Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(119)

Unified Diff: sky/tools/webkitpy/thirdparty/coverage/parser.py

Issue 946753002: Delete a bunch of dead python code. (Closed) Base URL: git@github.com:domokit/mojo.git@master
Patch Set: Created 5 years, 10 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
Index: sky/tools/webkitpy/thirdparty/coverage/parser.py
diff --git a/sky/tools/webkitpy/thirdparty/coverage/parser.py b/sky/tools/webkitpy/thirdparty/coverage/parser.py
deleted file mode 100644
index d380eda12baa9133ea0f03c82a7ffce84ac72af9..0000000000000000000000000000000000000000
--- a/sky/tools/webkitpy/thirdparty/coverage/parser.py
+++ /dev/null
@@ -1,650 +0,0 @@
-"""Code parsing for Coverage."""
-
-import opcode, re, sys, token, tokenize
-
-from coverage.backward import set, sorted, StringIO # pylint: disable=W0622
-from coverage.backward import open_source
-from coverage.bytecode import ByteCodes, CodeObjects
-from coverage.misc import nice_pair, expensive, join_regex
-from coverage.misc import CoverageException, NoSource, NotPython
-
-
-class CodeParser(object):
- """Parse code to find executable lines, excluded lines, etc."""
-
- def __init__(self, text=None, filename=None, exclude=None):
- """
- Source can be provided as `text`, the text itself, or `filename`, from
- which the text will be read. Excluded lines are those that match
- `exclude`, a regex.
-
- """
- assert text or filename, "CodeParser needs either text or filename"
- self.filename = filename or "<code>"
- self.text = text
- if not self.text:
- try:
- sourcef = open_source(self.filename)
- try:
- self.text = sourcef.read()
- finally:
- sourcef.close()
- except IOError:
- _, err, _ = sys.exc_info()
- raise NoSource(
- "No source for code: %r: %s" % (self.filename, err)
- )
-
- self.exclude = exclude
-
- self.show_tokens = False
-
- # The text lines of the parsed code.
- self.lines = self.text.split('\n')
-
- # The line numbers of excluded lines of code.
- self.excluded = set()
-
- # The line numbers of docstring lines.
- self.docstrings = set()
-
- # The line numbers of class definitions.
- self.classdefs = set()
-
- # A dict mapping line numbers to (lo,hi) for multi-line statements.
- self.multiline = {}
-
- # The line numbers that start statements.
- self.statement_starts = set()
-
- # Lazily-created ByteParser
- self._byte_parser = None
-
- def _get_byte_parser(self):
- """Create a ByteParser on demand."""
- if not self._byte_parser:
- self._byte_parser = \
- ByteParser(text=self.text, filename=self.filename)
- return self._byte_parser
- byte_parser = property(_get_byte_parser)
-
- def lines_matching(self, *regexes):
- """Find the lines matching one of a list of regexes.
-
- Returns a set of line numbers, the lines that contain a match for one
- of the regexes in `regexes`. The entire line needn't match, just a
- part of it.
-
- """
- regex_c = re.compile(join_regex(regexes))
- matches = set()
- for i, ltext in enumerate(self.lines):
- if regex_c.search(ltext):
- matches.add(i+1)
- return matches
-
- def _raw_parse(self):
- """Parse the source to find the interesting facts about its lines.
-
- A handful of member fields are updated.
-
- """
- # Find lines which match an exclusion pattern.
- if self.exclude:
- self.excluded = self.lines_matching(self.exclude)
-
- # Tokenize, to find excluded suites, to find docstrings, and to find
- # multi-line statements.
- indent = 0
- exclude_indent = 0
- excluding = False
- prev_toktype = token.INDENT
- first_line = None
- empty = True
-
- tokgen = tokenize.generate_tokens(StringIO(self.text).readline)
- for toktype, ttext, (slineno, _), (elineno, _), ltext in tokgen:
- if self.show_tokens: # pragma: no cover
- print("%10s %5s %-20r %r" % (
- tokenize.tok_name.get(toktype, toktype),
- nice_pair((slineno, elineno)), ttext, ltext
- ))
- if toktype == token.INDENT:
- indent += 1
- elif toktype == token.DEDENT:
- indent -= 1
- elif toktype == token.NAME and ttext == 'class':
- # Class definitions look like branches in the byte code, so
- # we need to exclude them. The simplest way is to note the
- # lines with the 'class' keyword.
