diff options
Diffstat (limited to 'coverage/inorout.py')
| -rw-r--r-- | coverage/inorout.py | 440 |
1 files changed, 440 insertions, 0 deletions
diff --git a/coverage/inorout.py b/coverage/inorout.py new file mode 100644 index 0000000..4fcec8e --- /dev/null +++ b/coverage/inorout.py @@ -0,0 +1,440 @@ +# Licensed under the Apache License: http://www.apache.org/licenses/LICENSE-2.0 +# For details: https://bitbucket.org/ned/coveragepy/src/default/NOTICE.txt + +"""Determining whether files are being measured/reported or not.""" + +# For finding the stdlib +import atexit +import inspect +import itertools +import os +import platform +import re +import sys +import traceback + +from coverage import env +from coverage.disposition import FileDisposition, disposition_init +from coverage.files import TreeMatcher, FnmatchMatcher, ModuleMatcher +from coverage.files import prep_patterns, find_python_files, canonical_filename +from coverage.misc import CoverageException +from coverage.python import source_for_file, source_for_morf + + +# Pypy has some unusual stuff in the "stdlib". Consider those locations +# when deciding where the stdlib is. These modules are not used for anything, +# they are modules importable from the pypy lib directories, so that we can +# find those directories. +_structseq = _pypy_irc_topic = None +if env.PYPY: + try: + import _structseq + except ImportError: + pass + + try: + import _pypy_irc_topic + except ImportError: + pass + + +def canonical_path(morf, directory=False): + """Return the canonical path of the module or file `morf`. + + If the module is a package, then return its directory. If it is a + module, then return its file, unless `directory` is True, in which + case return its enclosing directory. + + """ + morf_path = canonical_filename(source_for_morf(morf)) + if morf_path.endswith("__init__.py") or directory: + morf_path = os.path.split(morf_path)[0] + return morf_path + + +def name_for_module(module_globals, filename): + """Get the name of the module for a set of globals and file name. + + For configurability's sake, we allow __main__ modules to be matched by + their importable name. + + If loaded via runpy (aka -m), we can usually recover the "original" + full dotted module name, otherwise, we resort to interpreting the + file name to get the module's name. In the case that the module name + can't be determined, None is returned. + + """ + if module_globals is None: # pragma: only ironpython + # IronPython doesn't provide globals: https://github.com/IronLanguages/main/issues/1296 + module_globals = {} + + dunder_name = module_globals.get('__name__', None) + + if isinstance(dunder_name, str) and dunder_name != '__main__': + # This is the usual case: an imported module. + return dunder_name + + loader = module_globals.get('__loader__', None) + for attrname in ('fullname', 'name'): # attribute renamed in py3.2 + if hasattr(loader, attrname): + fullname = getattr(loader, attrname) + else: + continue + + if isinstance(fullname, str) and fullname != '__main__': + # Module loaded via: runpy -m + return fullname + + # Script as first argument to Python command line. + inspectedname = inspect.getmodulename(filename) + if inspectedname is not None: + return inspectedname + else: + return dunder_name + + +class InOrOut(object): + def __init__(self, warn): + self.warn = warn + + # The matchers for should_trace. + self.source_match = None + self.source_pkgs_match = None + self.pylib_paths = self.cover_paths = None + self.pylib_match = self.cover_match = None + self.include_match = self.omit_match = None + self.plugins = [] + self.disp_class = FileDisposition + + # The source argument can be directories or package names. + self.source = [] + self.source_pkgs = [] + self.omit = self.include = None + + def configure(self, config): + for src in config.source or []: + if os.path.isdir(src): + self.source.append(canonical_filename(src)) + else: + self.source_pkgs.append(src) + self.source_pkgs_unmatched = self.source_pkgs[:] + + self.omit = prep_patterns(config.run_omit) + self.include = prep_patterns(config.run_include) + + # The directories for files considered "installed with the interpreter". + self.pylib_paths = set() + if not config.cover_pylib: + # Look at where some standard modules are located. That's the + # indication for "installed with the interpreter". In some + # environments (virtualenv, for example), these modules may be + # spread across a few locations. Look at all the candidate modules + # we've imported, and take all the different ones. + for m in (atexit, inspect, os, platform, _pypy_irc_topic, re, _structseq, traceback): + if m is not None and hasattr(m, "__file__"): + self.pylib_paths.add(canonical_path(m, directory=True)) + + if _structseq and not hasattr(_structseq, '__file__'): + # PyPy 2.4 has no __file__ in the builtin modules, but the code + # objects still have the file names. So dig into one to find + # the path to exclude. + structseq_new = _structseq.structseq_new + try: + structseq_file = structseq_new.func_code.co_filename + except AttributeError: + structseq_file = structseq_new.