<form action="%s" method=get><center>Add a new category to this document: ' %
('/'.join(response.request_path.split('/')[:3]) + '/doc_add_category'))
fp.write('<input type=hidden name="doc_id" value="%s">\n' % doc.id)
fp.write('<input type=text name="tag" value="" size=40></form></center>\n')
note(4, "doc_categorize: retrieving repository categories... (%s)", time.ctime())
cats = repo.get_categories_with_docs()
note(4, "doc_categorize: have categories (%s)", time.ctime())
if cats:
fp.write("<hr>\n<center><small><i>All categories</i></small><br />")
avgsize = sum([len(x) for x in cats.values()]) / float(len(cats))
catkeys = cats.keys()
catkeys.sort(lambda x, y: cmp(x.lower(), y.lower()))
first = True
exclusions = _CONFIGURATION and _CONFIGURATION.get("exclusions")
for name in catkeys:
skip = False
for exclusion in exclusions:
if exclusion.match(name.lower()):
skip = True
break
if skip:
continue
if not first:
fp.write(" · ")
else:
first = False
size = max(0.5, figure_size(len(cats[name]), avgsize))
color = (name.lower() in doccats) and "red" or "black"
action = '/'.join(response.request_path.split('/')[:3]) + '/doc_%s_category?doc_id=%s&tag=%s' % (
(name.lower() in doccats) and "remove" or "add", doc.id, urllib.quote_plus(name))
actionsee = '/action/basic/repo_search?query=%s' % (
urllib.quote_plus('categories:"%s"' % name))
fp.write('<a style="font-size: %fem; color: %s;" href="%s" title="%s the \'%s\' category">%s</a>' % (
size, color, action,
(name.lower() in doccats) and "remove" or "add",
htmlescape(name), htmlescape(name)))
fp.write('<a style="font-size: %fem; color: %s; vertical-align: super;" href="%s" ' % (
max(0.4, size/2), STANDARD_LEGEND_COLOR, actionsee) +
'title="see the %s document%s in the \'%s\' category" target="_blank">%d</a>' % (
(len(cats[name]) == 1) and "one" or str(len(cats[name])),
(len(cats[name]) != 1) and "s" or "", htmlescape(name), len(cats[name])))
fp.write("</body>\n")
def related (repo, response, params):
"""
Find other documents related to the query document.
:param doc_id: the query document
:type doc_id: UpLib doc ID string
:param use-authorship: whether or not to use co-authorship as a measure of relatedness. Defaults to "true".
:type use-authorship: "true" or "false"
:param use-history: whether or not to to use the use history (most recently used list) as a factor in the calculation. Defaults to "true".
:type use-history: "true" or "false"
:param format: whether to return non-browser format results. Specifying "xml" will cause an XML document to be returned containing the results. Specifying "ziplist" will cause a zip file containing extra information about each document to be returned. If the ``format`` parameter is not specified, an HTML page showing the results broken down by category is returned.
:type format: "xml" or "ziplist" or none
:result: list of other documents related to the query document. See discussion of the ``format`` parameter.
:rtype: varies
"""
doc_id = params.get("doc_id")
if not doc_id:
response.error(HTTPCodes.BAD_REQUEST, "No doc_id specified.")