- self.classdefs.add(slineno)
- elif toktype == token.OP and ttext == ':':
- if not excluding and elineno in self.excluded:
- # Start excluding a suite. We trigger off of the colon
- # token so that the #pragma comment will be recognized on
- # the same line as the colon.
- exclude_indent = indent
- excluding = True
- elif toktype == token.STRING and prev_toktype == token.INDENT:
- # Strings that are first on an indented line are docstrings.
- # (a trick from trace.py in the stdlib.) This works for
- # 99.9999% of cases. For the rest (!) see:
- # http://stackoverflow.com/questions/1769332/x/1769794#1769794
- for i in range(slineno, elineno+1):
- self.docstrings.add(i)
- elif toktype == token.NEWLINE:
- if first_line is not None and elineno != first_line:
- # We're at the end of a line, and we've ended on a
- # different line than the first line of the statement,
- # so record a multi-line range.
- rng = (first_line, elineno)
- for l in range(first_line, elineno+1):
- self.multiline[l] = rng
- first_line = None
-
- if ttext.strip() and toktype != tokenize.COMMENT:
- # A non-whitespace token.
- empty = False
- if first_line is None:
- # The token is not whitespace, and is the first in a
- # statement.
- first_line = slineno
- # Check whether to end an excluded suite.
- if excluding and indent <= exclude_indent:
- excluding = False
- if excluding:
- self.excluded.add(elineno)
-
- prev_toktype = toktype
-
- # Find the starts of the executable statements.
- if not empty:
- self.statement_starts.update(self.byte_parser._find_statements())
-
- def first_line(self, line):
- """Return the first line number of the statement including `line`."""
- rng = self.multiline.get(line)
- if rng:
- first_line = rng[0]
- else:
- first_line = line
- return first_line
-
- def first_lines(self, lines, ignore=None):
- """Map the line numbers in `lines` to the correct first line of the
- statement.
-
- Skip any line mentioned in `ignore`.
-
- Returns a sorted list of the first lines.
-
- """
- ignore = ignore or []
- lset = set()
- for l in lines:
- if l in ignore:
- continue
- new_l = self.first_line(l)
- if new_l not in ignore:
- lset.add(new_l)
- return sorted(lset)
-
- def parse_source(self):
- """Parse source text to find executable lines, excluded lines, etc.
-
- Return values are 1) a sorted list of executable line numbers, and
- 2) a sorted list of excluded line numbers.
-
- Reported line numbers are normalized to the first line of multi-line
- statements.
-
- """
- self._raw_parse()
-
- excluded_lines = self.first_lines(self.excluded)
- ignore = excluded_lines + list(self.docstrings)
- lines = self.first_lines(self.statement_starts, ignore)
-
- return lines, excluded_lines
-
- def arcs(self):
- """Get information about the arcs available in the code.
-
- Returns a sorted list of line number pairs. Line numbers have been
- normalized to the first line of multiline statements.
-
- """
- all_arcs = []
- for l1, l2 in self.byte_parser._all_arcs():
- fl1 = self.first_line(l1)
- fl2 = self.first_line(l2)
- if fl1 != fl2:
- all_arcs.append((fl1, fl2))
- return sorted(all_arcs)
- arcs = expensive(arcs)
-
- def exit_counts(self):
- """Get a mapping from line numbers to count of exits from that line.
-
- Excluded lines are excluded.
-
- """
- excluded_lines = self.first_lines(self.excluded)
- exit_counts = {}
- for l1, l2 in self.arcs():
- if l1 < 0:
- # Don't ever report -1 as a line number
- continue
- if l1 in excluded_lines:
- # Don't report excluded lines as line numbers.
- continue
- if l2 in excluded_lines:
- # Arcs to excluded lines shouldn't count.
- continue
- if l1 not in exit_counts:
- exit_counts[l1] = 0
- exit_counts[l1] += 1
-
- # Class definitions have one extra exit, so remove one for each:
- for l in self.classdefs:
- # Ensure key is there: classdefs can include excluded lines.
- if l in exit_counts:
- exit_counts[l] -= 1
-
- return exit_counts
- exit_counts = expensive(exit_counts)
-
-
-## Opcodes that guide the ByteParser.
-
-def _opcode(name):
- """Return the opcode by name from the opcode module."""
- return opcode.opmap[name]
-
-def _opcode_set(*names):
- """Return a set of opcodes by the names in `names`."""