__code__.co_filename + self.pylib_paths.add(canonical_path(structseq_file)) + + # To avoid tracing the coverage.py code itself, we skip anything + # located where we are. + self.cover_paths = [canonical_path(__file__, directory=True)] + if env.TESTING: + # Don't include our own test code. + self.cover_paths.append(os.path.join(self.cover_paths[0], "tests")) + + # When testing, we use PyContracts, which should be considered + # part of coverage.py, and it uses six. Exclude those directories + # just as we exclude ourselves. + import contracts + import six + for mod in [contracts, six]: + self.cover_paths.append(canonical_path(mod)) + + # Create the matchers we need for should_trace + if self.source or self.source_pkgs: + self.source_match = TreeMatcher(self.source) + self.source_pkgs_match = ModuleMatcher(self.source_pkgs) + else: + if self.cover_paths: + self.cover_match = TreeMatcher(self.cover_paths) + if self.pylib_paths: + self.pylib_match = TreeMatcher(self.pylib_paths) + if self.include: + self.include_match = FnmatchMatcher(self.include) + if self.omit: + self.omit_match = FnmatchMatcher(self.omit) + + def should_trace(self, filename, frame=None): + """Decide whether to trace execution in `filename`, with a reason. + + This function is called from the trace function. As each new file name + is encountered, this function determines whether it is traced or not. + + Returns a FileDisposition object. + + """ + original_filename = filename + disp = disposition_init(self.disp_class, filename) + + def nope(disp, reason): + """Simple helper to make it easy to return NO.""" + disp.trace = False + disp.reason = reason + return disp + + if frame is not None: + # Compiled Python files have two file names: frame.f_code.co_filename is + # the file name at the time the .pyc was compiled. The second name is + # __file__, which is where the .pyc was actually loaded from. Since + # .pyc files can be moved after compilation (for example, by being + # installed), we look for __file__ in the frame and prefer it to the + # co_filename value. + dunder_file = frame.f_globals and frame.f_globals.get('__file__') + if dunder_file: + filename = source_for_file(dunder_file) + if original_filename and not original_filename.startswith('<'): + orig = os.path.basename(original_filename) + if orig != os.path.basename(filename): + # Files shouldn't be renamed when moved. This happens when + # exec'ing code. If it seems like something is wrong with + # the frame's file name, then just use the original. + filename = original_filename + + if not filename: + # Empty string is pretty useless. + return nope(disp, "empty string isn't a file name") + + if filename.startswith('memory:'): + return nope(disp, "memory isn't traceable") + + if filename.startswith('<'): + # Lots of non-file execution is represented with artificial + # file names like "<string>", "<doctest readme.txt[0]>", or + # "<exec_function>". Don't ever trace these executions, since we + # can't do anything with the data later anyway. + return nope(disp, "not a real file name") + + # pyexpat does a dumb thing, calling the trace function explicitly from + # C code with a C file name. + if re.search(r"[/\\]Modules[/\\]pyexpat.c", filename): + return nope(disp, "pyexpat lies about itself") + + # Jython reports the .class file to the tracer, use the source file. + if filename.endswith("$py.class"): + filename = filename[:-9] + ".py" + + canonical = canonical_filename(filename) + disp.canonical_filename = canonical + + # Try the plugins, see if they have an opinion about the file. + plugin = None + for plugin in self.plugins.file_tracers: + if not plugin._coverage_enabled: + continue + + try: + file_tracer = plugin.file_tracer(canonical) + if file_tracer is not None: + file_tracer._coverage_plugin = plugin + disp.trace = True + disp.file_tracer = file_tracer + if file_tracer.has_dynamic_source_filename(): + disp.has_dynamic_filename = True + else: + disp.source_filename = canonical_filename( + file_tracer.source_filename() + ) + break + except Exception: + self.warn( + "Disabling plug-in %r due to an exception:" % (plugin._coverage_plugin_name) + ) + traceback.print_exc() + plugin._coverage_enabled = False + continue + else: + # No plugin wanted it: it's Python. + disp.trace = True + disp.source_filename = canonical + + if not disp.has_dynamic_filename: + if not disp.source_filename: + raise CoverageException( + "Plugin %r didn't set source_filename for %r" % + (plugin, disp.original_filename) + ) + module_globals = frame.f_globals if frame is not None else {} + reason = self.check_include_omit_etc(disp.source_filename, module_globals) + if reason: + nope(disp, reason) + + return disp + + def check_include_omit_etc(self, filename, module_globals): + """Check a file name against the include, omit, etc, rules. + + Returns