def addexistingportalbox(req, colID, ln=CFG_SITE_LANG, pbxID=-1, score=0, position='', sel_ln='', callback='yes', confirm=0):
navtrail_previous_links = wsc.getnavtrail() + """> <a class="navtrail" href="%s/admin/websearch/websearchadmin.py/">WebSearch Admin</a> """ % (CFG_SITE_URL)
try:
uid = getUid(req)
except:
return error_page('Error', req)
auth = check_user(req,'cfgwebsearch')
if not auth[0]:
return page(title="Edit Collection",
body=wsc.perform_addexistingportalbox(colID=colID,
ln=ln,
pbxID=pbxID,
score=score,
position=position,
sel_ln=sel_ln,
callback=callback,
confirm=confirm),
uid=uid,
language=ln,
req=req,
navtrail = navtrail_previous_links,
lastupdated=__lastupdated__)
else:
return page(title='Authorization failure',
uid=uid,
body=adderrorbox('try to login first',
datalist=["""You are not a user authorized to perform admin tasks, try to
<a href="%s/youraccount/login?referer=%s/admin/websearch/websearchadmin.py/">login</a> with another account.""" % (CFG_SITE_SECURE_URL, CFG_SITE_URL)]),
navtrail= navtrail_previous_links,
lastupdated=__lastupdated__)
def addexistingportalbox(req, colID, ln=CFG_SITE_LANG, pbxID=-1, score=0, position='', sel_ln='', callback='yes', confirm=0):
navtrail_previous_links = wsc.getnavtrail() + """> <a class="navtrail" href="%s/admin/websearch/websearchadmin.py/">WebSearch Admin</a> """ % (CFG_SITE_URL)
try:
uid = getUid(req)
except:
return error_page('Error', req)
auth = check_user(req,'cfgwebsearch')
if not auth[0]:
return page(title="Edit Collection",
body=wsc.perform_addexistingportalbox(colID=colID,
ln=ln,
pbxID=pbxID,
score=score,
position=position,
sel_ln=sel_ln,
callback=callback,
confirm=confirm),
uid=uid,
language=ln,
req=req,
navtrail = navtrail_previous_links,
lastupdated=__lastupdated__)
else:
return page(title='Authorization failure',
uid=uid,
body=adderrorbox('try to login first',
datalist=["""You are not a user authorized to perform admin tasks, try to
<a href="%s/youraccount/login?referer=%s/admin/websearch/websearchadmin.py/">login</a> with another account.""" % (CFG_SITE_SECURE_URL, CFG_SITE_URL)]),
navtrail= navtrail_previous_links,
lastupdated=__lastupdated__)
def rank_records(rank_method_code, rank_limit_relevance, hitset_global, pattern=[], verbose=0, field='', rg=None, jrec=None):
"""rank_method_code, e.g. `jif' or `sbr' (word frequency vector model)
rank_limit_relevance, e.g. `23' for `nbc' (number of citations) or `0.10' for `vec'
hitset, search engine hits;
pattern, search engine query or record ID (you check the type)
verbose, verbose level
output:
list of records
list of rank values
prefix
postfix
verbose_output"""
voutput = ""
configcreated = ""
starttime = time.time()
afterfind = starttime - time.time()
aftermap = starttime - time.time()
try:
hitset = copy.deepcopy(hitset_global) #we are receiving a global hitset
if 'methods' not in globals():
create_rnkmethod_cache()
function = methods[rank_method_code]["function"]
#we get 'citation' method correctly here
func_object = globals().get(function)
if verbose > 0:
voutput += "function: %s <br/> " % function
voutput += "pattern: %s <br/>" % str(pattern)
if func_object and pattern and pattern[0][0:6] == "recid:" and function == "word_similarity":
result = find_similar(rank_method_code, pattern[0][6:], hitset, rank_limit_relevance, verbose, methods)
elif rank_method_code == "citation":
#we get rank_method_code correctly here. pattern[0] is the search word - not used by find_cit
p = ""
if pattern and pattern[0]:
p = pattern[0][6:]
result = find_citations(rank_method_code, p, hitset, verbose)
elif func_object:
if function == "word_similarity":
result = func_object(rank_method_code, pattern, hitset, rank_limit_relevance, verbose, methods)
elif function in ("word_similarity_solr", "word_similarity_xapian"):
if not rg:
rg = CFG_WEBSEARCH_DEF_RECORDS_IN_GROUPS
if not jrec:
jrec = 0
ranked_result_amount = rg + jrec
if verbose > 0:
voutput += "Ranked result amount: %s<br/><br/>" % ranked_result_amount
if verbose > 0:
voutput += "field: %s<br/>" % field
if function == "word_similarity_solr":
if verbose > 0:
voutput += "In Solr part:<br/>"
result = word_similarity_solr(pattern, hitset, methods[rank_method_code], verbose, field, ranked_result_amount)
if function == "word_similarity_xapian":
if verbose > 0:
voutput += "In Xapian part:<br/>"
result = word_similarity_xapian(pattern, hitset, methods[rank_method_code], verbose, field, ranked_result_amount)
else:
result = func_object(rank_method_code, pattern, hitset, rank_limit_relevance, verbose)
else:
result = rank_by_method(rank_method_code, pattern, hitset, rank_limit_relevance, verbose)
except Exception as e:
register_exception()
result = (None, "", adderrorbox("An error occured when trying to rank the search result "+rank_method_code, ["Unexpected error: %s<br />" % (e,)]), voutput)
afterfind = time.time() - starttime
if result[0] and result[1]: #split into two lists for search_engine
results_similar_recIDs = map(lambda x: x[0], result[0])
