def main():
global boiled_clues
args = get_args('create clue index')
outf = open_output()
boiled_clues = load_clues()
biggest_clues = "<li>%d total clues, which boil down to %d distinct clues" % (len(clues()), len(boiled_clues))
bcs = [ (len(v), bc, answers_from(v)) for bc, v in boiled_clues.items() ]
nreused = len([bc for n, bc, _ in bcs if n > 1])
biggest_clues += "<li>%d (%d%%) of these clues are used in more than one puzzle" % (nreused, nreused*100/len(boiled_clues))
cluepages_to_make = set()
# add all boiled clues from all input .xd files
for fn, contents in find_files(*args.inputs, ext='.xd'):
progress(fn)
xd = xdfile.xdfile(contents.decode('utf-8'), fn)
for pos, mainclue, mainanswer in xd.iterclues():
cluepages_to_make.add(boil(mainclue))
# add top 100 most used boiled clues from corpus
biggest_clues += '<h2>Most used clues</h2>'
biggest_clues += '<table class="clues most-used-clues">'
biggest_clues += th("clue", "# uses", "answers used with this clue")
for n, bc, ans in sorted(bcs, reverse=True)[:100]:
cluepages_to_make.add(bc)
biggest_clues += td(mkhref(unboil(bc), bc), n, html_select_options(ans))
biggest_clues += '</table>'
most_ambig = "<h2>Most ambiguous clues</h2>"
most_ambig += '(clues with the largest number of different answers)'
most_ambig += '<table class="clues most-different-answers">'
most_ambig += th("Clue", "answers")
for n, bc, ans in sorted(bcs, reverse=True, key=lambda x: len(set(x[2])))[:100]:
cluepages_to_make.add(bc)
clue = mkhref(unboil(bc), bc)
if 'quip' in bc or 'quote' in bc or 'theme' in bc or 'riddle' in bc:
most_ambig += td(clue, html_select_options(ans), rowclass="theme")
else:
most_ambig += td(clue, html_select_options(ans))
most_ambig += '</table>'
for bc in cluepages_to_make:
contents = mkwww_cluepage(bc)
if contents:
outf.write_html('pub/clue/%s/index.html' % bc, contents, title=bc)
outf.write_html('pub/clue/index.html', biggest_clues + most_ambig, title="Clues")
def mkwww_cluepage(bc):
bcs = boiled_clues[bc]
clue_html = ''
clue_html += '<div>Variants: ' + html_select_options([ ca.clue for ca in bcs ]) + '</div>'
clue_html += '<hr/>'
clue_html += '<div>Answers for this clue: ' + html_select_options([ ca.answer for ca in bcs ]) + '</div>'
clue_html += '<hr/>'
clue_html += pubyear.pubyear_html([ (ca.pubyear()[0], ca.pubyear()[1], 1) for ca in bcs ])
return clue_html
def main():
global all_uses
args = get_args('create word pages and index')
outf = open_output()
all_uses = {}
for ca in clues():
if ca.answer not in all_uses:
all_uses[ca.answer] = []
all_uses[ca.answer].append(ca)
h = '<li>%d different words</li>' % len(all_uses)
h += '<h2>Most used words</h2>'
h += '<table class="clues most-used-words">'
h += th("word", "# uses", "clues used with this answer")
wordpages_to_make = set(args.inputs)
for answer, uses in sorted(all_uses.items(), reverse=True, key=lambda x: len(x[1]))[:100]:
wordpages_to_make.add(answer)
h += td(mkhref(answer.upper(), answer.upper()),
len(uses),
html_select_options(uses, strmaker=lambda ca: ca.clue))
h += '</table>'
for word in wordpages_to_make:
outf.write_html('word/%s/index.html' % word.upper(), mkwww_wordpage(word), title=word)
outf.write_html('word/index.html', h, title="Words")
def mkwww_cluepage(bc):
if bc not in boiled_clues:
return ''
bcs = boiled_clues[bc]
clue_html = ''
clue_html += '<div>Variants: ' + html_select_options([ ca.clue for ca in bcs ]) + '</div>'
clue_html += '<hr/>'
clue_html += '<div>Answers for this clue: ' + html_select_options([ ca.answer for ca in bcs ]) + '</div>'
clue_html += '<hr/>'
