def log_lines(self, start_page, end_page):
"Returns the log lines and the previous/next timestamps, with images mixed in."
if end_page > (start_page + 5):
end_page = start_page + 5
# Collect the log lines
log_lines = []
done_closest = False
for page in range(start_page, end_page + 1):
log_lines += list(self.main_transcript_query().page(page))
for log_line in log_lines:
log_line.images = list(log_line.images())
log_line.lines = [(s, linkify(t, self.request)) for s, t in log_line.lines]
# If this is the first after the start time, add an anchor later
if log_line.timestamp > timestamp_to_seconds(self.kwargs.get("start", "00:00:00:00")) and not done_closest:
log_line.closest = True
done_closest = True
# Find all media that falls inside this same range, and add it onto the preceding line.
for image_line in self.media_transcript_query().range(log_lines[0].timestamp, log_lines[-1].timestamp):
# Find the line just before the images
last_line = None
for log_line in log_lines:
if log_line.timestamp > image_line.timestamp:
break
last_line = log_line
# Add the images to it
last_line.images += image_line.images()
# Find the previous log line from this, and then the beginning of its page
try:
previous_timestamp = self.main_transcript_query().page(start_page - 1).first().timestamp
except ValueError:
previous_timestamp = None
# Find the next log line and its timestamp
next_timestamp = log_lines[-1].next_timestamp()
# Return
return log_lines, previous_timestamp, next_timestamp, 0, None
def selection_url(start_seconds, end_seconds=None):
if end_seconds is None:
url = reverse("view_range", kwargs={"start": mission_time(start_seconds)})
else:
url = reverse("view_range", kwargs={"start": mission_time(start_seconds), "end": mission_time(end_seconds)})
if isinstance(start_seconds, basestring):
start_seconds = timestamp_to_seconds(start_seconds)
return '%s#log-line-%i' % ( url, start_seconds )
def selection_url(start_seconds, end_seconds=None):
if end_seconds is None:
url = reverse("view_range", kwargs={"start": mission_time(start_seconds)})
else:
url = reverse("view_range", kwargs={"start": mission_time(start_seconds), "end": mission_time(end_seconds)})
if isinstance(start_seconds, basestring):
start_seconds = timestamp_to_seconds(start_seconds)
return '%s#log-line-%i' % ( url, start_seconds )
def get_quote(self):
quote_timestamp = self.request.redis_conn.srandmember(
"mission:%s:homepage_quotes" % self.request.mission.name,
)
if quote_timestamp:
return LogLine(
self.request.redis_conn,
self.request.mission.main_transcript,
int(timestamp_to_seconds(quote_timestamp)),
)
def get_context_data(self, start=None, end=None):
if end is None:
end = start
# Get the content
log_lines, previous_timestamp, next_timestamp, max_highlight_index, first_highlighted_line = self.log_lines(
self.page_number(start),
self.page_number(end),
)
act = log_lines[0].act()
act_id = log_lines[0].act().number
acts = list(self.act_query().items())
previous_act = None
next_act = None
if act_id > 0:
previous_act = acts[act_id-1]
if act_id < len(acts) - 1:
next_act = acts[act_id+1]
for log_line in log_lines:
if log_line.transcript_page:
original_transcript_page = log_line.transcript_page
break
else:
original_transcript_page = None
if start:
permalink_fragment = '#log-line-%s' % timestamp_to_seconds(start)
else:
permalink_fragment = '#log-line-%s' % log_lines[0].timestamp
return {
'start' : start,
'log_lines': log_lines,
'next_timestamp': next_timestamp,
'previous_timestamp': previous_timestamp,
'acts': acts,
'act': act_id+1,
'current_act': act,
'previous_act': previous_act,
'next_act': next_act,
'max_highlight_index': max_highlight_index,
'first_highlighted_line': first_highlighted_line,
'original_transcript_page': original_transcript_page,
'permalink': 'http://%s%s%s' % (
self.request.META['HTTP_HOST'],
self.request.path,
permalink_fragment,
)
}
def render_to_response(self, context):
# Ensure that the request is always redirected to:
# - The first page (timestampless)
# - The timestamp for the start of an act
# - The timestamp for the start of an in-act page
# If the timestamp is already one of these, render as normal
requested_start = None
if context['start']:
requested_start = timestamp_to_seconds( context['start'] )
current_act = context['current_act']
first_log_line = context['log_lines'][0]
prior_log_line = first_log_line.previous()
# NOTE: is_act_first_page will be false for first act:
# that's handled by is_first_page
is_first_page = not prior_log_line
