def _format_line(self, line, column=0, rel_line=1):
"Formats a line from the data to be the appropriate length"
raw_data = self.data[line].rstrip()
with_ws = len(raw_data)
data = raw_data.lstrip()
line_length = len(data)
if line_length > 140:
if rel_line == 0:
# Trim from the beginning
data = "... %s" % data[-140:]
elif rel_line == 1:
# Trim surrounding the error position
if column < 70:
data = "%s ..." % data[:140]
elif column > line_length - 70:
data = "... %s" % data[-140:]
else:
data = "... %s ..." % data[column - 70:column + 70]
elif rel_line == 2:
# Trim from the end
data = "%s ..." % data[:140]
data = "%s%s" % (raw_data[0:with_ws - line_length], data)
data = textfilter.filter_ascii(data)
return data
def _save_message(self, stack, type_, message, context=None):
"Stores a message in the appropriate message stack."
uid = uuid.uuid4().hex
message["uid"] = uid
# Get the context for the message (if there's a context available)
if context is not None:
if isinstance(context, tuple):
message["context"] = context
else:
message["context"] = \
context.get_context(line=message["line"],
column=message["column"])
else:
message["context"] = None
message["message"] = filter_ascii(message["message"])
message["description"] = filter_ascii(message["description"])
stack.append(message)
# Mark the tier that the error occurred at
if message["tier"] is None:
message["tier"] = self.tier
if message["id"]:
tree = self.message_tree
last_id = None
for eid in message["id"]:
if last_id is not None:
tree = tree[last_id]
if eid not in tree:
tree[eid] = {"__errors": 0,
"__warnings": 0,
"__notices": 0,
"__messages": []}
tree[eid]["__%s" % type_] += 1
last_id = eid
tree[last_id]['__messages'].append(uid)
def decode(data):
"""
Decode data employing some charset detection and including unicode BOM
stripping.
"""
# Don't make more work than we have to.
if not isinstance(data, str):
return data
# Detect standard unicodes.
for bom, encoding in UNICODES:
if data.startswith(bom):
return unicode(data[len(bom):], encoding, "ignore")
# Try straight UTF-8
try:
return unicode(data, "utf-8")
except UnicodeDecodeError:
pass
# Test for latin_1, because it can be matched as UTF-16
# Somewhat of a hack, but it works and is about a thousand times faster
# than using chardet.
if all(ord(c) < 256 for c in data):
try:
return unicode(data, "latin_1")
except UnicodeDecodeError:
pass
# Test for various common encodings.
for encoding in COMMON_ENCODINGS:
try:
return unicode(data, encoding)
except UnicodeDecodeError:
pass
# Anything else gets filtered.
return unicode(textfilter.filter_ascii(data), errors="replace")
def decode(data):
"""
Decode data employing some charset detection and including unicode BOM
stripping.
"""
# Don't make more work than we have to.
if not isinstance(data, str):
return data
# Detect standard unicodes.
for bom, encoding in UNICODES:
if data.startswith(bom):
return unicode(data[len(bom):], encoding, "ignore")
# Try straight UTF-8
try:
return unicode(data, "utf-8")
except UnicodeDecodeError:
pass
# Test for latin_1, because it can be matched as UTF-16
# Somewhat of a hack, but it works and is about a thousand times faster
# than using chardet.
if all(ord(c) < 256 for c in data):
try:
return unicode(data, "latin_1")
except UnicodeDecodeError:
pass
# Test for various common encodings.
for encoding in COMMON_ENCODINGS:
try:
return unicode(data, encoding)
except UnicodeDecodeError:
pass
# Anything else gets filtered.
return unicode(textfilter.filter_ascii(data), errors="replace")