# TODO: maybe add pubyear chart back in, using stats.tsv as source data (by day-of-week)
# clue_html += pubyear.pubyear_html([ (ca.pubyear()[0], ca.pubyear()[1], 1) for ca in bcs ])
return clue_html
def main():
global boiled_clues
args = get_args('create clue index')
outf = open_output()
boiled_clues = load_clues()
biggest_clues = "<li>%d total clues, which boil down to %d distinct clues" % (len(clues()), len(boiled_clues))
bcs = [ (len(v), bc, answers_from(v)) for bc, v in boiled_clues.items() ]
nreused = len([bc for n, bc, _ in bcs if n > 1])
biggest_clues += "<li>%d (%d%%) of these clues are used in more than one puzzle" % (nreused, nreused*100/len(boiled_clues))
cluepages_to_make = set()
biggest_clues += '<h2>Most used clues</h2>'
biggest_clues += '<table class="clues most-used-clues">'
biggest_clues += th("clue", "# uses", "answers used with this clue")
for n, bc, ans in sorted(bcs, reverse=True)[:100]:
cluepages_to_make.add(bc)
biggest_clues += td(mkhref(unboil(bc), bc), n, html_select_options(ans))
biggest_clues += '</table>'
most_ambig = "<h2>Most ambiguous clues</h2>"
most_ambig += '(clues with the largest number of different answers)'
most_ambig += '<table class="clues most-different-answers">'
most_ambig += th("Clue", "answers")
for n, bc, ans in sorted(bcs, reverse=True, key=lambda x: len(set(x[2])))[:100]:
cluepages_to_make.add(bc)
clue = mkhref(unboil(bc), bc)
if 'quip' in bc or 'quote' in bc or 'theme' in bc or 'riddle' in bc:
most_ambig += td(clue, html_select_options(ans), rowclass="theme")
else:
most_ambig += td(clue, html_select_options(ans))
most_ambig += '</table>'
for bc in cluepages_to_make:
outf.write_html('pub/clue/%s/index.html' % bc, mkwww_cluepage(bc), title=bc)
outf.write_html('pub/clue/index.html', biggest_clues + most_ambig, title="Clues")
def mkwww_wordpage(answer):
uses = all_uses[answer]
h = ''
# h += pubyear.pubyear_html([ (ca.pubyear()[0], ca.pubyear()[1], 1) for ca in uses ])
h += '<hr/>'
h += '<div>Clued as: ' + html_select_options([ ca.clue for ca in uses ]) + '</div>'
h += '<h2>%d uses</h2>' % len(uses)
h += '<table>'
for ca in sorted(uses, reverse=True, key=lambda ca: ca.date):
try:
md = metadb.xd_puzzle(ca.xdid())
h += td(md.xdid, ca.clue, md.Author, md.Copyright)
except Exception as e:
h += td(ca.xdid, ca.clue, str(e))
if utils.get_args().debug:
raise
h += '</table>'
# h += '<hr/>'
# h += '<div>Mutations: '
# h +='</div>'
return h
def main():
args = utils.get_args('generates .html diffs with deep clues for all puzzles in similar.tsv')
outf = utils.open_output()
similars = utils.parse_tsv('gxd/similar.tsv', 'Similar')
xdids_todo = [ parse_pathname(fn).base for fn in args.inputs ]
if not xdids_todo:
xdids_todo = [ xdid for xdid, matches in metadb.get_similar_grids().items() if matches ]
for mainxdid in xdids_todo:
progress(mainxdid)
mainxd = xdfile.get_xd(mainxdid)
if not mainxd:
continue
matches = metadb.get_similar_grids().get(mainxdid, [])
xddates = {}
xddates[mainxdid] = mainxd.date() # Dict to store XD dates for further sort
html_grids = {}
html_clues = {}
# these are added directly to similar.tsv
nstaleclues = 0
nstaleanswers = 0
ntotalclues = 0
poss_answers = [] # TODO:
pub_uses = {} # [pubid] -> set(ClueAnswer)
dcl_html = ''
deepcl_html = [] # keep deep clues to parse later - per row
for pos, mainclue, mainanswer in mainxd.iterclues():
deepcl_html = [] # Temporary to be replaced late
mainca = ClueAnswer(mainxdid, mainxd.date(), mainanswer, mainclue)
# 'grid position' column
deepcl_html.append('<td class="pos">%s.</td>' % pos)
