#!python # encoding: utf-8 # ***** BEGIN LICENSE BLOCK ***** # Version: MPL 1.1/GPL 2.0/LGPL 2.1 # # The contents of this file are subject to the Mozilla Public License # Version 1.1 (the "License"); you may not use this file except in # compliance with the License. You may obtain a copy of the License at # http://www.mozilla.org/MPL/ # # Software distributed under the License is distributed on an "AS IS" # basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See the # License for the specific language governing rights and limitations # under the License. # # The Original Code is Komodo code. # # The Initial Developer of the Original Code is ActiveState Software Inc. # Portions created by ActiveState Software Inc are Copyright (C) 2000-2007 # ActiveState Software Inc. All Rights Reserved. # # Contributor(s): # ActiveState Software Inc # # Alternatively, the contents of this file may be used under the terms of # either the GNU General Public License Version 2 or later (the "GPL"), or # the GNU Lesser General Public License Version 2.1 or later (the "LGPL"), # in which case the provisions of the GPL or the LGPL are applicable instead # of those above. If you wish to allow use of your version of this file only # under the terms of either the GPL or the LGPL, and not to allow others to # use your version of this file under the terms of the MPL, indicate your # decision by deleting the provisions above and replace them with the notice # and other provisions required by the GPL or the LGPL. If you do not delete # the provisions above, a recipient may use your version of this file under # the terms of any one of the MPL, the GPL or the LGPL. # # ***** END LICENSE BLOCK ***** """Code Intelligence: utility functions""" import bisect import os from os.path import basename import sys import re import stat import textwrap import logging import types from pprint import pprint, pformat import time import codecs # Global dict for holding specific hotshot profilers hotshotProfilers = {} #---- general stuff def isident(char): char = ord(char) return ord('a') <= char <= ord('z') or ord('A') <= char <= ord('Z') or char == ord('_') def isdigit(char): char = ord(char) return ord('0') <= char <= ord('9') # A "safe" language name for the given language where safe generally # means safe for a file path. _safe_lang_from_lang_cache = { "C++": "cpp", } def safe_lang_from_lang(lang): global _safe_lang_from_lang_cache try: return _safe_lang_from_lang_cache[lang] except KeyError: safe_lang = lang.lower().replace(' ', '_') _safe_lang_from_lang_cache[lang] = safe_lang return safe_lang # @deprecated: Manager.buf_from_path now uses textinfo to guess lang. def guess_lang_from_path(path): lang_from_ext = { ".py": "Python", ".pl": "Perl", ".pm": "Perl", ".tcl": "Tcl", ".php": "PHP", ".inc": "PHP", ".rb": "Ruby", ".rhtml": "RHTML", ".html.erb": "RHTML", ".js": "JavaScript", ".java": "Java", ".css": "CSS", ".xul": "XUL", ".xbl": "XBL", ".html": "HTML", ".xml": "XML", ".tpl": "Smarty", ".django.html": "Django", ".mason.html": "Mason", ".ttkt.html": "TemplateToolkit", ".cxx": "C++", } idx = 0 base = basename(path) while base.find('.', idx) != -1: idx = base.find('.', idx) if idx == -1: break ext = base[idx:] if ext in lang_from_ext: return lang_from_ext[ext] idx += 1 from codeintel2.common import CodeIntelError raise CodeIntelError("couldn't guess lang for `%s'" % path) def gen_dirs_under_dirs(dirs, max_depth, interesting_file_patterns=None, skip_scc_control_dirs=True): """Generate all dirs under the given dirs (including the given dirs themselves). "max_depth" is an integer maximum number of sub-directories that this method with recurse. "file_patterns", if given, is a sequence of glob patterns for "interesting" files. Directories with no interesting files are not included (though sub-directories of these