results_similar_relevances = map(lambda x: x[1], result[0])
result = (results_similar_recIDs, results_similar_relevances, result[1], result[2], "%s" % configcreated + result[3])
aftermap = time.time() - starttime;
else:
result = (None, None, result[1], result[2], result[3])
#add stuff from here into voutput from result
tmp = voutput+result[4]
if verbose > 0:
tmp += "<br/>Elapsed time after finding: "+str(afterfind)+"\nElapsed after mapping: "+str(aftermap)
result = (result[0],result[1],result[2],result[3],tmp)
#dbg = string.join(map(str,methods[rank_method_code].items()))
#result = (None, "", adderrorbox("Debug ",rank_method_code+" "+dbg),"",voutput);
return result
def rank_records(rank_method_code, rank_limit_relevance, hitset, related_to=[], verbose=0, field='', rg=None, jrec=None):
"""Sorts given records or related records according to given method
Parameters:
- rank_method_code: Sort records using this method
e.g. `jif' or `sbr' (word frequency vector model)
- rank_limit_relevance: A parameter given to the sorting method
e.g. `23' for `nbc' (number of citations)
or `0.10' for `vec'
This is ignored when sorting by
citations. But I don't know what it means.
- hitset: records to sort
- related_to: if specified, instead of sorting given records,
we first fetch the related records ("related" being
defined by the method), then we sort these related
records
- verbose, verbose level
- field: stuff
- rg: more stuff
- jrec: even more stuff
Output:
- list of records
- list of rank values
- prefix, useless it is always '('
- postfix, useless it is always ')'
- verbose_output
"""
voutput = ""
configcreated = ""
starttime = time.time()
afterfind = starttime - time.time()
aftermap = starttime - time.time()
try:
# We are receiving a global hitset
hitset_global = hitset
hitset = intbitset(hitset_global)
if 'methods' not in globals():
create_rnkmethod_cache()
function = METHODS[rank_method_code]["function"]
# Check if we have specific function for sorting by this method
func_object = globals().get(function)
if verbose > 0:
voutput += "function: %s <br/> " % function
voutput += "related_to: %s <br/>" % str(related_to)
if func_object and related_to and related_to[0][0:6] == "recid:" and function == "word_similarity":
result = find_similar(rank_method_code, related_to[0][6:], hitset, rank_limit_relevance, verbose, METHODS)
elif func_object:
if function == "word_similarity":
result = func_object(rank_method_code, related_to, hitset, rank_limit_relevance, verbose, METHODS)
elif function in ("word_similarity_solr", "word_similarity_xapian"):
if not rg:
rg = CFG_WEBSEARCH_DEF_RECORDS_IN_GROUPS
if not jrec:
jrec = 0
ranked_result_amount = rg + jrec
if verbose > 0:
voutput += "Ranked result amount: %s<br/><br/>" % ranked_result_amount
if verbose > 0:
voutput += "field: %s<br/>" % field
if function == "word_similarity_solr":
if verbose > 0:
voutput += "In Solr part:<br/>"
result = word_similarity_solr(related_to, hitset, METHODS[rank_method_code], verbose, field, ranked_result_amount)
if function == "word_similarity_xapian":
if verbose > 0:
voutput += "In Xapian part:<br/>"
result = word_similarity_xapian(related_to, hitset, METHODS[rank_method_code], verbose, field, ranked_result_amount)
else:
result = func_object(rank_method_code, related_to, hitset, rank_limit_relevance, verbose)
else:
result = rank_by_method(rank_method_code, related_to, hitset, rank_limit_relevance, verbose)
except Exception as e:
register_exception()
from invenio.legacy.webpage import adderrorbox
result = (None, "", adderrorbox("An error occured when trying to rank the search result "+rank_method_code, ["Unexpected error: %s<br />" % (e,)]), voutput)
afterfind = time.time() - starttime
if result[0] and result[1]: #split into two lists for search_engine
results_similar_recIDs = [x[0] for x in result[0]]
results_similar_relevances = [x[1] for x in result[0]]
result = (results_similar_recIDs, results_similar_relevances, result[1], result[2], "%s%s" % (configcreated, result[3]))
aftermap = time.time() - starttime
else:
result = (None, None, result[1], result[2], result[3])
#add stuff from here into voutput from result
tmp = voutput+result[4]
if verbose > 0:
tmp += "<br/>Elapsed time after finding: %s\nElapsed after mapping: %s" % (afterfind, aftermap)
result = (result[0], result[1], result[2], result[3], tmp)
#dbg = string.join(map(str,methods[rank_method_code].items()))
#result = (None, "", adderrorbox("Debug ",rank_method_code+" "+dbg),"",voutput)
return result
def rank_records(rank_method_code, rank_limit_relevance, hitset, related_to=[], verbose=0, field='', rg=None, jrec=None):
"""Sorts given records or related records according to given method
Parameters:
- rank_method_code: Sort records using this method
e.g. `jif' or `sbr' (word frequency vector model)
- rank_limit_relevance: A parameter given to the sorting method
e.g. `23' for `nbc' (number of citations)
or `0.10' for `vec'
This is ignored when sorting by
citations. But I don't know what it means.