# find other uses of this clue, and other answers, in a single pass
for clueans in find_clue_variants(mainclue):
if clueans.answer != mainanswer:
poss_answers.append(clueans)
if clueans.answer == mainanswer:
if clueans.pubid in pub_uses:
otherpubs = pub_uses[clueans.pubid]
else:
otherpubs = set() # set of ClueAnswer
pub_uses[clueans.pubid] = otherpubs
otherpubs.add(clueans)
# add 'other uses' to clues_html
stale = False
deepcl_html.append('<td class="other-uses">')
if len(pub_uses) > 0:
sortable_uses = []
for pubid, uses in pub_uses.items():
# show the earliest unboiled clue
for u in sorted(uses, key=lambda x: x.date or ""):
# only show those published earlier
if u.date and u.date <= mainxd.date():
if pubid == mainxdid and u.date == mainxd.date():
pass
else:
stale = True
sortable_uses.append((u.date, u, 1))
deepcl_html.append(html_select([ (clue, nuses) for dt, clue, nuses in sorted(sortable_uses, key=lambda x: x[0], reverse=True) ], top_option=mainclue))
else:
deepcl_html.append('<div class="original">%s</div>' % esc(mainclue))
deepcl_html.append('</td>')
# add 'other answers' to clues_html
deepcl_html.append('<td class="other-answers">')
deepcl_html.append(html_select_options(poss_answers, strmaker=lambda ca: ca.answer, force_top=mainca))
deepcl_html.append('</td>')
# add 'other clues' to clues_html
deepcl_html.append('<td class="other-clues">')
# bclues is all boiled clues for this particular answer: { [bc] -> #uses }
bclues = load_answers().get(mainanswer, [])
stale_answer = False
if bclues:
uses = []
for bc, nuses in bclues.items():
# then find all clues besides this one
clue_usages = [ ca for ca in load_clues().get(bc, []) if ca.answer == mainanswer and ca.date < mainxd.date() ]
if clue_usages:
stale_answer = True
if nuses > 1:
# only use one (the most recent) ClueAnswer per boiled clue
# but use the clue only (no xdid)
ca = sorted(clue_usages, key=lambda ca: ca.date or "z")[-1].clue
else:
ca = sorted(clue_usages, key=lambda ca: ca.date or "z")[-1]
uses.append((ca, nuses))
if uses:
deepcl_html.append(html_select(uses))
deepcl_html.append('</td>') # end 'other-clues'
if stale_answer:
nstaleanswers += 1
if stale:
nstaleclues += 1
ntotalclues += 1
# Quick and dirty - to be replaced
dcl_html += '<tr>' + ' '.join(deepcl_html) + '</tr>'
# Store in list to make further formatting as html table easier
mainxd = xdfile.get_xd(mainxdid)
if mainxd:
html_grids[mainxdid] = grid_diff_html(mainxd)
# Add for main XD
diff_l = []
for pos, mainclue, mainanswer in mainxd.iterclues():
diff_h = mktag('div','fullgrid main') + '%s. ' %pos
diff_h += mainclue
diff_h += mktag('span', tagclass='main', inner=' ~ ' + mainanswer.upper())
diff_l.append(diff_h)
html_clues[mainxdid] = diff_l
# Process for all matches
for xdid in matches:
xd = xdfile.get_xd(xdid)
if not xd:
continue
xddates[xdid] = xd.date()
# output each grid
html_grids[xdid] = grid_diff_html(xd, compare_with=mainxd)
diff_l = []
# output comparison of each set of clues
for pos, clue, answer in xd.iterclues():
diff_h = mktag('div','fullgrid') + '%s. ' %pos
# Sometimes can return clue == None
sm = difflib.SequenceMatcher(lambda x: x == ' ', mainxd.get_clue(pos) or '', clue)
if sm.ratio() < 0.50:
diff_h += clue
else:
# Compare based on op codes
for opcode in sm.get_opcodes():
c, a1, a2, b1, b2 = opcode
if c == 'equal':
diff_h += '<span class="match">%s</span>' % clue[b1:b2]
else:
diff_h += '<span class="diff">%s</span>' % clue[b1:b2]
diff_h += mktag('span', tagclass=(answer == mainxd.get_answer(pos)) and 'match' or 'diff', inner=' ~ ' + answer.upper())