- s = set()
- for name in names:
- try:
- s.add(_opcode(name))
- except KeyError:
- pass
- return s
-
-# Opcodes that leave the code object.
-OPS_CODE_END = _opcode_set('RETURN_VALUE')
-
-# Opcodes that unconditionally end the code chunk.
-OPS_CHUNK_END = _opcode_set(
- 'JUMP_ABSOLUTE', 'JUMP_FORWARD', 'RETURN_VALUE', 'RAISE_VARARGS',
- 'BREAK_LOOP', 'CONTINUE_LOOP',
- )
-
-# Opcodes that unconditionally begin a new code chunk. By starting new chunks
-# with unconditional jump instructions, we neatly deal with jumps to jumps
-# properly.
-OPS_CHUNK_BEGIN = _opcode_set('JUMP_ABSOLUTE', 'JUMP_FORWARD')
-
-# Opcodes that push a block on the block stack.
-OPS_PUSH_BLOCK = _opcode_set(
- 'SETUP_LOOP', 'SETUP_EXCEPT', 'SETUP_FINALLY', 'SETUP_WITH'
- )
-
-# Block types for exception handling.
-OPS_EXCEPT_BLOCKS = _opcode_set('SETUP_EXCEPT', 'SETUP_FINALLY')
-
-# Opcodes that pop a block from the block stack.
-OPS_POP_BLOCK = _opcode_set('POP_BLOCK')
-
-# Opcodes that have a jump destination, but aren't really a jump.
-OPS_NO_JUMP = OPS_PUSH_BLOCK
-
-# Individual opcodes we need below.
-OP_BREAK_LOOP = _opcode('BREAK_LOOP')
-OP_END_FINALLY = _opcode('END_FINALLY')
-OP_COMPARE_OP = _opcode('COMPARE_OP')
-COMPARE_EXCEPTION = 10 # just have to get this const from the code.
-OP_LOAD_CONST = _opcode('LOAD_CONST')
-OP_RETURN_VALUE = _opcode('RETURN_VALUE')
-
-
-class ByteParser(object):
- """Parse byte codes to understand the structure of code."""
-
- def __init__(self, code=None, text=None, filename=None):
- if code:
- self.code = code
- self.text = text
- else:
- if not text:
- assert filename, "If no code or text, need a filename"
- sourcef = open_source(filename)
- try:
- text = sourcef.read()
- finally:
- sourcef.close()
- self.text = text
-
- try:
- # Python 2.3 and 2.4 don't like partial last lines, so be sure
- # the text ends nicely for them.
- self.code = compile(text + '\n', filename, "exec")
- except SyntaxError:
- _, synerr, _ = sys.exc_info()
- raise NotPython(
- "Couldn't parse '%s' as Python source: '%s' at line %d" %
- (filename, synerr.msg, synerr.lineno)
- )
-
- # Alternative Python implementations don't always provide all the
- # attributes on code objects that we need to do the analysis.
- for attr in ['co_lnotab', 'co_firstlineno', 'co_consts', 'co_code']:
- if not hasattr(self.code, attr):
- raise CoverageException(
- "This implementation of Python doesn't support code "
- "analysis.\n"
- "Run coverage.py under CPython for this command."
- )
-
- def child_parsers(self):
- """Iterate over all the code objects nested within this one.
-
- The iteration includes `self` as its first value.
-
- """
- children = CodeObjects(self.code)
- return [ByteParser(code=c, text=self.text) for c in children]
-
- # Getting numbers from the lnotab value changed in Py3.0.
- if sys.version_info >= (3, 0):
- def _lnotab_increments(self, lnotab):
- """Return a list of ints from the lnotab bytes in 3.x"""
- return list(lnotab)
- else:
- def _lnotab_increments(self, lnotab):
- """Return a list of ints from the lnotab string in 2.x"""
- return [ord(c) for c in lnotab]
-
- def _bytes_lines(self):
- """Map byte offsets to line numbers in `code`.
-
- Uses co_lnotab described in Python/compile.c to map byte offsets to
- line numbers. Returns a list: [(b0, l0), (b1, l1), ...]
-
- """
- # Adapted from dis.py in the standard library.
