def browserfindlinenumberfromperseus(citationlist: list, workobject: dbOpus,
resultmessage: str, dbcursor) -> tuple:
"""
here comes the fun part: alien format; inconsistent citation style; incorrect data...
sample url:
/browse/perseus/gr0016w001/7:130
is
Herodotus, Historiae, Book 7, section 130
:param citationlist:
:param workobject:
:param resultmessage:
:param dbcursor:
:return:
"""
try:
# dict does not always agree with our ids...
# do an imperfect test for this by inviting the exception
# you can still get a valid but wrong work, of course,
# but if you ask for w001 and only w003 exists, this is supposed to take care of that
returnfirstorlastlinenumber(workobject.universalid, dbcursor)
except:
# dict did not agree with our ids...: euripides, esp
# what follows is a 'hope for the best' approach
workid = perseusidmismatch(workobject.universalid, dbcursor)
workobject = workdict[workid]
# life=cal. or section=32
# this has already been nuked?
needscleaning = [True for c in citationlist if len(c.split('=')) > 1]
if True in needscleaning:
citationlist = perseusdelabeler(citationlist, workobject)
# another problem 'J. ' in sallust <bibl id="perseus/lt0631w001/J. 79:3" default="NO" valid="yes"><author>Sall.</author> J. 79, 3</bibl>
# 'lt0631w002/79:3' is what you need to send to finddblinefromincompletelocus()
# note that the work number is wrong, so the next is only a partial fix and valid only if wNNN has been set right
if ' ' in citationlist[-1]:
citationlist[-1] = citationlist[-1].split(' ')[-1]
p = finddblinefromincompletelocus(workobject, citationlist, dbcursor)
resultmessage = p['code']
thelinenumber = p['line']
return thelinenumber, resultmessage
def textsegmentfindstartandstop(authorobject, workobject, passageaslist,
cursor) -> dict:
"""
find the first and last lines of a work segment
:return:
"""
p = tuple(passageaslist)
lookforline = finddblinefromincompletelocus(workobject, p, cursor)
# assuming that lookforline['code'] == 'success'
# lookforline['code'] is (allegedly) only relevant to the Perseus lookup problem where a bad locus can be sent
foundline = lookforline['line']
line = grabonelinefromwork(authorobject.universalid, foundline, cursor)
lo = dblineintolineobject(line)
# let's say you looked for 'book 2' of something that has 'book, chapter, line'
# that means that you want everything that has the same level2 value as the lineobject
# build a where clause
passageaslist.reverse()
atloc = '|'.join(passageaslist)
selection = '{uid}_AT_{line}'.format(uid=workobject.universalid,
line=atloc)
w = atsignwhereclauses(selection, '=',
{authorobject.universalid: authorobject})
d = [workobject.universalid]
qw = str()
for i in range(0, len(w)):
qw += 'AND (' + w[i][0] + ') '
d.append(w[i][1])
query = 'SELECT index FROM {au} WHERE wkuniversalid=%s {whr} ORDER BY index DESC LIMIT 1'.format(
au=authorobject.universalid, whr=qw)
data = tuple(d)
cursor.execute(query, data)
found = cursor.fetchone()
startandstop = dict()
startandstop['startline'] = lo.index
startandstop['endline'] = found[0]
return startandstop
def browserfindlinenumberfromlocus(citationlist: list, workobject: dbOpus,
resultmessage: str, dbcursor) -> tuple:
"""
you were sent here by the citation builder autofill boxes
note that
browse/locus/gr0016w001/3|11|5
is
Herodotus, Historiae, Book 3, section 11, line 5
and
locus/lt1056w001/3|5|_0
is
Vitruvius, De Architectura, book 3, chapter 5, section 1, line 1
unfortunately you might need to find '300,19' as in 'Democritus, Fragmenta: Fragment 300,19, line 4'
'-' is used for something like Ar., Nubes 1510-1511
'_' is used for '_0' (which really means 'first available line at this level')
( ) and / should be converted to equivalents in the builder: they do us no good here
see dbswapoutbadcharsfromciations() in HipparchiaBuilder
see also citationcharacterset() in InputParsingObject()
:param citationlist:
:param workobject:
:param resultmessage:
:param dbcursor:
:return:
"""
ct = tuple(citationlist)