may be). "skip_scc_control_dirs" is a boolean (default True) indicating if svn and cvs control dirs should be skipped. """ from os.path import normpath, abspath, expanduser from fnmatch import fnmatch dirs_to_skip = (skip_scc_control_dirs and ["CVS", ".svn", ".hg", ".git", ".bzr"] or []) # We must keep track of the directories we have walked, as the list of dirs # can overlap - bug 90289. walked_these_dirs = {} for dir in dirs: norm_dir = normpath(abspath(expanduser(dir))) LEN_DIR = len(norm_dir) for dirpath, dirnames, filenames in walk2(norm_dir): if dirpath in walked_these_dirs: dirnames[:] = [] # Already walked - no need to do it again. continue if dirpath[LEN_DIR:].count(os.sep) >= max_depth: dirnames[:] = [] # hit max_depth else: walked_these_dirs[dirpath] = True for dir_to_skip in dirs_to_skip: if dir_to_skip in dirnames: dirnames.remove(dir_to_skip) if interesting_file_patterns: for pat, filename in ( (p, f) for p in interesting_file_patterns for f in filenames): if fnmatch(filename, pat): break else: # No interesting files in this dir. continue yield dirpath #---- standard module/class/function doc parsing LINE_LIMIT = 5 # limit full number of lines this number LINE_WIDTH = 60 # wrap doc summaries to this width # Examples matches to this pattern: # foo(args) # foo(args) -> retval # foo(args) -- description # retval = foo(args) # retval = foo(args) -- description _gPySigLinePat = re.compile( r"^((?P[^=]+?)\s*=|class)?\s*(?P[\w\.]+\s?\(.*?\))\s*(?P[:<>=-]*)\s*(?P.*)$") _gSentenceSepPat = re.compile(r"(?<=\.)\s+", re.M) # split on sentence bndry def parseDocSummary(doclines, limit=LINE_LIMIT, width=LINE_WIDTH): """Parse out a short summary from the given doclines. "doclines" is a list of lines (without trailing newlines) to parse. "limit" is the number of lines to which to limit the summary. The "short summary" is the first sentence limited by (1) the "limit" number of lines and (2) one paragraph. If the first *two* sentences fit on the first line, then use both. Returns a list of summary lines. """ # Skip blank lines. start = 0 while start < len(doclines): if doclines[start].strip(): break start += 1 desclines = [] for i in range(start, len(doclines)): if len(desclines) >= limit: break stripped = doclines[i].strip() if not stripped: break sentences = _gSentenceSepPat.split(stripped) if sentences and not sentences[-1].endswith('.'): del sentences[-1] # last bit might not be a complete sentence if not sentences: desclines.append(stripped + ' ') continue elif i == start and len(sentences) > 1: desclines.append(' '.join([s.strip() for s in sentences[:2]])) else: desclines.append(sentences[0].strip()) break if desclines: if desclines[-1][-1] == ' ': # If terminated at non-sentence boundary then have extraneous # trailing space. desclines[-1] = desclines[-1][:-1] desclines = textwrap.wrap(''.join(desclines), width) return desclines def parsePyFuncDoc(doc, fallbackCallSig=None, scope="?", funcname="?"): """Parse the given Python function/method doc-string into call-signature and description bits. "doc" is the function doc string. "fallbackCallSig" (optional) is a list of call signature lines to fallback to if one cannot be determined from the doc string. "scope" (optional) is the module/class parent scope name. This is just used for better error/log reporting. "funcname" (optional) is the function name. This is just used for better error/log reporting. Examples of doc strings with call-signature info: close(): explicitly release resources held. x.