- byte_increments = self._lnotab_increments(self.code.co_lnotab[0::2])
- line_increments = self._lnotab_increments(self.code.co_lnotab[1::2])
-
- bytes_lines = []
- last_line_num = None
- line_num = self.code.co_firstlineno
- byte_num = 0
- for byte_incr, line_incr in zip(byte_increments, line_increments):
- if byte_incr:
- if line_num != last_line_num:
- bytes_lines.append((byte_num, line_num))
- last_line_num = line_num
- byte_num += byte_incr
- line_num += line_incr
- if line_num != last_line_num:
- bytes_lines.append((byte_num, line_num))
- return bytes_lines
-
- def _find_statements(self):
- """Find the statements in `self.code`.
-
- Return a set of line numbers that start statements. Recurses into all
- code objects reachable from `self.code`.
-
- """
- stmts = set()
- for bp in self.child_parsers():
- # Get all of the lineno information from this code.
- for _, l in bp._bytes_lines():
- stmts.add(l)
- return stmts
-
- def _split_into_chunks(self):
- """Split the code object into a list of `Chunk` objects.
-
- Each chunk is only entered at its first instruction, though there can
- be many exits from a chunk.
-
- Returns a list of `Chunk` objects.
-
- """
-
- # The list of chunks so far, and the one we're working on.
- chunks = []
- chunk = None
- bytes_lines_map = dict(self._bytes_lines())
-
- # The block stack: loops and try blocks get pushed here for the
- # implicit jumps that can occur.
- # Each entry is a tuple: (block type, destination)
- block_stack = []
-
- # Some op codes are followed by branches that should be ignored. This
- # is a count of how many ignores are left.
- ignore_branch = 0
-
- # We have to handle the last two bytecodes specially.
- ult = penult = None
-
- for bc in ByteCodes(self.code.co_code):
- # Maybe have to start a new chunk
- if bc.offset in bytes_lines_map:
- # Start a new chunk for each source line number.
- if chunk:
- chunk.exits.add(bc.offset)
- chunk = Chunk(bc.offset, bytes_lines_map[bc.offset])
- chunks.append(chunk)
- elif bc.op in OPS_CHUNK_BEGIN:
- # Jumps deserve their own unnumbered chunk. This fixes
- # problems with jumps to jumps getting confused.
- if chunk:
- chunk.exits.add(bc.offset)
- chunk = Chunk(bc.offset)
- chunks.append(chunk)
-
- if not chunk:
- chunk = Chunk(bc.offset)
- chunks.append(chunk)
-
- # Look at the opcode
- if bc.jump_to >= 0 and bc.op not in OPS_NO_JUMP:
- if ignore_branch:
- # Someone earlier wanted us to ignore this branch.
- ignore_branch -= 1
- else:
- # The opcode has a jump, it's an exit for this chunk.
- chunk.exits.add(bc.jump_to)
-
- if bc.op in OPS_CODE_END:
- # The opcode can exit the code object.
- chunk.exits.add(-self.code.co_firstlineno)
- if bc.op in OPS_PUSH_BLOCK:
- # The opcode adds a block to the block_stack.
- block_stack.append((bc.op, bc.jump_to))
- if bc.op in OPS_POP_BLOCK:
- # The opcode pops a block from the block stack.
- block_stack.pop()
- if bc.op in OPS_CHUNK_END:
- # This opcode forces the end of the chunk.
- if bc.op == OP_BREAK_LOOP:
- # A break is implicit: jump where the top of the
- # block_stack points.
- chunk.exits.add(block_stack[-1][1])
- chunk = None
- if bc.op == OP_END_FINALLY:
- if block_stack:
- # A break that goes through a finally will jump to whatever
- # block is on top of the stack.
- chunk.exits.add(block_stack[-1][1])
- # For the finally clause we need to find the closest exception
- # block, and use its jump target as an exit.
- for iblock in range(len(block_stack)-1, -1, -1):
- if block_stack[iblock][0] in OPS_EXCEPT_BLOCKS:
- chunk.exits.add(block_stack[iblock][1])
- break
- if bc.op == OP_COMPARE_OP and bc.arg == COMPARE_EXCEPTION:
- # This is an except clause. We want to overlook the next
- # branch, so that except's don't count as branches.
- ignore_branch += 1
-
- penult = ult
- ult = bc
-
- if chunks:
- # The last two bytecodes could be a dummy "return None" that
- # shouldn't be counted as real code. Every Python code object seems
- # to end with a return, and a "return None" is inserted if there
- # isn't an explicit return in the source.