if len(ct) == workobject.availablelevels:
thelinenumber = finddblinefromlocus(workobject, ct, dbcursor)
else:
elements = finddblinefromincompletelocus(workobject, citationlist,
dbcursor)
resultmessage = elements['code']
thelinenumber = elements['line']
return thelinenumber, resultmessage
def textmaker(author: str,
work=None,
passage=None,
endpoint=None,
citationdelimiter='|') -> JSON_STR:
"""
build a text suitable for display
"GET /textof/lt0474/024/20/30"
:return:
"""
probeforsessionvariables()
dbconnection = ConnectionObject('autocommit')
dbcursor = dbconnection.cursor()
linesevery = hipparchia.config['SHOWLINENUMBERSEVERY']
po = TextmakerInputParsingObject(author, work, passage, endpoint,
citationdelimiter)
ao = po.authorobject
wo = po.workobject
segmenttext = str()
# consolewarning('po.passageaslist: {p}'.format(p=po.passageaslist))
if ao and wo:
# we have both an author and a work, maybe we also have a subset of the work
if endpoint:
firstlinenumber = finddblinefromincompletelocus(
wo, po.passageaslist, dbcursor)
lastlinenumber = finddblinefromincompletelocus(wo,
po.endpointlist,
dbcursor,
findlastline=True)
if firstlinenumber['code'] == 'success' and lastlinenumber[
'code'] == 'success':
startline = firstlinenumber['line']
endline = lastlinenumber['line']
startlnobj = dblineintolineobject(
grabonelinefromwork(ao.universalid, startline, dbcursor))
stoplnobj = dblineintolineobject(
grabonelinefromwork(ao.universalid, endline, dbcursor))
else:
msg = '"buildtexttospan/" could not find first and last: {a}w{b} - {c} TO {d}'
consolewarning(
msg.format(a=author, b=work, c=passage, d=endpoint))
startlnobj = makeablankline(work, 0)
stoplnobj = makeablankline(work, 1)
startline = 0
endline = 1
segmenttext = 'from {a} to {b}'.format(a=startlnobj.shortlocus(),
b=stoplnobj.shortlocus())
elif not po.passageaslist:
# whole work
startline = wo.starts
endline = wo.ends
else:
startandstop = textsegmentfindstartandstop(ao, wo,
po.passageaslist,
dbcursor)
startline = startandstop['startline']
endline = startandstop['endline']
texthtml = buildtext(wo.universalid, startline, endline, linesevery,
dbcursor)
else:
texthtml = str()
if hipparchia.config['INSISTUPONSTANDARDANGLEBRACKETS']:
texthtml = gtltsubstitutes(texthtml)
if not segmenttext:
segmenttext = '.'.join(po.passageaslist)
if not ao or not wo:
ao = makeanemptyauthor('gr0000')
wo = makeanemptywork('gr0000w000')
results = dict()
results['authorname'] = avoidsmallvariants(ao.shortname)
results['title'] = avoidsmallvariants(wo.title)
results['structure'] = avoidsmallvariants(wo.citation())
results['worksegment'] = segmenttext
results['texthtml'] = texthtml
results = json.dumps(results)
dbconnection.connectioncleanup()
return results
def buildindexto(searchid: str,
author: str,
work=None,
passage=None,
endpoint=None,
citationdelimiter='|',
justvocab=False) -> JSON_STR:
"""
build a complete index to a an author, work, or segment of a work
:return:
"""
probeforsessionvariables()
pollid = validatepollid(searchid)
starttime = time.time()
progresspolldict[pollid] = ProgressPoll(pollid)
progresspolldict[pollid].activate()
dbconnection = ConnectionObject('autocommit')
dbcursor = dbconnection.cursor()
po = IndexmakerInputParsingObject(author, work, passage, endpoint,
citationdelimiter)
ao = po.authorobject
wo = po.workobject
psg = po.passageaslist
stop = po.endpointlist
if not work:
wo = makeanemptywork('gr0000w000')
# bool
useheadwords = session['headwordindexing']
allworks = list()
output = list()
cdict = dict()
segmenttext = str()
valid = True
if ao and work and psg and stop:
start = psg
firstlinenumber = finddblinefromincompletelocus(wo, start, dbcursor)
lastlinenumber = finddblinefromincompletelocus(wo,
stop,
dbcursor,
findlastline=True)
if firstlinenumber['code'] == 'success' and lastlinenumber[
'code'] == 'success':
cdict = {
wo.universalid:
(firstlinenumber['line'], lastlinenumber['line'])
}
startln = dblineintolineobject(
grabonelinefromwork(ao.universalid, firstlinenumber['line'],
dbcursor))