a string or None. String means, don't trace, and is the reason + why. None means no reason found to not trace. + + """ + modulename = name_for_module(module_globals, filename) + + # If the user specified source or include, then that's authoritative + # about the outer bound of what to measure and we don't have to apply + # any canned exclusions. If they didn't, then we have to exclude the + # stdlib and coverage.py directories. + if self.source_match: + if self.source_pkgs_match.match(modulename): + if modulename in self.source_pkgs_unmatched: + self.source_pkgs_unmatched.remove(modulename) + elif not self.source_match.match(filename): + return "falls outside the --source trees" + elif self.include_match: + if not self.include_match.match(filename): + return "falls outside the --include trees" + else: + # If we aren't supposed to trace installed code, then check if this + # is near the Python standard library and skip it if so. + if self.pylib_match and self.pylib_match.match(filename): + return "is in the stdlib" + + # We exclude the coverage.py code itself, since a little of it + # will be measured otherwise. + if self.cover_match and self.cover_match.match(filename): + return "is part of coverage.py" + + # Check the file against the omit pattern. + if self.omit_match and self.omit_match.match(filename): + return "is inside an --omit pattern" + + # No reason found to skip this file. + return None + + def warn_conflicting_settings(self): + if self.include: + if self.source or self.source_pkgs: + self.warn("--include is ignored because --source is set", slug="include-ignored") + + def warn_already_imported_files(self): + if self.include or self.source or self.source_pkgs: + warned = set() + for mod in list(sys.modules.values()): + filename = getattr(mod, "__file__", None) + if filename is None: + continue + if filename in warned: + continue + + disp = self.should_trace(filename) + if disp.trace: + msg = "Already imported a file that will be measured: {0}".format(filename) + self.warn(msg, slug="already-imported") + warned.add(filename) + + return True + return False + + def warn_unimported_source(self): + for pkg in self.source_pkgs_unmatched: + self.warn_about_unmeasured_code(pkg) + + def warn_about_unmeasured_code(self, pkg): + """Warn about a package or module that we never traced. + + `pkg` is a string, the name of the package or module. + + """ + mod = sys.modules.get(pkg) + if mod is None: + self.warn("Module %s was never imported." % pkg, slug="module-not-imported") + return + + is_namespace = hasattr(mod, '__path__') and not hasattr(mod, '__file__') + has_file = hasattr(mod, '__file__') and os.path.exists(mod.__file__) + + if is_namespace: + # A namespace package. It's OK for this not to have been traced, + # since there is no code directly in it. + return + + if not has_file: + self.warn("Module %s has no Python source." % pkg, slug="module-not-python") + return + + # The module was in sys.modules, and seems like a module with code, but + # we never measured it. I guess that means it was imported before + # coverage even started. + self.warn( + "Module %s was previously imported, but not measured" % pkg, + slug="module-not-measured", + ) + + def find_unexecuted_files(self): + for pkg in self.source_pkgs: + if (not pkg in sys.modules or + not hasattr(sys.modules[pkg], '__file__') or + not os.path.exists(sys.modules[pkg].__file__)): + continue + pkg_file = source_for_file(sys.modules[pkg].__file__) + for ret in self._find_unexecuted_files(canonical_path(pkg_file)): + yield ret + + for src in self.source: + for ret in self._find_unexecuted_files(src): + yield ret + + def _find_plugin_files(self, src_dir): + """Get executable files from the plugins.""" + for plugin in self.plugins.file_tracers: + for x_file in plugin.find_executable_files(src_dir): + yield x_file, plugin._coverage_plugin_name + + def _find_unexecuted_files(self, src_dir): + """Find unexecuted files in `src_dir`. + + Search for files in `src_dir` that are probably importable, + and add them as unexecuted files in `self.data`. + + """ + py_files = ((py_file, None) for py_file in find_python_files(src_dir)) + plugin_files = self._find_plugin_files(src_dir) + + for file_path, plugin_name in itertools.chain(py_files, plugin_files): + file_path = canonical_filename(file_path) + if self.omit_match and self.omit_match.match(file_path): + # Turns out this file was omitted, so don't pull it back + # in as unexecuted. + continue + yield file_path, plugin_name + + def sys_info(self): + info = [ + ('cover_paths', self.cover_paths), + ('pylib_paths', self.pylib_paths), + ] + + matcher_names = [ + 'source_match', 'source_pkgs_match', + 'include_match', 'omit_match', + 'cover_match', 'pylib_match', + ] + + for matcher_name in matcher_names: + matcher = getattr(self, matcher_name) + if matcher: + matcher_info = matcher.info() + else: + matcher_info = '-none-' + info.append((matcher_name, matcher_info)) + + return info |