__repr__() <==> repr(x) read([s]) -- Read s characters, or the rest of the string recv(buffersize[, flags]) -> data replace (str, old, new[, maxsplit]) -> string class StringIO([buffer]) Returns a 2-tuple: (, ) """ if doc is None or not doc.strip(): return ([], []) limit = LINE_LIMIT if not isinstance(doc, str): # try to convert from utf8 to unicode; if we fail, too bad. try: doc = codecs.utf_8_decode(doc)[0] except UnicodeDecodeError: pass doclines = doc.splitlines(0) index = 0 siglines = [] desclines = [] # Skip leading blank lines. while index < len(doclines): if doclines[index].strip(): break index += 1 # Parse out the call signature block, if it looks like there is one. if index >= len(doclines): match = None else: first = doclines[index].strip() match = _gPySigLinePat.match(first) if match: # The 'doc' looks like it starts with a call signature block. for i, line in enumerate(doclines[index:]): if len(siglines) >= limit: index = i break stripped = line.strip() if not stripped: index = i break match = _gPySigLinePat.match(stripped) if not match: index = i break # Now parse off what may be description content on the same line. # ":", "-" or "--" separator: tail is description # "-->" or "->" separator: tail if part of call sig # "<==>" separator: tail if part of call sig # other separtor: leave as part of call sig for now descSeps = ("-", "--", ":") groupd = match.groupdict() retval, head, sep, tail = ( groupd.get("retval"), groupd.get("head"), groupd.get("sep"), groupd.get("tail")) if retval: siglines.append(head + " -> " + retval) if tail and sep in descSeps: desclines.append(tail) elif tail and sep in descSeps: siglines.append(head) desclines.append(tail) else: siglines.append(stripped) else: index = len(doclines) if not siglines and fallbackCallSig: siglines = fallbackCallSig # Parse out the description block. if desclines: # Use what we have already. Just need to wrap it. desclines = textwrap.wrap(' '.join(desclines), LINE_WIDTH) else: doclines = doclines[index:] # strip leading empty lines while len(doclines) > 0 and not doclines[0].rstrip(): del doclines[0] try: skip_first_line = (doclines[0][0] not in (" \t")) except IndexError: skip_first_line = False # no lines, or first line is empty desclines = dedent("\n".join( doclines), skip_first_line=skip_first_line) desclines = desclines.splitlines(0) ## debug logging # f = open("parsePyFuncDoc.log", "a") # if 0: # f.write("\n---- %s:\n" % funcname) # f.write(pformat(siglines)+"\n") # f.write(pformat(desclines)+"\n") # else: # f.write("\n") # if siglines: # f.write("\n".join(siglines)+"\n") # else: # f.write("\n" % (scope, funcname)) # for descline in desclines: # f.write("\t%s\n" % descline) # f.close() return (siglines, desclines) #---- debugging utilities def unmark_text(markedup_text): """Parse text with potential markup as follows and return (, ). "<|>" indicates the current position (pos), defaults to the end of the text. "<+>" indicates the trigger position (trg_pos), if present. "<$>" indicates the start position (start_pos) for some kind of of processing, if present. "" is a numbered marker. N can be any of 0-99. These positions are returned as the associate number key in . Note that the positions are in UTF-8 byte counts, not character counts. This matches the behaviour of Scintilla positions. E.g.: >>> unmark_text('foo.<|>') ('foo.', {'pos': 4}) >>> unmark_text('foo.<|><+>') ('foo.', {'trg_pos': 4, 'pos': 4}) >>> unmark_text('foo.<+>ba<|>') ('foo.ba', {'trg_pos': 4, 'pos': 6}) >>> unmark_text('fo<|>o.<+>ba') ('foo.ba', {'trg_pos': 4, 'pos': 2}) >>> unmark_text('os.path.join<$>(<|>') ('os.path.join(', {'pos': 13, 'start_pos': 12}) >>> unmark_text('abc<3>defghi<2>jk<4>lm<1>nopqrstuvwxyz') ('abcdefghijklmnopqrstuvwxyz', {1: 13, 2: 9, 3: 3, 4: 11, 'pos': 26}) >>> unmark_text('Ůɳíčóďé<|>') ('Ůɳíčóďé', {'pos': 14}) See the matching markup_text() below. """ splitter = re.compile(r"(<(?:[\|\+\$\[\]<]|\d+)>)") text = "" if isinstance(markup_text, str) else "" data = {} posNameFromSymbol = { "<|>": "pos", "<+>": "trg_pos", "<$>": "start_pos", "<[>": "start_selection", "<]>": "end_selection", } def byte_length(text): if isinstance(text, str): return