stopln = dblineintolineobject(
grabonelinefromwork(ao.universalid, lastlinenumber['line'],
dbcursor))
else:
msg = '"indexspan/" could not find first and last: {a}w{b} - {c} TO {d}'
consolewarning(msg.format(a=author, b=work, c=passage, d=endpoint))
startln = makeablankline(work, 0)
stopln = makeablankline(work, 1)
valid = False
segmenttext = 'from {a} to {b}'.format(a=startln.shortlocus(),
b=stopln.shortlocus())
elif ao and work and psg:
# subsection of a work of an author
progresspolldict[pollid].statusis(
'Preparing a partial index to {t}'.format(t=wo.title))
startandstop = textsegmentfindstartandstop(ao, wo, psg, dbcursor)
startline = startandstop['startline']
endline = startandstop['endline']
cdict = {wo.universalid: (startline, endline)}
elif ao and work:
# one work
progresspolldict[pollid].statusis(
'Preparing an index to {t}'.format(t=wo.title))
startline = wo.starts
endline = wo.ends
cdict = {wo.universalid: (startline, endline)}
elif ao:
# whole author
allworks = [
'{w} ⇒ {t}'.format(w=w.universalid[6:10], t=w.title)
for w in ao.listofworks
]
allworks.sort()
progresspolldict[pollid].statusis(
'Preparing an index to the works of {a}'.format(a=ao.shortname))
for wkid in ao.listworkids():
cdict[wkid] = (workdict[wkid].starts, workdict[wkid].ends)
else:
# we do not have a valid selection
valid = False
output = ['invalid input']
if not stop:
segmenttext = '.'.join(psg)
if valid and justvocab:
dbconnection.connectioncleanup()
del progresspolldict[pollid]
return cdict
if valid:
output = buildindextowork(cdict, progresspolldict[pollid],
useheadwords, dbcursor)
# get ready to send stuff to the page
count = len(output)
try:
locale.setlocale(locale.LC_ALL, 'en_US')
count = locale.format_string('%d', count, grouping=True)
except locale.Error:
count = str(count)
progresspolldict[pollid].statusis('Preparing the index HTML')
indexhtml = wordindextohtmltable(output, useheadwords)
buildtime = time.time() - starttime
buildtime = round(buildtime, 2)
progresspolldict[pollid].deactivate()
if not ao:
ao = makeanemptyauthor('gr0000')
results = dict()
results['authorname'] = avoidsmallvariants(ao.shortname)
results['title'] = avoidsmallvariants(wo.title)
results['structure'] = avoidsmallvariants(wo.citation())
results['worksegment'] = segmenttext
results['elapsed'] = buildtime
results['wordsfound'] = count
results['indexhtml'] = indexhtml
results['keytoworks'] = allworks
results['newjs'] = supplementalindexjs()
results = json.dumps(results)
dbconnection.connectioncleanup()
del progresspolldict[pollid]
return results
def selectionmade(requestargs: MultiDict) -> JSON_STR:
"""
once a choice is made, parse and register it inside session['selections']
then return the human readable version of the same for display on the page
'_AT_' syntax is used to restrict the scope of a search
"GET /selection/make/_?auth=lt0474&work=001&locus=13|4&endpoint= HTTP/1.1"
request.args ImmutableMultiDict([('auth', 'lt0474'), ('work', '001'), ('locus', '13|4'), ('endpoint', '')])
"GET /selection/make/_?auth=lt0474&work=001&locus=10&endpoint=20&raw=t HTTP/1.1"
request.args ImmutableMultiDict([('auth', 'lt0474'), ('work', '001'), ('locus', '10'), ('endpoint', '20'), ('raw', 't')])
"GET /selection/make/_?auth=lt0474&work=001&exclude=t HTTP/1.1"
request.args ImmutableMultiDict([('auth', 'lt0474'), ('work', '001'), ('exclude', 't')])
:return:
"""
probeforsessionvariables()
uid = depunct(requestargs.get('auth', str()))
workid = depunct(requestargs.get('work', str()))
genre = depunct(requestargs.get('genre', str()))
auloc = depunct(requestargs.get('auloc', str()))
rawdataentry = re.sub('[^tf]', str(), requestargs.get('raw', str()))
exclude = re.sub('[^tf]', str(), requestargs.get('exclude', str()))
allowedpunct = '|,.'
locus = depunct(requestargs.get('locus', str()), allowedpunct)
endpoint = depunct(requestargs.get('endpoint', str()), allowedpunct)
allowedpunct = '.-?():'
wkprov = depunct(requestargs.get('wkprov', str()), allowedpunct)
allowedpunct = '.'