is_act_first_page = False
if prior_log_line:
is_act_first_page = prior_log_line.timestamp < current_act.start \
<= first_log_line.timestamp
page_start_url = None
# If we're on the first page, but have a timestamp,
# redirect to the bare page URL
if requested_start and is_first_page:
if context['transcript_name'] != context['mission_main_transcript']:
# Split transcript name from [mission]/[transcript]
transcript = context['transcript_name'].split('/')[1]
page_start_url = reverse("view_page", kwargs={"transcript": transcript})
else:
page_start_url = reverse("view_page")
# If we're on the first page of an act,
# but not on the act-start timestamp, redirect to that
elif is_act_first_page \
and requested_start != current_act.start:
page_start_url = timestamp_to_url( context, current_act.start )
# If we're on any other page and the timestamp doesn't match
# the timestamp of the first item, redirect to that item's timestamp
elif requested_start and not is_act_first_page \
and requested_start != first_log_line.timestamp:
page_start_url = timestamp_to_url(
context,
first_log_line.timestamp
)
# Redirect to the URL we found
if page_start_url:
if self.request.GET:
page_start_url += '?%s' % self.request.GET.urlencode()
return HttpResponseRedirect( page_start_url )
return super( PageView, self ).render_to_response( context )
def render_to_response(self, context):
# Ensure that the request is always redirected to:
# - The first page (timestampless)
# - The timestamp for the start of an act
# - The timestamp for the start of an in-act page
# If the timestamp is already one of these, render as normal
requested_start = None
if context['start']:
requested_start = timestamp_to_seconds(context['start'])
current_act = context['current_act']
first_log_line = context['log_lines'][0]
prior_log_line = first_log_line.previous()
# NOTE: is_act_first_page will be false for first act:
# that's handled by is_first_page
is_first_page = not prior_log_line
is_act_first_page = False
if prior_log_line:
is_act_first_page = prior_log_line.timestamp < current_act.start \
<= first_log_line.timestamp
page_start_url = None
# If we're on the first page, but have a timestamp,
# redirect to the bare page URL
if requested_start and is_first_page:
if context['transcript_name'] != context['mission_main_transcript']:
# Split transcript name from [mission]/[transcript]
transcript = context['transcript_name'].split('/')[1]
page_start_url = reverse("view_page",
kwargs={"transcript": transcript})
else:
page_start_url = reverse("view_page")
# If we're on the first page of an act,
# but not on the act-start timestamp, redirect to that
elif is_act_first_page \
and requested_start != current_act.start:
page_start_url = timestamp_to_url(context, current_act.start)
# If we're on any other page and the timestamp doesn't match
# the timestamp of the first item, redirect to that item's timestamp
elif requested_start and not is_act_first_page \
and requested_start != first_log_line.timestamp:
page_start_url = timestamp_to_url(context,
first_log_line.timestamp)
# Redirect to the URL we found
if page_start_url:
if self.request.GET:
page_start_url += '?%s' % self.request.GET.urlencode()
return HttpResponseRedirect(page_start_url)
return super(PageView, self).render_to_response(context)
def stream(request, start):
bitrate = 48000
offset = 555
file_path = os.path.join(settings.SITE_ROOT, '../missions/mr3/audio/ATG.mp3')
start = timestamp_to_seconds(start)
offset = int((start + offset) * bitrate / 8)
file_size = os.path.getsize(file_path)
if offset > file_size or offset < 0:
raise Http404
fh = open(file_path, 'r')
fh.seek(offset)
response = HttpResponse(ProgressiveFileWrapper(fh, int(bitrate / 8), 1))
response['Content-Type'] = 'audio/mpeg'
return response
def stream(request, start):
bitrate = 48000
offset = 555
file_path = os.path.join(settings.SITE_ROOT,
'../missions/mr3/audio/ATG.mp3')
start = timestamp_to_seconds(start)
offset = int((start + offset) * bitrate / 8)
file_size = os.path.getsize(file_path)
if offset > file_size or offset < 0:
raise Http404
fh = open(file_path, 'r')
fh.seek(offset)
response = HttpResponse(ProgressiveFileWrapper(fh, int(bitrate / 8), 1))
response['Content-Type'] = 'audio/mpeg'
return response
def get_context_data(self, start=None, end=None):
if end is None:
end = start
# Get the content
log_lines, previous_timestamp, next_timestamp, max_highlight_index, first_highlighted_line = self.log_lines(