diff_h += mktag('/div')
diff_l.append(diff_h)
html_clues[xdid] = diff_l
# Wrap into table
diff_h = mktag('table') + mktag('tr')
# Sort by date
sortedkeys = sorted(xddates.items(), key=operator.itemgetter(1))
for w, dt in sortedkeys:
# Wrap into table
diff_h += mktag('td') + html_grids[w] + mktag('/td')
diff_h += mktag('/tr')
for i, clue in enumerate(html_clues[sortedkeys[0][0]]):
diff_h += mktag('tr')
for w, dt in sortedkeys:
if i < len(html_clues[w]):
diff_h += mktag('td') + html_clues[w][i] + mktag('/td')
diff_h += mktag('/tr')
# Process deepclues
diff_h += mktag('table') + dcl_html + mktag('/table')
diff_h += mktag('/table')
outf.write_html('pub/deep/%s/index.html' % mainxdid, diff_h,
title='Deep clue comparison for ' + mainxdid)
def main():
args = utils.get_args(
'generates .html diffs with deep clues for all puzzles in similar.tsv')
outf = utils.open_output()
similars = utils.parse_tsv('gxd/similar.tsv', 'Similar')
xds_todo = []
for fn, contents in find_files(*args.inputs, ext='.xd'):
xd = xdfile.xdfile(contents.decode('utf-8'), fn)
xds_todo.append(xd)
for mainxd in xds_todo:
mainxdid = mainxd.xdid()
progress(mainxdid)
matches = metadb.xd_similar(mainxdid)
xddates = {}
xddates[mainxdid] = mainxd.date(
) # Dict to store XD dates for further sort
html_grids = {}
# these are added directly to similar.tsv
nstaleclues = 0
nstaleanswers = 0
ntotalclues = 0
dcl_html = '<tr>'
dcl_html += '<th></th>'
dcl_html += '<th>Clue</th>'
dcl_html += '<th>ANSWERs</th>'
dcl_html += '<th>Alt. clue possibilities</th>'
dcl_html += '</tr>'
deepcl_html = [] # keep deep clues to parse later - per row
for pos, mainclue, mainanswer in mainxd.iterclues():
if not pos:
continue
poss_answers = [] # TODO:
pub_uses = {} # [pubid] -> set(ClueAnswer)
deepcl_html = [] # Temporary to be replaced late
mainca = ClueAnswer(mainxdid, mainxd.date(), mainanswer, mainclue)
# 'grid position' column
deepcl_html.append('<td class="pos">%s.</td>' % pos)
# find other uses of this clue, and other answers, in a single pass
for clueans in find_clue_variants(mainclue):
if clueans.answer != mainanswer:
poss_answers.append(clueans)
if clueans.answer == mainanswer:
if clueans.pubid in pub_uses:
otherpubs = pub_uses[clueans.pubid]
else:
otherpubs = set() # set of ClueAnswer
pub_uses[clueans.pubid] = otherpubs
otherpubs.add(clueans)
# add 'other uses' to clues_html
deepcl_html.append('<td class="other-uses">')
prev = prev_uses(pub_uses, mainxd, mainclue)
if prev:
deepcl_html.append('<a href="/pub/clue/%s">%s [x%s]</a>' %
(boil(mainclue), mainclue, len(prev)))
nstaleclues += 1
else:
deepcl_html.append(mainclue)
deepcl_html.append('</td>')
# add 'other answers' to clues_html
deepcl_html.append('<td class="other-answers">')
deepcl_html.append(
html_select_options(poss_answers,
strmaker=lambda ca: ca.answer,
force_top=mainca,
add_total=False))
deepcl_html.append('</td>')
# add 'other clues' to clues_html
deepcl_html.append('<td class="other-clues">')
other_clues = html_other_clues(mainanswer, mainclue, mainxd)
if other_clues:
deepcl_html.append(other_clues)
nstaleanswers += 1
deepcl_html.append('</td>') # end 'other-clues'
ntotalclues += 1
# Quick and dirty - to be replaced
dcl_html += '<tr>' + ' '.join(deepcl_html) + '</tr>'
# Process deepclues
diff_h = '<div class="main-container">'
diff_h += grid_to_html(mainxd)
diff_h += mktag('table', 'deepclues') + dcl_html + mktag('/table')
diff_h += '</div>'
info('writing deepclues for %s' % mainxdid)
outf.write_html('pub/deep/%s/index.html' % mainxdid,
diff_h,
title='Deep clue analysis for ' + mainxdid)