len(text.encode("utf-8")) return len(text) bracketed_digits_re = re.compile(r'<\d+>$') for token in splitter.split(markedup_text): if token in posNameFromSymbol: data[posNameFromSymbol[token]] = byte_length(text) elif token == "<<>": # escape sequence text += "<" elif bracketed_digits_re.match(token): data[int(token[1:-1])] = byte_length(text) else: text += token if "pos" not in data: data["pos"] = byte_length(text) # sys.stderr.write(">> text:%r, data:%s\n" % (text, data)) return text, data def markup_text(text, pos=None, trg_pos=None, start_pos=None): """Markup text with position markers. See the matching unmark_text() above. """ positions_and_markers = [] if pos is not None: positions_and_markers.append((pos, '<|>')) if trg_pos is not None: positions_and_markers.append((trg_pos, '<+>')) if start_pos is not None: positions_and_markers.append((start_pos, '<$>')) positions_and_markers.sort() if not isinstance(text, bytes): text = text.encode("utf-8") m_text = "" m_pos = 0 for position, marker in positions_and_markers: m_text += text[m_pos:position].decode('utf-8', 'ignore') + marker m_pos = position m_text += text[m_pos:].decode('utf-8', 'ignore') return m_text def lines_from_pos(unmarked_text, positions): """Get 1-based line numbers from positions @param unmarked_text {str} The text to examine @param positions {dict or list of int} Byte positions to look up @returns {dict or list of int} Matching line numbers (1-based) Given some text and either a list of positions, or a dict containing positions as values, return a matching data structure with positions replaced with the line number of the lines the positions are on. Positions after the last line are assumed to be on a hypothetical line. E.g.: Assuming the following text with \n line endings, where each line is exactly 20 characters long: >>> text = ''' ... line one ... line two ... line three ... '''.lstrip() >>> lines_from_pos(text, [5, 15, 25, 55, 999]) [1, 1, 2, 3, 4] >>> lines_from_pos(text, {"hello": 10, "moo": 20, "not": "an int"}) {'moo': 1, 'hello': 1} """ lines = str(unmarked_text).splitlines(True) offsets = [0] for line in lines: offsets.append(offsets[-1] + len(line.encode("utf-8"))) try: # assume a dict keys = iter(positions.keys()) values = {} except AttributeError: # assume a list/tuple keys = list(range(len(positions))) values = [] for key in keys: try: position = positions[key] - 0 except TypeError: continue # not a number line_no = bisect.bisect_left(offsets, position) try: values[key] = line_no except IndexError: if key == len(values): values.append(line_no) else: raise return values # Recipe: banner (1.0.1) in C:\trentm\tm\recipes\cookbook def banner(text, ch='=', length=78): """Return a banner line centering the given text. "text" is the text to show in the banner. None can be given to have no text. "ch" (optional, default '=') is the banner line character (can also be a short string to repeat). "length" (optional, default 78) is the length of banner to make. Examples: >>> banner("Peggy Sue") '================================= Peggy Sue ==================================' >>> banner("Peggy Sue", ch='-', length=50) '------------------- Peggy Sue --------------------' >>> banner("Pretty pretty pretty pretty Peggy Sue", length=40) 'Pretty pretty pretty pretty Peggy Sue' """ if text is None: return ch * length elif len(text) + 2 + len(ch) * 2 > length: # Not enough space for even one line char (plus space) around text. return text else: remain = length - (len(text) + 2) prefix_len = remain // 2 suffix_len = remain - prefix_len if len(ch) == 1: prefix = ch * prefix_len suffix = ch * suffix_len else: prefix = ch * (prefix_len // len(ch)) + ch[:prefix_len % len(ch)] suffix = ch * (suffix_len // len(ch)) + ch[:suffix_len % len(ch)] return prefix + ' ' + text + ' ' + suffix # Recipe: dedent (0.1.2) in C:\trentm\tm\recipes\cookbook def _dedentlines(lines, tabsize=8, skip_first_line=False): """_dedentlines(lines, tabsize=8, skip_first_line=False) -> dedented lines "lines" is a list of lines to dedent. "tabsize" is the tab width to use for indent width calculations. "skip_first_line" is a boolean indicating if the first line should be skipped for calculating the indent width and for dedenting. This is sometimes useful for docstrings and similar. Same as dedent() except operates on a sequence of lines. Note: the lines list is modified **in-place**. """ DEBUG = False if DEBUG: print("dedent: dedent(..., tabsize=%d, skip_first_line=%r)"\ % (tabsize, skip_first_line)) indents = [] margin = None for i, line in enumerate(lines): if i == 0 and skip_first_line: continue indent = 0 for ch in line: if ch == ' ': indent += 1 elif ch == '\t': indent += tabsize - (indent % tabsize) elif ch in '\r\n': continue # skip all-whitespace lines else: break else: continue # skip all-whitespace lines if DEBUG: print("dedent: indent=%d: %r" % (indent, line)) if margin is None: margin = indent else: margin = min(margin, indent) if DEBUG: print("dedent: margin=%r" % margin) if margin is not None and margin > 0: for i, line in enumerate(lines): if i == 0 and skip_first_line: continue removed = 0 for j, ch in enumerate(line): if ch == ' ': removed += 1 elif ch == '\t': removed += tabsize - (removed % tabsize) elif ch in '\r\n': if DEBUG: print("dedent: %r: EOL -> strip up to EOL" % line) lines[i] = lines[i][j:] break else: raise ValueError("unexpected non-whitespace char %r in " "line %r while removing %d-space margin" % (ch, line, margin)) if DEBUG: print("dedent: %r: %r -> removed %d/%d"\ % (line, ch, removed, margin)) if removed == margin: lines[i] = lines[i][j+1:] break elif removed > margin: lines[i] = ' '*(removed-margin) + lines[i][j+1:] break else: if removed: lines[i] = lines[i][removed:] return lines def dedent(text, tabsize=8, skip_first_line=False): """dedent(text, tabsize=8, skip_first_line=False) -> dedented text "text" is the text to dedent. "tabsize" is the tab width to use for indent width calculations. "skip_first_line" is a boolean indicating if the first line should be skipped for calculating the indent width and for dedenting. This is sometimes useful for docstrings and similar. textwrap.dedent(s), but don't expand tabs to spaces """ lines = text.splitlines(1) _dedentlines(lines, tabsize=tabsize, skip_first_line=skip_first_line) return ''.join(lines) # Recipe: indent (0.2.1) in C:\trentm\tm\recipes\cookbook def indent(s, width=4, skip_first_line=False): """indent(s, [width=4]) -> 's' indented by 'width' spaces The optional "skip_first_line" argument is a boolean (default False) indicating if the first line should NOT be indented. """ lines = s.splitlines(1) indentstr = ' '*width if skip_first_line: return indentstr.join(lines) else: return indentstr + indentstr.join(lines) def walk2(top, topdown=True, onerror=None, followlinks=False, ondecodeerror=None): """A version of `os.walk` that adds support for handling errors for files that cannot be decoded with the default encoding. (See bug 82268.) By default `UnicodeDecodeError`s from the os.listdir() call are ignored. If optional arg 'ondecodeerror' is specified, it should be a function; it will be called with one argument, the `UnicodeDecodeError` instance. It can report the error to continue with the walk, or raise the exception to abort the walk. """ from os.path import join, isdir, islink # We may not have read permission for top, in which case we can't # get a list of the files the directory contains. os.path.walk # always suppressed the exception then, rather than blow up for a # minor reason when (say) a thousand readable directories are still # left to visit. That logic is copied here. try: # Note that listdir