- if ult and penult:
- if penult.op == OP_LOAD_CONST and ult.op == OP_RETURN_VALUE:
- if self.code.co_consts[penult.arg] is None:
- # This is "return None", but is it dummy? A real line
- # would be a last chunk all by itself.
- if chunks[-1].byte != penult.offset:
- ex = -self.code.co_firstlineno
- # Split the last chunk
- last_chunk = chunks[-1]
- last_chunk.exits.remove(ex)
- last_chunk.exits.add(penult.offset)
- chunk = Chunk(penult.offset)
- chunk.exits.add(ex)
- chunks.append(chunk)
-
- # Give all the chunks a length.
- chunks[-1].length = bc.next_offset - chunks[-1].byte
- for i in range(len(chunks)-1):
- chunks[i].length = chunks[i+1].byte - chunks[i].byte
-
- return chunks
-
- def _arcs(self):
- """Find the executable arcs in the code.
-
- Returns a set of pairs, (from,to). From and to are integer line
- numbers. If from is < 0, then the arc is an entrance into the code
- object. If to is < 0, the arc is an exit from the code object.
-
- """
- chunks = self._split_into_chunks()
-
- # A map from byte offsets to chunks jumped into.
- byte_chunks = dict([(c.byte, c) for c in chunks])
-
- # Build a map from byte offsets to actual lines reached.
- byte_lines = {}
- bytes_to_add = set([c.byte for c in chunks])
-
- while bytes_to_add:
- byte_to_add = bytes_to_add.pop()
- if byte_to_add in byte_lines or byte_to_add < 0:
- continue
-
- # Which lines does this chunk lead to?
- bytes_considered = set()
- bytes_to_consider = [byte_to_add]
- lines = set()
-
- while bytes_to_consider:
- byte = bytes_to_consider.pop()
- bytes_considered.add(byte)
-
- # Find chunk for byte
- try:
- ch = byte_chunks[byte]
- except KeyError:
- for ch in chunks:
- if ch.byte <= byte < ch.byte+ch.length:
- break
- else:
- # No chunk for this byte!
- raise Exception("Couldn't find chunk @ %d" % byte)
- byte_chunks[byte] = ch
-
- if ch.line:
- lines.add(ch.line)
- else:
- for ex in ch.exits:
- if ex < 0:
- lines.add(ex)
- elif ex not in bytes_considered:
- bytes_to_consider.append(ex)
-
- bytes_to_add.update(ch.exits)
-
- byte_lines[byte_to_add] = lines
-
- # Figure out for each chunk where the exits go.
- arcs = set()
- for chunk in chunks:
- if chunk.line:
- for ex in chunk.exits:
- if ex < 0:
- exit_lines = [ex]
- else:
- exit_lines = byte_lines[ex]
- for exit_line in exit_lines:
- if chunk.line != exit_line:
- arcs.add((chunk.line, exit_line))
- for line in byte_lines[0]:
- arcs.add((-1, line))
-
- return arcs
-
- def _all_chunks(self):
- """Returns a list of `Chunk` objects for this code and its children.
-
- See `_split_into_chunks` for details.
-
- """
- chunks = []
- for bp in self.child_parsers():
- chunks.extend(bp._split_into_chunks())
-
- return chunks
-
- def _all_arcs(self):
- """Get the set of all arcs in this code object and its children.
-
- See `_arcs` for details.
-
- """
- arcs = set()
- for bp in self.child_parsers():
- arcs.update(bp._arcs())
-
- return arcs
-
-
-class Chunk(object):
- """A sequence of bytecodes with a single entrance.
-
- To analyze byte code, we have to divide it into chunks, sequences of byte
- codes such that each basic block has only one entrance, the first
- instruction in the block.
-
- This is almost the CS concept of `basic block`_, except that we're willing
- to have many exits from a chunk, and "basic block" is a more cumbersome
- term.
-
- .. _basic block: http://en.wikipedia.org/wiki/Basic_block
-
- An exit < 0 means the chunk can leave the code (return). The exit is
- the negative of the starting line number of the code block.
-
- """
- def __init__(self, byte, line=0):
- self.byte = byte
- self.line = line
- self.length = 0
- self.exits = set()
-
- def __repr__(self):
- return "<%d+%d @%d %r>" % (
- self.byte, self.length, self.line, list(self.exits)
- )
« no previous file with comments | « sky/tools/webkitpy/thirdparty/coverage/misc.py ('k') | sky/tools/webkitpy/thirdparty/coverage/phystokens.py » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698