wkgenre = depunct(requestargs.get('wkgenre', str()), allowedpunct)
if exclude != 't':
suffix = 'selections'
other = 'exclusions'
else:
suffix = 'exclusions'
other = 'selections'
if rawdataentry == 't':
locus = re.sub(r'\.', '|', locus)
endpoint = re.sub(r'\.', '|', endpoint)
# the selection box might contain stale info if you deselect a corpus while items are still in the box
uid = selectionisactive(uid)
if genre and genre not in returnactivelist(authorgenresdict):
genre = str()
if wkgenre and wkgenre not in returnactivelist(workgenresdict):
wkgenre = str()
if auloc and auloc not in returnactivelist(authorlocationdict):
auloc = str()
if wkprov and wkprov not in returnactivelist(workprovenancedict):
wkprov = str()
# you have validated the input, now do something with it...
if uid and workid and locus and endpoint:
# a span in an author: 3 verrine orations, e.g. [note that the selection is 'greedy': 1start - 3end]
# http://127.0.0.1:5000/makeselection?auth=lt0474&work=005&locus=2|1&endpoint=2|3
# convert this into a 'firstline' through 'lastline' format
emptycursor = None
workobject = None
try:
workobject = workdict['{a}w{b}'.format(a=uid, b=workid)]
except KeyError:
consolewarning('"/selection/make/" sent a bad workuniversalid: {a}w{b}'.format(a=uid, b=workid))
start = locus.split('|')
stop = endpoint.split('|')
start.reverse()
stop.reverse()
if workobject:
firstline = finddblinefromincompletelocus(workobject, start, emptycursor)
lastline = finddblinefromincompletelocus(workobject, stop, emptycursor, findlastline=True)
citationtemplate = '{a}w{b}_FROM_{c}_TO_{d}'
if firstline['code'] == 'success' and lastline['code'] == 'success':
fl = firstline['line']
ll = lastline['line']
loc = citationtemplate.format(a=uid, b=workid, c=fl, d=ll)
# print('span selected:', loc)
# span selected: lt0474w005_FROM_4501_TO_11915
# Cicero, In Verrem: 2.1.t.1
# Cicero, In Verrem: 2.3.228.15
if ll > fl:
session['psg' + suffix].append(loc)
session['psg' + suffix] = tidyuplist(session['psg' + suffix])
else:
msg = '"makeselection/" sent a firstline greater than the lastine value: {a} > {b} [{c}; {d}]'
consolewarning(msg.format(a=fl, b=ll, c=locus, d=endpoint))
rationalizeselections(loc, suffix)
else:
msg = '"makeselection/" could not find first and last: {a}w{b} - {c} TO {d}'
consolewarning(msg.format(a=uid, b=workid, c=locus, d=endpoint))
elif uid and workid and locus:
# a specific passage
session['psg' + suffix].append(uid + 'w' + workid + '_AT_' + locus)
session['psg' + suffix] = tidyuplist(session['psg' + suffix])
rationalizeselections(uid + 'w' + workid + '_AT_' + locus, suffix)
elif uid and workid:
# a specific work
session['wk' + suffix].append(uid + 'w' + workid)
session['wk' + suffix] = tidyuplist(session['wk' + suffix])
rationalizeselections(uid + 'w' + workid, suffix)
elif uid and not workid:
# a specific author
session['au' + suffix].append(uid)
session['au' + suffix] = tidyuplist(session['au' + suffix])
rationalizeselections(uid, suffix)
# if vs elif: allow multiple simultaneous instance
if genre:
# add to the +/- genre list and then subtract from the -/+ list
session['agn' + suffix].append(genre)
session['agn' + suffix] = tidyuplist(session['agn' + suffix])
session['agn' + other] = dropdupes(session['agn' + other], session['agn' + suffix])
if wkgenre:
# add to the +/- genre list and then subtract from the -/+ list
session['wkgn' + suffix].append(wkgenre)
session['wkgn' + suffix] = tidyuplist(session['wkgn' + suffix])
session['wkgn' + other] = dropdupes(session['wkgn' + other], session['wkgn' + suffix])
if auloc:
# add to the +/- locations list and then subtract from the -/+ list
session['aloc' + suffix].append(auloc)
session['aloc' + suffix] = tidyuplist(session['aloc' + suffix])
session['aloc' + other] = dropdupes(session['aloc' + other], session['aloc' + suffix])
if wkprov:
# add to the +/- locations list and then subtract from the -/+ list
session['wloc' + suffix].append(wkprov)
session['wloc' + suffix] = tidyuplist(session['wloc' + suffix])
session['wloc' + other] = dropdupes(session['wloc' + other], session['wloc' + suffix])
# after the update to the session, you need to update the page html to reflect the changes
# print('session["psgselections"]=', session['psgselections'])
# print('session["psgexclusions"]=', session['psgexclusions'])
return getcurrentselections()