and error are globals in this module due # to earlier import-*. names = os.listdir(top) except os.error as err: if onerror is not None: onerror(err) return dirs, nondirs = [], [] for name in names: try: if isdir(join(top, name)): dirs.append(name) else: nondirs.append(name) except UnicodeDecodeError as err: if ondecodeerror is not None: ondecodeerror(err) if topdown: yield top, dirs, nondirs for name in dirs: path = join(top, name) if followlinks or not islink(path): for x in walk2(path, topdown, onerror, followlinks): yield x if not topdown: yield top, dirs, nondirs # Decorators useful for timing and profiling specific functions. # # timeit usage: # Decorate the desired function and you'll get a print for how long # each call to the function took. # # hotspotit usage: # 1. decorate the desired function # 2. run your code # 3. run: # python .../codeintel/support/show_stats.py .../.prof # def timeit(func): clock = (sys.platform == "win32" and time.clock or time.time) def wrapper(*args, **kw): start_time = clock() try: return func(*args, **kw) finally: total_time = clock() - start_time print("%s took %.3fs" % (func.__name__, total_time)) return wrapper def hotshotit(func): def wrapper(*args, **kw): import hotshot global hotshotProfilers prof_name = func.__name__+".prof" profiler = hotshotProfilers.get(prof_name) if profiler is None: profiler = hotshot.Profile(prof_name) hotshotProfilers[prof_name] = profiler return profiler.runcall(func, *args, **kw) return wrapper _koCProfiler = None def getProfiler(): global _koCProfiler if _koCProfiler is None: class _KoCProfileManager(object): def __init__(self): import atexit import cProfile from codeintel2.common import _xpcom_ self.prof = cProfile.Profile() if _xpcom_: from xpcom import components _KoCProfileManager._com_interfaces_ = [ components.interfaces.nsIObserver] obsSvc = components.classes["@mozilla.org/observer-service;1"].\ getService( components.interfaces.nsIObserverService) obsSvc.addObserver(self, 'xpcom-shutdown', False) else: atexit.register(self.atexit_handler) def atexit_handler(self): self.prof.print_stats(sort="time") def observe(self, subject, topic, data): if topic == "xpcom-shutdown": self.atexit_handler() _koCProfiler = _KoCProfileManager() return _koCProfiler.prof def profile_method(func): def wrapper(*args, **kw): return getProfiler().runcall(func, *args, **kw) return wrapper # Utility functions to perform sorting the same way as scintilla does it # for the code-completion list. def OrdPunctLast(value): result = [] value = value.upper() for ch in value: i = ord(ch) if i >= 0x21 and i <= 0x2F: # ch >= '!' && ch <= '/' result.append(chr(i - ord("!") + ord('['))) # ch - '!' + '[' elif i >= 0x3A and i <= 0x40: # ch >= ':' && ch <= '@' result.append(chr(i - ord(":") + ord('['))) # ch - ':' + '[' else: result.append(ch) return "".join(result) def CompareNPunctLast(value1, value2): # value 1 is smaller, return negative # value 1 is equal, return 0 # value 1 is larger, return positive return cmp(OrdPunctLast(value1), OrdPunctLast(value2)) # Utility function to make a lookup dictionary def make_short_name_dict(names, length=3): outdict = {} for name in names: if len(name) >= length: shortname = name[:length] l = outdict.get(shortname) if not l: outdict[shortname] = [name] else: l.append(name) # pprint(outdict) for values in list(outdict.values()): values.sort(key=OrdPunctLast) return outdict def makePerformantLogger(logger): """Replaces the info() and debug() methods with dummy methods. Assumes that the logging level does not change during runtime. """ if not logger.isEnabledFor(logging.INFO): def _log_ignore(self, *args, **kwargs): pass logger.info = _log_ignore if not logger.isEnabledFor(logging.DEBUG): logger.debug = _log_ignore #---- mainline self-test if __name__ == "__main__": import doctest doctest.testmod()