- hitset: records to sort
- related_to: if specified, instead of sorting given records,
we first fetch the related records ("related" being
defined by the method), then we sort these related
records
- verbose, verbose level
- field: stuff
- rg: more stuff
- jrec: even more stuff
Output:
- list of records
- list of rank values
- prefix, useless it is always '('
- postfix, useless it is always ')'
- verbose_output
"""
voutput = ""
configcreated = ""
starttime = time.time()
afterfind = starttime - time.time()
aftermap = starttime - time.time()
try:
# We are receiving a global hitset
hitset_global = hitset
hitset = intbitset(hitset_global)
if 'methods' not in globals():
create_rnkmethod_cache()
function = METHODS[rank_method_code]["function"]
# Check if we have specific function for sorting by this method
func_object = globals().get(function)
if verbose > 0:
voutput += "function: %s <br/> " % function
voutput += "related_to: %s <br/>" % str(related_to)
if func_object and related_to and related_to[0][0:6] == "recid:" and function == "word_similarity":
result = find_similar(rank_method_code, related_to[0][6:], hitset, rank_limit_relevance, verbose, METHODS)
elif func_object:
if function == "word_similarity":
result = func_object(rank_method_code, related_to, hitset, rank_limit_relevance, verbose, METHODS)
elif function in ("word_similarity_solr", "word_similarity_xapian"):
if not rg:
rg = CFG_WEBSEARCH_DEF_RECORDS_IN_GROUPS
if not jrec:
jrec = 0
ranked_result_amount = rg + jrec
if verbose > 0:
voutput += "Ranked result amount: %s<br/><br/>" % ranked_result_amount
if verbose > 0:
voutput += "field: %s<br/>" % field
if function == "word_similarity_solr":
if verbose > 0:
voutput += "In Solr part:<br/>"
result = word_similarity_solr(related_to, hitset, METHODS[rank_method_code], verbose, field, ranked_result_amount)
if function == "word_similarity_xapian":
if verbose > 0:
voutput += "In Xapian part:<br/>"
result = word_similarity_xapian(related_to, hitset, METHODS[rank_method_code], verbose, field, ranked_result_amount)
else:
result = func_object(rank_method_code, related_to, hitset, rank_limit_relevance, verbose)
else:
result = rank_by_method(rank_method_code, related_to, hitset, rank_limit_relevance, verbose)
except Exception as e:
register_exception()
from invenio.legacy.webpage import adderrorbox
result = (None, "", adderrorbox("An error occured when trying to rank the search result "+rank_method_code, ["Unexpected error: %s<br />" % (e,)]), voutput)
afterfind = time.time() - starttime
if result[0] and result[1]: #split into two lists for search_engine
results_similar_recIDs = [x[0] for x in result[0]]
results_similar_relevances = [x[1] for x in result[0]]
result = (results_similar_recIDs, results_similar_relevances, result[1], result[2], "%s%s" % (configcreated, result[3]))
aftermap = time.time() - starttime
else:
result = (None, None, result[1], result[2], result[3])
#add stuff from here into voutput from result
tmp = voutput+result[4]
if verbose > 0:
tmp += "<br/>Elapsed time after finding: %s\nElapsed after mapping: %s" % (afterfind, aftermap)
result = (result[0], result[1], result[2], result[3], tmp)
#dbg = string.join(map(str,methods[rank_method_code].items()))
#result = (None, "", adderrorbox("Debug ",rank_method_code+" "+dbg),"",voutput)
return result