self.page_number(start), self.page_number(end)
)
act = log_lines[0].act()
act_id = log_lines[0].act().number
acts = list(self.act_query().items())
previous_act = None
next_act = None
if act_id > 0:
previous_act = acts[act_id - 1]
if act_id < len(acts) - 1:
next_act = acts[act_id + 1]
for log_line in log_lines:
if log_line.transcript_page:
original_transcript_page = log_line.transcript_page
break
else:
original_transcript_page = None
if start:
permalink_fragment = "#log-line-%s" % timestamp_to_seconds(start)
else:
permalink_fragment = "#log-line-%s" % log_lines[0].timestamp
return {
"start": start,
"log_lines": log_lines,
"next_timestamp": next_timestamp,
"previous_timestamp": previous_timestamp,
"acts": acts,
"act": act_id + 1,
"current_act": act,
"previous_act": previous_act,
"next_act": next_act,
"max_highlight_index": max_highlight_index,
"first_highlighted_line": first_highlighted_line,
"original_transcript_page": original_transcript_page,
"permalink": "http://%s%s%s" % (self.request.META["HTTP_HOST"], self.request.path, permalink_fragment),
}
def get_quote(self):
quote_timestamp = self.request.redis_conn.srandmember(
"mission:%s:homepage_quotes" % self.request.mission.name,
)
if quote_timestamp:
if '/' in quote_timestamp:
transcript, timestamp = quote_timestamp.rsplit('/', 1)
transcript = "%s/%s" % (self.request.mission.name, transcript)
else:
transcript = self.request.mission.main_transcript
timestamp = quote_timestamp
return LogLine(
self.request.redis_conn,
transcript,
int(timestamp_to_seconds(timestamp)),
)
def selection_url_in_transcript(context, start_seconds, transcript, end_seconds=None):
url_args = {
"start": mission_time(start_seconds)
}
if end_seconds:
url_args["end"] = mission_time(end_seconds)
if transcript and transcript != context['request'].mission.main_transcript:
# Split transcript name from [mission]/[transcript]
url_args["transcript"] = transcript.split('/')[1]
# Render the URL
url = reverse("view_range", kwargs=url_args)
if isinstance(start_seconds, basestring):
start_seconds = timestamp_to_seconds(start_seconds)
return '%s#log-line-%i' % ( url, start_seconds )
def selection_url_in_transcript(context,
start_seconds,
transcript,
end_seconds=None):
url_args = {"start": mission_time(start_seconds)}
if end_seconds:
url_args["end"] = mission_time(end_seconds)
if transcript and transcript != context['request'].mission.main_transcript:
# Split transcript name from [mission]/[transcript]
url_args["transcript"] = transcript.split('/')[1]
# Render the URL
url = reverse("view_range", kwargs=url_args)
if isinstance(start_seconds, basestring):
start_seconds = timestamp_to_seconds(start_seconds)
return '%s#log-line-%i' % (url, start_seconds)
def render_to_response(self, context):
# Ensure that the request is always redirected to:
# - The first page (timestampless)
# - The timestamp for the start of an act
# - The timestamp for the start of an in-act page
# If the timestamp is already one of these, render as normal
requested_start = None
if context["start"]:
requested_start = timestamp_to_seconds(context["start"])
current_act = context["current_act"]
first_log_line = context["log_lines"][0]
prior_log_line = first_log_line.previous()
# NOTE: is_act_first_page will be false for first act:
# that's handled by is_first_page
is_first_page = not prior_log_line
is_act_first_page = False
if prior_log_line:
is_act_first_page = prior_log_line.timestamp < current_act.start <= first_log_line.timestamp
page_start_url = None
# If we're on the first page, but have a timestamp,
# redirect to the bare page URL
if requested_start and is_first_page:
page_start_url = reverse("view_page")
# If we're on the first page of an act,
# but not on the act-start timestamp, redirect to that
elif is_act_first_page and requested_start != current_act.start:
page_start_url = timestamp_to_url(current_act.start)
# If we're on any other page and the timestamp doesn't match
# the timestamp of the first item, redirect to that item's timestamp
elif requested_start and not is_act_first_page and requested_start != first_log_line.timestamp:
page_start_url = timestamp_to_url(first_log_line.timestamp)
# Redirect to the URL we found
if page_start_url:
if self.request.GET:
page_start_url += "?%s" % self.request.GET.urlencode()
return HttpResponseRedirect(page_start_url)
return super(PageView, self).render_to_response(context)
def log_lines(self, start_page, end_page):
"Returns the log lines and the previous/next timestamps, with images mixed in."