return
if not repo.valid_doc_id(doc_id):
response.error(HTTPCodes.NOT_FOUND, "Invalid doc_id %s specified." % doc_id)
return
doc = repo.get_document(doc_id)
use_authorship = (params.get("use-authorship") or "true") == "true"
use_history = (params.get("use-history") or "true") == "true"
docs, likethis, authored, recent, others, qstring = find_related(doc, True, use_history=use_history, use_authorship=use_authorship)
if response.xml_request or (params.get("format") == "xml"):
retval = getDOMImplementation().createDocument(None, "result", None)
e = retval.createElement('paradigm')
e.setAttribute('id', doc_id)
e.setAttribute("title", _safe_title(doc))
e.setAttribute('use-history', use_history and "true" or "false")
e.setAttribute('use-authorship', use_authorship and "true" or "false")
retval.documentElement.appendChild(e)
g = retval.createElement('similar')
g.setAttribute('query', qstring)
for doc, score in likethis:
e = retval.createElement('document')
e.setAttribute('doc_id', doc.id)
e.setAttribute('score', str(score))
title = _safe_title(doc)
note("title is %s", repr(title))
e.setAttribute('title', title)
g.appendChild(e)
retval.documentElement.appendChild(g)
g = retval.createElement('co-authored')
for doc, score in authored:
e = retval.createElement('document')
e.setAttribute('doc_id', doc.id)
e.setAttribute('score', str(score))
title = _safe_title(doc)
note("title is %s", repr(title))
e.setAttribute('title', title)
g.appendChild(e)
retval.documentElement.appendChild(g)
g = retval.createElement('recent')
for doc, score in recent:
e = retval.createElement('document')
e.setAttribute('doc_id', doc.id)
e.setAttribute('score', str(score))
title = _safe_title(doc)
note("title is %s", repr(title))
e.setAttribute('title', title)
g.appendChild(e)
retval.documentElement.appendChild(g)
g = retval.createElement('linked')
for doc, score in others:
e = retval.createElement('document')
e.setAttribute('doc_id', doc.id)
e.setAttribute('score', str(score))
title = _safe_title(doc)
note("title is %s", repr(title))
e.setAttribute('title', title)
g.appendChild(e)
retval.documentElement.appendChild(g)
g = retval.createElement('combined')
for doc, score in docs:
e = retval.createElement('document')
e.setAttribute('doc_id', doc.id)
e.setAttribute('score', str(score))
title = _safe_title(doc)
note("title is %s", repr(title))
e.setAttribute('title', title)
g.appendChild(e)
retval.documentElement.appendChild(g)
fp = response.open("application/xml;charset=utf-8")
fp.write(retval.toxml("UTF-8") + "\n")
fp.close()
elif params.get("format") == "ziplist":
no_icon = (params.get("no-icon") == "true")
include_doc_functions = params.get("include-doc-functions")
tpath = tempfile.mktemp()
zf = zipfile.ZipFile(tpath, "w")
try:
try:
for doc, score in docs:
zf.writestr(doc.id.encode("ASCII", "strict") + "/", "")
zf.writestr(doc.id.encode("ASCII", "strict") + "/score", str(score))
if not no_icon:
zf.writestr(doc.id.encode("ASCII", "strict") + "/first.png", doc.document_icon())
if include_doc_functions:
zf.writestr(doc.id.encode("ASCII", "strict") + "/doc_functions.txt", get_doc_functions(doc))
zf.writestr(doc.id.encode("ASCII", "strict") + "/metadata.txt", doc.metadata_text())
finally:
zf.close()
response.return_file("application/x-uplib-searchresults-zipped", tpath, True)
except:
msg = string.join(traceback.format_exception(*sys.exc_info()))
os.remove(tpath)
note("Exception building zipfile for search results:\n%s", msg)
response.error(HTTPCodes.INTERNAL_SERVER_ERROR, "Can't build zipfile for search results:\n%s\n" % htmlescape(msg))
else:
fp = response.open()
title = "Documents related to %s" % repr(doc.get_metadata("title") or doc.id)
fp.write("<head><title>%s</title>\n" % htmlescape(title))
fp.write('<meta http-equiv="Content-Script-Type" content="text/javascript">\n')
fp.write('<link REL="SHORTCUT ICON" HREF="/favicon.ico">\n')
fp.write('<link REL="ICON" type="image/ico" HREF="/favicon.ico">\n')
issue_javascript_head_boilerplate(fp)
issue_title_styles(fp)
fp.write('</head><body bgcolor="%s">\n' % STANDARD_BACKGROUND_COLOR)
issue_menu_definition(fp)
fp.write('<h2>%s</h2><br>\n' % htmlescape(title))
show_abstract(repo, doc, fp, True, showpagesearch=False)
fp.write('<p><hr><b>Context documents:</b><br>')
for related, score in docs:
show_title (fp, related, {related.id: score}, True)
fp.write('<p><hr><b>Like this:</b><br>')
for related, score in likethis:
show_title (fp, related, { related.id: score }, True)
fp.write('<p><i>query was: %s</i>\n' % htmlescape(qstring))
fp.write('<p><hr><b>Co-authored:</b><br>')
for related, score in authored:
show_title (fp, related, { related.id: score }, True)
fp.write('<p><hr><b>Recently consulted:</b><br>')
for related, score in recent:
show_title (fp, related, { related.id: score }, True)
fp.write('<p><hr><b>Other considerations:</b><br>')
for explanation, related, score in others:
fp.write('<p><i>%s</i><br>\n' % htmlescape(explanation))
show_title (fp, related, { related.id: score }, True)
fp.write('</body>\n')
fp.close()