if end_page > (start_page + 5):
end_page = start_page + 5
# Collect the log lines
log_lines = []
done_closest = False
for page in range(start_page, end_page + 1):
log_lines += list(self.main_transcript_query().page(page))
for log_line in log_lines:
log_line.images = list(log_line.images())
log_line.lines = [(s, linkify(t, self.request))
for s, t in log_line.lines]
# If this is the first after the start time, add an anchor later
if log_line.timestamp > timestamp_to_seconds(
self.kwargs.get('start',
"00:00:00:00")) and not done_closest:
log_line.closest = True
done_closest = True
# Find all media that falls inside this same range, and add it onto the preceding line.
for image_line in self.media_transcript_query().range(
log_lines[0].timestamp, log_lines[-1].timestamp):
# Find the line just before the images
last_line = None
for log_line in log_lines:
if log_line.timestamp > image_line.timestamp:
break
last_line = log_line
# Add the images to it
last_line.images += image_line.images()
# Find the previous log line from this, and then the beginning of its page
try:
previous_timestamp = self.main_transcript_query().page(
start_page - 1).first().timestamp
except ValueError:
previous_timestamp = None
# Find the next log line and its timestamp
next_timestamp = log_lines[-1].next_timestamp()
# Return
return log_lines, previous_timestamp, next_timestamp, 0, None
def get_context_data(self, start=None, end=None, transcript=None):
if end is None:
end = start
# Get the content
log_lines, previous_timestamp, next_timestamp, max_highlight_index, first_highlighted_line = self.log_lines(
self.page_number(start),
self.page_number(end),
)
act = log_lines[0].act()
act_id = log_lines[0].act().number
acts = list(self.act_query().items())
previous_act = None
next_act = None
if act_id > 0:
previous_act = acts[act_id - 1]
if act_id < len(acts) - 1:
next_act = acts[act_id + 1]
for log_line in log_lines:
if log_line.transcript_page:
original_transcript_page = log_line.transcript_page
break
else:
original_transcript_page = None
if start:
permalink_fragment = '#log-line-%s' % timestamp_to_seconds(start)
else:
permalink_fragment = '#log-line-%s' % log_lines[0].timestamp
return {
# HACK: Force request into context. Not sure why it's not here.
'request':
self.request,
'mission_name':
self.request.mission.name,
'mission_main_transcript':
self.request.mission.main_transcript,
'transcript_name':
self.get_transcript_name(),
'transcript_short_name':
self.get_transcript_name().split('/')[1],
'start':
start,
'log_lines':
log_lines,
'next_timestamp':
next_timestamp,
'previous_timestamp':
previous_timestamp,
'acts':
acts,
'act':
act_id + 1,
'current_act':
act,
'previous_act':
previous_act,
'next_act':
next_act,
'max_highlight_index':
max_highlight_index,
'first_highlighted_line':
first_highlighted_line,
'original_transcript_page':
original_transcript_page,
'permalink':
'http://%s%s%s' % (
self.request.META['HTTP_HOST'],
self.request.path,
permalink_fragment,
),
'other_transcripts':
self.other_transcripts(
log_lines[0].timestamp,
log_lines[-1].timestamp,
),
}
def get_chunks(self, offset=0):
"""
Reads the log lines from the file in order and yields them.
"""
current_chunk = None
reuse_line = None
lines = iter(self.get_lines(offset))
while lines or reuse_line:
# If there's a line to reuse, use that, else read a new
# line from the file.
if reuse_line:
line = reuse_line
reuse_line = None
else:
try:
line = lines.next()
except StopIteration:
break
offset += len(line)
line = line.decode("utf8")
# If it's a comment or empty line, ignore it.
if not line.strip() or line.strip()[0] == "#":
continue
# If it's a timestamp header, make a new chunk object.
elif line[0] == "[":
# Read the timestamp
try:
timestamp = int(line[1:].split("]")[0])
except ValueError:
timestamp = timestamp_to_seconds(line[1:].split("]")[0])
if current_chunk:
yield current_chunk
# Start a new log line item
current_chunk = {
"timestamp": timestamp,
"lines": [],
"meta": {},
"offset": offset - len(line),
}
# If it's metadata, read the entire thing.
elif line[0] == "_":
# Meta item
name, blob = line.split(":", 1)
while True:
try:
line = lines.next()
except StopIteration:
break
offset += len(line)
line = line.decode("utf8")
if not line.strip() or line.strip()[0] == "#":
continue
if line[0] in string.whitespace:
blob += line
else:
reuse_line = line
break
# Parse the blob
blob = blob.strip()
if blob:
try:
data = json.loads(blob)
except ValueError:
try:
data = json.loads('"%s"' % blob)
except ValueError:
print "Error: Invalid json at timestamp %s, key %s" % \
(seconds_to_timestamp(timestamp), name)
continue
current_chunk['meta'][name.strip()] = data
# If it's a continuation, append to the current line
elif line[0] in string.whitespace:
# Continuation line
if not current_chunk:
print "Error: Continuation line before first timestamp header. Line: %s" % \
(line)
elif not current_chunk['lines']:
print "Error: Continuation line before first speaker name. Timestamp %s" % \
(seconds_to_timestamp(timestamp))
else:
current_chunk['lines'][-1]['text'] += " " + line.strip()
# If it's a new line, start a new line. Shock.
else:
# New line of speech
try:
speaker, text = line.split(":", 1)
except ValueError:
print "Error: First speaker line not in Name: Text format: %s. Timestamp %s" % \
(line, seconds_to_timestamp(timestamp))
else:
line = {
"speaker": speaker.strip(),
"text": text.strip(),
}
current_chunk['lines'].append(line)
# Finally, if there's one last chunk, yield it.
if current_chunk:
yield current_chunk
def get_chunks(self, offset=0):
"""
Reads the log lines from the file in order and yields them.
"""
current_chunk = None
reuse_line = None
lines = iter(self.get_lines(offset))
while lines or reuse_line:
# If there's a line to reuse, use that, else read a new
# line from the file.
if reuse_line:
line = reuse_line
reuse_line = None
else:
try:
line = lines.next()
except StopIteration:
break
offset += len(line)
line = line.decode("utf8")
# If it's a comment or empty line, ignore it.
if not line.strip() or line.strip()[0] == "#":
continue
# If it's a timestamp header, make a new chunk object.
elif line[0] == "[":
# Read the timestamp
try:
timestamp = int(line[1:].split("]")[0])
except ValueError:
try:
timestamp = timestamp_to_seconds(line[1:].split("]")[0])
except ValueError:
print "Error: invalid timestamp %s" % (line[1:], )
raise
if current_chunk:
yield current_chunk
# Start a new log line item
current_chunk = {
"timestamp": timestamp,
"lines": [],
"meta": {},
"offset": offset - len(line),
}
# If it's metadata, read the entire thing.
elif line[0] == "_":
# Meta item
name, blob = line.split(":", 1)
while True:
try:
line = lines.next()
except StopIteration:
break
offset += len(line)
line = line.decode("utf8")
if not line.strip() or line.strip()[0] == "#":
continue
if line[0] in string.whitespace:
blob += line
else:
reuse_line = line
break
# Parse the blob
blob = blob.strip()
if blob:
try:
data = json.loads(blob)
except ValueError:
try:
data = json.loads('"%s"' % blob)
except ValueError:
print "Error: Invalid json at timestamp %s, key %s" % \
(seconds_to_timestamp(timestamp), name)
continue
current_chunk['meta'][name.strip()] = data
# If it's a continuation, append to the current line
elif line[0] in string.whitespace:
# Continuation line
if not current_chunk:
print "Error: Continuation line before first timestamp header. Line: %s" % \
(line)
elif not current_chunk['lines']:
print "Error: Continuation line before first speaker name."
else:
current_chunk['lines'][-1]['text'] += " " + line.strip()
# If it's a new line, start a new line. Shock.
else:
# New line of speech
try:
speaker, text = line.split(":", 1)
except ValueError:
print "Error: First speaker line not in Name: Text format: %s." % (line,)
else:
line = {
"speaker": speaker.strip(),
"text": text.strip(),
}
current_chunk['lines'].append(line)
# Finally, if there's one last chunk, yield it.
if current_chunk:
yield current_chunk
def get_context_data(self, start=None, end=None, transcript=None):
if end is None:
end = start
# Get the content
log_lines, previous_timestamp, next_timestamp, max_highlight_index, first_highlighted_line = self.log_lines(
self.page_number(start),
self.page_number(end),
)
act = log_lines[0].act()
act_id = log_lines[0].act().number
acts = list(self.act_query().items())
previous_act = None
next_act = None
if act_id > 0:
previous_act = acts[act_id-1]
if act_id < len(acts) - 1:
next_act = acts[act_id+1]
for log_line in log_lines:
if log_line.transcript_page:
original_transcript_page = log_line.transcript_page
break
else:
original_transcript_page = None
if start:
permalink_fragment = '#log-line-%s' % timestamp_to_seconds(start)
else:
permalink_fragment = '#log-line-%s' % log_lines[0].timestamp
return {
# HACK: Force request into context. Not sure why it's not here.
'request': self.request,
'mission_name': self.request.mission.name,
'mission_main_transcript': self.request.mission.main_transcript,
'transcript_name': self.get_transcript_name(),
'transcript_short_name': self.get_transcript_name().split('/')[1],
'start' : start,
'log_lines': log_lines,
'next_timestamp': next_timestamp,
'previous_timestamp': previous_timestamp,
'acts': acts,
'act': act_id+1,
'current_act': act,
'previous_act': previous_act,
'next_act': next_act,
'max_highlight_index': max_highlight_index,
'first_highlighted_line': first_highlighted_line,
'original_transcript_page': original_transcript_page,
'permalink': 'http://%s%s%s' % (
self.request.META['HTTP_HOST'],
self.request.path,
permalink_fragment,
),
'other_transcripts': self.other_transcripts(
log_lines[0].timestamp,
log_lines[-1].timestamp,
),
}
def get_context_data(self):
# Get the query text
q = self.request.GET.get('q', '')
# Get the offset value
try:
offset = int(self.request.GET.get('offset', '0'))
if offset < 0:
offset = 0
except ValueError:
offset = 0
# Is it a special search?
special_value = self.request.redis_conn.get("special_search:%s:%s" % (
self.request.mission.name,
q,
))
if special_value:
self.template_name = "search/special.html"
return {
"q": q,
"text": special_value,
}
# Get the results from Xapian
db = xappy.SearchConnection(
os.path.join(
settings.SITE_ROOT,
'..',
"xappydb",
), )
query = db.query_parse(
q,
default_op=db.OP_OR,
deny=["mission"],
)
query = db.query_filter(
query,
db.query_composite(db.OP_AND, [
db.query_field("mission", self.request.mission.name),
db.query_field("transcript",
self.request.mission.main_transcript),
]))
results = db.search(
query=query,
startrank=offset,
endrank=offset + PAGESIZE,
checkatleast=
-1, # everything (entire xapian db fits in memory, so this should be fine)
sortby="-weight",
)
# Go through the results, building a list of LogLine objects
redis_conn = self.request.redis_conn
log_lines = []
for result in results:
transcript_name, timestamp = result.id.split(":", 1)
log_line = LogLine(redis_conn, transcript_name, int(timestamp))
log_line.speaker = Character(redis_conn,
transcript_name.split('/')[0],
result.data['speaker_identifier'][0])
log_line.title = mark_safe(
result.summarise("text",
maxlen=50,
ellipsis='…',
strict_length=True,
hl=None))
log_line.summary = mark_safe(
result.summarise("text",
maxlen=600,
ellipsis='…',
hl=('<mark>', '</mark>')))
log_lines.append(log_line)
def page_url(offset):
return reverse("search") + '?' + urllib.urlencode(
{
'q': q.encode('utf-8'),
'offset': offset,
})
if offset == 0:
previous_page = False
else:
previous_page = page_url(offset - PAGESIZE)
if offset + PAGESIZE > results.matches_estimated:
next_page = False
else:
next_page = page_url(offset + PAGESIZE)
thispage = offset / PAGESIZE
maxpage = results.matches_estimated / PAGESIZE
pages_to_show = set([0]) | set([thispage - 1, thispage, thispage + 1
]) | set([maxpage])
if 0 == thispage:
pages_to_show.remove(thispage - 1)
if maxpage == thispage:
pages_to_show.remove(thispage + 1)
pages = []
class Page(object):
def __init__(self, number, url, selected=False):
self.number = number
self.url = url
self.selected = selected
pages_in_order = list(pages_to_show)
pages_in_order.sort()
for page in pages_in_order:
if len(pages) > 0 and page != pages[-1].number:
pages.append('...')
pages.append(
Page(page + 1, page_url(page * PAGESIZE), page == thispage))
error_info = self.request.redis_conn.hgetall(
"error_page:%s:%s" % (
self.request.mission.name,
'no_search_results',
), )
if not error_info:
error_info = {}
if error_info.has_key('classic_moment_quote'):
error_quote = LogLine(
self.request.redis_conn, self.request.mission.main_transcript,
timestamp_to_seconds(error_info['classic_moment_quote']))
else:
error_quote = None
return {
'log_lines': log_lines,
'result': results,
'q': q,
'previous_page': previous_page,
'next_page': next_page,
'pages': pages,
'debug': {
'query': query,
},
'error': {
'info': error_info,
'quote': error_quote,
}
}
def get_context_data(self):
# Get the query text
q = self.request.GET.get("q", "")
# Get the offset value
try:
offset = int(self.request.GET.get("offset", "0"))
if offset < 0:
offset = 0
except ValueError:
offset = 0
# Is it a special search?
special_value = self.request.redis_conn.get("special_search:%s:%s" % (self.request.mission.name, q))
if special_value:
self.template_name = "search/special.html"
return {"q": q, "text": special_value}
# Get the results from Xapian
db = xappy.SearchConnection(os.path.join(settings.SITE_ROOT, "..", "xappydb"))
query = db.query_parse(q, default_op=db.OP_OR, deny=["mission"])
query = db.query_filter(
query,
db.query_composite(
db.OP_AND,
[
db.query_field("mission", self.request.mission.name),
db.query_field("transcript", "%s/TEC" % self.request.mission.name),
],
),
)
results = db.search(
query=query,
startrank=offset,
endrank=offset + PAGESIZE,
checkatleast=-1, # everything (entire xapian db fits in memory, so this should be fine)
sortby="-weight",
)
# Go through the results, building a list of LogLine objects
redis_conn = self.request.redis_conn
log_lines = []
for result in results:
transcript_name, timestamp = result.id.split(":", 1)
log_line = LogLine(redis_conn, transcript_name, int(timestamp))
log_line.speaker = Character(
redis_conn, transcript_name.split("/")[0], result.data["speaker_identifier"][0]
)
log_line.title = mark_safe(
result.summarise("text", maxlen=50, ellipsis="…", strict_length=True, hl=None)
)
log_line.summary = mark_safe(
result.summarise("text", maxlen=600, ellipsis="…", hl=("<mark>", "</mark>"))
)
log_lines.append(log_line)
def page_url(offset):
return reverse("search") + "?" + urllib.urlencode({"q": q, "offset": offset})
if offset == 0:
previous_page = False
else:
previous_page = page_url(offset - PAGESIZE)
if offset + PAGESIZE > results.matches_estimated:
next_page = False
else:
next_page = page_url(offset + PAGESIZE)
thispage = offset / PAGESIZE
maxpage = results.matches_estimated / PAGESIZE
pages_to_show = set([0]) | set([thispage - 1, thispage, thispage + 1]) | set([maxpage])
if 0 == thispage:
pages_to_show.remove(thispage - 1)
if maxpage == thispage:
pages_to_show.remove(thispage + 1)
pages = []
class Page(object):
def __init__(self, number, url, selected=False):
self.number = number
self.url = url
self.selected = selected
pages_in_order = list(pages_to_show)
pages_in_order.sort()
for page in pages_in_order:
if len(pages) > 0 and page != pages[-1].number:
pages.append("...")
pages.append(Page(page + 1, page_url(page * PAGESIZE), page == thispage))
error_info = self.request.redis_conn.hgetall(
"error_page:%s:%s" % (self.request.mission.name, "no_search_results")
)
if not error_info:
error_info = {}
if error_info.has_key("classic_moment_quote"):
error_quote = LogLine(
self.request.redis_conn,
self.request.mission.main_transcript,
timestamp_to_seconds(error_info["classic_moment_quote"]),
)
else:
error_quote = None
return {
"log_lines": log_lines,
"result": results,
"q": q,
"previous_page": previous_page,
"next_page": next_page,
"pages": pages,
"debug": {"query": query},
"error": {"info": error_info, "quote": error_quote},
}
def parse_mission_time(self, mission_time):
"Parses a mission timestamp from a URL and converts it to a number of seconds"
# d, h, m, s = [ int(x) for x in mission_time.split(':') ]
# print mission_time
# return s + m*60 + h*3600 + d*86400
return timestamp_to_seconds( mission_time )
def parse_mission_time(self, mission_time):
"Parses a mission timestamp from a URL and converts it to a number of seconds"
# d, h, m, s = [ int(x) for x in mission_time.split(':') ]
# print mission_time
# return s + m*60 + h*3600 + d*86400
return timestamp_to_seconds(mission_time)
def get_context_data(self):
# Get the query text
q = self.request.GET.get('q', '')
# Get the offset value
try:
offset = int(
self.request.GET.get('offset', '0')
)
if offset < 0:
offset = 0
except ValueError:
offset = 0
# Is it a special search?
special_value = self.request.redis_conn.get("special_search:%s:%s" % (
self.request.mission.name,
q,
))
if special_value:
self.template_name = "search/special.html"
return {
"q": q,
"text": special_value,
}
# Get the results from Xapian
db = xappy.SearchConnection(
os.path.join(
settings.SITE_ROOT,
'..',
"xappydb",
),
)
query = db.query_parse(
q,
default_op=db.OP_OR,
deny = [ "mission" ],
)
query=db.query_filter(
query,
db.query_composite(db.OP_AND, [
db.query_field("mission", self.request.mission.name),
db.query_field("transcript", self.request.mission.main_transcript),
])
)
results = db.search(
query=query,
startrank=offset,
endrank=offset+PAGESIZE,
checkatleast=-1, # everything (entire xapian db fits in memory, so this should be fine)
sortby="-weight",
)
# Go through the results, building a list of LogLine objects
redis_conn = self.request.redis_conn
log_lines = []
for result in results:
transcript_name, timestamp = result.id.split(":", 1)
log_line = LogLine(redis_conn, transcript_name, int(timestamp))
log_line.speaker = Character(redis_conn, transcript_name.split('/')[0], result.data['speaker_identifier'][0])
log_line.title = mark_safe(result.summarise("text", maxlen=50, ellipsis='…', strict_length=True, hl=None))
log_line.summary = mark_safe(result.summarise("text", maxlen=600, ellipsis='…', hl=('<mark>', '</mark>')))
log_lines.append(log_line)
def page_url(offset):
return reverse("search") + '?' + urllib.urlencode({
'q': q.encode('utf-8'),
'offset': offset,
})
if offset==0:
previous_page = False
else:
previous_page = page_url(offset - PAGESIZE)
if offset+PAGESIZE > results.matches_estimated:
next_page = False
else:
next_page = page_url(offset + PAGESIZE)
thispage = offset / PAGESIZE
maxpage = results.matches_estimated / PAGESIZE
pages_to_show = set([0]) | set([thispage-1, thispage, thispage+1]) | set([maxpage])
if 0 == thispage:
pages_to_show.remove(thispage-1)
if maxpage == thispage:
pages_to_show.remove(thispage+1)
pages = []
class Page(object):
def __init__(self, number, url, selected=False):
self.number = number
self.url = url
self.selected = selected
pages_in_order = list(pages_to_show)
pages_in_order.sort()
for page in pages_in_order:
if len(pages)>0 and page != pages[-1].number:
pages.append('...')
pages.append(Page(page+1, page_url(page*PAGESIZE), page==thispage))
error_info = self.request.redis_conn.hgetall(
"error_page:%s:%s" % (
self.request.mission.name,
'no_search_results',
),
)
if not error_info:
error_info = {}
if error_info.has_key('classic_moment_quote'):
error_quote = LogLine(
self.request.redis_conn,
self.request.mission.main_transcript,
timestamp_to_seconds(error_info['classic_moment_quote'])
)
else:
error_quote = None
return {
'log_lines': log_lines,
'result': results,
'q': q,
'previous_page': previous_page,
'next_page': next_page,
'pages': pages,
'debug': {
'query': query,
},
'error': {
'info': error_info,
'quote': error_quote,
}
}
def get_context_data(self, start=None, end=None):
if end is None:
end = start
# Get the content
log_lines, previous_timestamp, next_timestamp, max_highlight_index, first_highlighted_line = self.log_lines(
self.page_number(start),
self.page_number(end),
)
act = log_lines[0].act()
act_id = log_lines[0].act().number
acts = list(self.act_query().items())
previous_act = None
next_act = None
if act_id > 0:
previous_act = acts[act_id - 1]
if act_id < len(acts) - 1:
next_act = acts[act_id + 1]
for log_line in log_lines:
if log_line.transcript_page:
original_transcript_page = log_line.transcript_page
break
else:
original_transcript_page = None
if start:
permalink_fragment = '#log-line-%s' % timestamp_to_seconds(start)
else:
permalink_fragment = '#log-line-%s' % log_lines[0].timestamp
return {
'start':
start,
'log_lines':
log_lines,
'next_timestamp':
next_timestamp,
'previous_timestamp':
previous_timestamp,
'acts':
acts,
'act':
act_id + 1,
'current_act':
act,
'previous_act':
previous_act,
'next_act':
next_act,
'max_highlight_index':
max_highlight_index,
'first_highlighted_line':
first_highlighted_line,
'original_transcript_page':
original_transcript_page,
'permalink':
'http://%s%s%s' % (
self.request.META['HTTP_HOST'],
self.request.path,
permalink_fragment,
)
}